diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,56034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.4, + "eval_steps": 500, + "global_step": 80000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 5e-05, + "grad_norm": 2.3343162536621094, + "learning_rate": 2.25e-06, + "loss": 9.078922271728516, + "step": 10 + }, + { + "epoch": 0.0001, + "grad_norm": 2.257415533065796, + "learning_rate": 4.75e-06, + "loss": 9.042113494873046, + "step": 20 + }, + { + "epoch": 0.00015, + "grad_norm": 2.5989434719085693, + "learning_rate": 7.250000000000001e-06, + "loss": 8.959950256347657, + "step": 30 + }, + { + "epoch": 0.0002, + "grad_norm": 2.651371717453003, + "learning_rate": 9.75e-06, + "loss": 8.819075012207032, + "step": 40 + }, + { + "epoch": 0.00025, + "grad_norm": 2.047802209854126, + "learning_rate": 1.2250000000000001e-05, + "loss": 8.612620544433593, + "step": 50 + }, + { + "epoch": 0.0003, + "grad_norm": 1.632555365562439, + "learning_rate": 1.475e-05, + "loss": 8.42562255859375, + "step": 60 + }, + { + "epoch": 0.00035, + "grad_norm": 1.4303501844406128, + "learning_rate": 1.7250000000000003e-05, + "loss": 8.29529266357422, + "step": 70 + }, + { + "epoch": 0.0004, + "grad_norm": 1.397361159324646, + "learning_rate": 1.975e-05, + "loss": 8.194226837158203, + "step": 80 + }, + { + "epoch": 0.00045, + "grad_norm": 1.376143217086792, + "learning_rate": 2.225e-05, + "loss": 8.092294311523437, + "step": 90 + }, + { + "epoch": 0.0005, + "grad_norm": 1.2575387954711914, + "learning_rate": 2.4750000000000002e-05, + "loss": 7.982857513427734, + "step": 100 + }, + { + "epoch": 0.00055, + "grad_norm": 1.1970226764678955, + "learning_rate": 2.725e-05, + "loss": 7.897898101806641, + "step": 110 + }, + { + "epoch": 0.0006, + "grad_norm": 1.1119627952575684, + "learning_rate": 2.9749999999999998e-05, + "loss": 7.7918754577636715, + "step": 120 + }, + { + "epoch": 0.00065, + "grad_norm": 1.0283393859863281, + "learning_rate": 3.2250000000000005e-05, + "loss": 7.689813232421875, + "step": 130 + }, + { + "epoch": 0.0007, + "grad_norm": 0.8330586552619934, + "learning_rate": 3.4750000000000004e-05, + "loss": 7.600614166259765, + "step": 140 + }, + { + "epoch": 0.00075, + "grad_norm": 0.819470226764679, + "learning_rate": 3.725e-05, + "loss": 7.517872619628906, + "step": 150 + }, + { + "epoch": 0.0008, + "grad_norm": 0.6170676350593567, + "learning_rate": 3.9750000000000004e-05, + "loss": 7.4453590393066404, + "step": 160 + }, + { + "epoch": 0.00085, + "grad_norm": 0.6371561884880066, + "learning_rate": 4.2250000000000004e-05, + "loss": 7.3723396301269535, + "step": 170 + }, + { + "epoch": 0.0009, + "grad_norm": 0.4984106123447418, + "learning_rate": 4.475e-05, + "loss": 7.3142822265625, + "step": 180 + }, + { + "epoch": 0.00095, + "grad_norm": 0.4508594870567322, + "learning_rate": 4.725e-05, + "loss": 7.2722938537597654, + "step": 190 + }, + { + "epoch": 0.001, + "grad_norm": 0.4540443420410156, + "learning_rate": 4.975e-05, + "loss": 7.25008544921875, + "step": 200 + }, + { + "epoch": 0.00105, + "grad_norm": 0.7292822003364563, + "learning_rate": 5.2249999999999996e-05, + "loss": 7.183808898925781, + "step": 210 + }, + { + "epoch": 0.0011, + "grad_norm": 0.5947257876396179, + "learning_rate": 5.475e-05, + "loss": 7.181377410888672, + "step": 220 + }, + { + "epoch": 0.00115, + "grad_norm": 0.6903289556503296, + "learning_rate": 5.725e-05, + "loss": 7.116062164306641, + "step": 230 + }, + { + "epoch": 0.0012, + "grad_norm": 0.5764403939247131, + "learning_rate": 5.9749999999999995e-05, + "loss": 7.085305786132812, + "step": 240 + }, + { + "epoch": 0.00125, + "grad_norm": 0.7021174430847168, + "learning_rate": 6.225e-05, + "loss": 7.051827239990234, + "step": 250 + }, + { + "epoch": 0.0013, + "grad_norm": 0.7887938022613525, + "learning_rate": 6.475e-05, + "loss": 7.002043914794922, + "step": 260 + }, + { + "epoch": 0.00135, + "grad_norm": 1.1115361452102661, + "learning_rate": 6.725000000000001e-05, + "loss": 6.945926666259766, + "step": 270 + }, + { + "epoch": 0.0014, + "grad_norm": 1.2276875972747803, + "learning_rate": 6.975e-05, + "loss": 6.907218170166016, + "step": 280 + }, + { + "epoch": 0.00145, + "grad_norm": 0.7439374327659607, + "learning_rate": 7.225e-05, + "loss": 6.885887145996094, + "step": 290 + }, + { + "epoch": 0.0015, + "grad_norm": 1.1284825801849365, + "learning_rate": 7.475e-05, + "loss": 6.812039184570312, + "step": 300 + }, + { + "epoch": 0.00155, + "grad_norm": 0.9264832735061646, + "learning_rate": 7.725000000000001e-05, + "loss": 6.757131195068359, + "step": 310 + }, + { + "epoch": 0.0016, + "grad_norm": 0.9103614091873169, + "learning_rate": 7.975e-05, + "loss": 6.683384704589844, + "step": 320 + }, + { + "epoch": 0.00165, + "grad_norm": 0.8917039632797241, + "learning_rate": 8.225000000000001e-05, + "loss": 6.653724670410156, + "step": 330 + }, + { + "epoch": 0.0017, + "grad_norm": 1.1469368934631348, + "learning_rate": 8.475000000000001e-05, + "loss": 6.583345031738281, + "step": 340 + }, + { + "epoch": 0.00175, + "grad_norm": 0.9747764468193054, + "learning_rate": 8.724999999999999e-05, + "loss": 6.557423400878906, + "step": 350 + }, + { + "epoch": 0.0018, + "grad_norm": 1.4919801950454712, + "learning_rate": 8.975e-05, + "loss": 6.460021209716797, + "step": 360 + }, + { + "epoch": 0.00185, + "grad_norm": 1.528967261314392, + "learning_rate": 9.225e-05, + "loss": 6.4359130859375, + "step": 370 + }, + { + "epoch": 0.0019, + "grad_norm": 0.9848876595497131, + "learning_rate": 9.475e-05, + "loss": 6.369429397583008, + "step": 380 + }, + { + "epoch": 0.00195, + "grad_norm": 1.3570668697357178, + "learning_rate": 9.725e-05, + "loss": 6.35345344543457, + "step": 390 + }, + { + "epoch": 0.002, + "grad_norm": 1.023625135421753, + "learning_rate": 9.975000000000001e-05, + "loss": 6.303249740600586, + "step": 400 + }, + { + "epoch": 0.00205, + "grad_norm": 0.9506981372833252, + "learning_rate": 0.00010224999999999999, + "loss": 6.273816299438477, + "step": 410 + }, + { + "epoch": 0.0021, + "grad_norm": 1.207095742225647, + "learning_rate": 0.00010475, + "loss": 6.219621658325195, + "step": 420 + }, + { + "epoch": 0.00215, + "grad_norm": 1.339953064918518, + "learning_rate": 0.00010725, + "loss": 6.237022781372071, + "step": 430 + }, + { + "epoch": 0.0022, + "grad_norm": 1.1513832807540894, + "learning_rate": 0.00010975, + "loss": 6.1426136016845705, + "step": 440 + }, + { + "epoch": 0.00225, + "grad_norm": 1.417026162147522, + "learning_rate": 0.00011225, + "loss": 6.1436103820800785, + "step": 450 + }, + { + "epoch": 0.0023, + "grad_norm": 1.1084486246109009, + "learning_rate": 0.00011475000000000001, + "loss": 6.0862174987792965, + "step": 460 + }, + { + "epoch": 0.00235, + "grad_norm": 1.1368422508239746, + "learning_rate": 0.00011724999999999999, + "loss": 6.076276016235352, + "step": 470 + }, + { + "epoch": 0.0024, + "grad_norm": 1.2283637523651123, + "learning_rate": 0.00011975, + "loss": 6.028984832763672, + "step": 480 + }, + { + "epoch": 0.00245, + "grad_norm": 1.2255951166152954, + "learning_rate": 0.00012225, + "loss": 6.009500122070312, + "step": 490 + }, + { + "epoch": 0.0025, + "grad_norm": 1.0867785215377808, + "learning_rate": 0.00012475, + "loss": 5.959277725219726, + "step": 500 + }, + { + "epoch": 0.00255, + "grad_norm": 1.3084031343460083, + "learning_rate": 0.00012725, + "loss": 5.943234634399414, + "step": 510 + }, + { + "epoch": 0.0026, + "grad_norm": 1.0814982652664185, + "learning_rate": 0.00012975, + "loss": 5.909140396118164, + "step": 520 + }, + { + "epoch": 0.00265, + "grad_norm": 1.337663173675537, + "learning_rate": 0.00013225000000000002, + "loss": 5.863074874877929, + "step": 530 + }, + { + "epoch": 0.0027, + "grad_norm": 0.9246945381164551, + "learning_rate": 0.00013475000000000002, + "loss": 5.882530212402344, + "step": 540 + }, + { + "epoch": 0.00275, + "grad_norm": 1.1996309757232666, + "learning_rate": 0.00013725, + "loss": 5.83284912109375, + "step": 550 + }, + { + "epoch": 0.0028, + "grad_norm": 1.2468345165252686, + "learning_rate": 0.00013975, + "loss": 5.825077056884766, + "step": 560 + }, + { + "epoch": 0.00285, + "grad_norm": 0.8803932070732117, + "learning_rate": 0.00014225, + "loss": 5.762301635742188, + "step": 570 + }, + { + "epoch": 0.0029, + "grad_norm": 1.1504731178283691, + "learning_rate": 0.00014475, + "loss": 5.776547241210937, + "step": 580 + }, + { + "epoch": 0.00295, + "grad_norm": 1.0328158140182495, + "learning_rate": 0.00014725, + "loss": 5.740188217163086, + "step": 590 + }, + { + "epoch": 0.003, + "grad_norm": 1.5344996452331543, + "learning_rate": 0.00014975, + "loss": 5.711307525634766, + "step": 600 + }, + { + "epoch": 0.00305, + "grad_norm": 1.369181752204895, + "learning_rate": 0.00015225, + "loss": 5.698526763916016, + "step": 610 + }, + { + "epoch": 0.0031, + "grad_norm": 1.2572104930877686, + "learning_rate": 0.00015475, + "loss": 5.650114059448242, + "step": 620 + }, + { + "epoch": 0.00315, + "grad_norm": 1.2770206928253174, + "learning_rate": 0.00015725, + "loss": 5.63488655090332, + "step": 630 + }, + { + "epoch": 0.0032, + "grad_norm": 1.2198976278305054, + "learning_rate": 0.00015975, + "loss": 5.577988052368164, + "step": 640 + }, + { + "epoch": 0.00325, + "grad_norm": 1.4904592037200928, + "learning_rate": 0.00016225000000000001, + "loss": 5.599241256713867, + "step": 650 + }, + { + "epoch": 0.0033, + "grad_norm": 1.2390382289886475, + "learning_rate": 0.00016475000000000002, + "loss": 5.574135971069336, + "step": 660 + }, + { + "epoch": 0.00335, + "grad_norm": 1.3068722486495972, + "learning_rate": 0.00016725000000000003, + "loss": 5.5108482360839846, + "step": 670 + }, + { + "epoch": 0.0034, + "grad_norm": 1.4691752195358276, + "learning_rate": 0.00016975, + "loss": 5.519583129882813, + "step": 680 + }, + { + "epoch": 0.00345, + "grad_norm": 1.3841990232467651, + "learning_rate": 0.00017224999999999999, + "loss": 5.513322448730468, + "step": 690 + }, + { + "epoch": 0.0035, + "grad_norm": 1.281400203704834, + "learning_rate": 0.00017475, + "loss": 5.438730239868164, + "step": 700 + }, + { + "epoch": 0.00355, + "grad_norm": 1.0936923027038574, + "learning_rate": 0.00017725, + "loss": 5.462588119506836, + "step": 710 + }, + { + "epoch": 0.0036, + "grad_norm": 1.2100780010223389, + "learning_rate": 0.00017975, + "loss": 5.431028747558594, + "step": 720 + }, + { + "epoch": 0.00365, + "grad_norm": 1.2060819864273071, + "learning_rate": 0.00018225, + "loss": 5.40233268737793, + "step": 730 + }, + { + "epoch": 0.0037, + "grad_norm": 1.3502517938613892, + "learning_rate": 0.00018475, + "loss": 5.364802551269531, + "step": 740 + }, + { + "epoch": 0.00375, + "grad_norm": 1.3670893907546997, + "learning_rate": 0.00018725, + "loss": 5.394402313232422, + "step": 750 + }, + { + "epoch": 0.0038, + "grad_norm": 1.259055495262146, + "learning_rate": 0.00018975, + "loss": 5.328908538818359, + "step": 760 + }, + { + "epoch": 0.00385, + "grad_norm": 1.1133089065551758, + "learning_rate": 0.00019225, + "loss": 5.274171447753906, + "step": 770 + }, + { + "epoch": 0.0039, + "grad_norm": 1.1152657270431519, + "learning_rate": 0.00019475000000000002, + "loss": 5.282376861572265, + "step": 780 + }, + { + "epoch": 0.00395, + "grad_norm": 1.1693851947784424, + "learning_rate": 0.00019725000000000002, + "loss": 5.270669174194336, + "step": 790 + }, + { + "epoch": 0.004, + "grad_norm": 1.185726284980774, + "learning_rate": 0.00019975, + "loss": 5.285358047485351, + "step": 800 + }, + { + "epoch": 0.00405, + "grad_norm": 1.2114828824996948, + "learning_rate": 0.00020225, + "loss": 5.22089958190918, + "step": 810 + }, + { + "epoch": 0.0041, + "grad_norm": 1.764055848121643, + "learning_rate": 0.00020475, + "loss": 5.1827880859375, + "step": 820 + }, + { + "epoch": 0.00415, + "grad_norm": 1.1757320165634155, + "learning_rate": 0.00020725, + "loss": 5.193819427490235, + "step": 830 + }, + { + "epoch": 0.0042, + "grad_norm": 1.2490066289901733, + "learning_rate": 0.00020975, + "loss": 5.146733093261719, + "step": 840 + }, + { + "epoch": 0.00425, + "grad_norm": 1.3350523710250854, + "learning_rate": 0.00021225, + "loss": 5.143191528320313, + "step": 850 + }, + { + "epoch": 0.0043, + "grad_norm": 1.1717677116394043, + "learning_rate": 0.00021475, + "loss": 5.085927963256836, + "step": 860 + }, + { + "epoch": 0.00435, + "grad_norm": 1.159227728843689, + "learning_rate": 0.00021725, + "loss": 5.087388229370117, + "step": 870 + }, + { + "epoch": 0.0044, + "grad_norm": 1.273136854171753, + "learning_rate": 0.00021975, + "loss": 5.081501770019531, + "step": 880 + }, + { + "epoch": 0.00445, + "grad_norm": 1.1953152418136597, + "learning_rate": 0.00022225, + "loss": 5.046857070922852, + "step": 890 + }, + { + "epoch": 0.0045, + "grad_norm": 1.040152907371521, + "learning_rate": 0.00022475000000000001, + "loss": 5.06628189086914, + "step": 900 + }, + { + "epoch": 0.00455, + "grad_norm": 1.2100125551223755, + "learning_rate": 0.00022725000000000002, + "loss": 5.018717956542969, + "step": 910 + }, + { + "epoch": 0.0046, + "grad_norm": 1.4811681509017944, + "learning_rate": 0.00022975000000000003, + "loss": 4.991384887695313, + "step": 920 + }, + { + "epoch": 0.00465, + "grad_norm": 1.1811575889587402, + "learning_rate": 0.00023225, + "loss": 4.962765502929687, + "step": 930 + }, + { + "epoch": 0.0047, + "grad_norm": 1.1551120281219482, + "learning_rate": 0.00023475, + "loss": 4.994625091552734, + "step": 940 + }, + { + "epoch": 0.00475, + "grad_norm": 1.3614791631698608, + "learning_rate": 0.00023725, + "loss": 4.942966461181641, + "step": 950 + }, + { + "epoch": 0.0048, + "grad_norm": 1.091278314590454, + "learning_rate": 0.00023975, + "loss": 4.937471389770508, + "step": 960 + }, + { + "epoch": 0.00485, + "grad_norm": 1.4918296337127686, + "learning_rate": 0.00024225, + "loss": 4.940725326538086, + "step": 970 + }, + { + "epoch": 0.0049, + "grad_norm": 1.2006937265396118, + "learning_rate": 0.00024475, + "loss": 4.896607208251953, + "step": 980 + }, + { + "epoch": 0.00495, + "grad_norm": 1.0817883014678955, + "learning_rate": 0.00024725, + "loss": 4.889888000488281, + "step": 990 + }, + { + "epoch": 0.005, + "grad_norm": 1.1135517358779907, + "learning_rate": 0.00024975, + "loss": 4.894549560546875, + "step": 1000 + }, + { + "epoch": 0.00505, + "grad_norm": 1.363113284111023, + "learning_rate": 0.00025225, + "loss": 4.853108596801758, + "step": 1010 + }, + { + "epoch": 0.0051, + "grad_norm": 1.0389511585235596, + "learning_rate": 0.00025475, + "loss": 4.806407165527344, + "step": 1020 + }, + { + "epoch": 0.00515, + "grad_norm": 1.491611123085022, + "learning_rate": 0.00025725, + "loss": 4.87169075012207, + "step": 1030 + }, + { + "epoch": 0.0052, + "grad_norm": 1.2202712297439575, + "learning_rate": 0.00025975, + "loss": 4.798575973510742, + "step": 1040 + }, + { + "epoch": 0.00525, + "grad_norm": 1.2685787677764893, + "learning_rate": 0.00026225, + "loss": 4.787151336669922, + "step": 1050 + }, + { + "epoch": 0.0053, + "grad_norm": 1.3609626293182373, + "learning_rate": 0.00026475, + "loss": 4.8220054626464846, + "step": 1060 + }, + { + "epoch": 0.00535, + "grad_norm": 0.9951308965682983, + "learning_rate": 0.00026725, + "loss": 4.778158569335938, + "step": 1070 + }, + { + "epoch": 0.0054, + "grad_norm": 1.1353764533996582, + "learning_rate": 0.00026974999999999997, + "loss": 4.73405876159668, + "step": 1080 + }, + { + "epoch": 0.00545, + "grad_norm": 1.0133823156356812, + "learning_rate": 0.00027225, + "loss": 4.759817123413086, + "step": 1090 + }, + { + "epoch": 0.0055, + "grad_norm": 1.1280566453933716, + "learning_rate": 0.00027475, + "loss": 4.739881896972657, + "step": 1100 + }, + { + "epoch": 0.00555, + "grad_norm": 1.1176393032073975, + "learning_rate": 0.00027725, + "loss": 4.728521347045898, + "step": 1110 + }, + { + "epoch": 0.0056, + "grad_norm": 1.0539538860321045, + "learning_rate": 0.00027975, + "loss": 4.690623474121094, + "step": 1120 + }, + { + "epoch": 0.00565, + "grad_norm": 1.1005791425704956, + "learning_rate": 0.00028225, + "loss": 4.668269348144531, + "step": 1130 + }, + { + "epoch": 0.0057, + "grad_norm": 1.0565214157104492, + "learning_rate": 0.00028475, + "loss": 4.623973083496094, + "step": 1140 + }, + { + "epoch": 0.00575, + "grad_norm": 1.1037288904190063, + "learning_rate": 0.00028725, + "loss": 4.682107543945312, + "step": 1150 + }, + { + "epoch": 0.0058, + "grad_norm": 1.0827094316482544, + "learning_rate": 0.00028975, + "loss": 4.628668212890625, + "step": 1160 + }, + { + "epoch": 0.00585, + "grad_norm": 1.1619179248809814, + "learning_rate": 0.00029225000000000003, + "loss": 4.6247306823730465, + "step": 1170 + }, + { + "epoch": 0.0059, + "grad_norm": 1.028248906135559, + "learning_rate": 0.00029475000000000004, + "loss": 4.612622451782227, + "step": 1180 + }, + { + "epoch": 0.00595, + "grad_norm": 1.2123770713806152, + "learning_rate": 0.00029725000000000004, + "loss": 4.5824028015136715, + "step": 1190 + }, + { + "epoch": 0.006, + "grad_norm": 1.0007641315460205, + "learning_rate": 0.00029975000000000005, + "loss": 4.621369171142578, + "step": 1200 + }, + { + "epoch": 0.00605, + "grad_norm": 1.146349310874939, + "learning_rate": 0.00030225, + "loss": 4.593479156494141, + "step": 1210 + }, + { + "epoch": 0.0061, + "grad_norm": 1.0729461908340454, + "learning_rate": 0.00030475, + "loss": 4.567705154418945, + "step": 1220 + }, + { + "epoch": 0.00615, + "grad_norm": 1.1311410665512085, + "learning_rate": 0.00030725, + "loss": 4.543987655639649, + "step": 1230 + }, + { + "epoch": 0.0062, + "grad_norm": 0.9540078043937683, + "learning_rate": 0.00030975, + "loss": 4.574246978759765, + "step": 1240 + }, + { + "epoch": 0.00625, + "grad_norm": 1.12738835811615, + "learning_rate": 0.00031225000000000003, + "loss": 4.522233200073242, + "step": 1250 + }, + { + "epoch": 0.0063, + "grad_norm": 1.1553608179092407, + "learning_rate": 0.00031475, + "loss": 4.525931167602539, + "step": 1260 + }, + { + "epoch": 0.00635, + "grad_norm": 1.1553597450256348, + "learning_rate": 0.00031725, + "loss": 4.520528411865234, + "step": 1270 + }, + { + "epoch": 0.0064, + "grad_norm": 1.070203185081482, + "learning_rate": 0.00031975, + "loss": 4.509066390991211, + "step": 1280 + }, + { + "epoch": 0.00645, + "grad_norm": 1.0271059274673462, + "learning_rate": 0.00032225, + "loss": 4.476989364624023, + "step": 1290 + }, + { + "epoch": 0.0065, + "grad_norm": 0.9966850876808167, + "learning_rate": 0.00032475, + "loss": 4.4804534912109375, + "step": 1300 + }, + { + "epoch": 0.00655, + "grad_norm": 1.0687849521636963, + "learning_rate": 0.00032725, + "loss": 4.432140350341797, + "step": 1310 + }, + { + "epoch": 0.0066, + "grad_norm": 1.1434048414230347, + "learning_rate": 0.00032975, + "loss": 4.456147766113281, + "step": 1320 + }, + { + "epoch": 0.00665, + "grad_norm": 0.895724892616272, + "learning_rate": 0.00033224999999999997, + "loss": 4.405745315551758, + "step": 1330 + }, + { + "epoch": 0.0067, + "grad_norm": 1.0732239484786987, + "learning_rate": 0.00033475, + "loss": 4.395022201538086, + "step": 1340 + }, + { + "epoch": 0.00675, + "grad_norm": 1.1442835330963135, + "learning_rate": 0.00033725, + "loss": 4.4049835205078125, + "step": 1350 + }, + { + "epoch": 0.0068, + "grad_norm": 0.9532873034477234, + "learning_rate": 0.00033975, + "loss": 4.395751190185547, + "step": 1360 + }, + { + "epoch": 0.00685, + "grad_norm": 1.072645664215088, + "learning_rate": 0.00034225, + "loss": 4.355947113037109, + "step": 1370 + }, + { + "epoch": 0.0069, + "grad_norm": 1.0552868843078613, + "learning_rate": 0.00034475, + "loss": 4.313367080688477, + "step": 1380 + }, + { + "epoch": 0.00695, + "grad_norm": 1.1490488052368164, + "learning_rate": 0.00034725, + "loss": 4.288275146484375, + "step": 1390 + }, + { + "epoch": 0.007, + "grad_norm": 1.0485172271728516, + "learning_rate": 0.00034975, + "loss": 4.2844291687011715, + "step": 1400 + }, + { + "epoch": 0.00705, + "grad_norm": 0.9210212230682373, + "learning_rate": 0.00035225, + "loss": 4.305771255493164, + "step": 1410 + }, + { + "epoch": 0.0071, + "grad_norm": 0.9192109107971191, + "learning_rate": 0.00035475000000000003, + "loss": 4.30635986328125, + "step": 1420 + }, + { + "epoch": 0.00715, + "grad_norm": 0.9939811825752258, + "learning_rate": 0.00035725000000000004, + "loss": 4.2772071838378904, + "step": 1430 + }, + { + "epoch": 0.0072, + "grad_norm": 1.1245921850204468, + "learning_rate": 0.00035975000000000004, + "loss": 4.271445083618164, + "step": 1440 + }, + { + "epoch": 0.00725, + "grad_norm": 0.9126390814781189, + "learning_rate": 0.00036225000000000005, + "loss": 4.236653900146484, + "step": 1450 + }, + { + "epoch": 0.0073, + "grad_norm": 0.9839813709259033, + "learning_rate": 0.00036475, + "loss": 4.2448585510253904, + "step": 1460 + }, + { + "epoch": 0.00735, + "grad_norm": 0.9173403978347778, + "learning_rate": 0.00036725, + "loss": 4.190647888183594, + "step": 1470 + }, + { + "epoch": 0.0074, + "grad_norm": 0.9106733798980713, + "learning_rate": 0.00036975, + "loss": 4.231861114501953, + "step": 1480 + }, + { + "epoch": 0.00745, + "grad_norm": 0.9079275131225586, + "learning_rate": 0.00037225, + "loss": 4.2070762634277346, + "step": 1490 + }, + { + "epoch": 0.0075, + "grad_norm": 0.9809225797653198, + "learning_rate": 0.00037475000000000003, + "loss": 4.127109146118164, + "step": 1500 + }, + { + "epoch": 0.00755, + "grad_norm": 1.0052870512008667, + "learning_rate": 0.00037725, + "loss": 4.232234191894531, + "step": 1510 + }, + { + "epoch": 0.0076, + "grad_norm": 0.906537652015686, + "learning_rate": 0.00037975, + "loss": 4.1622314453125, + "step": 1520 + }, + { + "epoch": 0.00765, + "grad_norm": 1.0689282417297363, + "learning_rate": 0.00038225, + "loss": 4.135288238525391, + "step": 1530 + }, + { + "epoch": 0.0077, + "grad_norm": 0.820002555847168, + "learning_rate": 0.00038475, + "loss": 4.1841175079345705, + "step": 1540 + }, + { + "epoch": 0.00775, + "grad_norm": 0.8886502981185913, + "learning_rate": 0.00038725, + "loss": 4.110052871704101, + "step": 1550 + }, + { + "epoch": 0.0078, + "grad_norm": 0.8644295930862427, + "learning_rate": 0.00038975, + "loss": 4.16099967956543, + "step": 1560 + }, + { + "epoch": 0.00785, + "grad_norm": 0.9768062829971313, + "learning_rate": 0.00039225, + "loss": 4.136515426635742, + "step": 1570 + }, + { + "epoch": 0.0079, + "grad_norm": 0.8727269172668457, + "learning_rate": 0.00039474999999999997, + "loss": 4.084750747680664, + "step": 1580 + }, + { + "epoch": 0.00795, + "grad_norm": 0.8942521214485168, + "learning_rate": 0.00039725, + "loss": 4.082816314697266, + "step": 1590 + }, + { + "epoch": 0.008, + "grad_norm": 0.844272792339325, + "learning_rate": 0.00039975, + "loss": 4.084886932373047, + "step": 1600 + }, + { + "epoch": 0.00805, + "grad_norm": 0.8976606726646423, + "learning_rate": 0.00040225, + "loss": 4.046478271484375, + "step": 1610 + }, + { + "epoch": 0.0081, + "grad_norm": 0.8926315307617188, + "learning_rate": 0.00040475, + "loss": 4.093951416015625, + "step": 1620 + }, + { + "epoch": 0.00815, + "grad_norm": 1.0100226402282715, + "learning_rate": 0.00040725, + "loss": 4.065290069580078, + "step": 1630 + }, + { + "epoch": 0.0082, + "grad_norm": 0.8588587641716003, + "learning_rate": 0.00040975, + "loss": 4.076088333129883, + "step": 1640 + }, + { + "epoch": 0.00825, + "grad_norm": 0.978767991065979, + "learning_rate": 0.00041225, + "loss": 4.048191070556641, + "step": 1650 + }, + { + "epoch": 0.0083, + "grad_norm": 0.8871893882751465, + "learning_rate": 0.00041475, + "loss": 4.033791732788086, + "step": 1660 + }, + { + "epoch": 0.00835, + "grad_norm": 1.6964958906173706, + "learning_rate": 0.00041725000000000003, + "loss": 4.047354507446289, + "step": 1670 + }, + { + "epoch": 0.0084, + "grad_norm": 0.8660668134689331, + "learning_rate": 0.00041975000000000004, + "loss": 4.099149322509765, + "step": 1680 + }, + { + "epoch": 0.00845, + "grad_norm": 0.893744707107544, + "learning_rate": 0.00042225000000000005, + "loss": 4.010662460327149, + "step": 1690 + }, + { + "epoch": 0.0085, + "grad_norm": 0.8851510286331177, + "learning_rate": 0.00042475000000000005, + "loss": 4.044161605834961, + "step": 1700 + }, + { + "epoch": 0.00855, + "grad_norm": 0.8414413332939148, + "learning_rate": 0.00042725, + "loss": 4.028578567504883, + "step": 1710 + }, + { + "epoch": 0.0086, + "grad_norm": 0.7908818125724792, + "learning_rate": 0.00042975, + "loss": 4.05084228515625, + "step": 1720 + }, + { + "epoch": 0.00865, + "grad_norm": 0.8644911646842957, + "learning_rate": 0.00043225, + "loss": 4.023723220825195, + "step": 1730 + }, + { + "epoch": 0.0087, + "grad_norm": 0.9518306851387024, + "learning_rate": 0.00043475, + "loss": 4.001881790161133, + "step": 1740 + }, + { + "epoch": 0.00875, + "grad_norm": 0.818665087223053, + "learning_rate": 0.00043725000000000003, + "loss": 3.9804187774658204, + "step": 1750 + }, + { + "epoch": 0.0088, + "grad_norm": 0.8849632143974304, + "learning_rate": 0.00043975, + "loss": 3.9829757690429686, + "step": 1760 + }, + { + "epoch": 0.00885, + "grad_norm": 0.8573206067085266, + "learning_rate": 0.00044225, + "loss": 3.993016815185547, + "step": 1770 + }, + { + "epoch": 0.0089, + "grad_norm": 0.841739296913147, + "learning_rate": 0.00044475, + "loss": 3.9397891998291015, + "step": 1780 + }, + { + "epoch": 0.00895, + "grad_norm": 0.8425242304801941, + "learning_rate": 0.00044725, + "loss": 3.956979751586914, + "step": 1790 + }, + { + "epoch": 0.009, + "grad_norm": 0.7534254789352417, + "learning_rate": 0.00044975, + "loss": 3.956821823120117, + "step": 1800 + }, + { + "epoch": 0.00905, + "grad_norm": 0.8840813040733337, + "learning_rate": 0.00045225, + "loss": 3.9565155029296877, + "step": 1810 + }, + { + "epoch": 0.0091, + "grad_norm": 0.8877798318862915, + "learning_rate": 0.00045475, + "loss": 3.9765884399414064, + "step": 1820 + }, + { + "epoch": 0.00915, + "grad_norm": 1.0531920194625854, + "learning_rate": 0.00045725, + "loss": 3.933223342895508, + "step": 1830 + }, + { + "epoch": 0.0092, + "grad_norm": 0.8447864055633545, + "learning_rate": 0.00045975, + "loss": 3.9445892333984376, + "step": 1840 + }, + { + "epoch": 0.00925, + "grad_norm": 0.9871447682380676, + "learning_rate": 0.00046225, + "loss": 3.9233097076416015, + "step": 1850 + }, + { + "epoch": 0.0093, + "grad_norm": 0.8590869903564453, + "learning_rate": 0.00046475, + "loss": 3.950835037231445, + "step": 1860 + }, + { + "epoch": 0.00935, + "grad_norm": 0.779175877571106, + "learning_rate": 0.00046725, + "loss": 3.956658935546875, + "step": 1870 + }, + { + "epoch": 0.0094, + "grad_norm": 0.853615403175354, + "learning_rate": 0.00046975, + "loss": 3.9342689514160156, + "step": 1880 + }, + { + "epoch": 0.00945, + "grad_norm": 0.800541341304779, + "learning_rate": 0.00047225, + "loss": 3.921456146240234, + "step": 1890 + }, + { + "epoch": 0.0095, + "grad_norm": 1.5777511596679688, + "learning_rate": 0.00047475, + "loss": 3.894826126098633, + "step": 1900 + }, + { + "epoch": 0.00955, + "grad_norm": 0.8786264657974243, + "learning_rate": 0.00047725, + "loss": 3.92449951171875, + "step": 1910 + }, + { + "epoch": 0.0096, + "grad_norm": 0.8234652280807495, + "learning_rate": 0.00047975000000000003, + "loss": 3.9034229278564454, + "step": 1920 + }, + { + "epoch": 0.00965, + "grad_norm": 0.825477123260498, + "learning_rate": 0.00048225000000000004, + "loss": 3.9134811401367187, + "step": 1930 + }, + { + "epoch": 0.0097, + "grad_norm": 0.8801007866859436, + "learning_rate": 0.00048475000000000005, + "loss": 3.9166229248046873, + "step": 1940 + }, + { + "epoch": 0.00975, + "grad_norm": 0.8332860469818115, + "learning_rate": 0.00048725000000000005, + "loss": 3.911066436767578, + "step": 1950 + }, + { + "epoch": 0.0098, + "grad_norm": 0.8440579175949097, + "learning_rate": 0.0004897500000000001, + "loss": 3.9266098022460936, + "step": 1960 + }, + { + "epoch": 0.00985, + "grad_norm": 0.859488844871521, + "learning_rate": 0.0004922500000000001, + "loss": 3.854235076904297, + "step": 1970 + }, + { + "epoch": 0.0099, + "grad_norm": 0.9074660539627075, + "learning_rate": 0.0004947500000000001, + "loss": 3.883164978027344, + "step": 1980 + }, + { + "epoch": 0.00995, + "grad_norm": 0.8589808940887451, + "learning_rate": 0.0004972500000000001, + "loss": 3.8719593048095704, + "step": 1990 + }, + { + "epoch": 0.01, + "grad_norm": 0.7976661920547485, + "learning_rate": 0.0004997500000000001, + "loss": 3.8751861572265627, + "step": 2000 + }, + { + "epoch": 0.01005, + "grad_norm": 0.855452835559845, + "learning_rate": 0.0004999999974510319, + "loss": 3.884695816040039, + "step": 2010 + }, + { + "epoch": 0.0101, + "grad_norm": 0.8024077415466309, + "learning_rate": 0.0004999999886397844, + "loss": 3.8703163146972654, + "step": 2020 + }, + { + "epoch": 0.01015, + "grad_norm": 0.7268491387367249, + "learning_rate": 0.0004999999735347887, + "loss": 3.830159378051758, + "step": 2030 + }, + { + "epoch": 0.0102, + "grad_norm": 0.7871618270874023, + "learning_rate": 0.0004999999521360455, + "loss": 3.8416542053222655, + "step": 2040 + }, + { + "epoch": 0.01025, + "grad_norm": 0.9002761840820312, + "learning_rate": 0.000499999924443555, + "loss": 3.86944580078125, + "step": 2050 + }, + { + "epoch": 0.0103, + "grad_norm": 0.8270666599273682, + "learning_rate": 0.0004999998904573181, + "loss": 3.8234973907470704, + "step": 2060 + }, + { + "epoch": 0.01035, + "grad_norm": 1.476311445236206, + "learning_rate": 0.0004999998501773356, + "loss": 3.8667575836181642, + "step": 2070 + }, + { + "epoch": 0.0104, + "grad_norm": 0.8321850895881653, + "learning_rate": 0.0004999998036036087, + "loss": 3.814508819580078, + "step": 2080 + }, + { + "epoch": 0.01045, + "grad_norm": 0.8316356539726257, + "learning_rate": 0.0004999997507361384, + "loss": 3.8478622436523438, + "step": 2090 + }, + { + "epoch": 0.0105, + "grad_norm": 0.7881673574447632, + "learning_rate": 0.0004999996915749259, + "loss": 3.8270721435546875, + "step": 2100 + }, + { + "epoch": 0.01055, + "grad_norm": 0.7225477695465088, + "learning_rate": 0.0004999996261199729, + "loss": 3.791330337524414, + "step": 2110 + }, + { + "epoch": 0.0106, + "grad_norm": 0.725504994392395, + "learning_rate": 0.000499999554371281, + "loss": 3.814841461181641, + "step": 2120 + }, + { + "epoch": 0.01065, + "grad_norm": 0.7578139305114746, + "learning_rate": 0.000499999476328852, + "loss": 3.824485015869141, + "step": 2130 + }, + { + "epoch": 0.0107, + "grad_norm": 0.8052675724029541, + "learning_rate": 0.0004999993919926879, + "loss": 3.7955345153808593, + "step": 2140 + }, + { + "epoch": 0.01075, + "grad_norm": 0.8439462780952454, + "learning_rate": 0.0004999993013627906, + "loss": 3.822450637817383, + "step": 2150 + }, + { + "epoch": 0.0108, + "grad_norm": 0.9002898931503296, + "learning_rate": 0.0004999992044391626, + "loss": 3.7820838928222655, + "step": 2160 + }, + { + "epoch": 0.01085, + "grad_norm": 0.7959500551223755, + "learning_rate": 0.0004999991012218064, + "loss": 3.7834346771240233, + "step": 2170 + }, + { + "epoch": 0.0109, + "grad_norm": 0.8416097164154053, + "learning_rate": 0.0004999989917107245, + "loss": 3.800525665283203, + "step": 2180 + }, + { + "epoch": 0.01095, + "grad_norm": 0.7272082567214966, + "learning_rate": 0.0004999988759059196, + "loss": 3.785325622558594, + "step": 2190 + }, + { + "epoch": 0.011, + "grad_norm": 0.798922061920166, + "learning_rate": 0.0004999987538073947, + "loss": 3.8025848388671877, + "step": 2200 + }, + { + "epoch": 0.01105, + "grad_norm": 0.8258434534072876, + "learning_rate": 0.0004999986254151528, + "loss": 3.8028076171875, + "step": 2210 + }, + { + "epoch": 0.0111, + "grad_norm": 0.7578490972518921, + "learning_rate": 0.0004999984907291973, + "loss": 3.814771270751953, + "step": 2220 + }, + { + "epoch": 0.01115, + "grad_norm": 0.7296440005302429, + "learning_rate": 0.0004999983497495314, + "loss": 3.828553009033203, + "step": 2230 + }, + { + "epoch": 0.0112, + "grad_norm": 0.6956039667129517, + "learning_rate": 0.0004999982024761586, + "loss": 3.7805644989013674, + "step": 2240 + }, + { + "epoch": 0.01125, + "grad_norm": 0.8552576303482056, + "learning_rate": 0.0004999980489090828, + "loss": 3.7358455657958984, + "step": 2250 + }, + { + "epoch": 0.0113, + "grad_norm": 0.6542779207229614, + "learning_rate": 0.0004999978890483079, + "loss": 3.7660770416259766, + "step": 2260 + }, + { + "epoch": 0.01135, + "grad_norm": 0.7167996764183044, + "learning_rate": 0.0004999977228938377, + "loss": 3.7541370391845703, + "step": 2270 + }, + { + "epoch": 0.0114, + "grad_norm": 0.6436230540275574, + "learning_rate": 0.0004999975504456765, + "loss": 3.7451408386230467, + "step": 2280 + }, + { + "epoch": 0.01145, + "grad_norm": 0.7030497789382935, + "learning_rate": 0.0004999973717038286, + "loss": 3.7742000579833985, + "step": 2290 + }, + { + "epoch": 0.0115, + "grad_norm": 0.7394648790359497, + "learning_rate": 0.0004999971866682986, + "loss": 3.7445747375488283, + "step": 2300 + }, + { + "epoch": 0.01155, + "grad_norm": 0.6508715748786926, + "learning_rate": 0.0004999969953390912, + "loss": 3.762335205078125, + "step": 2310 + }, + { + "epoch": 0.0116, + "grad_norm": 0.7359210252761841, + "learning_rate": 0.000499996797716211, + "loss": 3.7320266723632813, + "step": 2320 + }, + { + "epoch": 0.01165, + "grad_norm": 0.6742557883262634, + "learning_rate": 0.000499996593799663, + "loss": 3.7738414764404298, + "step": 2330 + }, + { + "epoch": 0.0117, + "grad_norm": 0.7306770086288452, + "learning_rate": 0.0004999963835894526, + "loss": 3.745998764038086, + "step": 2340 + }, + { + "epoch": 0.01175, + "grad_norm": 0.6674320101737976, + "learning_rate": 0.0004999961670855848, + "loss": 3.720348358154297, + "step": 2350 + }, + { + "epoch": 0.0118, + "grad_norm": 0.7425591349601746, + "learning_rate": 0.0004999959442880652, + "loss": 3.734689712524414, + "step": 2360 + }, + { + "epoch": 0.01185, + "grad_norm": 0.7246630191802979, + "learning_rate": 0.0004999957151968992, + "loss": 3.730331039428711, + "step": 2370 + }, + { + "epoch": 0.0119, + "grad_norm": 0.8148749470710754, + "learning_rate": 0.000499995479812093, + "loss": 3.7183677673339846, + "step": 2380 + }, + { + "epoch": 0.01195, + "grad_norm": 0.6541326642036438, + "learning_rate": 0.0004999952381336522, + "loss": 3.7176422119140624, + "step": 2390 + }, + { + "epoch": 0.012, + "grad_norm": 0.6819788813591003, + "learning_rate": 0.0004999949901615828, + "loss": 3.687652587890625, + "step": 2400 + }, + { + "epoch": 0.01205, + "grad_norm": 0.7168387770652771, + "learning_rate": 0.0004999947358958914, + "loss": 3.6707527160644533, + "step": 2410 + }, + { + "epoch": 0.0121, + "grad_norm": 0.7159331440925598, + "learning_rate": 0.0004999944753365841, + "loss": 3.7234470367431642, + "step": 2420 + }, + { + "epoch": 0.01215, + "grad_norm": 0.6596474051475525, + "learning_rate": 0.0004999942084836675, + "loss": 3.6825366973876954, + "step": 2430 + }, + { + "epoch": 0.0122, + "grad_norm": 0.7244632840156555, + "learning_rate": 0.0004999939353371485, + "loss": 3.6678745269775392, + "step": 2440 + }, + { + "epoch": 0.01225, + "grad_norm": 0.7652741074562073, + "learning_rate": 0.0004999936558970337, + "loss": 3.6948467254638673, + "step": 2450 + }, + { + "epoch": 0.0123, + "grad_norm": 0.7884398698806763, + "learning_rate": 0.0004999933701633304, + "loss": 3.687468338012695, + "step": 2460 + }, + { + "epoch": 0.01235, + "grad_norm": 0.7135241627693176, + "learning_rate": 0.0004999930781360457, + "loss": 3.6970123291015624, + "step": 2470 + }, + { + "epoch": 0.0124, + "grad_norm": 0.6481680274009705, + "learning_rate": 0.0004999927798151869, + "loss": 3.699774169921875, + "step": 2480 + }, + { + "epoch": 0.01245, + "grad_norm": 0.670994758605957, + "learning_rate": 0.0004999924752007614, + "loss": 3.640522766113281, + "step": 2490 + }, + { + "epoch": 0.0125, + "grad_norm": 0.7799317240715027, + "learning_rate": 0.0004999921642927772, + "loss": 3.6726100921630858, + "step": 2500 + }, + { + "epoch": 0.01255, + "grad_norm": 0.6819888353347778, + "learning_rate": 0.0004999918470912417, + "loss": 3.669649124145508, + "step": 2510 + }, + { + "epoch": 0.0126, + "grad_norm": 0.7387822270393372, + "learning_rate": 0.0004999915235961634, + "loss": 3.664703369140625, + "step": 2520 + }, + { + "epoch": 0.01265, + "grad_norm": 0.6641786694526672, + "learning_rate": 0.00049999119380755, + "loss": 3.7187923431396483, + "step": 2530 + }, + { + "epoch": 0.0127, + "grad_norm": 0.7271946668624878, + "learning_rate": 0.00049999085772541, + "loss": 3.6911380767822264, + "step": 2540 + }, + { + "epoch": 0.01275, + "grad_norm": 0.6493403911590576, + "learning_rate": 0.0004999905153497518, + "loss": 3.6872806549072266, + "step": 2550 + }, + { + "epoch": 0.0128, + "grad_norm": 0.6692785620689392, + "learning_rate": 0.000499990166680584, + "loss": 3.6697696685791015, + "step": 2560 + }, + { + "epoch": 0.01285, + "grad_norm": 0.7098038196563721, + "learning_rate": 0.0004999898117179156, + "loss": 3.6384334564208984, + "step": 2570 + }, + { + "epoch": 0.0129, + "grad_norm": 0.7367982864379883, + "learning_rate": 0.0004999894504617553, + "loss": 3.6720401763916017, + "step": 2580 + }, + { + "epoch": 0.01295, + "grad_norm": 0.7480196356773376, + "learning_rate": 0.0004999890829121121, + "loss": 3.722239685058594, + "step": 2590 + }, + { + "epoch": 0.013, + "grad_norm": 0.6685135364532471, + "learning_rate": 0.0004999887090689956, + "loss": 3.667375946044922, + "step": 2600 + }, + { + "epoch": 0.01305, + "grad_norm": 0.6777262687683105, + "learning_rate": 0.0004999883289324149, + "loss": 3.7057205200195313, + "step": 2610 + }, + { + "epoch": 0.0131, + "grad_norm": 0.7621126174926758, + "learning_rate": 0.0004999879425023798, + "loss": 3.623689651489258, + "step": 2620 + }, + { + "epoch": 0.01315, + "grad_norm": 0.7537278532981873, + "learning_rate": 0.0004999875497788998, + "loss": 3.63403434753418, + "step": 2630 + }, + { + "epoch": 0.0132, + "grad_norm": 0.6767253875732422, + "learning_rate": 0.000499987150761985, + "loss": 3.6611797332763674, + "step": 2640 + }, + { + "epoch": 0.01325, + "grad_norm": 0.6853849291801453, + "learning_rate": 0.0004999867454516452, + "loss": 3.611134338378906, + "step": 2650 + }, + { + "epoch": 0.0133, + "grad_norm": 0.7755582928657532, + "learning_rate": 0.0004999863338478909, + "loss": 3.6205188751220705, + "step": 2660 + }, + { + "epoch": 0.01335, + "grad_norm": 0.6784001588821411, + "learning_rate": 0.0004999859159507322, + "loss": 3.6452457427978517, + "step": 2670 + }, + { + "epoch": 0.0134, + "grad_norm": 0.6832279562950134, + "learning_rate": 0.0004999854917601798, + "loss": 3.6245983123779295, + "step": 2680 + }, + { + "epoch": 0.01345, + "grad_norm": 0.6772762537002563, + "learning_rate": 0.0004999850612762444, + "loss": 3.6952835083007813, + "step": 2690 + }, + { + "epoch": 0.0135, + "grad_norm": 0.6490644216537476, + "learning_rate": 0.0004999846244989367, + "loss": 3.6287288665771484, + "step": 2700 + }, + { + "epoch": 0.01355, + "grad_norm": 0.6805738806724548, + "learning_rate": 0.0004999841814282677, + "loss": 3.689727783203125, + "step": 2710 + }, + { + "epoch": 0.0136, + "grad_norm": 0.6843099594116211, + "learning_rate": 0.0004999837320642486, + "loss": 3.6035720825195314, + "step": 2720 + }, + { + "epoch": 0.01365, + "grad_norm": 0.6835290193557739, + "learning_rate": 0.0004999832764068908, + "loss": 3.648077392578125, + "step": 2730 + }, + { + "epoch": 0.0137, + "grad_norm": 0.6242978572845459, + "learning_rate": 0.0004999828144562055, + "loss": 3.6532615661621093, + "step": 2740 + }, + { + "epoch": 0.01375, + "grad_norm": 0.6319617629051208, + "learning_rate": 0.0004999823462122047, + "loss": 3.596415328979492, + "step": 2750 + }, + { + "epoch": 0.0138, + "grad_norm": 0.6294659972190857, + "learning_rate": 0.0004999818716748999, + "loss": 3.6159149169921876, + "step": 2760 + }, + { + "epoch": 0.01385, + "grad_norm": 0.6812044978141785, + "learning_rate": 0.0004999813908443032, + "loss": 3.606142044067383, + "step": 2770 + }, + { + "epoch": 0.0139, + "grad_norm": 0.6877551674842834, + "learning_rate": 0.0004999809037204267, + "loss": 3.5921886444091795, + "step": 2780 + }, + { + "epoch": 0.01395, + "grad_norm": 0.7487320303916931, + "learning_rate": 0.0004999804103032826, + "loss": 3.607263946533203, + "step": 2790 + }, + { + "epoch": 0.014, + "grad_norm": 0.7017234563827515, + "learning_rate": 0.0004999799105928833, + "loss": 3.618475341796875, + "step": 2800 + }, + { + "epoch": 0.01405, + "grad_norm": 0.6794924139976501, + "learning_rate": 0.0004999794045892415, + "loss": 3.6255321502685547, + "step": 2810 + }, + { + "epoch": 0.0141, + "grad_norm": 0.6516705751419067, + "learning_rate": 0.0004999788922923698, + "loss": 3.6404697418212892, + "step": 2820 + }, + { + "epoch": 0.01415, + "grad_norm": 0.6393859386444092, + "learning_rate": 0.0004999783737022811, + "loss": 3.61108512878418, + "step": 2830 + }, + { + "epoch": 0.0142, + "grad_norm": 0.63853520154953, + "learning_rate": 0.0004999778488189886, + "loss": 3.671839141845703, + "step": 2840 + }, + { + "epoch": 0.01425, + "grad_norm": 0.6674492359161377, + "learning_rate": 0.0004999773176425054, + "loss": 3.6105899810791016, + "step": 2850 + }, + { + "epoch": 0.0143, + "grad_norm": 0.6288520693778992, + "learning_rate": 0.0004999767801728449, + "loss": 3.614085006713867, + "step": 2860 + }, + { + "epoch": 0.01435, + "grad_norm": 0.7506749033927917, + "learning_rate": 0.0004999762364100205, + "loss": 3.642609786987305, + "step": 2870 + }, + { + "epoch": 0.0144, + "grad_norm": 0.7595744729042053, + "learning_rate": 0.0004999756863540462, + "loss": 3.5758941650390623, + "step": 2880 + }, + { + "epoch": 0.01445, + "grad_norm": 0.6687138676643372, + "learning_rate": 0.0004999751300049357, + "loss": 3.583455276489258, + "step": 2890 + }, + { + "epoch": 0.0145, + "grad_norm": 0.742243766784668, + "learning_rate": 0.0004999745673627028, + "loss": 3.6163925170898437, + "step": 2900 + }, + { + "epoch": 0.01455, + "grad_norm": 0.6841875314712524, + "learning_rate": 0.0004999739984273619, + "loss": 3.590650177001953, + "step": 2910 + }, + { + "epoch": 0.0146, + "grad_norm": 0.7439877986907959, + "learning_rate": 0.0004999734231989273, + "loss": 3.587427520751953, + "step": 2920 + }, + { + "epoch": 0.01465, + "grad_norm": 0.642848551273346, + "learning_rate": 0.0004999728416774134, + "loss": 3.5803817749023437, + "step": 2930 + }, + { + "epoch": 0.0147, + "grad_norm": 0.6418166756629944, + "learning_rate": 0.0004999722538628349, + "loss": 3.5845123291015626, + "step": 2940 + }, + { + "epoch": 0.01475, + "grad_norm": 0.6803068518638611, + "learning_rate": 0.0004999716597552066, + "loss": 3.6186405181884767, + "step": 2950 + }, + { + "epoch": 0.0148, + "grad_norm": 0.6172810792922974, + "learning_rate": 0.0004999710593545434, + "loss": 3.559121322631836, + "step": 2960 + }, + { + "epoch": 0.01485, + "grad_norm": 0.6860426664352417, + "learning_rate": 0.0004999704526608605, + "loss": 3.613996887207031, + "step": 2970 + }, + { + "epoch": 0.0149, + "grad_norm": 0.6330755949020386, + "learning_rate": 0.0004999698396741731, + "loss": 3.6051406860351562, + "step": 2980 + }, + { + "epoch": 0.01495, + "grad_norm": 0.6370383501052856, + "learning_rate": 0.0004999692203944966, + "loss": 3.5669055938720704, + "step": 2990 + }, + { + "epoch": 0.015, + "grad_norm": 0.6261282563209534, + "learning_rate": 0.0004999685948218467, + "loss": 3.5868515014648437, + "step": 3000 + }, + { + "epoch": 0.01505, + "grad_norm": 0.6380876898765564, + "learning_rate": 0.0004999679629562391, + "loss": 3.5831539154052736, + "step": 3010 + }, + { + "epoch": 0.0151, + "grad_norm": 0.6675437092781067, + "learning_rate": 0.0004999673247976896, + "loss": 3.5860305786132813, + "step": 3020 + }, + { + "epoch": 0.01515, + "grad_norm": 0.5933724045753479, + "learning_rate": 0.0004999666803462145, + "loss": 3.5695167541503907, + "step": 3030 + }, + { + "epoch": 0.0152, + "grad_norm": 0.64336758852005, + "learning_rate": 0.0004999660296018299, + "loss": 3.585354232788086, + "step": 3040 + }, + { + "epoch": 0.01525, + "grad_norm": 0.6326108574867249, + "learning_rate": 0.0004999653725645521, + "loss": 3.5107154846191406, + "step": 3050 + }, + { + "epoch": 0.0153, + "grad_norm": 0.6805384159088135, + "learning_rate": 0.0004999647092343978, + "loss": 3.5842018127441406, + "step": 3060 + }, + { + "epoch": 0.01535, + "grad_norm": 0.6318498253822327, + "learning_rate": 0.0004999640396113835, + "loss": 3.5976116180419924, + "step": 3070 + }, + { + "epoch": 0.0154, + "grad_norm": 0.6757373809814453, + "learning_rate": 0.0004999633636955262, + "loss": 3.596676254272461, + "step": 3080 + }, + { + "epoch": 0.01545, + "grad_norm": 0.6330369114875793, + "learning_rate": 0.0004999626814868429, + "loss": 3.6001823425292967, + "step": 3090 + }, + { + "epoch": 0.0155, + "grad_norm": 0.6700558066368103, + "learning_rate": 0.0004999619929853508, + "loss": 3.53560791015625, + "step": 3100 + }, + { + "epoch": 0.01555, + "grad_norm": 0.5963563919067383, + "learning_rate": 0.0004999612981910673, + "loss": 3.588658905029297, + "step": 3110 + }, + { + "epoch": 0.0156, + "grad_norm": 0.6764097809791565, + "learning_rate": 0.0004999605971040096, + "loss": 3.5285282135009766, + "step": 3120 + }, + { + "epoch": 0.01565, + "grad_norm": 0.5837964415550232, + "learning_rate": 0.0004999598897241956, + "loss": 3.516434097290039, + "step": 3130 + }, + { + "epoch": 0.0157, + "grad_norm": 0.7300524115562439, + "learning_rate": 0.0004999591760516431, + "loss": 3.484610366821289, + "step": 3140 + }, + { + "epoch": 0.01575, + "grad_norm": 0.6900105476379395, + "learning_rate": 0.00049995845608637, + "loss": 3.535774993896484, + "step": 3150 + }, + { + "epoch": 0.0158, + "grad_norm": 0.6509477496147156, + "learning_rate": 0.0004999577298283943, + "loss": 3.5543285369873048, + "step": 3160 + }, + { + "epoch": 0.01585, + "grad_norm": 0.6126559972763062, + "learning_rate": 0.0004999569972777345, + "loss": 3.554057312011719, + "step": 3170 + }, + { + "epoch": 0.0159, + "grad_norm": 0.6393418908119202, + "learning_rate": 0.000499956258434409, + "loss": 3.5865543365478514, + "step": 3180 + }, + { + "epoch": 0.01595, + "grad_norm": 0.6159090995788574, + "learning_rate": 0.0004999555132984364, + "loss": 3.5893169403076173, + "step": 3190 + }, + { + "epoch": 0.016, + "grad_norm": 0.5819607377052307, + "learning_rate": 0.0004999547618698354, + "loss": 3.5406970977783203, + "step": 3200 + }, + { + "epoch": 0.01605, + "grad_norm": 0.640669584274292, + "learning_rate": 0.000499954004148625, + "loss": 3.559315490722656, + "step": 3210 + }, + { + "epoch": 0.0161, + "grad_norm": 0.6256479024887085, + "learning_rate": 0.0004999532401348241, + "loss": 3.552470016479492, + "step": 3220 + }, + { + "epoch": 0.01615, + "grad_norm": 0.5943206548690796, + "learning_rate": 0.000499952469828452, + "loss": 3.5505302429199217, + "step": 3230 + }, + { + "epoch": 0.0162, + "grad_norm": 0.6300965547561646, + "learning_rate": 0.0004999516932295282, + "loss": 3.5221412658691404, + "step": 3240 + }, + { + "epoch": 0.01625, + "grad_norm": 0.6127278208732605, + "learning_rate": 0.0004999509103380723, + "loss": 3.5795822143554688, + "step": 3250 + }, + { + "epoch": 0.0163, + "grad_norm": 0.6127208471298218, + "learning_rate": 0.0004999501211541038, + "loss": 3.5408397674560548, + "step": 3260 + }, + { + "epoch": 0.01635, + "grad_norm": 0.5766595005989075, + "learning_rate": 0.0004999493256776427, + "loss": 3.5035457611083984, + "step": 3270 + }, + { + "epoch": 0.0164, + "grad_norm": 0.6019328832626343, + "learning_rate": 0.0004999485239087089, + "loss": 3.5107486724853514, + "step": 3280 + }, + { + "epoch": 0.01645, + "grad_norm": 0.6682644486427307, + "learning_rate": 0.0004999477158473229, + "loss": 3.511515808105469, + "step": 3290 + }, + { + "epoch": 0.0165, + "grad_norm": 0.6373304724693298, + "learning_rate": 0.0004999469014935047, + "loss": 3.5440528869628904, + "step": 3300 + }, + { + "epoch": 0.01655, + "grad_norm": 0.6540581583976746, + "learning_rate": 0.0004999460808472749, + "loss": 3.544932556152344, + "step": 3310 + }, + { + "epoch": 0.0166, + "grad_norm": 0.725408136844635, + "learning_rate": 0.0004999452539086541, + "loss": 3.5245697021484377, + "step": 3320 + }, + { + "epoch": 0.01665, + "grad_norm": 0.6191024780273438, + "learning_rate": 0.0004999444206776633, + "loss": 3.5504871368408204, + "step": 3330 + }, + { + "epoch": 0.0167, + "grad_norm": 0.5994449853897095, + "learning_rate": 0.0004999435811543234, + "loss": 3.544473648071289, + "step": 3340 + }, + { + "epoch": 0.01675, + "grad_norm": 0.6154484748840332, + "learning_rate": 0.0004999427353386555, + "loss": 3.5766544342041016, + "step": 3350 + }, + { + "epoch": 0.0168, + "grad_norm": 0.6386438012123108, + "learning_rate": 0.0004999418832306809, + "loss": 3.4755970001220704, + "step": 3360 + }, + { + "epoch": 0.01685, + "grad_norm": 0.6584908962249756, + "learning_rate": 0.0004999410248304212, + "loss": 3.5235107421875, + "step": 3370 + }, + { + "epoch": 0.0169, + "grad_norm": 0.6079191565513611, + "learning_rate": 0.0004999401601378977, + "loss": 3.5458160400390626, + "step": 3380 + }, + { + "epoch": 0.01695, + "grad_norm": 0.6491190791130066, + "learning_rate": 0.0004999392891531323, + "loss": 3.5086013793945314, + "step": 3390 + }, + { + "epoch": 0.017, + "grad_norm": 0.6182178854942322, + "learning_rate": 0.0004999384118761471, + "loss": 3.521654510498047, + "step": 3400 + }, + { + "epoch": 0.01705, + "grad_norm": 0.6033272743225098, + "learning_rate": 0.0004999375283069639, + "loss": 3.505342483520508, + "step": 3410 + }, + { + "epoch": 0.0171, + "grad_norm": 0.6302740573883057, + "learning_rate": 0.0004999366384456052, + "loss": 3.4622074127197267, + "step": 3420 + }, + { + "epoch": 0.01715, + "grad_norm": 0.5898723006248474, + "learning_rate": 0.0004999357422920933, + "loss": 3.4894100189208985, + "step": 3430 + }, + { + "epoch": 0.0172, + "grad_norm": 0.5746996402740479, + "learning_rate": 0.0004999348398464508, + "loss": 3.524348831176758, + "step": 3440 + }, + { + "epoch": 0.01725, + "grad_norm": 0.6024063229560852, + "learning_rate": 0.0004999339311087004, + "loss": 3.465243911743164, + "step": 3450 + }, + { + "epoch": 0.0173, + "grad_norm": 0.6816478967666626, + "learning_rate": 0.0004999330160788647, + "loss": 3.577271270751953, + "step": 3460 + }, + { + "epoch": 0.01735, + "grad_norm": 0.6810399293899536, + "learning_rate": 0.0004999320947569672, + "loss": 3.4485542297363283, + "step": 3470 + }, + { + "epoch": 0.0174, + "grad_norm": 0.6252614259719849, + "learning_rate": 0.0004999311671430309, + "loss": 3.476985549926758, + "step": 3480 + }, + { + "epoch": 0.01745, + "grad_norm": 0.6355000138282776, + "learning_rate": 0.0004999302332370791, + "loss": 3.476367950439453, + "step": 3490 + }, + { + "epoch": 0.0175, + "grad_norm": 0.6477369666099548, + "learning_rate": 0.0004999292930391354, + "loss": 3.473195266723633, + "step": 3500 + }, + { + "epoch": 0.01755, + "grad_norm": 0.5500322580337524, + "learning_rate": 0.0004999283465492233, + "loss": 3.4632278442382813, + "step": 3510 + }, + { + "epoch": 0.0176, + "grad_norm": 0.6324060559272766, + "learning_rate": 0.0004999273937673667, + "loss": 3.4542987823486326, + "step": 3520 + }, + { + "epoch": 0.01765, + "grad_norm": 0.5894066095352173, + "learning_rate": 0.0004999264346935898, + "loss": 3.4843379974365236, + "step": 3530 + }, + { + "epoch": 0.0177, + "grad_norm": 0.646304190158844, + "learning_rate": 0.0004999254693279165, + "loss": 3.5141651153564455, + "step": 3540 + }, + { + "epoch": 0.01775, + "grad_norm": 0.6445115804672241, + "learning_rate": 0.0004999244976703712, + "loss": 3.5036289215087892, + "step": 3550 + }, + { + "epoch": 0.0178, + "grad_norm": 0.6237940192222595, + "learning_rate": 0.0004999235197209783, + "loss": 3.487467956542969, + "step": 3560 + }, + { + "epoch": 0.01785, + "grad_norm": 0.6349276304244995, + "learning_rate": 0.0004999225354797625, + "loss": 3.517232894897461, + "step": 3570 + }, + { + "epoch": 0.0179, + "grad_norm": 0.5693162083625793, + "learning_rate": 0.0004999215449467485, + "loss": 3.494275665283203, + "step": 3580 + }, + { + "epoch": 0.01795, + "grad_norm": 0.654822587966919, + "learning_rate": 0.0004999205481219613, + "loss": 3.508022689819336, + "step": 3590 + }, + { + "epoch": 0.018, + "grad_norm": 0.6147361397743225, + "learning_rate": 0.000499919545005426, + "loss": 3.479543685913086, + "step": 3600 + }, + { + "epoch": 0.01805, + "grad_norm": 0.6028027534484863, + "learning_rate": 0.0004999185355971679, + "loss": 3.5227497100830076, + "step": 3610 + }, + { + "epoch": 0.0181, + "grad_norm": 0.5791961550712585, + "learning_rate": 0.0004999175198972121, + "loss": 3.5154701232910157, + "step": 3620 + }, + { + "epoch": 0.01815, + "grad_norm": 0.6521320343017578, + "learning_rate": 0.0004999164979055846, + "loss": 3.4626533508300783, + "step": 3630 + }, + { + "epoch": 0.0182, + "grad_norm": 0.5789424180984497, + "learning_rate": 0.0004999154696223108, + "loss": 3.47105712890625, + "step": 3640 + }, + { + "epoch": 0.01825, + "grad_norm": 0.6314758062362671, + "learning_rate": 0.0004999144350474169, + "loss": 3.4776729583740233, + "step": 3650 + }, + { + "epoch": 0.0183, + "grad_norm": 0.5984445810317993, + "learning_rate": 0.0004999133941809286, + "loss": 3.526297378540039, + "step": 3660 + }, + { + "epoch": 0.01835, + "grad_norm": 0.6039959788322449, + "learning_rate": 0.0004999123470228724, + "loss": 3.4368518829345702, + "step": 3670 + }, + { + "epoch": 0.0184, + "grad_norm": 0.6400973200798035, + "learning_rate": 0.0004999112935732744, + "loss": 3.4787437438964846, + "step": 3680 + }, + { + "epoch": 0.01845, + "grad_norm": 0.6096590757369995, + "learning_rate": 0.0004999102338321614, + "loss": 3.5046970367431642, + "step": 3690 + }, + { + "epoch": 0.0185, + "grad_norm": 0.5478902459144592, + "learning_rate": 0.0004999091677995598, + "loss": 3.4350818634033202, + "step": 3700 + }, + { + "epoch": 0.01855, + "grad_norm": 0.6681373715400696, + "learning_rate": 0.0004999080954754966, + "loss": 3.4954544067382813, + "step": 3710 + }, + { + "epoch": 0.0186, + "grad_norm": 0.6630448698997498, + "learning_rate": 0.0004999070168599988, + "loss": 3.494060516357422, + "step": 3720 + }, + { + "epoch": 0.01865, + "grad_norm": 0.6019526124000549, + "learning_rate": 0.0004999059319530935, + "loss": 3.4460498809814455, + "step": 3730 + }, + { + "epoch": 0.0187, + "grad_norm": 0.616905689239502, + "learning_rate": 0.0004999048407548081, + "loss": 3.5202560424804688, + "step": 3740 + }, + { + "epoch": 0.01875, + "grad_norm": 0.5977236032485962, + "learning_rate": 0.00049990374326517, + "loss": 3.4559837341308595, + "step": 3750 + }, + { + "epoch": 0.0188, + "grad_norm": 0.6482720375061035, + "learning_rate": 0.0004999026394842067, + "loss": 3.422872543334961, + "step": 3760 + }, + { + "epoch": 0.01885, + "grad_norm": 0.5901925563812256, + "learning_rate": 0.0004999015294119462, + "loss": 3.524945068359375, + "step": 3770 + }, + { + "epoch": 0.0189, + "grad_norm": 0.581281840801239, + "learning_rate": 0.0004999004130484164, + "loss": 3.495956039428711, + "step": 3780 + }, + { + "epoch": 0.01895, + "grad_norm": 0.5675626993179321, + "learning_rate": 0.0004998992903936454, + "loss": 3.4204673767089844, + "step": 3790 + }, + { + "epoch": 0.019, + "grad_norm": 0.6054050922393799, + "learning_rate": 0.0004998981614476614, + "loss": 3.4799049377441404, + "step": 3800 + }, + { + "epoch": 0.01905, + "grad_norm": 0.6581118702888489, + "learning_rate": 0.0004998970262104928, + "loss": 3.444287872314453, + "step": 3810 + }, + { + "epoch": 0.0191, + "grad_norm": 0.6280413269996643, + "learning_rate": 0.0004998958846821682, + "loss": 3.439754867553711, + "step": 3820 + }, + { + "epoch": 0.01915, + "grad_norm": 0.5735342502593994, + "learning_rate": 0.0004998947368627164, + "loss": 3.4775081634521485, + "step": 3830 + }, + { + "epoch": 0.0192, + "grad_norm": 0.6043553352355957, + "learning_rate": 0.0004998935827521664, + "loss": 3.4760643005371095, + "step": 3840 + }, + { + "epoch": 0.01925, + "grad_norm": 0.6038229465484619, + "learning_rate": 0.0004998924223505469, + "loss": 3.47470703125, + "step": 3850 + }, + { + "epoch": 0.0193, + "grad_norm": 0.6006757616996765, + "learning_rate": 0.0004998912556578876, + "loss": 3.501897430419922, + "step": 3860 + }, + { + "epoch": 0.01935, + "grad_norm": 0.6147339940071106, + "learning_rate": 0.0004998900826742175, + "loss": 3.420754241943359, + "step": 3870 + }, + { + "epoch": 0.0194, + "grad_norm": 0.6941632628440857, + "learning_rate": 0.0004998889033995662, + "loss": 3.4571887969970705, + "step": 3880 + }, + { + "epoch": 0.01945, + "grad_norm": 0.5504199862480164, + "learning_rate": 0.0004998877178339635, + "loss": 3.488801956176758, + "step": 3890 + }, + { + "epoch": 0.0195, + "grad_norm": 0.5912291407585144, + "learning_rate": 0.0004998865259774391, + "loss": 3.460232162475586, + "step": 3900 + }, + { + "epoch": 0.01955, + "grad_norm": 0.6000552773475647, + "learning_rate": 0.000499885327830023, + "loss": 3.4617469787597654, + "step": 3910 + }, + { + "epoch": 0.0196, + "grad_norm": 0.6254057288169861, + "learning_rate": 0.0004998841233917456, + "loss": 3.3998645782470702, + "step": 3920 + }, + { + "epoch": 0.01965, + "grad_norm": 0.672609269618988, + "learning_rate": 0.000499882912662637, + "loss": 3.429425811767578, + "step": 3930 + }, + { + "epoch": 0.0197, + "grad_norm": 0.5990504622459412, + "learning_rate": 0.0004998816956427278, + "loss": 3.4547550201416017, + "step": 3940 + }, + { + "epoch": 0.01975, + "grad_norm": 0.5760524868965149, + "learning_rate": 0.0004998804723320484, + "loss": 3.4571720123291017, + "step": 3950 + }, + { + "epoch": 0.0198, + "grad_norm": 0.5720862150192261, + "learning_rate": 0.00049987924273063, + "loss": 3.4800418853759765, + "step": 3960 + }, + { + "epoch": 0.01985, + "grad_norm": 0.5871039032936096, + "learning_rate": 0.0004998780068385032, + "loss": 3.457152557373047, + "step": 3970 + }, + { + "epoch": 0.0199, + "grad_norm": 0.5872801542282104, + "learning_rate": 0.0004998767646556993, + "loss": 3.4540611267089845, + "step": 3980 + }, + { + "epoch": 0.01995, + "grad_norm": 0.6100932359695435, + "learning_rate": 0.0004998755161822496, + "loss": 3.4853626251220704, + "step": 3990 + }, + { + "epoch": 0.02, + "grad_norm": 0.5599991083145142, + "learning_rate": 0.0004998742614181853, + "loss": 3.4525306701660154, + "step": 4000 + }, + { + "epoch": 0.02005, + "grad_norm": 0.5894349217414856, + "learning_rate": 0.0004998730003635383, + "loss": 3.453520965576172, + "step": 4010 + }, + { + "epoch": 0.0201, + "grad_norm": 0.6467536687850952, + "learning_rate": 0.00049987173301834, + "loss": 3.4074188232421876, + "step": 4020 + }, + { + "epoch": 0.02015, + "grad_norm": 0.7268823981285095, + "learning_rate": 0.0004998704593826226, + "loss": 3.4183631896972657, + "step": 4030 + }, + { + "epoch": 0.0202, + "grad_norm": 0.5635799169540405, + "learning_rate": 0.0004998691794564179, + "loss": 3.4707740783691405, + "step": 4040 + }, + { + "epoch": 0.02025, + "grad_norm": 0.5924507975578308, + "learning_rate": 0.0004998678932397585, + "loss": 3.459727096557617, + "step": 4050 + }, + { + "epoch": 0.0203, + "grad_norm": 0.6172810792922974, + "learning_rate": 0.0004998666007326765, + "loss": 3.4237899780273438, + "step": 4060 + }, + { + "epoch": 0.02035, + "grad_norm": 0.5856655240058899, + "learning_rate": 0.0004998653019352044, + "loss": 3.4462814331054688, + "step": 4070 + }, + { + "epoch": 0.0204, + "grad_norm": 0.5529986023902893, + "learning_rate": 0.0004998639968473751, + "loss": 3.444235610961914, + "step": 4080 + }, + { + "epoch": 0.02045, + "grad_norm": 0.5573139190673828, + "learning_rate": 0.0004998626854692212, + "loss": 3.424711227416992, + "step": 4090 + }, + { + "epoch": 0.0205, + "grad_norm": 0.6657218337059021, + "learning_rate": 0.000499861367800776, + "loss": 3.420772933959961, + "step": 4100 + }, + { + "epoch": 0.02055, + "grad_norm": 0.6016212701797485, + "learning_rate": 0.0004998600438420725, + "loss": 3.4227878570556642, + "step": 4110 + }, + { + "epoch": 0.0206, + "grad_norm": 0.6032813191413879, + "learning_rate": 0.000499858713593144, + "loss": 3.4662796020507813, + "step": 4120 + }, + { + "epoch": 0.02065, + "grad_norm": 0.5646603107452393, + "learning_rate": 0.0004998573770540242, + "loss": 3.374886703491211, + "step": 4130 + }, + { + "epoch": 0.0207, + "grad_norm": 0.5694288015365601, + "learning_rate": 0.0004998560342247465, + "loss": 3.402272033691406, + "step": 4140 + }, + { + "epoch": 0.02075, + "grad_norm": 0.5621424913406372, + "learning_rate": 0.0004998546851053448, + "loss": 3.406237030029297, + "step": 4150 + }, + { + "epoch": 0.0208, + "grad_norm": 0.60035640001297, + "learning_rate": 0.0004998533296958532, + "loss": 3.4060001373291016, + "step": 4160 + }, + { + "epoch": 0.02085, + "grad_norm": 0.6091265082359314, + "learning_rate": 0.0004998519679963056, + "loss": 3.430144500732422, + "step": 4170 + }, + { + "epoch": 0.0209, + "grad_norm": 0.5506263971328735, + "learning_rate": 0.0004998506000067363, + "loss": 3.4125076293945313, + "step": 4180 + }, + { + "epoch": 0.02095, + "grad_norm": 0.5959593653678894, + "learning_rate": 0.0004998492257271799, + "loss": 3.44761962890625, + "step": 4190 + }, + { + "epoch": 0.021, + "grad_norm": 0.6202824115753174, + "learning_rate": 0.0004998478451576709, + "loss": 3.4453659057617188, + "step": 4200 + }, + { + "epoch": 0.02105, + "grad_norm": 0.5680320858955383, + "learning_rate": 0.000499846458298244, + "loss": 3.4352569580078125, + "step": 4210 + }, + { + "epoch": 0.0211, + "grad_norm": 0.5744881629943848, + "learning_rate": 0.0004998450651489343, + "loss": 3.4182552337646483, + "step": 4220 + }, + { + "epoch": 0.02115, + "grad_norm": 0.5880019068717957, + "learning_rate": 0.0004998436657097767, + "loss": 3.3939498901367187, + "step": 4230 + }, + { + "epoch": 0.0212, + "grad_norm": 0.5629853010177612, + "learning_rate": 0.0004998422599808064, + "loss": 3.3730587005615233, + "step": 4240 + }, + { + "epoch": 0.02125, + "grad_norm": 0.5585372447967529, + "learning_rate": 0.0004998408479620591, + "loss": 3.395060348510742, + "step": 4250 + }, + { + "epoch": 0.0213, + "grad_norm": 0.602639377117157, + "learning_rate": 0.00049983942965357, + "loss": 3.376165008544922, + "step": 4260 + }, + { + "epoch": 0.02135, + "grad_norm": 0.5435888767242432, + "learning_rate": 0.0004998380050553749, + "loss": 3.4497444152832033, + "step": 4270 + }, + { + "epoch": 0.0214, + "grad_norm": 0.5394490957260132, + "learning_rate": 0.0004998365741675097, + "loss": 3.377367401123047, + "step": 4280 + }, + { + "epoch": 0.02145, + "grad_norm": 0.5523499846458435, + "learning_rate": 0.0004998351369900104, + "loss": 3.4086681365966798, + "step": 4290 + }, + { + "epoch": 0.0215, + "grad_norm": 0.5249232053756714, + "learning_rate": 0.0004998336935229132, + "loss": 3.445172119140625, + "step": 4300 + }, + { + "epoch": 0.02155, + "grad_norm": 0.6035228371620178, + "learning_rate": 0.0004998322437662544, + "loss": 3.385270690917969, + "step": 4310 + }, + { + "epoch": 0.0216, + "grad_norm": 0.5993716716766357, + "learning_rate": 0.0004998307877200706, + "loss": 3.3881988525390625, + "step": 4320 + }, + { + "epoch": 0.02165, + "grad_norm": 0.5337836146354675, + "learning_rate": 0.0004998293253843984, + "loss": 3.430636978149414, + "step": 4330 + }, + { + "epoch": 0.0217, + "grad_norm": 0.5783175826072693, + "learning_rate": 0.0004998278567592745, + "loss": 3.394705963134766, + "step": 4340 + }, + { + "epoch": 0.02175, + "grad_norm": 0.562882125377655, + "learning_rate": 0.0004998263818447362, + "loss": 3.330760192871094, + "step": 4350 + }, + { + "epoch": 0.0218, + "grad_norm": 0.5970223546028137, + "learning_rate": 0.0004998249006408201, + "loss": 3.392806625366211, + "step": 4360 + }, + { + "epoch": 0.02185, + "grad_norm": 0.533291757106781, + "learning_rate": 0.0004998234131475639, + "loss": 3.4166580200195313, + "step": 4370 + }, + { + "epoch": 0.0219, + "grad_norm": 0.5964130759239197, + "learning_rate": 0.0004998219193650051, + "loss": 3.431548309326172, + "step": 4380 + }, + { + "epoch": 0.02195, + "grad_norm": 0.5541797280311584, + "learning_rate": 0.000499820419293181, + "loss": 3.417970657348633, + "step": 4390 + }, + { + "epoch": 0.022, + "grad_norm": 0.5302194356918335, + "learning_rate": 0.0004998189129321296, + "loss": 3.4232574462890626, + "step": 4400 + }, + { + "epoch": 0.02205, + "grad_norm": 0.5958655476570129, + "learning_rate": 0.0004998174002818887, + "loss": 3.4420162200927735, + "step": 4410 + }, + { + "epoch": 0.0221, + "grad_norm": 0.5308816432952881, + "learning_rate": 0.0004998158813424964, + "loss": 3.3628673553466797, + "step": 4420 + }, + { + "epoch": 0.02215, + "grad_norm": 0.597850501537323, + "learning_rate": 0.000499814356113991, + "loss": 3.4038185119628905, + "step": 4430 + }, + { + "epoch": 0.0222, + "grad_norm": 0.5634832978248596, + "learning_rate": 0.0004998128245964108, + "loss": 3.4190078735351563, + "step": 4440 + }, + { + "epoch": 0.02225, + "grad_norm": 0.5729529857635498, + "learning_rate": 0.0004998112867897943, + "loss": 3.3864978790283202, + "step": 4450 + }, + { + "epoch": 0.0223, + "grad_norm": 0.5532680749893188, + "learning_rate": 0.0004998097426941806, + "loss": 3.379507827758789, + "step": 4460 + }, + { + "epoch": 0.02235, + "grad_norm": 0.5703796744346619, + "learning_rate": 0.0004998081923096081, + "loss": 3.3984169006347655, + "step": 4470 + }, + { + "epoch": 0.0224, + "grad_norm": 0.5448310971260071, + "learning_rate": 0.0004998066356361161, + "loss": 3.4070724487304687, + "step": 4480 + }, + { + "epoch": 0.02245, + "grad_norm": 0.5454280972480774, + "learning_rate": 0.0004998050726737437, + "loss": 3.355885314941406, + "step": 4490 + }, + { + "epoch": 0.0225, + "grad_norm": 0.525452196598053, + "learning_rate": 0.0004998035034225303, + "loss": 3.4095619201660154, + "step": 4500 + }, + { + "epoch": 0.02255, + "grad_norm": 0.5875637531280518, + "learning_rate": 0.0004998019278825153, + "loss": 3.369589614868164, + "step": 4510 + }, + { + "epoch": 0.0226, + "grad_norm": 0.6186979413032532, + "learning_rate": 0.0004998003460537385, + "loss": 3.448271560668945, + "step": 4520 + }, + { + "epoch": 0.02265, + "grad_norm": 0.5718867182731628, + "learning_rate": 0.0004997987579362396, + "loss": 3.416722869873047, + "step": 4530 + }, + { + "epoch": 0.0227, + "grad_norm": 0.5502358078956604, + "learning_rate": 0.0004997971635300587, + "loss": 3.385459136962891, + "step": 4540 + }, + { + "epoch": 0.02275, + "grad_norm": 0.6429696679115295, + "learning_rate": 0.0004997955628352357, + "loss": 3.4048225402832033, + "step": 4550 + }, + { + "epoch": 0.0228, + "grad_norm": 0.5708484649658203, + "learning_rate": 0.0004997939558518113, + "loss": 3.386952590942383, + "step": 4560 + }, + { + "epoch": 0.02285, + "grad_norm": 0.5098332166671753, + "learning_rate": 0.0004997923425798256, + "loss": 3.3808101654052733, + "step": 4570 + }, + { + "epoch": 0.0229, + "grad_norm": 0.543864905834198, + "learning_rate": 0.0004997907230193194, + "loss": 3.369842529296875, + "step": 4580 + }, + { + "epoch": 0.02295, + "grad_norm": 0.5955133438110352, + "learning_rate": 0.0004997890971703334, + "loss": 3.4258468627929686, + "step": 4590 + }, + { + "epoch": 0.023, + "grad_norm": 0.5688019394874573, + "learning_rate": 0.0004997874650329086, + "loss": 3.3727138519287108, + "step": 4600 + }, + { + "epoch": 0.02305, + "grad_norm": 0.7661529779434204, + "learning_rate": 0.0004997858266070859, + "loss": 3.3554248809814453, + "step": 4610 + }, + { + "epoch": 0.0231, + "grad_norm": 0.5965901613235474, + "learning_rate": 0.0004997841818929068, + "loss": 3.3623687744140627, + "step": 4620 + }, + { + "epoch": 0.02315, + "grad_norm": 0.575777530670166, + "learning_rate": 0.0004997825308904126, + "loss": 3.352273941040039, + "step": 4630 + }, + { + "epoch": 0.0232, + "grad_norm": 0.6090394854545593, + "learning_rate": 0.0004997808735996448, + "loss": 3.3748565673828126, + "step": 4640 + }, + { + "epoch": 0.02325, + "grad_norm": 0.613265335559845, + "learning_rate": 0.0004997792100206452, + "loss": 3.349831390380859, + "step": 4650 + }, + { + "epoch": 0.0233, + "grad_norm": 0.5631701946258545, + "learning_rate": 0.0004997775401534557, + "loss": 3.398875427246094, + "step": 4660 + }, + { + "epoch": 0.02335, + "grad_norm": 0.5359291434288025, + "learning_rate": 0.0004997758639981182, + "loss": 3.3583789825439454, + "step": 4670 + }, + { + "epoch": 0.0234, + "grad_norm": 0.5559905171394348, + "learning_rate": 0.000499774181554675, + "loss": 3.3322887420654297, + "step": 4680 + }, + { + "epoch": 0.02345, + "grad_norm": 0.5420958995819092, + "learning_rate": 0.0004997724928231684, + "loss": 3.3934837341308595, + "step": 4690 + }, + { + "epoch": 0.0235, + "grad_norm": 0.541562020778656, + "learning_rate": 0.0004997707978036413, + "loss": 3.3532943725585938, + "step": 4700 + }, + { + "epoch": 0.02355, + "grad_norm": 0.5963573455810547, + "learning_rate": 0.0004997690964961357, + "loss": 3.3899742126464845, + "step": 4710 + }, + { + "epoch": 0.0236, + "grad_norm": 0.5850399136543274, + "learning_rate": 0.0004997673889006949, + "loss": 3.3891578674316407, + "step": 4720 + }, + { + "epoch": 0.02365, + "grad_norm": 0.5471701622009277, + "learning_rate": 0.0004997656750173616, + "loss": 3.415765380859375, + "step": 4730 + }, + { + "epoch": 0.0237, + "grad_norm": 0.6228621602058411, + "learning_rate": 0.0004997639548461792, + "loss": 3.3721908569335937, + "step": 4740 + }, + { + "epoch": 0.02375, + "grad_norm": 0.5328726172447205, + "learning_rate": 0.0004997622283871909, + "loss": 3.4010467529296875, + "step": 4750 + }, + { + "epoch": 0.0238, + "grad_norm": 0.5387725234031677, + "learning_rate": 0.0004997604956404403, + "loss": 3.358773422241211, + "step": 4760 + }, + { + "epoch": 0.02385, + "grad_norm": 0.5747665762901306, + "learning_rate": 0.0004997587566059707, + "loss": 3.381891632080078, + "step": 4770 + }, + { + "epoch": 0.0239, + "grad_norm": 0.5500767230987549, + "learning_rate": 0.0004997570112838261, + "loss": 3.389537048339844, + "step": 4780 + }, + { + "epoch": 0.02395, + "grad_norm": 0.5584102272987366, + "learning_rate": 0.0004997552596740504, + "loss": 3.371202850341797, + "step": 4790 + }, + { + "epoch": 0.024, + "grad_norm": 0.5645338892936707, + "learning_rate": 0.0004997535017766878, + "loss": 3.3755672454833983, + "step": 4800 + }, + { + "epoch": 0.02405, + "grad_norm": 0.546771228313446, + "learning_rate": 0.0004997517375917825, + "loss": 3.345697784423828, + "step": 4810 + }, + { + "epoch": 0.0241, + "grad_norm": 0.5296199321746826, + "learning_rate": 0.0004997499671193787, + "loss": 3.4195510864257814, + "step": 4820 + }, + { + "epoch": 0.02415, + "grad_norm": 0.5072226524353027, + "learning_rate": 0.0004997481903595213, + "loss": 3.321868133544922, + "step": 4830 + }, + { + "epoch": 0.0242, + "grad_norm": 0.5472995042800903, + "learning_rate": 0.0004997464073122548, + "loss": 3.3572242736816404, + "step": 4840 + }, + { + "epoch": 0.02425, + "grad_norm": 0.5193058252334595, + "learning_rate": 0.0004997446179776241, + "loss": 3.342325210571289, + "step": 4850 + }, + { + "epoch": 0.0243, + "grad_norm": 0.5555915236473083, + "learning_rate": 0.0004997428223556745, + "loss": 3.310529327392578, + "step": 4860 + }, + { + "epoch": 0.02435, + "grad_norm": 0.5621734857559204, + "learning_rate": 0.0004997410204464509, + "loss": 3.323847198486328, + "step": 4870 + }, + { + "epoch": 0.0244, + "grad_norm": 0.5624080896377563, + "learning_rate": 0.0004997392122499987, + "loss": 3.3494342803955077, + "step": 4880 + }, + { + "epoch": 0.02445, + "grad_norm": 0.6030921936035156, + "learning_rate": 0.0004997373977663637, + "loss": 3.394601058959961, + "step": 4890 + }, + { + "epoch": 0.0245, + "grad_norm": 0.5362664461135864, + "learning_rate": 0.0004997355769955912, + "loss": 3.380575942993164, + "step": 4900 + }, + { + "epoch": 0.02455, + "grad_norm": 0.5660577416419983, + "learning_rate": 0.0004997337499377273, + "loss": 3.389097213745117, + "step": 4910 + }, + { + "epoch": 0.0246, + "grad_norm": 0.5238977074623108, + "learning_rate": 0.0004997319165928179, + "loss": 3.335544967651367, + "step": 4920 + }, + { + "epoch": 0.02465, + "grad_norm": 0.5447637438774109, + "learning_rate": 0.0004997300769609092, + "loss": 3.337771987915039, + "step": 4930 + }, + { + "epoch": 0.0247, + "grad_norm": 0.5461397171020508, + "learning_rate": 0.0004997282310420474, + "loss": 3.347020721435547, + "step": 4940 + }, + { + "epoch": 0.02475, + "grad_norm": 0.5890595316886902, + "learning_rate": 0.0004997263788362791, + "loss": 3.3059432983398436, + "step": 4950 + }, + { + "epoch": 0.0248, + "grad_norm": 0.5955908298492432, + "learning_rate": 0.0004997245203436509, + "loss": 3.3813755035400392, + "step": 4960 + }, + { + "epoch": 0.02485, + "grad_norm": 0.5834391713142395, + "learning_rate": 0.0004997226555642095, + "loss": 3.3516002655029298, + "step": 4970 + }, + { + "epoch": 0.0249, + "grad_norm": 0.542777955532074, + "learning_rate": 0.000499720784498002, + "loss": 3.324716567993164, + "step": 4980 + }, + { + "epoch": 0.02495, + "grad_norm": 0.5612271428108215, + "learning_rate": 0.0004997189071450754, + "loss": 3.337554168701172, + "step": 4990 + }, + { + "epoch": 0.025, + "grad_norm": 0.5587865710258484, + "learning_rate": 0.000499717023505477, + "loss": 3.3515819549560546, + "step": 5000 + }, + { + "epoch": 0.02505, + "grad_norm": 0.5721408128738403, + "learning_rate": 0.0004997151335792542, + "loss": 3.370501708984375, + "step": 5010 + }, + { + "epoch": 0.0251, + "grad_norm": 0.5328549146652222, + "learning_rate": 0.0004997132373664544, + "loss": 3.4016963958740236, + "step": 5020 + }, + { + "epoch": 0.02515, + "grad_norm": 0.6018338799476624, + "learning_rate": 0.0004997113348671257, + "loss": 3.4178356170654296, + "step": 5030 + }, + { + "epoch": 0.0252, + "grad_norm": 0.6293740272521973, + "learning_rate": 0.0004997094260813158, + "loss": 3.3435684204101563, + "step": 5040 + }, + { + "epoch": 0.02525, + "grad_norm": 0.5415748953819275, + "learning_rate": 0.0004997075110090728, + "loss": 3.3703983306884764, + "step": 5050 + }, + { + "epoch": 0.0253, + "grad_norm": 0.5283946394920349, + "learning_rate": 0.0004997055896504448, + "loss": 3.348424530029297, + "step": 5060 + }, + { + "epoch": 0.02535, + "grad_norm": 0.5401787757873535, + "learning_rate": 0.0004997036620054803, + "loss": 3.3652698516845705, + "step": 5070 + }, + { + "epoch": 0.0254, + "grad_norm": 0.5331237316131592, + "learning_rate": 0.0004997017280742277, + "loss": 3.3250370025634766, + "step": 5080 + }, + { + "epoch": 0.02545, + "grad_norm": 0.5199392437934875, + "learning_rate": 0.0004996997878567358, + "loss": 3.3795284271240233, + "step": 5090 + }, + { + "epoch": 0.0255, + "grad_norm": 0.5445046424865723, + "learning_rate": 0.0004996978413530535, + "loss": 3.299697494506836, + "step": 5100 + }, + { + "epoch": 0.02555, + "grad_norm": 0.5484325289726257, + "learning_rate": 0.0004996958885632296, + "loss": 3.3918502807617186, + "step": 5110 + }, + { + "epoch": 0.0256, + "grad_norm": 0.5142081379890442, + "learning_rate": 0.0004996939294873135, + "loss": 3.343300628662109, + "step": 5120 + }, + { + "epoch": 0.02565, + "grad_norm": 0.512411892414093, + "learning_rate": 0.0004996919641253542, + "loss": 3.324250030517578, + "step": 5130 + }, + { + "epoch": 0.0257, + "grad_norm": 0.5514593124389648, + "learning_rate": 0.0004996899924774016, + "loss": 3.3378204345703124, + "step": 5140 + }, + { + "epoch": 0.02575, + "grad_norm": 0.6022439002990723, + "learning_rate": 0.0004996880145435048, + "loss": 3.3298824310302733, + "step": 5150 + }, + { + "epoch": 0.0258, + "grad_norm": 0.5651016235351562, + "learning_rate": 0.0004996860303237142, + "loss": 3.3646209716796873, + "step": 5160 + }, + { + "epoch": 0.02585, + "grad_norm": 0.5258949995040894, + "learning_rate": 0.0004996840398180793, + "loss": 3.3781143188476563, + "step": 5170 + }, + { + "epoch": 0.0259, + "grad_norm": 0.5325801968574524, + "learning_rate": 0.0004996820430266505, + "loss": 3.3289825439453127, + "step": 5180 + }, + { + "epoch": 0.02595, + "grad_norm": 0.560452401638031, + "learning_rate": 0.0004996800399494778, + "loss": 3.337248992919922, + "step": 5190 + }, + { + "epoch": 0.026, + "grad_norm": 0.5500338077545166, + "learning_rate": 0.0004996780305866117, + "loss": 3.2971576690673827, + "step": 5200 + }, + { + "epoch": 0.02605, + "grad_norm": 0.5112565159797668, + "learning_rate": 0.0004996760149381031, + "loss": 3.3154132843017576, + "step": 5210 + }, + { + "epoch": 0.0261, + "grad_norm": 0.519473135471344, + "learning_rate": 0.0004996739930040023, + "loss": 3.3483524322509766, + "step": 5220 + }, + { + "epoch": 0.02615, + "grad_norm": 0.5626395344734192, + "learning_rate": 0.0004996719647843604, + "loss": 3.3469623565673827, + "step": 5230 + }, + { + "epoch": 0.0262, + "grad_norm": 0.5947638154029846, + "learning_rate": 0.0004996699302792285, + "loss": 3.315066909790039, + "step": 5240 + }, + { + "epoch": 0.02625, + "grad_norm": 0.546932578086853, + "learning_rate": 0.0004996678894886578, + "loss": 3.354358673095703, + "step": 5250 + }, + { + "epoch": 0.0263, + "grad_norm": 0.5316585898399353, + "learning_rate": 0.0004996658424126997, + "loss": 3.3153141021728514, + "step": 5260 + }, + { + "epoch": 0.02635, + "grad_norm": 0.5336185693740845, + "learning_rate": 0.0004996637890514055, + "loss": 3.3387420654296873, + "step": 5270 + }, + { + "epoch": 0.0264, + "grad_norm": 0.5515270233154297, + "learning_rate": 0.0004996617294048272, + "loss": 3.3405010223388674, + "step": 5280 + }, + { + "epoch": 0.02645, + "grad_norm": 0.5438713431358337, + "learning_rate": 0.0004996596634730165, + "loss": 3.3046504974365236, + "step": 5290 + }, + { + "epoch": 0.0265, + "grad_norm": 0.5312449336051941, + "learning_rate": 0.0004996575912560255, + "loss": 3.3710407257080077, + "step": 5300 + }, + { + "epoch": 0.02655, + "grad_norm": 0.5835204124450684, + "learning_rate": 0.0004996555127539063, + "loss": 3.287685012817383, + "step": 5310 + }, + { + "epoch": 0.0266, + "grad_norm": 0.5348700881004333, + "learning_rate": 0.0004996534279667112, + "loss": 3.3438148498535156, + "step": 5320 + }, + { + "epoch": 0.02665, + "grad_norm": 0.5361063480377197, + "learning_rate": 0.0004996513368944927, + "loss": 3.352157211303711, + "step": 5330 + }, + { + "epoch": 0.0267, + "grad_norm": 0.58526211977005, + "learning_rate": 0.0004996492395373034, + "loss": 3.3187511444091795, + "step": 5340 + }, + { + "epoch": 0.02675, + "grad_norm": 0.5395480394363403, + "learning_rate": 0.0004996471358951962, + "loss": 3.346028137207031, + "step": 5350 + }, + { + "epoch": 0.0268, + "grad_norm": 0.5473170876502991, + "learning_rate": 0.0004996450259682242, + "loss": 3.3017204284667967, + "step": 5360 + }, + { + "epoch": 0.02685, + "grad_norm": 0.5243493318557739, + "learning_rate": 0.0004996429097564401, + "loss": 3.36591796875, + "step": 5370 + }, + { + "epoch": 0.0269, + "grad_norm": 0.49968254566192627, + "learning_rate": 0.0004996407872598976, + "loss": 3.2803108215332033, + "step": 5380 + }, + { + "epoch": 0.02695, + "grad_norm": 0.5156917572021484, + "learning_rate": 0.0004996386584786499, + "loss": 3.367662048339844, + "step": 5390 + }, + { + "epoch": 0.027, + "grad_norm": 0.5220332145690918, + "learning_rate": 0.0004996365234127506, + "loss": 3.316374969482422, + "step": 5400 + }, + { + "epoch": 0.02705, + "grad_norm": 0.5972458720207214, + "learning_rate": 0.0004996343820622535, + "loss": 3.364453125, + "step": 5410 + }, + { + "epoch": 0.0271, + "grad_norm": 0.5501435995101929, + "learning_rate": 0.0004996322344272124, + "loss": 3.3758991241455076, + "step": 5420 + }, + { + "epoch": 0.02715, + "grad_norm": 0.5560432076454163, + "learning_rate": 0.0004996300805076816, + "loss": 3.320022201538086, + "step": 5430 + }, + { + "epoch": 0.0272, + "grad_norm": 0.557008683681488, + "learning_rate": 0.0004996279203037152, + "loss": 3.327789306640625, + "step": 5440 + }, + { + "epoch": 0.02725, + "grad_norm": 0.5233785510063171, + "learning_rate": 0.0004996257538153675, + "loss": 3.337006378173828, + "step": 5450 + }, + { + "epoch": 0.0273, + "grad_norm": 0.5185394287109375, + "learning_rate": 0.0004996235810426932, + "loss": 3.3326675415039064, + "step": 5460 + }, + { + "epoch": 0.02735, + "grad_norm": 0.5394155383110046, + "learning_rate": 0.0004996214019857468, + "loss": 3.3142547607421875, + "step": 5470 + }, + { + "epoch": 0.0274, + "grad_norm": 0.7663456201553345, + "learning_rate": 0.0004996192166445833, + "loss": 3.2972366333007814, + "step": 5480 + }, + { + "epoch": 0.02745, + "grad_norm": 0.5015429258346558, + "learning_rate": 0.0004996170250192578, + "loss": 3.3144386291503904, + "step": 5490 + }, + { + "epoch": 0.0275, + "grad_norm": 0.5444296598434448, + "learning_rate": 0.0004996148271098253, + "loss": 3.298097610473633, + "step": 5500 + }, + { + "epoch": 0.02755, + "grad_norm": 0.4968287944793701, + "learning_rate": 0.0004996126229163411, + "loss": 3.3259471893310546, + "step": 5510 + }, + { + "epoch": 0.0276, + "grad_norm": 0.602137565612793, + "learning_rate": 0.000499610412438861, + "loss": 3.3213661193847654, + "step": 5520 + }, + { + "epoch": 0.02765, + "grad_norm": 0.5279825329780579, + "learning_rate": 0.0004996081956774402, + "loss": 3.2942222595214843, + "step": 5530 + }, + { + "epoch": 0.0277, + "grad_norm": 0.5516049265861511, + "learning_rate": 0.0004996059726321349, + "loss": 3.3561817169189454, + "step": 5540 + }, + { + "epoch": 0.02775, + "grad_norm": 0.5174930691719055, + "learning_rate": 0.0004996037433030008, + "loss": 3.3343955993652346, + "step": 5550 + }, + { + "epoch": 0.0278, + "grad_norm": 0.5527567267417908, + "learning_rate": 0.0004996015076900942, + "loss": 3.3083194732666015, + "step": 5560 + }, + { + "epoch": 0.02785, + "grad_norm": 0.5275153517723083, + "learning_rate": 0.0004995992657934713, + "loss": 3.2909008026123048, + "step": 5570 + }, + { + "epoch": 0.0279, + "grad_norm": 0.5090030431747437, + "learning_rate": 0.0004995970176131885, + "loss": 3.2781532287597654, + "step": 5580 + }, + { + "epoch": 0.02795, + "grad_norm": 0.5101607441902161, + "learning_rate": 0.0004995947631493026, + "loss": 3.322576141357422, + "step": 5590 + }, + { + "epoch": 0.028, + "grad_norm": 0.542224109172821, + "learning_rate": 0.0004995925024018701, + "loss": 3.3109146118164063, + "step": 5600 + }, + { + "epoch": 0.02805, + "grad_norm": 0.4783613383769989, + "learning_rate": 0.0004995902353709479, + "loss": 3.2997344970703124, + "step": 5610 + }, + { + "epoch": 0.0281, + "grad_norm": 0.4911406636238098, + "learning_rate": 0.0004995879620565934, + "loss": 3.3286956787109374, + "step": 5620 + }, + { + "epoch": 0.02815, + "grad_norm": 0.5052363872528076, + "learning_rate": 0.0004995856824588635, + "loss": 3.3202182769775392, + "step": 5630 + }, + { + "epoch": 0.0282, + "grad_norm": 0.5725943446159363, + "learning_rate": 0.0004995833965778158, + "loss": 3.3066280364990233, + "step": 5640 + }, + { + "epoch": 0.02825, + "grad_norm": 0.5519552230834961, + "learning_rate": 0.0004995811044135077, + "loss": 3.32977294921875, + "step": 5650 + }, + { + "epoch": 0.0283, + "grad_norm": 0.5240675210952759, + "learning_rate": 0.0004995788059659969, + "loss": 3.2978183746337892, + "step": 5660 + }, + { + "epoch": 0.02835, + "grad_norm": 0.527763307094574, + "learning_rate": 0.0004995765012353415, + "loss": 3.2522998809814454, + "step": 5670 + }, + { + "epoch": 0.0284, + "grad_norm": 0.508025586605072, + "learning_rate": 0.0004995741902215992, + "loss": 3.297067642211914, + "step": 5680 + }, + { + "epoch": 0.02845, + "grad_norm": 0.5200346112251282, + "learning_rate": 0.0004995718729248283, + "loss": 3.299541473388672, + "step": 5690 + }, + { + "epoch": 0.0285, + "grad_norm": 0.4993706941604614, + "learning_rate": 0.0004995695493450873, + "loss": 3.281297302246094, + "step": 5700 + }, + { + "epoch": 0.02855, + "grad_norm": 0.5262503623962402, + "learning_rate": 0.0004995672194824345, + "loss": 3.2850521087646483, + "step": 5710 + }, + { + "epoch": 0.0286, + "grad_norm": 0.5106416344642639, + "learning_rate": 0.0004995648833369286, + "loss": 3.323222351074219, + "step": 5720 + }, + { + "epoch": 0.02865, + "grad_norm": 0.5121148824691772, + "learning_rate": 0.0004995625409086285, + "loss": 3.380869674682617, + "step": 5730 + }, + { + "epoch": 0.0287, + "grad_norm": 0.5575656294822693, + "learning_rate": 0.000499560192197593, + "loss": 3.287732696533203, + "step": 5740 + }, + { + "epoch": 0.02875, + "grad_norm": 0.5861167907714844, + "learning_rate": 0.0004995578372038814, + "loss": 3.352260208129883, + "step": 5750 + }, + { + "epoch": 0.0288, + "grad_norm": 0.49247583746910095, + "learning_rate": 0.000499555475927553, + "loss": 3.3240116119384764, + "step": 5760 + }, + { + "epoch": 0.02885, + "grad_norm": 0.5201653838157654, + "learning_rate": 0.0004995531083686669, + "loss": 3.321155548095703, + "step": 5770 + }, + { + "epoch": 0.0289, + "grad_norm": 0.5006564259529114, + "learning_rate": 0.0004995507345272832, + "loss": 3.3171451568603514, + "step": 5780 + }, + { + "epoch": 0.02895, + "grad_norm": 0.5203686356544495, + "learning_rate": 0.0004995483544034614, + "loss": 3.324483108520508, + "step": 5790 + }, + { + "epoch": 0.029, + "grad_norm": 0.5225478410720825, + "learning_rate": 0.0004995459679972613, + "loss": 3.3051406860351564, + "step": 5800 + }, + { + "epoch": 0.02905, + "grad_norm": 0.5420324206352234, + "learning_rate": 0.0004995435753087433, + "loss": 3.2843505859375, + "step": 5810 + }, + { + "epoch": 0.0291, + "grad_norm": 0.5187715291976929, + "learning_rate": 0.0004995411763379674, + "loss": 3.3219398498535155, + "step": 5820 + }, + { + "epoch": 0.02915, + "grad_norm": 0.52501380443573, + "learning_rate": 0.0004995387710849939, + "loss": 3.2794036865234375, + "step": 5830 + }, + { + "epoch": 0.0292, + "grad_norm": 0.5337326526641846, + "learning_rate": 0.0004995363595498837, + "loss": 3.3209949493408204, + "step": 5840 + }, + { + "epoch": 0.02925, + "grad_norm": 0.5138839483261108, + "learning_rate": 0.0004995339417326973, + "loss": 3.3103240966796874, + "step": 5850 + }, + { + "epoch": 0.0293, + "grad_norm": 0.4882403016090393, + "learning_rate": 0.0004995315176334954, + "loss": 3.3514450073242186, + "step": 5860 + }, + { + "epoch": 0.02935, + "grad_norm": 0.5009715557098389, + "learning_rate": 0.0004995290872523394, + "loss": 3.2508968353271483, + "step": 5870 + }, + { + "epoch": 0.0294, + "grad_norm": 0.5526741743087769, + "learning_rate": 0.0004995266505892903, + "loss": 3.317081832885742, + "step": 5880 + }, + { + "epoch": 0.02945, + "grad_norm": 0.5285725593566895, + "learning_rate": 0.0004995242076444093, + "loss": 3.265087127685547, + "step": 5890 + }, + { + "epoch": 0.0295, + "grad_norm": 0.5580034852027893, + "learning_rate": 0.0004995217584177581, + "loss": 3.306072998046875, + "step": 5900 + }, + { + "epoch": 0.02955, + "grad_norm": 0.5201782584190369, + "learning_rate": 0.0004995193029093984, + "loss": 3.2790176391601564, + "step": 5910 + }, + { + "epoch": 0.0296, + "grad_norm": 0.5060904622077942, + "learning_rate": 0.0004995168411193918, + "loss": 3.273830032348633, + "step": 5920 + }, + { + "epoch": 0.02965, + "grad_norm": 0.5688839554786682, + "learning_rate": 0.0004995143730478005, + "loss": 3.262922668457031, + "step": 5930 + }, + { + "epoch": 0.0297, + "grad_norm": 0.5308981537818909, + "learning_rate": 0.0004995118986946864, + "loss": 3.3189670562744142, + "step": 5940 + }, + { + "epoch": 0.02975, + "grad_norm": 0.524124264717102, + "learning_rate": 0.000499509418060112, + "loss": 3.3045177459716797, + "step": 5950 + }, + { + "epoch": 0.0298, + "grad_norm": 0.5645061731338501, + "learning_rate": 0.0004995069311441396, + "loss": 3.311348724365234, + "step": 5960 + }, + { + "epoch": 0.02985, + "grad_norm": 0.49135470390319824, + "learning_rate": 0.0004995044379468319, + "loss": 3.276051712036133, + "step": 5970 + }, + { + "epoch": 0.0299, + "grad_norm": 0.5229732990264893, + "learning_rate": 0.0004995019384682516, + "loss": 3.318883514404297, + "step": 5980 + }, + { + "epoch": 0.02995, + "grad_norm": 0.5387836694717407, + "learning_rate": 0.0004994994327084619, + "loss": 3.3334293365478516, + "step": 5990 + }, + { + "epoch": 0.03, + "grad_norm": 0.5268911719322205, + "learning_rate": 0.0004994969206675254, + "loss": 3.348115158081055, + "step": 6000 + }, + { + "epoch": 0.03005, + "grad_norm": 0.533720076084137, + "learning_rate": 0.0004994944023455056, + "loss": 3.2566577911376955, + "step": 6010 + }, + { + "epoch": 0.0301, + "grad_norm": 0.5218222737312317, + "learning_rate": 0.000499491877742466, + "loss": 3.3181602478027346, + "step": 6020 + }, + { + "epoch": 0.03015, + "grad_norm": 0.4921139180660248, + "learning_rate": 0.00049948934685847, + "loss": 3.3189598083496095, + "step": 6030 + }, + { + "epoch": 0.0302, + "grad_norm": 0.5054450631141663, + "learning_rate": 0.0004994868096935813, + "loss": 3.2642593383789062, + "step": 6040 + }, + { + "epoch": 0.03025, + "grad_norm": 0.5111818313598633, + "learning_rate": 0.0004994842662478639, + "loss": 3.3180168151855467, + "step": 6050 + }, + { + "epoch": 0.0303, + "grad_norm": 0.5112853050231934, + "learning_rate": 0.0004994817165213817, + "loss": 3.2967681884765625, + "step": 6060 + }, + { + "epoch": 0.03035, + "grad_norm": 0.5122900605201721, + "learning_rate": 0.0004994791605141991, + "loss": 3.242235565185547, + "step": 6070 + }, + { + "epoch": 0.0304, + "grad_norm": 0.4978131353855133, + "learning_rate": 0.0004994765982263801, + "loss": 3.3159523010253906, + "step": 6080 + }, + { + "epoch": 0.03045, + "grad_norm": 0.49992668628692627, + "learning_rate": 0.0004994740296579896, + "loss": 3.305986785888672, + "step": 6090 + }, + { + "epoch": 0.0305, + "grad_norm": 0.49126267433166504, + "learning_rate": 0.000499471454809092, + "loss": 3.232387924194336, + "step": 6100 + }, + { + "epoch": 0.03055, + "grad_norm": 0.4805096387863159, + "learning_rate": 0.0004994688736797521, + "loss": 3.26964111328125, + "step": 6110 + }, + { + "epoch": 0.0306, + "grad_norm": 0.5200438499450684, + "learning_rate": 0.0004994662862700351, + "loss": 3.273984909057617, + "step": 6120 + }, + { + "epoch": 0.03065, + "grad_norm": 0.5100458264350891, + "learning_rate": 0.0004994636925800059, + "loss": 3.289195251464844, + "step": 6130 + }, + { + "epoch": 0.0307, + "grad_norm": 0.502159595489502, + "learning_rate": 0.0004994610926097301, + "loss": 3.2619098663330077, + "step": 6140 + }, + { + "epoch": 0.03075, + "grad_norm": 0.5062483549118042, + "learning_rate": 0.0004994584863592727, + "loss": 3.319671630859375, + "step": 6150 + }, + { + "epoch": 0.0308, + "grad_norm": 0.49435099959373474, + "learning_rate": 0.0004994558738286997, + "loss": 3.297564697265625, + "step": 6160 + }, + { + "epoch": 0.03085, + "grad_norm": 0.4899642765522003, + "learning_rate": 0.0004994532550180768, + "loss": 3.2695159912109375, + "step": 6170 + }, + { + "epoch": 0.0309, + "grad_norm": 0.5279033780097961, + "learning_rate": 0.0004994506299274698, + "loss": 3.288330078125, + "step": 6180 + }, + { + "epoch": 0.03095, + "grad_norm": 0.48805439472198486, + "learning_rate": 0.0004994479985569449, + "loss": 3.2908885955810545, + "step": 6190 + }, + { + "epoch": 0.031, + "grad_norm": 0.6219847202301025, + "learning_rate": 0.0004994453609065682, + "loss": 3.275014877319336, + "step": 6200 + }, + { + "epoch": 0.03105, + "grad_norm": 0.4954379200935364, + "learning_rate": 0.0004994427169764063, + "loss": 3.2329742431640627, + "step": 6210 + }, + { + "epoch": 0.0311, + "grad_norm": 0.5530985593795776, + "learning_rate": 0.0004994400667665257, + "loss": 3.291941452026367, + "step": 6220 + }, + { + "epoch": 0.03115, + "grad_norm": 0.5238752365112305, + "learning_rate": 0.000499437410276993, + "loss": 3.2442073822021484, + "step": 6230 + }, + { + "epoch": 0.0312, + "grad_norm": 0.5220026969909668, + "learning_rate": 0.0004994347475078752, + "loss": 3.2727874755859374, + "step": 6240 + }, + { + "epoch": 0.03125, + "grad_norm": 0.5084006190299988, + "learning_rate": 0.0004994320784592393, + "loss": 3.2786075592041017, + "step": 6250 + }, + { + "epoch": 0.0313, + "grad_norm": 0.5238522291183472, + "learning_rate": 0.0004994294031311525, + "loss": 3.2862674713134767, + "step": 6260 + }, + { + "epoch": 0.03135, + "grad_norm": 0.5116362571716309, + "learning_rate": 0.000499426721523682, + "loss": 3.2732200622558594, + "step": 6270 + }, + { + "epoch": 0.0314, + "grad_norm": 0.5625714063644409, + "learning_rate": 0.0004994240336368957, + "loss": 3.277906036376953, + "step": 6280 + }, + { + "epoch": 0.03145, + "grad_norm": 0.5552911758422852, + "learning_rate": 0.0004994213394708609, + "loss": 3.2867942810058595, + "step": 6290 + }, + { + "epoch": 0.0315, + "grad_norm": 0.5355042815208435, + "learning_rate": 0.0004994186390256455, + "loss": 3.2476322174072267, + "step": 6300 + }, + { + "epoch": 0.03155, + "grad_norm": 0.5392954349517822, + "learning_rate": 0.0004994159323013175, + "loss": 3.2763442993164062, + "step": 6310 + }, + { + "epoch": 0.0316, + "grad_norm": 0.5052130818367004, + "learning_rate": 0.0004994132192979452, + "loss": 3.248729705810547, + "step": 6320 + }, + { + "epoch": 0.03165, + "grad_norm": 0.4915483891963959, + "learning_rate": 0.0004994105000155968, + "loss": 3.2349502563476564, + "step": 6330 + }, + { + "epoch": 0.0317, + "grad_norm": 0.5317585468292236, + "learning_rate": 0.0004994077744543406, + "loss": 3.296322250366211, + "step": 6340 + }, + { + "epoch": 0.03175, + "grad_norm": 0.5325540900230408, + "learning_rate": 0.0004994050426142453, + "loss": 3.2660888671875, + "step": 6350 + }, + { + "epoch": 0.0318, + "grad_norm": 0.4836825728416443, + "learning_rate": 0.0004994023044953799, + "loss": 3.278049087524414, + "step": 6360 + }, + { + "epoch": 0.03185, + "grad_norm": 0.48468777537345886, + "learning_rate": 0.000499399560097813, + "loss": 3.2285030364990233, + "step": 6370 + }, + { + "epoch": 0.0319, + "grad_norm": 0.5050444602966309, + "learning_rate": 0.0004993968094216139, + "loss": 3.2519603729248048, + "step": 6380 + }, + { + "epoch": 0.03195, + "grad_norm": 0.5457867980003357, + "learning_rate": 0.0004993940524668518, + "loss": 3.2818103790283204, + "step": 6390 + }, + { + "epoch": 0.032, + "grad_norm": 0.510442852973938, + "learning_rate": 0.0004993912892335961, + "loss": 3.303788757324219, + "step": 6400 + }, + { + "epoch": 0.03205, + "grad_norm": 0.5095071792602539, + "learning_rate": 0.0004993885197219164, + "loss": 3.245164489746094, + "step": 6410 + }, + { + "epoch": 0.0321, + "grad_norm": 0.48452892899513245, + "learning_rate": 0.0004993857439318823, + "loss": 3.276142120361328, + "step": 6420 + }, + { + "epoch": 0.03215, + "grad_norm": 0.5133600234985352, + "learning_rate": 0.0004993829618635638, + "loss": 3.25024299621582, + "step": 6430 + }, + { + "epoch": 0.0322, + "grad_norm": 0.4911234676837921, + "learning_rate": 0.0004993801735170308, + "loss": 3.2562419891357424, + "step": 6440 + }, + { + "epoch": 0.03225, + "grad_norm": 0.5619009733200073, + "learning_rate": 0.0004993773788923538, + "loss": 3.245888519287109, + "step": 6450 + }, + { + "epoch": 0.0323, + "grad_norm": 0.4921376705169678, + "learning_rate": 0.0004993745779896028, + "loss": 3.282980728149414, + "step": 6460 + }, + { + "epoch": 0.03235, + "grad_norm": 0.5232189893722534, + "learning_rate": 0.0004993717708088486, + "loss": 3.2775272369384765, + "step": 6470 + }, + { + "epoch": 0.0324, + "grad_norm": 0.5013182163238525, + "learning_rate": 0.0004993689573501615, + "loss": 3.2703956604003905, + "step": 6480 + }, + { + "epoch": 0.03245, + "grad_norm": 0.49869999289512634, + "learning_rate": 0.0004993661376136128, + "loss": 3.2950801849365234, + "step": 6490 + }, + { + "epoch": 0.0325, + "grad_norm": 0.5136047601699829, + "learning_rate": 0.0004993633115992731, + "loss": 3.2700271606445312, + "step": 6500 + }, + { + "epoch": 0.03255, + "grad_norm": 0.4794382154941559, + "learning_rate": 0.0004993604793072138, + "loss": 3.2879711151123048, + "step": 6510 + }, + { + "epoch": 0.0326, + "grad_norm": 0.5391140580177307, + "learning_rate": 0.000499357640737506, + "loss": 3.2888607025146483, + "step": 6520 + }, + { + "epoch": 0.03265, + "grad_norm": 0.46451038122177124, + "learning_rate": 0.0004993547958902214, + "loss": 3.253022003173828, + "step": 6530 + }, + { + "epoch": 0.0327, + "grad_norm": 0.47015851736068726, + "learning_rate": 0.0004993519447654313, + "loss": 3.2001136779785155, + "step": 6540 + }, + { + "epoch": 0.03275, + "grad_norm": 0.5482874512672424, + "learning_rate": 0.0004993490873632079, + "loss": 3.2537071228027346, + "step": 6550 + }, + { + "epoch": 0.0328, + "grad_norm": 0.5338917374610901, + "learning_rate": 0.0004993462236836226, + "loss": 3.2528221130371096, + "step": 6560 + }, + { + "epoch": 0.03285, + "grad_norm": 0.5224674940109253, + "learning_rate": 0.0004993433537267479, + "loss": 3.26834716796875, + "step": 6570 + }, + { + "epoch": 0.0329, + "grad_norm": 0.5648828148841858, + "learning_rate": 0.000499340477492656, + "loss": 3.2703411102294924, + "step": 6580 + }, + { + "epoch": 0.03295, + "grad_norm": 0.47461557388305664, + "learning_rate": 0.0004993375949814193, + "loss": 3.315426254272461, + "step": 6590 + }, + { + "epoch": 0.033, + "grad_norm": 0.5208790302276611, + "learning_rate": 0.0004993347061931102, + "loss": 3.2754776000976564, + "step": 6600 + }, + { + "epoch": 0.03305, + "grad_norm": 0.4847155511379242, + "learning_rate": 0.0004993318111278015, + "loss": 3.2533458709716796, + "step": 6610 + }, + { + "epoch": 0.0331, + "grad_norm": 0.5042962431907654, + "learning_rate": 0.0004993289097855663, + "loss": 3.2328304290771483, + "step": 6620 + }, + { + "epoch": 0.03315, + "grad_norm": 0.4863143563270569, + "learning_rate": 0.0004993260021664773, + "loss": 3.2165931701660155, + "step": 6630 + }, + { + "epoch": 0.0332, + "grad_norm": 0.5163339376449585, + "learning_rate": 0.0004993230882706079, + "loss": 3.2574859619140626, + "step": 6640 + }, + { + "epoch": 0.03325, + "grad_norm": 0.49472928047180176, + "learning_rate": 0.0004993201680980314, + "loss": 3.205926513671875, + "step": 6650 + }, + { + "epoch": 0.0333, + "grad_norm": 0.5075984001159668, + "learning_rate": 0.0004993172416488214, + "loss": 3.243471145629883, + "step": 6660 + }, + { + "epoch": 0.03335, + "grad_norm": 0.5120320916175842, + "learning_rate": 0.0004993143089230514, + "loss": 3.2755123138427735, + "step": 6670 + }, + { + "epoch": 0.0334, + "grad_norm": 0.4763045012950897, + "learning_rate": 0.0004993113699207954, + "loss": 3.279844284057617, + "step": 6680 + }, + { + "epoch": 0.03345, + "grad_norm": 0.48032042384147644, + "learning_rate": 0.0004993084246421273, + "loss": 3.2587844848632814, + "step": 6690 + }, + { + "epoch": 0.0335, + "grad_norm": 0.49343982338905334, + "learning_rate": 0.0004993054730871214, + "loss": 3.2762332916259767, + "step": 6700 + }, + { + "epoch": 0.03355, + "grad_norm": 0.47082820534706116, + "learning_rate": 0.0004993025152558518, + "loss": 3.24307861328125, + "step": 6710 + }, + { + "epoch": 0.0336, + "grad_norm": 0.5307480692863464, + "learning_rate": 0.000499299551148393, + "loss": 3.1941551208496093, + "step": 6720 + }, + { + "epoch": 0.03365, + "grad_norm": 0.5009476542472839, + "learning_rate": 0.0004992965807648197, + "loss": 3.2805355072021483, + "step": 6730 + }, + { + "epoch": 0.0337, + "grad_norm": 0.5025913119316101, + "learning_rate": 0.0004992936041052067, + "loss": 3.2021499633789063, + "step": 6740 + }, + { + "epoch": 0.03375, + "grad_norm": 0.498792827129364, + "learning_rate": 0.0004992906211696288, + "loss": 3.209592819213867, + "step": 6750 + }, + { + "epoch": 0.0338, + "grad_norm": 0.4788338840007782, + "learning_rate": 0.0004992876319581612, + "loss": 3.2201515197753907, + "step": 6760 + }, + { + "epoch": 0.03385, + "grad_norm": 0.4831807613372803, + "learning_rate": 0.000499284636470879, + "loss": 3.24774055480957, + "step": 6770 + }, + { + "epoch": 0.0339, + "grad_norm": 0.4945288300514221, + "learning_rate": 0.0004992816347078581, + "loss": 3.300674819946289, + "step": 6780 + }, + { + "epoch": 0.03395, + "grad_norm": 0.4917133152484894, + "learning_rate": 0.0004992786266691734, + "loss": 3.245574951171875, + "step": 6790 + }, + { + "epoch": 0.034, + "grad_norm": 0.4901202321052551, + "learning_rate": 0.000499275612354901, + "loss": 3.247610092163086, + "step": 6800 + }, + { + "epoch": 0.03405, + "grad_norm": 0.5339242219924927, + "learning_rate": 0.0004992725917651168, + "loss": 3.2915828704833983, + "step": 6810 + }, + { + "epoch": 0.0341, + "grad_norm": 0.5356748104095459, + "learning_rate": 0.0004992695648998967, + "loss": 3.2269500732421874, + "step": 6820 + }, + { + "epoch": 0.03415, + "grad_norm": 0.5013329386711121, + "learning_rate": 0.000499266531759317, + "loss": 3.2411808013916015, + "step": 6830 + }, + { + "epoch": 0.0342, + "grad_norm": 1.5115830898284912, + "learning_rate": 0.000499263492343454, + "loss": 3.2329681396484373, + "step": 6840 + }, + { + "epoch": 0.03425, + "grad_norm": 0.5123684406280518, + "learning_rate": 0.0004992604466523842, + "loss": 3.1909128189086915, + "step": 6850 + }, + { + "epoch": 0.0343, + "grad_norm": 0.49313297867774963, + "learning_rate": 0.0004992573946861845, + "loss": 3.2709373474121093, + "step": 6860 + }, + { + "epoch": 0.03435, + "grad_norm": 0.48508724570274353, + "learning_rate": 0.0004992543364449314, + "loss": 3.2271133422851563, + "step": 6870 + }, + { + "epoch": 0.0344, + "grad_norm": 0.5724422931671143, + "learning_rate": 0.0004992512719287021, + "loss": 3.2603755950927735, + "step": 6880 + }, + { + "epoch": 0.03445, + "grad_norm": 0.4949094355106354, + "learning_rate": 0.0004992482011375736, + "loss": 3.2478958129882813, + "step": 6890 + }, + { + "epoch": 0.0345, + "grad_norm": 0.48316216468811035, + "learning_rate": 0.0004992451240716235, + "loss": 3.2388153076171875, + "step": 6900 + }, + { + "epoch": 0.03455, + "grad_norm": 0.4677082896232605, + "learning_rate": 0.0004992420407309289, + "loss": 3.2624778747558594, + "step": 6910 + }, + { + "epoch": 0.0346, + "grad_norm": 0.4864460527896881, + "learning_rate": 0.0004992389511155676, + "loss": 3.246590805053711, + "step": 6920 + }, + { + "epoch": 0.03465, + "grad_norm": 0.5285934209823608, + "learning_rate": 0.0004992358552256174, + "loss": 3.1736331939697267, + "step": 6930 + }, + { + "epoch": 0.0347, + "grad_norm": 0.49750179052352905, + "learning_rate": 0.0004992327530611562, + "loss": 3.223249053955078, + "step": 6940 + }, + { + "epoch": 0.03475, + "grad_norm": 0.4802745282649994, + "learning_rate": 0.0004992296446222622, + "loss": 3.2537120819091796, + "step": 6950 + }, + { + "epoch": 0.0348, + "grad_norm": 0.5321634411811829, + "learning_rate": 0.0004992265299090136, + "loss": 3.239070510864258, + "step": 6960 + }, + { + "epoch": 0.03485, + "grad_norm": 0.49372953176498413, + "learning_rate": 0.0004992234089214885, + "loss": 3.2400741577148438, + "step": 6970 + }, + { + "epoch": 0.0349, + "grad_norm": 0.5358172655105591, + "learning_rate": 0.0004992202816597659, + "loss": 3.3270233154296873, + "step": 6980 + }, + { + "epoch": 0.03495, + "grad_norm": 0.5094825625419617, + "learning_rate": 0.0004992171481239245, + "loss": 3.2704368591308595, + "step": 6990 + }, + { + "epoch": 0.035, + "grad_norm": 0.4792259931564331, + "learning_rate": 0.0004992140083140429, + "loss": 3.266705322265625, + "step": 7000 + }, + { + "epoch": 0.03505, + "grad_norm": 0.47373342514038086, + "learning_rate": 0.0004992108622302004, + "loss": 3.2925643920898438, + "step": 7010 + }, + { + "epoch": 0.0351, + "grad_norm": 0.4856434166431427, + "learning_rate": 0.000499207709872476, + "loss": 3.2224048614501952, + "step": 7020 + }, + { + "epoch": 0.03515, + "grad_norm": 0.48476263880729675, + "learning_rate": 0.0004992045512409493, + "loss": 3.2432369232177733, + "step": 7030 + }, + { + "epoch": 0.0352, + "grad_norm": 0.4937853515148163, + "learning_rate": 0.0004992013863356995, + "loss": 3.261783218383789, + "step": 7040 + }, + { + "epoch": 0.03525, + "grad_norm": 0.4836219549179077, + "learning_rate": 0.0004991982151568066, + "loss": 3.1651046752929686, + "step": 7050 + }, + { + "epoch": 0.0353, + "grad_norm": 0.48047834634780884, + "learning_rate": 0.0004991950377043504, + "loss": 3.2807167053222654, + "step": 7060 + }, + { + "epoch": 0.03535, + "grad_norm": 0.49931254982948303, + "learning_rate": 0.0004991918539784106, + "loss": 3.232622528076172, + "step": 7070 + }, + { + "epoch": 0.0354, + "grad_norm": 0.5312271118164062, + "learning_rate": 0.0004991886639790676, + "loss": 3.2711376190185546, + "step": 7080 + }, + { + "epoch": 0.03545, + "grad_norm": 0.5021573305130005, + "learning_rate": 0.0004991854677064017, + "loss": 3.244460678100586, + "step": 7090 + }, + { + "epoch": 0.0355, + "grad_norm": 0.4539817273616791, + "learning_rate": 0.0004991822651604932, + "loss": 3.236097717285156, + "step": 7100 + }, + { + "epoch": 0.03555, + "grad_norm": 0.47527721524238586, + "learning_rate": 0.000499179056341423, + "loss": 3.2103286743164063, + "step": 7110 + }, + { + "epoch": 0.0356, + "grad_norm": 0.49326497316360474, + "learning_rate": 0.0004991758412492717, + "loss": 3.2810848236083983, + "step": 7120 + }, + { + "epoch": 0.03565, + "grad_norm": 0.5137958526611328, + "learning_rate": 0.0004991726198841201, + "loss": 3.226940155029297, + "step": 7130 + }, + { + "epoch": 0.0357, + "grad_norm": 0.4870864450931549, + "learning_rate": 0.0004991693922460496, + "loss": 3.2021297454833983, + "step": 7140 + }, + { + "epoch": 0.03575, + "grad_norm": 0.4792308509349823, + "learning_rate": 0.0004991661583351412, + "loss": 3.2091800689697267, + "step": 7150 + }, + { + "epoch": 0.0358, + "grad_norm": 0.4589584767818451, + "learning_rate": 0.0004991629181514765, + "loss": 3.2232219696044924, + "step": 7160 + }, + { + "epoch": 0.03585, + "grad_norm": 0.47441065311431885, + "learning_rate": 0.0004991596716951371, + "loss": 3.2480442047119142, + "step": 7170 + }, + { + "epoch": 0.0359, + "grad_norm": 0.5039315223693848, + "learning_rate": 0.0004991564189662046, + "loss": 3.1960277557373047, + "step": 7180 + }, + { + "epoch": 0.03595, + "grad_norm": 0.49738526344299316, + "learning_rate": 0.0004991531599647607, + "loss": 3.31766357421875, + "step": 7190 + }, + { + "epoch": 0.036, + "grad_norm": 0.4779835641384125, + "learning_rate": 0.000499149894690888, + "loss": 3.271509552001953, + "step": 7200 + }, + { + "epoch": 0.03605, + "grad_norm": 0.5294590592384338, + "learning_rate": 0.000499146623144668, + "loss": 3.2204833984375, + "step": 7210 + }, + { + "epoch": 0.0361, + "grad_norm": 0.4992515742778778, + "learning_rate": 0.0004991433453261836, + "loss": 3.209004211425781, + "step": 7220 + }, + { + "epoch": 0.03615, + "grad_norm": 0.5233227610588074, + "learning_rate": 0.0004991400612355172, + "loss": 3.2482990264892577, + "step": 7230 + }, + { + "epoch": 0.0362, + "grad_norm": 0.495282918214798, + "learning_rate": 0.0004991367708727514, + "loss": 3.213483428955078, + "step": 7240 + }, + { + "epoch": 0.03625, + "grad_norm": 0.4923434853553772, + "learning_rate": 0.0004991334742379691, + "loss": 3.228572463989258, + "step": 7250 + }, + { + "epoch": 0.0363, + "grad_norm": 0.4879579246044159, + "learning_rate": 0.0004991301713312532, + "loss": 3.1901361465454103, + "step": 7260 + }, + { + "epoch": 0.03635, + "grad_norm": 0.46293842792510986, + "learning_rate": 0.0004991268621526868, + "loss": 3.1946006774902345, + "step": 7270 + }, + { + "epoch": 0.0364, + "grad_norm": 0.4598213732242584, + "learning_rate": 0.0004991235467023534, + "loss": 3.2230091094970703, + "step": 7280 + }, + { + "epoch": 0.03645, + "grad_norm": 0.46436646580696106, + "learning_rate": 0.0004991202249803364, + "loss": 3.1835960388183593, + "step": 7290 + }, + { + "epoch": 0.0365, + "grad_norm": 0.5189260244369507, + "learning_rate": 0.0004991168969867194, + "loss": 3.2238300323486326, + "step": 7300 + }, + { + "epoch": 0.03655, + "grad_norm": 0.5279228687286377, + "learning_rate": 0.0004991135627215862, + "loss": 3.2716575622558595, + "step": 7310 + }, + { + "epoch": 0.0366, + "grad_norm": 0.47355177998542786, + "learning_rate": 0.0004991102221850206, + "loss": 3.225547027587891, + "step": 7320 + }, + { + "epoch": 0.03665, + "grad_norm": 0.4564456343650818, + "learning_rate": 0.0004991068753771069, + "loss": 3.2203292846679688, + "step": 7330 + }, + { + "epoch": 0.0367, + "grad_norm": 0.528670072555542, + "learning_rate": 0.0004991035222979293, + "loss": 3.213151550292969, + "step": 7340 + }, + { + "epoch": 0.03675, + "grad_norm": 0.4890163540840149, + "learning_rate": 0.0004991001629475721, + "loss": 3.224197769165039, + "step": 7350 + }, + { + "epoch": 0.0368, + "grad_norm": 0.4717147648334503, + "learning_rate": 0.0004990967973261201, + "loss": 3.2003841400146484, + "step": 7360 + }, + { + "epoch": 0.03685, + "grad_norm": 0.4791920781135559, + "learning_rate": 0.0004990934254336579, + "loss": 3.2163047790527344, + "step": 7370 + }, + { + "epoch": 0.0369, + "grad_norm": 0.49152177572250366, + "learning_rate": 0.0004990900472702702, + "loss": 3.1819284439086912, + "step": 7380 + }, + { + "epoch": 0.03695, + "grad_norm": 0.4812125861644745, + "learning_rate": 0.0004990866628360423, + "loss": 3.199116516113281, + "step": 7390 + }, + { + "epoch": 0.037, + "grad_norm": 0.48500382900238037, + "learning_rate": 0.0004990832721310594, + "loss": 3.243368911743164, + "step": 7400 + }, + { + "epoch": 0.03705, + "grad_norm": 0.4728527069091797, + "learning_rate": 0.0004990798751554067, + "loss": 3.2324337005615233, + "step": 7410 + }, + { + "epoch": 0.0371, + "grad_norm": 0.4938664436340332, + "learning_rate": 0.0004990764719091699, + "loss": 3.2224071502685545, + "step": 7420 + }, + { + "epoch": 0.03715, + "grad_norm": 0.4866945743560791, + "learning_rate": 0.0004990730623924346, + "loss": 3.2091888427734374, + "step": 7430 + }, + { + "epoch": 0.0372, + "grad_norm": 0.516192615032196, + "learning_rate": 0.0004990696466052864, + "loss": 3.2389488220214844, + "step": 7440 + }, + { + "epoch": 0.03725, + "grad_norm": 0.4919445514678955, + "learning_rate": 0.0004990662245478117, + "loss": 3.186039924621582, + "step": 7450 + }, + { + "epoch": 0.0373, + "grad_norm": 0.4867427945137024, + "learning_rate": 0.0004990627962200965, + "loss": 3.1708776473999025, + "step": 7460 + }, + { + "epoch": 0.03735, + "grad_norm": 0.49825319647789, + "learning_rate": 0.000499059361622227, + "loss": 3.1951290130615235, + "step": 7470 + }, + { + "epoch": 0.0374, + "grad_norm": 0.511611819267273, + "learning_rate": 0.0004990559207542896, + "loss": 3.2481945037841795, + "step": 7480 + }, + { + "epoch": 0.03745, + "grad_norm": 0.4825802445411682, + "learning_rate": 0.0004990524736163713, + "loss": 3.1918725967407227, + "step": 7490 + }, + { + "epoch": 0.0375, + "grad_norm": 0.49114444851875305, + "learning_rate": 0.0004990490202085586, + "loss": 3.145640754699707, + "step": 7500 + }, + { + "epoch": 0.03755, + "grad_norm": 0.5187509059906006, + "learning_rate": 0.0004990455605309383, + "loss": 3.259341812133789, + "step": 7510 + }, + { + "epoch": 0.0376, + "grad_norm": 0.4948999583721161, + "learning_rate": 0.0004990420945835978, + "loss": 3.202710723876953, + "step": 7520 + }, + { + "epoch": 0.03765, + "grad_norm": 0.5042881369590759, + "learning_rate": 0.0004990386223666242, + "loss": 3.170927810668945, + "step": 7530 + }, + { + "epoch": 0.0377, + "grad_norm": 0.48680421710014343, + "learning_rate": 0.000499035143880105, + "loss": 3.223421096801758, + "step": 7540 + }, + { + "epoch": 0.03775, + "grad_norm": 0.4766632914543152, + "learning_rate": 0.0004990316591241277, + "loss": 3.176899528503418, + "step": 7550 + }, + { + "epoch": 0.0378, + "grad_norm": 0.48786455392837524, + "learning_rate": 0.00049902816809878, + "loss": 3.189167594909668, + "step": 7560 + }, + { + "epoch": 0.03785, + "grad_norm": 0.46467486023902893, + "learning_rate": 0.0004990246708041499, + "loss": 3.19119930267334, + "step": 7570 + }, + { + "epoch": 0.0379, + "grad_norm": 0.4480521082878113, + "learning_rate": 0.0004990211672403253, + "loss": 3.2240291595458985, + "step": 7580 + }, + { + "epoch": 0.03795, + "grad_norm": 0.48638153076171875, + "learning_rate": 0.0004990176574073946, + "loss": 3.242887496948242, + "step": 7590 + }, + { + "epoch": 0.038, + "grad_norm": 0.5008899569511414, + "learning_rate": 0.0004990141413054458, + "loss": 3.2070663452148436, + "step": 7600 + }, + { + "epoch": 0.03805, + "grad_norm": 0.49910885095596313, + "learning_rate": 0.0004990106189345679, + "loss": 3.2050395965576173, + "step": 7610 + }, + { + "epoch": 0.0381, + "grad_norm": 0.47233179211616516, + "learning_rate": 0.0004990070902948493, + "loss": 3.194321060180664, + "step": 7620 + }, + { + "epoch": 0.03815, + "grad_norm": 0.5117303133010864, + "learning_rate": 0.0004990035553863787, + "loss": 3.183699607849121, + "step": 7630 + }, + { + "epoch": 0.0382, + "grad_norm": 0.498078852891922, + "learning_rate": 0.0004990000142092454, + "loss": 3.2045738220214846, + "step": 7640 + }, + { + "epoch": 0.03825, + "grad_norm": 0.5052784085273743, + "learning_rate": 0.0004989964667635383, + "loss": 3.2328865051269533, + "step": 7650 + }, + { + "epoch": 0.0383, + "grad_norm": 0.5305922031402588, + "learning_rate": 0.000498992913049347, + "loss": 3.2302021026611327, + "step": 7660 + }, + { + "epoch": 0.03835, + "grad_norm": 0.49253398180007935, + "learning_rate": 0.0004989893530667606, + "loss": 3.2207775115966797, + "step": 7670 + }, + { + "epoch": 0.0384, + "grad_norm": 0.4628109633922577, + "learning_rate": 0.000498985786815869, + "loss": 3.171371650695801, + "step": 7680 + }, + { + "epoch": 0.03845, + "grad_norm": 0.47889015078544617, + "learning_rate": 0.0004989822142967618, + "loss": 3.22485237121582, + "step": 7690 + }, + { + "epoch": 0.0385, + "grad_norm": 0.44255518913269043, + "learning_rate": 0.0004989786355095292, + "loss": 3.247653579711914, + "step": 7700 + }, + { + "epoch": 0.03855, + "grad_norm": 0.5055277943611145, + "learning_rate": 0.0004989750504542609, + "loss": 3.232309341430664, + "step": 7710 + }, + { + "epoch": 0.0386, + "grad_norm": 0.45164525508880615, + "learning_rate": 0.0004989714591310475, + "loss": 3.220487594604492, + "step": 7720 + }, + { + "epoch": 0.03865, + "grad_norm": 0.47939571738243103, + "learning_rate": 0.0004989678615399792, + "loss": 3.1935367584228516, + "step": 7730 + }, + { + "epoch": 0.0387, + "grad_norm": 0.5174024105072021, + "learning_rate": 0.0004989642576811467, + "loss": 3.192740631103516, + "step": 7740 + }, + { + "epoch": 0.03875, + "grad_norm": 0.46976742148399353, + "learning_rate": 0.0004989606475546406, + "loss": 3.2493206024169923, + "step": 7750 + }, + { + "epoch": 0.0388, + "grad_norm": 0.5179904103279114, + "learning_rate": 0.000498957031160552, + "loss": 3.2292903900146483, + "step": 7760 + }, + { + "epoch": 0.03885, + "grad_norm": 0.4879210889339447, + "learning_rate": 0.0004989534084989716, + "loss": 3.166905403137207, + "step": 7770 + }, + { + "epoch": 0.0389, + "grad_norm": 0.4663355052471161, + "learning_rate": 0.000498949779569991, + "loss": 3.1985023498535154, + "step": 7780 + }, + { + "epoch": 0.03895, + "grad_norm": 0.48129352927207947, + "learning_rate": 0.0004989461443737013, + "loss": 3.2262107849121096, + "step": 7790 + }, + { + "epoch": 0.039, + "grad_norm": 0.4751749038696289, + "learning_rate": 0.000498942502910194, + "loss": 3.198856735229492, + "step": 7800 + }, + { + "epoch": 0.03905, + "grad_norm": 0.4726470410823822, + "learning_rate": 0.0004989388551795609, + "loss": 3.2094451904296877, + "step": 7810 + }, + { + "epoch": 0.0391, + "grad_norm": 0.5053130984306335, + "learning_rate": 0.0004989352011818939, + "loss": 3.2072391510009766, + "step": 7820 + }, + { + "epoch": 0.03915, + "grad_norm": 0.49643540382385254, + "learning_rate": 0.0004989315409172847, + "loss": 3.174172782897949, + "step": 7830 + }, + { + "epoch": 0.0392, + "grad_norm": 0.4497954845428467, + "learning_rate": 0.0004989278743858257, + "loss": 3.154504585266113, + "step": 7840 + }, + { + "epoch": 0.03925, + "grad_norm": 0.5007823705673218, + "learning_rate": 0.0004989242015876091, + "loss": 3.185009002685547, + "step": 7850 + }, + { + "epoch": 0.0393, + "grad_norm": 0.5103575587272644, + "learning_rate": 0.0004989205225227275, + "loss": 3.2249782562255858, + "step": 7860 + }, + { + "epoch": 0.03935, + "grad_norm": 0.495313435792923, + "learning_rate": 0.0004989168371912732, + "loss": 3.1945751190185545, + "step": 7870 + }, + { + "epoch": 0.0394, + "grad_norm": 0.44986823201179504, + "learning_rate": 0.0004989131455933394, + "loss": 3.1957509994506834, + "step": 7880 + }, + { + "epoch": 0.03945, + "grad_norm": 0.4652824103832245, + "learning_rate": 0.0004989094477290186, + "loss": 3.179241180419922, + "step": 7890 + }, + { + "epoch": 0.0395, + "grad_norm": 0.5021610260009766, + "learning_rate": 0.0004989057435984043, + "loss": 3.242484283447266, + "step": 7900 + }, + { + "epoch": 0.03955, + "grad_norm": 0.4753408432006836, + "learning_rate": 0.0004989020332015896, + "loss": 3.214987945556641, + "step": 7910 + }, + { + "epoch": 0.0396, + "grad_norm": 0.481865257024765, + "learning_rate": 0.0004988983165386678, + "loss": 3.2566688537597654, + "step": 7920 + }, + { + "epoch": 0.03965, + "grad_norm": 0.47750943899154663, + "learning_rate": 0.0004988945936097325, + "loss": 3.1665376663208007, + "step": 7930 + }, + { + "epoch": 0.0397, + "grad_norm": 0.5159562826156616, + "learning_rate": 0.0004988908644148776, + "loss": 3.196044158935547, + "step": 7940 + }, + { + "epoch": 0.03975, + "grad_norm": 0.47829440236091614, + "learning_rate": 0.0004988871289541967, + "loss": 3.1601608276367186, + "step": 7950 + }, + { + "epoch": 0.0398, + "grad_norm": 0.48772871494293213, + "learning_rate": 0.000498883387227784, + "loss": 3.198003578186035, + "step": 7960 + }, + { + "epoch": 0.03985, + "grad_norm": 0.5004878640174866, + "learning_rate": 0.0004988796392357337, + "loss": 3.2155906677246096, + "step": 7970 + }, + { + "epoch": 0.0399, + "grad_norm": 0.4770375192165375, + "learning_rate": 0.0004988758849781402, + "loss": 3.226980209350586, + "step": 7980 + }, + { + "epoch": 0.03995, + "grad_norm": 0.5024970173835754, + "learning_rate": 0.000498872124455098, + "loss": 3.189903450012207, + "step": 7990 + }, + { + "epoch": 0.04, + "grad_norm": 0.5264940857887268, + "learning_rate": 0.0004988683576667016, + "loss": 3.207796096801758, + "step": 8000 + }, + { + "epoch": 0.04005, + "grad_norm": 0.4478391110897064, + "learning_rate": 0.000498864584613046, + "loss": 3.228920745849609, + "step": 8010 + }, + { + "epoch": 0.0401, + "grad_norm": 0.43791306018829346, + "learning_rate": 0.0004988608052942262, + "loss": 3.21553955078125, + "step": 8020 + }, + { + "epoch": 0.04015, + "grad_norm": 0.512967050075531, + "learning_rate": 0.0004988570197103372, + "loss": 3.1639360427856444, + "step": 8030 + }, + { + "epoch": 0.0402, + "grad_norm": 0.4745829105377197, + "learning_rate": 0.0004988532278614745, + "loss": 3.1789928436279298, + "step": 8040 + }, + { + "epoch": 0.04025, + "grad_norm": 0.49086299538612366, + "learning_rate": 0.0004988494297477335, + "loss": 3.187704849243164, + "step": 8050 + }, + { + "epoch": 0.0403, + "grad_norm": 0.44906172156333923, + "learning_rate": 0.0004988456253692096, + "loss": 3.1720056533813477, + "step": 8060 + }, + { + "epoch": 0.04035, + "grad_norm": 0.48861899971961975, + "learning_rate": 0.0004988418147259989, + "loss": 3.2126617431640625, + "step": 8070 + }, + { + "epoch": 0.0404, + "grad_norm": 0.5241518616676331, + "learning_rate": 0.000498837997818197, + "loss": 3.253704071044922, + "step": 8080 + }, + { + "epoch": 0.04045, + "grad_norm": 0.49537423253059387, + "learning_rate": 0.0004988341746459003, + "loss": 3.1824726104736327, + "step": 8090 + }, + { + "epoch": 0.0405, + "grad_norm": 0.44281241297721863, + "learning_rate": 0.0004988303452092049, + "loss": 3.2190208435058594, + "step": 8100 + }, + { + "epoch": 0.04055, + "grad_norm": 0.49280691146850586, + "learning_rate": 0.0004988265095082072, + "loss": 3.188656044006348, + "step": 8110 + }, + { + "epoch": 0.0406, + "grad_norm": 0.477510005235672, + "learning_rate": 0.0004988226675430037, + "loss": 3.1533061981201174, + "step": 8120 + }, + { + "epoch": 0.04065, + "grad_norm": 0.4721514582633972, + "learning_rate": 0.0004988188193136914, + "loss": 3.2003250122070312, + "step": 8130 + }, + { + "epoch": 0.0407, + "grad_norm": 0.45115435123443604, + "learning_rate": 0.0004988149648203668, + "loss": 3.1746553421020507, + "step": 8140 + }, + { + "epoch": 0.04075, + "grad_norm": 0.441228985786438, + "learning_rate": 0.0004988111040631272, + "loss": 3.17890625, + "step": 8150 + }, + { + "epoch": 0.0408, + "grad_norm": 0.4961966574192047, + "learning_rate": 0.0004988072370420698, + "loss": 3.195034217834473, + "step": 8160 + }, + { + "epoch": 0.04085, + "grad_norm": 0.5196247696876526, + "learning_rate": 0.0004988033637572918, + "loss": 3.1872684478759767, + "step": 8170 + }, + { + "epoch": 0.0409, + "grad_norm": 0.43423622846603394, + "learning_rate": 0.0004987994842088907, + "loss": 3.2337421417236327, + "step": 8180 + }, + { + "epoch": 0.04095, + "grad_norm": 0.5099454522132874, + "learning_rate": 0.0004987955983969644, + "loss": 3.2432933807373048, + "step": 8190 + }, + { + "epoch": 0.041, + "grad_norm": 0.4863505959510803, + "learning_rate": 0.0004987917063216105, + "loss": 3.2176727294921874, + "step": 8200 + }, + { + "epoch": 0.04105, + "grad_norm": 0.5485396981239319, + "learning_rate": 0.000498787807982927, + "loss": 3.1979175567626954, + "step": 8210 + }, + { + "epoch": 0.0411, + "grad_norm": 0.4897323250770569, + "learning_rate": 0.0004987839033810122, + "loss": 3.176752281188965, + "step": 8220 + }, + { + "epoch": 0.04115, + "grad_norm": 0.48155128955841064, + "learning_rate": 0.0004987799925159641, + "loss": 3.1694107055664062, + "step": 8230 + }, + { + "epoch": 0.0412, + "grad_norm": 0.47420260310173035, + "learning_rate": 0.0004987760753878816, + "loss": 3.179531288146973, + "step": 8240 + }, + { + "epoch": 0.04125, + "grad_norm": 0.4835508465766907, + "learning_rate": 0.000498772151996863, + "loss": 3.211697006225586, + "step": 8250 + }, + { + "epoch": 0.0413, + "grad_norm": 0.5134413242340088, + "learning_rate": 0.0004987682223430071, + "loss": 3.2388690948486327, + "step": 8260 + }, + { + "epoch": 0.04135, + "grad_norm": 0.45428788661956787, + "learning_rate": 0.0004987642864264128, + "loss": 3.2121723175048826, + "step": 8270 + }, + { + "epoch": 0.0414, + "grad_norm": 0.46207815408706665, + "learning_rate": 0.0004987603442471793, + "loss": 3.1860191345214846, + "step": 8280 + }, + { + "epoch": 0.04145, + "grad_norm": 0.4950490891933441, + "learning_rate": 0.0004987563958054059, + "loss": 3.1937427520751953, + "step": 8290 + }, + { + "epoch": 0.0415, + "grad_norm": 0.5140568017959595, + "learning_rate": 0.0004987524411011917, + "loss": 3.171319580078125, + "step": 8300 + }, + { + "epoch": 0.04155, + "grad_norm": 0.4580945074558258, + "learning_rate": 0.0004987484801346366, + "loss": 3.1947460174560547, + "step": 8310 + }, + { + "epoch": 0.0416, + "grad_norm": 0.474616676568985, + "learning_rate": 0.0004987445129058402, + "loss": 3.1680988311767577, + "step": 8320 + }, + { + "epoch": 0.04165, + "grad_norm": 0.49910998344421387, + "learning_rate": 0.0004987405394149022, + "loss": 3.1660444259643556, + "step": 8330 + }, + { + "epoch": 0.0417, + "grad_norm": 0.46230077743530273, + "learning_rate": 0.0004987365596619229, + "loss": 3.1908626556396484, + "step": 8340 + }, + { + "epoch": 0.04175, + "grad_norm": 0.4711679220199585, + "learning_rate": 0.0004987325736470023, + "loss": 3.187510871887207, + "step": 8350 + }, + { + "epoch": 0.0418, + "grad_norm": 0.4694119095802307, + "learning_rate": 0.0004987285813702408, + "loss": 3.244455337524414, + "step": 8360 + }, + { + "epoch": 0.04185, + "grad_norm": 0.47847452759742737, + "learning_rate": 0.000498724582831739, + "loss": 3.158231735229492, + "step": 8370 + }, + { + "epoch": 0.0419, + "grad_norm": 0.49473607540130615, + "learning_rate": 0.0004987205780315975, + "loss": 3.1924915313720703, + "step": 8380 + }, + { + "epoch": 0.04195, + "grad_norm": 0.4969825744628906, + "learning_rate": 0.000498716566969917, + "loss": 3.1896041870117187, + "step": 8390 + }, + { + "epoch": 0.042, + "grad_norm": 0.47029972076416016, + "learning_rate": 0.0004987125496467988, + "loss": 3.218551254272461, + "step": 8400 + }, + { + "epoch": 0.04205, + "grad_norm": 0.5104393362998962, + "learning_rate": 0.0004987085260623437, + "loss": 3.1943470001220704, + "step": 8410 + }, + { + "epoch": 0.0421, + "grad_norm": 0.5051606297492981, + "learning_rate": 0.000498704496216653, + "loss": 3.170995903015137, + "step": 8420 + }, + { + "epoch": 0.04215, + "grad_norm": 0.4941001832485199, + "learning_rate": 0.0004987004601098285, + "loss": 3.180413246154785, + "step": 8430 + }, + { + "epoch": 0.0422, + "grad_norm": 0.48504188656806946, + "learning_rate": 0.0004986964177419714, + "loss": 3.1566673278808595, + "step": 8440 + }, + { + "epoch": 0.04225, + "grad_norm": 0.4373553693294525, + "learning_rate": 0.0004986923691131837, + "loss": 3.160191535949707, + "step": 8450 + }, + { + "epoch": 0.0423, + "grad_norm": 0.49672383069992065, + "learning_rate": 0.0004986883142235673, + "loss": 3.2302066802978517, + "step": 8460 + }, + { + "epoch": 0.04235, + "grad_norm": 0.4408532679080963, + "learning_rate": 0.0004986842530732243, + "loss": 3.230965805053711, + "step": 8470 + }, + { + "epoch": 0.0424, + "grad_norm": 0.48300543427467346, + "learning_rate": 0.0004986801856622568, + "loss": 3.1637229919433594, + "step": 8480 + }, + { + "epoch": 0.04245, + "grad_norm": 0.48256465792655945, + "learning_rate": 0.0004986761119907672, + "loss": 3.1872787475585938, + "step": 8490 + }, + { + "epoch": 0.0425, + "grad_norm": 0.5214313864707947, + "learning_rate": 0.0004986720320588583, + "loss": 3.1958826065063475, + "step": 8500 + }, + { + "epoch": 0.04255, + "grad_norm": 0.48056840896606445, + "learning_rate": 0.0004986679458666327, + "loss": 3.1689069747924803, + "step": 8510 + }, + { + "epoch": 0.0426, + "grad_norm": 0.4667593538761139, + "learning_rate": 0.000498663853414193, + "loss": 3.2324119567871095, + "step": 8520 + }, + { + "epoch": 0.04265, + "grad_norm": 0.4160764813423157, + "learning_rate": 0.0004986597547016426, + "loss": 3.1597890853881836, + "step": 8530 + }, + { + "epoch": 0.0427, + "grad_norm": 0.4717496335506439, + "learning_rate": 0.0004986556497290845, + "loss": 3.2089996337890625, + "step": 8540 + }, + { + "epoch": 0.04275, + "grad_norm": 0.4735127091407776, + "learning_rate": 0.0004986515384966221, + "loss": 3.203990936279297, + "step": 8550 + }, + { + "epoch": 0.0428, + "grad_norm": 0.4572399854660034, + "learning_rate": 0.0004986474210043589, + "loss": 3.12225227355957, + "step": 8560 + }, + { + "epoch": 0.04285, + "grad_norm": 0.4615976810455322, + "learning_rate": 0.0004986432972523986, + "loss": 3.2253887176513674, + "step": 8570 + }, + { + "epoch": 0.0429, + "grad_norm": 0.4790298044681549, + "learning_rate": 0.0004986391672408448, + "loss": 3.2347080230712892, + "step": 8580 + }, + { + "epoch": 0.04295, + "grad_norm": 0.4623195230960846, + "learning_rate": 0.0004986350309698017, + "loss": 3.2130195617675783, + "step": 8590 + }, + { + "epoch": 0.043, + "grad_norm": 0.5045721530914307, + "learning_rate": 0.0004986308884393733, + "loss": 3.150187683105469, + "step": 8600 + }, + { + "epoch": 0.04305, + "grad_norm": 0.44986870884895325, + "learning_rate": 0.0004986267396496641, + "loss": 3.1823823928833006, + "step": 8610 + }, + { + "epoch": 0.0431, + "grad_norm": 0.4488791227340698, + "learning_rate": 0.0004986225846007783, + "loss": 3.173758888244629, + "step": 8620 + }, + { + "epoch": 0.04315, + "grad_norm": 0.4780832529067993, + "learning_rate": 0.0004986184232928204, + "loss": 3.1506847381591796, + "step": 8630 + }, + { + "epoch": 0.0432, + "grad_norm": 0.48588043451309204, + "learning_rate": 0.0004986142557258956, + "loss": 3.228723907470703, + "step": 8640 + }, + { + "epoch": 0.04325, + "grad_norm": 0.46934494376182556, + "learning_rate": 0.0004986100819001086, + "loss": 3.189278793334961, + "step": 8650 + }, + { + "epoch": 0.0433, + "grad_norm": 0.5075275897979736, + "learning_rate": 0.0004986059018155643, + "loss": 3.1659372329711912, + "step": 8660 + }, + { + "epoch": 0.04335, + "grad_norm": 0.45489445328712463, + "learning_rate": 0.0004986017154723682, + "loss": 3.2025218963623048, + "step": 8670 + }, + { + "epoch": 0.0434, + "grad_norm": 0.4383278489112854, + "learning_rate": 0.0004985975228706254, + "loss": 3.2197311401367186, + "step": 8680 + }, + { + "epoch": 0.04345, + "grad_norm": 0.4586508274078369, + "learning_rate": 0.0004985933240104418, + "loss": 3.1314653396606444, + "step": 8690 + }, + { + "epoch": 0.0435, + "grad_norm": 0.4438301622867584, + "learning_rate": 0.0004985891188919229, + "loss": 3.176669692993164, + "step": 8700 + }, + { + "epoch": 0.04355, + "grad_norm": 0.4582520127296448, + "learning_rate": 0.0004985849075151745, + "loss": 3.1596294403076173, + "step": 8710 + }, + { + "epoch": 0.0436, + "grad_norm": 0.4565725326538086, + "learning_rate": 0.0004985806898803028, + "loss": 3.1581663131713866, + "step": 8720 + }, + { + "epoch": 0.04365, + "grad_norm": 0.45854610204696655, + "learning_rate": 0.0004985764659874139, + "loss": 3.1613229751586913, + "step": 8730 + }, + { + "epoch": 0.0437, + "grad_norm": 0.4610767960548401, + "learning_rate": 0.000498572235836614, + "loss": 3.1517505645751953, + "step": 8740 + }, + { + "epoch": 0.04375, + "grad_norm": 0.44236859679222107, + "learning_rate": 0.0004985679994280099, + "loss": 3.1448314666748045, + "step": 8750 + }, + { + "epoch": 0.0438, + "grad_norm": 0.4517967700958252, + "learning_rate": 0.0004985637567617079, + "loss": 3.210121917724609, + "step": 8760 + }, + { + "epoch": 0.04385, + "grad_norm": 0.4288327395915985, + "learning_rate": 0.0004985595078378152, + "loss": 3.2162425994873045, + "step": 8770 + }, + { + "epoch": 0.0439, + "grad_norm": 0.4756951630115509, + "learning_rate": 0.0004985552526564384, + "loss": 3.143497085571289, + "step": 8780 + }, + { + "epoch": 0.04395, + "grad_norm": 0.5115601420402527, + "learning_rate": 0.0004985509912176848, + "loss": 3.208416748046875, + "step": 8790 + }, + { + "epoch": 0.044, + "grad_norm": 0.47615137696266174, + "learning_rate": 0.0004985467235216616, + "loss": 3.1686588287353517, + "step": 8800 + }, + { + "epoch": 0.04405, + "grad_norm": 0.45866504311561584, + "learning_rate": 0.0004985424495684765, + "loss": 3.142284393310547, + "step": 8810 + }, + { + "epoch": 0.0441, + "grad_norm": 0.4642805755138397, + "learning_rate": 0.0004985381693582367, + "loss": 3.1494525909423827, + "step": 8820 + }, + { + "epoch": 0.04415, + "grad_norm": 0.4460066258907318, + "learning_rate": 0.0004985338828910502, + "loss": 3.1885353088378907, + "step": 8830 + }, + { + "epoch": 0.0442, + "grad_norm": 0.4527949392795563, + "learning_rate": 0.0004985295901670248, + "loss": 3.1968292236328124, + "step": 8840 + }, + { + "epoch": 0.04425, + "grad_norm": 0.48217228055000305, + "learning_rate": 0.0004985252911862688, + "loss": 3.1719280242919923, + "step": 8850 + }, + { + "epoch": 0.0443, + "grad_norm": 0.44714465737342834, + "learning_rate": 0.00049852098594889, + "loss": 3.1482439041137695, + "step": 8860 + }, + { + "epoch": 0.04435, + "grad_norm": 0.4518544375896454, + "learning_rate": 0.0004985166744549972, + "loss": 3.1676105499267577, + "step": 8870 + }, + { + "epoch": 0.0444, + "grad_norm": 0.4798972010612488, + "learning_rate": 0.0004985123567046989, + "loss": 3.1607086181640627, + "step": 8880 + }, + { + "epoch": 0.04445, + "grad_norm": 0.4583210349082947, + "learning_rate": 0.0004985080326981035, + "loss": 3.217298126220703, + "step": 8890 + }, + { + "epoch": 0.0445, + "grad_norm": 0.4706636369228363, + "learning_rate": 0.0004985037024353201, + "loss": 3.1845428466796877, + "step": 8900 + }, + { + "epoch": 0.04455, + "grad_norm": 0.4885094463825226, + "learning_rate": 0.0004984993659164576, + "loss": 3.2011890411376953, + "step": 8910 + }, + { + "epoch": 0.0446, + "grad_norm": 0.479494571685791, + "learning_rate": 0.0004984950231416252, + "loss": 3.163376045227051, + "step": 8920 + }, + { + "epoch": 0.04465, + "grad_norm": 0.4421994388103485, + "learning_rate": 0.0004984906741109323, + "loss": 3.180369567871094, + "step": 8930 + }, + { + "epoch": 0.0447, + "grad_norm": 0.45548033714294434, + "learning_rate": 0.0004984863188244883, + "loss": 3.160597229003906, + "step": 8940 + }, + { + "epoch": 0.04475, + "grad_norm": 0.4687221944332123, + "learning_rate": 0.000498481957282403, + "loss": 3.1894569396972656, + "step": 8950 + }, + { + "epoch": 0.0448, + "grad_norm": 0.46652963757514954, + "learning_rate": 0.0004984775894847861, + "loss": 3.147208023071289, + "step": 8960 + }, + { + "epoch": 0.04485, + "grad_norm": 0.4670984447002411, + "learning_rate": 0.0004984732154317473, + "loss": 3.1225109100341797, + "step": 8970 + }, + { + "epoch": 0.0449, + "grad_norm": 0.4656585454940796, + "learning_rate": 0.0004984688351233973, + "loss": 3.136271095275879, + "step": 8980 + }, + { + "epoch": 0.04495, + "grad_norm": 0.5161711573600769, + "learning_rate": 0.0004984644485598458, + "loss": 3.1514142990112304, + "step": 8990 + }, + { + "epoch": 0.045, + "grad_norm": 0.4666406512260437, + "learning_rate": 0.0004984600557412035, + "loss": 3.130951499938965, + "step": 9000 + }, + { + "epoch": 0.04505, + "grad_norm": 0.48298704624176025, + "learning_rate": 0.000498455656667581, + "loss": 3.1868284225463865, + "step": 9010 + }, + { + "epoch": 0.0451, + "grad_norm": 0.46987617015838623, + "learning_rate": 0.000498451251339089, + "loss": 3.162750244140625, + "step": 9020 + }, + { + "epoch": 0.04515, + "grad_norm": 0.49695655703544617, + "learning_rate": 0.0004984468397558384, + "loss": 3.1335426330566407, + "step": 9030 + }, + { + "epoch": 0.0452, + "grad_norm": 0.47324106097221375, + "learning_rate": 0.0004984424219179402, + "loss": 3.1983768463134767, + "step": 9040 + }, + { + "epoch": 0.04525, + "grad_norm": 0.4681551158428192, + "learning_rate": 0.0004984379978255058, + "loss": 3.169388198852539, + "step": 9050 + }, + { + "epoch": 0.0453, + "grad_norm": 0.465307354927063, + "learning_rate": 0.0004984335674786464, + "loss": 3.156377601623535, + "step": 9060 + }, + { + "epoch": 0.04535, + "grad_norm": 0.46597254276275635, + "learning_rate": 0.0004984291308774737, + "loss": 3.1542333602905273, + "step": 9070 + }, + { + "epoch": 0.0454, + "grad_norm": 0.49214160442352295, + "learning_rate": 0.0004984246880220992, + "loss": 3.146438217163086, + "step": 9080 + }, + { + "epoch": 0.04545, + "grad_norm": 0.4822292625904083, + "learning_rate": 0.0004984202389126348, + "loss": 3.1200899124145507, + "step": 9090 + }, + { + "epoch": 0.0455, + "grad_norm": 0.4341776669025421, + "learning_rate": 0.0004984157835491926, + "loss": 3.1669944763183593, + "step": 9100 + }, + { + "epoch": 0.04555, + "grad_norm": 0.4656651020050049, + "learning_rate": 0.0004984113219318846, + "loss": 3.124550628662109, + "step": 9110 + }, + { + "epoch": 0.0456, + "grad_norm": 0.49773067235946655, + "learning_rate": 0.0004984068540608234, + "loss": 3.1915782928466796, + "step": 9120 + }, + { + "epoch": 0.04565, + "grad_norm": 0.43588635325431824, + "learning_rate": 0.0004984023799361213, + "loss": 3.2002723693847654, + "step": 9130 + }, + { + "epoch": 0.0457, + "grad_norm": 0.47583580017089844, + "learning_rate": 0.0004983978995578909, + "loss": 3.1738697052001954, + "step": 9140 + }, + { + "epoch": 0.04575, + "grad_norm": 0.43456465005874634, + "learning_rate": 0.0004983934129262449, + "loss": 3.175614929199219, + "step": 9150 + }, + { + "epoch": 0.0458, + "grad_norm": 0.4774296283721924, + "learning_rate": 0.0004983889200412966, + "loss": 3.1233564376831056, + "step": 9160 + }, + { + "epoch": 0.04585, + "grad_norm": 0.4676629304885864, + "learning_rate": 0.0004983844209031588, + "loss": 3.157203483581543, + "step": 9170 + }, + { + "epoch": 0.0459, + "grad_norm": 0.44385552406311035, + "learning_rate": 0.0004983799155119449, + "loss": 3.1758047103881837, + "step": 9180 + }, + { + "epoch": 0.04595, + "grad_norm": 0.4954364001750946, + "learning_rate": 0.0004983754038677682, + "loss": 3.146318054199219, + "step": 9190 + }, + { + "epoch": 0.046, + "grad_norm": 0.43064701557159424, + "learning_rate": 0.0004983708859707424, + "loss": 3.172922897338867, + "step": 9200 + }, + { + "epoch": 0.04605, + "grad_norm": 0.473246306180954, + "learning_rate": 0.0004983663618209812, + "loss": 3.183860015869141, + "step": 9210 + }, + { + "epoch": 0.0461, + "grad_norm": 0.44938579201698303, + "learning_rate": 0.0004983618314185987, + "loss": 3.2038631439208984, + "step": 9220 + }, + { + "epoch": 0.04615, + "grad_norm": 0.4906012713909149, + "learning_rate": 0.0004983572947637085, + "loss": 3.1435977935791017, + "step": 9230 + }, + { + "epoch": 0.0462, + "grad_norm": 0.44056573510169983, + "learning_rate": 0.0004983527518564252, + "loss": 3.17974967956543, + "step": 9240 + }, + { + "epoch": 0.04625, + "grad_norm": 0.4435023367404938, + "learning_rate": 0.000498348202696863, + "loss": 3.159469795227051, + "step": 9250 + }, + { + "epoch": 0.0463, + "grad_norm": 0.44911468029022217, + "learning_rate": 0.0004983436472851363, + "loss": 3.1221521377563475, + "step": 9260 + }, + { + "epoch": 0.04635, + "grad_norm": 0.4753960072994232, + "learning_rate": 0.0004983390856213602, + "loss": 3.1625490188598633, + "step": 9270 + }, + { + "epoch": 0.0464, + "grad_norm": 0.43610942363739014, + "learning_rate": 0.0004983345177056491, + "loss": 3.1648799896240236, + "step": 9280 + }, + { + "epoch": 0.04645, + "grad_norm": 0.5077446103096008, + "learning_rate": 0.0004983299435381182, + "loss": 3.084033393859863, + "step": 9290 + }, + { + "epoch": 0.0465, + "grad_norm": 0.47832268476486206, + "learning_rate": 0.0004983253631188828, + "loss": 3.170730781555176, + "step": 9300 + }, + { + "epoch": 0.04655, + "grad_norm": 0.4493735432624817, + "learning_rate": 0.0004983207764480579, + "loss": 3.1903718948364257, + "step": 9310 + }, + { + "epoch": 0.0466, + "grad_norm": 0.5186095237731934, + "learning_rate": 0.000498316183525759, + "loss": 3.140734100341797, + "step": 9320 + }, + { + "epoch": 0.04665, + "grad_norm": 0.4616861641407013, + "learning_rate": 0.000498311584352102, + "loss": 3.199073600769043, + "step": 9330 + }, + { + "epoch": 0.0467, + "grad_norm": 0.4488754868507385, + "learning_rate": 0.0004983069789272025, + "loss": 3.2065086364746094, + "step": 9340 + }, + { + "epoch": 0.04675, + "grad_norm": 0.46984371542930603, + "learning_rate": 0.0004983023672511765, + "loss": 3.154161071777344, + "step": 9350 + }, + { + "epoch": 0.0468, + "grad_norm": 0.48453396558761597, + "learning_rate": 0.0004982977493241399, + "loss": 3.1652700424194338, + "step": 9360 + }, + { + "epoch": 0.04685, + "grad_norm": 0.4598109722137451, + "learning_rate": 0.0004982931251462093, + "loss": 3.1189956665039062, + "step": 9370 + }, + { + "epoch": 0.0469, + "grad_norm": 0.42392611503601074, + "learning_rate": 0.0004982884947175008, + "loss": 3.1411985397338866, + "step": 9380 + }, + { + "epoch": 0.04695, + "grad_norm": 0.4499303996562958, + "learning_rate": 0.0004982838580381313, + "loss": 3.143824577331543, + "step": 9390 + }, + { + "epoch": 0.047, + "grad_norm": 0.46810421347618103, + "learning_rate": 0.0004982792151082171, + "loss": 3.1978601455688476, + "step": 9400 + }, + { + "epoch": 0.04705, + "grad_norm": 0.4723028242588043, + "learning_rate": 0.0004982745659278755, + "loss": 3.187900733947754, + "step": 9410 + }, + { + "epoch": 0.0471, + "grad_norm": 0.45987626910209656, + "learning_rate": 0.0004982699104972233, + "loss": 3.1270910263061524, + "step": 9420 + }, + { + "epoch": 0.04715, + "grad_norm": 0.4601254165172577, + "learning_rate": 0.0004982652488163777, + "loss": 3.221453857421875, + "step": 9430 + }, + { + "epoch": 0.0472, + "grad_norm": 0.4647272527217865, + "learning_rate": 0.0004982605808854561, + "loss": 3.136343765258789, + "step": 9440 + }, + { + "epoch": 0.04725, + "grad_norm": 0.47629663348197937, + "learning_rate": 0.0004982559067045761, + "loss": 3.1871618270874023, + "step": 9450 + }, + { + "epoch": 0.0473, + "grad_norm": 0.4741997718811035, + "learning_rate": 0.0004982512262738553, + "loss": 3.169599151611328, + "step": 9460 + }, + { + "epoch": 0.04735, + "grad_norm": 0.446980744600296, + "learning_rate": 0.0004982465395934116, + "loss": 3.1348052978515626, + "step": 9470 + }, + { + "epoch": 0.0474, + "grad_norm": 0.46821969747543335, + "learning_rate": 0.0004982418466633629, + "loss": 3.0810340881347655, + "step": 9480 + }, + { + "epoch": 0.04745, + "grad_norm": 0.5000423192977905, + "learning_rate": 0.0004982371474838273, + "loss": 3.152790641784668, + "step": 9490 + }, + { + "epoch": 0.0475, + "grad_norm": 0.4642346203327179, + "learning_rate": 0.0004982324420549232, + "loss": 3.1314273834228517, + "step": 9500 + }, + { + "epoch": 0.04755, + "grad_norm": 0.44183972477912903, + "learning_rate": 0.000498227730376769, + "loss": 3.129351806640625, + "step": 9510 + }, + { + "epoch": 0.0476, + "grad_norm": 0.4796864986419678, + "learning_rate": 0.0004982230124494834, + "loss": 3.1135072708129883, + "step": 9520 + }, + { + "epoch": 0.04765, + "grad_norm": 0.5090488195419312, + "learning_rate": 0.0004982182882731852, + "loss": 3.1708940505981444, + "step": 9530 + }, + { + "epoch": 0.0477, + "grad_norm": 0.48335006833076477, + "learning_rate": 0.000498213557847993, + "loss": 3.1296825408935547, + "step": 9540 + }, + { + "epoch": 0.04775, + "grad_norm": 0.44532644748687744, + "learning_rate": 0.0004982088211740263, + "loss": 3.1444740295410156, + "step": 9550 + }, + { + "epoch": 0.0478, + "grad_norm": 0.4348602890968323, + "learning_rate": 0.0004982040782514042, + "loss": 3.146635818481445, + "step": 9560 + }, + { + "epoch": 0.04785, + "grad_norm": 0.4683625102043152, + "learning_rate": 0.0004981993290802461, + "loss": 3.19100399017334, + "step": 9570 + }, + { + "epoch": 0.0479, + "grad_norm": 0.48038575053215027, + "learning_rate": 0.0004981945736606716, + "loss": 3.1122241973876954, + "step": 9580 + }, + { + "epoch": 0.04795, + "grad_norm": 0.4651840031147003, + "learning_rate": 0.0004981898119928003, + "loss": 3.1572896957397463, + "step": 9590 + }, + { + "epoch": 0.048, + "grad_norm": 0.49913492798805237, + "learning_rate": 0.0004981850440767521, + "loss": 3.138683319091797, + "step": 9600 + }, + { + "epoch": 0.04805, + "grad_norm": 0.43197235465049744, + "learning_rate": 0.0004981802699126471, + "loss": 3.1174150466918946, + "step": 9610 + }, + { + "epoch": 0.0481, + "grad_norm": 0.4621870815753937, + "learning_rate": 0.0004981754895006054, + "loss": 3.154473876953125, + "step": 9620 + }, + { + "epoch": 0.04815, + "grad_norm": 0.4482491910457611, + "learning_rate": 0.0004981707028407476, + "loss": 3.127729034423828, + "step": 9630 + }, + { + "epoch": 0.0482, + "grad_norm": 0.46307626366615295, + "learning_rate": 0.0004981659099331937, + "loss": 3.1659677505493162, + "step": 9640 + }, + { + "epoch": 0.04825, + "grad_norm": 0.46798089146614075, + "learning_rate": 0.0004981611107780649, + "loss": 3.1130056381225586, + "step": 9650 + }, + { + "epoch": 0.0483, + "grad_norm": 0.462826669216156, + "learning_rate": 0.0004981563053754819, + "loss": 3.1170122146606447, + "step": 9660 + }, + { + "epoch": 0.04835, + "grad_norm": 0.48662203550338745, + "learning_rate": 0.0004981514937255654, + "loss": 3.1794395446777344, + "step": 9670 + }, + { + "epoch": 0.0484, + "grad_norm": 0.4743303656578064, + "learning_rate": 0.0004981466758284366, + "loss": 3.1294204711914064, + "step": 9680 + }, + { + "epoch": 0.04845, + "grad_norm": 0.44033724069595337, + "learning_rate": 0.0004981418516842171, + "loss": 3.1464229583740235, + "step": 9690 + }, + { + "epoch": 0.0485, + "grad_norm": 0.46795007586479187, + "learning_rate": 0.0004981370212930281, + "loss": 3.1095577239990235, + "step": 9700 + }, + { + "epoch": 0.04855, + "grad_norm": 0.44767820835113525, + "learning_rate": 0.0004981321846549913, + "loss": 3.173564910888672, + "step": 9710 + }, + { + "epoch": 0.0486, + "grad_norm": 0.4682691991329193, + "learning_rate": 0.0004981273417702282, + "loss": 3.125366973876953, + "step": 9720 + }, + { + "epoch": 0.04865, + "grad_norm": 0.4793137013912201, + "learning_rate": 0.0004981224926388611, + "loss": 3.1432727813720702, + "step": 9730 + }, + { + "epoch": 0.0487, + "grad_norm": 0.4479905664920807, + "learning_rate": 0.0004981176372610118, + "loss": 3.1420108795166017, + "step": 9740 + }, + { + "epoch": 0.04875, + "grad_norm": 0.4606311023235321, + "learning_rate": 0.0004981127756368027, + "loss": 3.1381200790405273, + "step": 9750 + }, + { + "epoch": 0.0488, + "grad_norm": 0.4481891989707947, + "learning_rate": 0.0004981079077663561, + "loss": 3.0982547760009767, + "step": 9760 + }, + { + "epoch": 0.04885, + "grad_norm": 0.475312203168869, + "learning_rate": 0.0004981030336497946, + "loss": 3.110179328918457, + "step": 9770 + }, + { + "epoch": 0.0489, + "grad_norm": 0.47322967648506165, + "learning_rate": 0.000498098153287241, + "loss": 3.1205921173095703, + "step": 9780 + }, + { + "epoch": 0.04895, + "grad_norm": 0.45136526226997375, + "learning_rate": 0.0004980932666788179, + "loss": 3.1918619155883787, + "step": 9790 + }, + { + "epoch": 0.049, + "grad_norm": 0.4928245544433594, + "learning_rate": 0.0004980883738246485, + "loss": 3.126162528991699, + "step": 9800 + }, + { + "epoch": 0.04905, + "grad_norm": 0.42997628450393677, + "learning_rate": 0.000498083474724856, + "loss": 3.098705863952637, + "step": 9810 + }, + { + "epoch": 0.0491, + "grad_norm": 0.49567344784736633, + "learning_rate": 0.0004980785693795637, + "loss": 3.1715114593505858, + "step": 9820 + }, + { + "epoch": 0.04915, + "grad_norm": 0.4609673023223877, + "learning_rate": 0.0004980736577888949, + "loss": 3.110226058959961, + "step": 9830 + }, + { + "epoch": 0.0492, + "grad_norm": 0.4655815064907074, + "learning_rate": 0.0004980687399529735, + "loss": 3.102264404296875, + "step": 9840 + }, + { + "epoch": 0.04925, + "grad_norm": 0.4333123564720154, + "learning_rate": 0.0004980638158719234, + "loss": 3.1059314727783205, + "step": 9850 + }, + { + "epoch": 0.0493, + "grad_norm": 0.47606518864631653, + "learning_rate": 0.0004980588855458682, + "loss": 3.1884363174438475, + "step": 9860 + }, + { + "epoch": 0.04935, + "grad_norm": 0.45604875683784485, + "learning_rate": 0.0004980539489749323, + "loss": 3.1141437530517577, + "step": 9870 + }, + { + "epoch": 0.0494, + "grad_norm": 0.44792354106903076, + "learning_rate": 0.0004980490061592399, + "loss": 3.1420080184936525, + "step": 9880 + }, + { + "epoch": 0.04945, + "grad_norm": 0.4336458742618561, + "learning_rate": 0.0004980440570989153, + "loss": 3.1274736404418944, + "step": 9890 + }, + { + "epoch": 0.0495, + "grad_norm": 0.4494745135307312, + "learning_rate": 0.0004980391017940834, + "loss": 3.1162479400634764, + "step": 9900 + }, + { + "epoch": 0.04955, + "grad_norm": 0.4632682204246521, + "learning_rate": 0.0004980341402448686, + "loss": 3.122186851501465, + "step": 9910 + }, + { + "epoch": 0.0496, + "grad_norm": 0.4392301142215729, + "learning_rate": 0.0004980291724513961, + "loss": 3.1191726684570313, + "step": 9920 + }, + { + "epoch": 0.04965, + "grad_norm": 0.4833999574184418, + "learning_rate": 0.0004980241984137907, + "loss": 3.152467155456543, + "step": 9930 + }, + { + "epoch": 0.0497, + "grad_norm": 0.4630514979362488, + "learning_rate": 0.000498019218132178, + "loss": 3.1465633392333983, + "step": 9940 + }, + { + "epoch": 0.04975, + "grad_norm": 0.47820279002189636, + "learning_rate": 0.000498014231606683, + "loss": 3.1434974670410156, + "step": 9950 + }, + { + "epoch": 0.0498, + "grad_norm": 0.43320852518081665, + "learning_rate": 0.0004980092388374313, + "loss": 3.126866912841797, + "step": 9960 + }, + { + "epoch": 0.04985, + "grad_norm": 0.43913236260414124, + "learning_rate": 0.0004980042398245488, + "loss": 3.096966361999512, + "step": 9970 + }, + { + "epoch": 0.0499, + "grad_norm": 0.45619794726371765, + "learning_rate": 0.0004979992345681612, + "loss": 3.157577896118164, + "step": 9980 + }, + { + "epoch": 0.04995, + "grad_norm": 0.44136855006217957, + "learning_rate": 0.0004979942230683945, + "loss": 3.1897405624389648, + "step": 9990 + }, + { + "epoch": 0.05, + "grad_norm": 0.47601768374443054, + "learning_rate": 0.0004979892053253749, + "loss": 3.1573501586914063, + "step": 10000 + }, + { + "epoch": 0.05005, + "grad_norm": 0.46735256910324097, + "learning_rate": 0.0004979841813392286, + "loss": 3.2196956634521485, + "step": 10010 + }, + { + "epoch": 0.0501, + "grad_norm": 0.4888245463371277, + "learning_rate": 0.0004979791511100823, + "loss": 3.1438331604003906, + "step": 10020 + }, + { + "epoch": 0.05015, + "grad_norm": 0.4527634382247925, + "learning_rate": 0.0004979741146380625, + "loss": 3.1816883087158203, + "step": 10030 + }, + { + "epoch": 0.0502, + "grad_norm": 0.4709538519382477, + "learning_rate": 0.0004979690719232961, + "loss": 3.121601676940918, + "step": 10040 + }, + { + "epoch": 0.05025, + "grad_norm": 0.4551320970058441, + "learning_rate": 0.0004979640229659099, + "loss": 3.175899314880371, + "step": 10050 + }, + { + "epoch": 0.0503, + "grad_norm": 0.47166213393211365, + "learning_rate": 0.0004979589677660311, + "loss": 3.1516786575317384, + "step": 10060 + }, + { + "epoch": 0.05035, + "grad_norm": 0.45336854457855225, + "learning_rate": 0.0004979539063237869, + "loss": 3.1167551040649415, + "step": 10070 + }, + { + "epoch": 0.0504, + "grad_norm": 0.4455978274345398, + "learning_rate": 0.0004979488386393048, + "loss": 3.153247261047363, + "step": 10080 + }, + { + "epoch": 0.05045, + "grad_norm": 0.43548882007598877, + "learning_rate": 0.0004979437647127125, + "loss": 3.142081642150879, + "step": 10090 + }, + { + "epoch": 0.0505, + "grad_norm": 0.4301036596298218, + "learning_rate": 0.0004979386845441374, + "loss": 3.1151012420654296, + "step": 10100 + }, + { + "epoch": 0.05055, + "grad_norm": 0.4644053280353546, + "learning_rate": 0.0004979335981337075, + "loss": 3.099431037902832, + "step": 10110 + }, + { + "epoch": 0.0506, + "grad_norm": 0.4449102282524109, + "learning_rate": 0.000497928505481551, + "loss": 3.123041534423828, + "step": 10120 + }, + { + "epoch": 0.05065, + "grad_norm": 0.4521437883377075, + "learning_rate": 0.0004979234065877961, + "loss": 3.162443733215332, + "step": 10130 + }, + { + "epoch": 0.0507, + "grad_norm": 0.420913428068161, + "learning_rate": 0.000497918301452571, + "loss": 3.1176223754882812, + "step": 10140 + }, + { + "epoch": 0.05075, + "grad_norm": 0.4713296592235565, + "learning_rate": 0.0004979131900760044, + "loss": 3.136612129211426, + "step": 10150 + }, + { + "epoch": 0.0508, + "grad_norm": 0.4489203691482544, + "learning_rate": 0.0004979080724582249, + "loss": 3.2235206604003905, + "step": 10160 + }, + { + "epoch": 0.05085, + "grad_norm": 0.4542710781097412, + "learning_rate": 0.0004979029485993612, + "loss": 3.0903194427490233, + "step": 10170 + }, + { + "epoch": 0.0509, + "grad_norm": 0.4545997083187103, + "learning_rate": 0.0004978978184995425, + "loss": 3.076351356506348, + "step": 10180 + }, + { + "epoch": 0.05095, + "grad_norm": 0.44490206241607666, + "learning_rate": 0.0004978926821588979, + "loss": 3.0963388442993165, + "step": 10190 + }, + { + "epoch": 0.051, + "grad_norm": 0.46630752086639404, + "learning_rate": 0.0004978875395775567, + "loss": 3.170342445373535, + "step": 10200 + }, + { + "epoch": 0.05105, + "grad_norm": 0.5092012882232666, + "learning_rate": 0.0004978823907556484, + "loss": 3.118005561828613, + "step": 10210 + }, + { + "epoch": 0.0511, + "grad_norm": 0.48260194063186646, + "learning_rate": 0.0004978772356933024, + "loss": 3.16711311340332, + "step": 10220 + }, + { + "epoch": 0.05115, + "grad_norm": 0.4346385896205902, + "learning_rate": 0.0004978720743906488, + "loss": 3.137546730041504, + "step": 10230 + }, + { + "epoch": 0.0512, + "grad_norm": 0.46634021401405334, + "learning_rate": 0.0004978669068478173, + "loss": 3.1567102432250977, + "step": 10240 + }, + { + "epoch": 0.05125, + "grad_norm": 0.43062540888786316, + "learning_rate": 0.0004978617330649381, + "loss": 3.113858222961426, + "step": 10250 + }, + { + "epoch": 0.0513, + "grad_norm": 0.47711533308029175, + "learning_rate": 0.0004978565530421415, + "loss": 3.0628761291503905, + "step": 10260 + }, + { + "epoch": 0.05135, + "grad_norm": 0.47512751817703247, + "learning_rate": 0.0004978513667795577, + "loss": 3.1160711288452148, + "step": 10270 + }, + { + "epoch": 0.0514, + "grad_norm": 0.49649766087532043, + "learning_rate": 0.0004978461742773174, + "loss": 3.1114694595336916, + "step": 10280 + }, + { + "epoch": 0.05145, + "grad_norm": 0.4633052349090576, + "learning_rate": 0.0004978409755355515, + "loss": 3.15592041015625, + "step": 10290 + }, + { + "epoch": 0.0515, + "grad_norm": 0.4716596007347107, + "learning_rate": 0.0004978357705543906, + "loss": 3.130977249145508, + "step": 10300 + }, + { + "epoch": 0.05155, + "grad_norm": 0.4475041627883911, + "learning_rate": 0.0004978305593339657, + "loss": 3.1257179260253904, + "step": 10310 + }, + { + "epoch": 0.0516, + "grad_norm": 0.4297730326652527, + "learning_rate": 0.0004978253418744083, + "loss": 3.165741539001465, + "step": 10320 + }, + { + "epoch": 0.05165, + "grad_norm": 0.5081633925437927, + "learning_rate": 0.0004978201181758495, + "loss": 3.148785400390625, + "step": 10330 + }, + { + "epoch": 0.0517, + "grad_norm": 0.4439052641391754, + "learning_rate": 0.0004978148882384209, + "loss": 3.1132671356201174, + "step": 10340 + }, + { + "epoch": 0.05175, + "grad_norm": 0.44513624906539917, + "learning_rate": 0.0004978096520622541, + "loss": 3.1179723739624023, + "step": 10350 + }, + { + "epoch": 0.0518, + "grad_norm": 0.46803104877471924, + "learning_rate": 0.000497804409647481, + "loss": 3.1514823913574217, + "step": 10360 + }, + { + "epoch": 0.05185, + "grad_norm": 0.44299083948135376, + "learning_rate": 0.0004977991609942335, + "loss": 3.1690086364746093, + "step": 10370 + }, + { + "epoch": 0.0519, + "grad_norm": 0.481351375579834, + "learning_rate": 0.0004977939061026438, + "loss": 3.1178579330444336, + "step": 10380 + }, + { + "epoch": 0.05195, + "grad_norm": 0.46323496103286743, + "learning_rate": 0.0004977886449728442, + "loss": 3.1988492965698243, + "step": 10390 + }, + { + "epoch": 0.052, + "grad_norm": 0.42991703748703003, + "learning_rate": 0.0004977833776049672, + "loss": 3.1830417633056642, + "step": 10400 + }, + { + "epoch": 0.05205, + "grad_norm": 0.45019909739494324, + "learning_rate": 0.0004977781039991452, + "loss": 3.1153478622436523, + "step": 10410 + }, + { + "epoch": 0.0521, + "grad_norm": 0.45817631483078003, + "learning_rate": 0.0004977728241555111, + "loss": 3.0962575912475585, + "step": 10420 + }, + { + "epoch": 0.05215, + "grad_norm": 0.4269137680530548, + "learning_rate": 0.0004977675380741978, + "loss": 3.069341278076172, + "step": 10430 + }, + { + "epoch": 0.0522, + "grad_norm": 0.4315447509288788, + "learning_rate": 0.0004977622457553385, + "loss": 3.0743148803710936, + "step": 10440 + }, + { + "epoch": 0.05225, + "grad_norm": 0.4580780565738678, + "learning_rate": 0.0004977569471990662, + "loss": 3.132797050476074, + "step": 10450 + }, + { + "epoch": 0.0523, + "grad_norm": 0.43326619267463684, + "learning_rate": 0.0004977516424055144, + "loss": 3.0980857849121093, + "step": 10460 + }, + { + "epoch": 0.05235, + "grad_norm": 0.4674171209335327, + "learning_rate": 0.0004977463313748168, + "loss": 3.1178348541259764, + "step": 10470 + }, + { + "epoch": 0.0524, + "grad_norm": 0.4720475971698761, + "learning_rate": 0.0004977410141071068, + "loss": 3.157978630065918, + "step": 10480 + }, + { + "epoch": 0.05245, + "grad_norm": 0.4488123059272766, + "learning_rate": 0.0004977356906025184, + "loss": 3.1341386795043946, + "step": 10490 + }, + { + "epoch": 0.0525, + "grad_norm": 0.4450739026069641, + "learning_rate": 0.0004977303608611858, + "loss": 3.104084014892578, + "step": 10500 + }, + { + "epoch": 0.05255, + "grad_norm": 0.43123775720596313, + "learning_rate": 0.0004977250248832429, + "loss": 3.137073516845703, + "step": 10510 + }, + { + "epoch": 0.0526, + "grad_norm": 0.4580446779727936, + "learning_rate": 0.0004977196826688242, + "loss": 3.1203731536865233, + "step": 10520 + }, + { + "epoch": 0.05265, + "grad_norm": 0.49189940094947815, + "learning_rate": 0.0004977143342180643, + "loss": 3.1264890670776366, + "step": 10530 + }, + { + "epoch": 0.0527, + "grad_norm": 0.47779935598373413, + "learning_rate": 0.0004977089795310974, + "loss": 3.1408477783203126, + "step": 10540 + }, + { + "epoch": 0.05275, + "grad_norm": 0.4622478485107422, + "learning_rate": 0.0004977036186080588, + "loss": 3.129569244384766, + "step": 10550 + }, + { + "epoch": 0.0528, + "grad_norm": 0.4581221342086792, + "learning_rate": 0.0004976982514490832, + "loss": 3.1510597229003907, + "step": 10560 + }, + { + "epoch": 0.05285, + "grad_norm": 0.4686342179775238, + "learning_rate": 0.0004976928780543058, + "loss": 3.1692018508911133, + "step": 10570 + }, + { + "epoch": 0.0529, + "grad_norm": 0.4100887179374695, + "learning_rate": 0.0004976874984238618, + "loss": 3.078816604614258, + "step": 10580 + }, + { + "epoch": 0.05295, + "grad_norm": 0.44600728154182434, + "learning_rate": 0.0004976821125578867, + "loss": 3.105412483215332, + "step": 10590 + }, + { + "epoch": 0.053, + "grad_norm": 0.44593143463134766, + "learning_rate": 0.000497676720456516, + "loss": 3.1173164367675783, + "step": 10600 + }, + { + "epoch": 0.05305, + "grad_norm": 0.44827836751937866, + "learning_rate": 0.0004976713221198857, + "loss": 3.1582052230834963, + "step": 10610 + }, + { + "epoch": 0.0531, + "grad_norm": 0.4115162491798401, + "learning_rate": 0.0004976659175481313, + "loss": 3.1177928924560545, + "step": 10620 + }, + { + "epoch": 0.05315, + "grad_norm": 0.45012423396110535, + "learning_rate": 0.0004976605067413892, + "loss": 3.048428535461426, + "step": 10630 + }, + { + "epoch": 0.0532, + "grad_norm": 0.47237685322761536, + "learning_rate": 0.0004976550896997955, + "loss": 3.0643192291259767, + "step": 10640 + }, + { + "epoch": 0.05325, + "grad_norm": 0.4639837145805359, + "learning_rate": 0.0004976496664234866, + "loss": 3.1085338592529297, + "step": 10650 + }, + { + "epoch": 0.0533, + "grad_norm": 0.4418773651123047, + "learning_rate": 0.0004976442369125989, + "loss": 3.112244415283203, + "step": 10660 + }, + { + "epoch": 0.05335, + "grad_norm": 0.42964956164360046, + "learning_rate": 0.0004976388011672693, + "loss": 3.0980808258056642, + "step": 10670 + }, + { + "epoch": 0.0534, + "grad_norm": 0.4648042917251587, + "learning_rate": 0.0004976333591876344, + "loss": 3.123680305480957, + "step": 10680 + }, + { + "epoch": 0.05345, + "grad_norm": 0.41698747873306274, + "learning_rate": 0.0004976279109738315, + "loss": 3.1205434799194336, + "step": 10690 + }, + { + "epoch": 0.0535, + "grad_norm": 0.4486411213874817, + "learning_rate": 0.0004976224565259974, + "loss": 3.0940483093261717, + "step": 10700 + }, + { + "epoch": 0.05355, + "grad_norm": 0.46960002183914185, + "learning_rate": 0.0004976169958442699, + "loss": 3.1081212997436523, + "step": 10710 + }, + { + "epoch": 0.0536, + "grad_norm": 0.5034363865852356, + "learning_rate": 0.0004976115289287859, + "loss": 3.157314682006836, + "step": 10720 + }, + { + "epoch": 0.05365, + "grad_norm": 0.4539710581302643, + "learning_rate": 0.0004976060557796834, + "loss": 3.1331735610961915, + "step": 10730 + }, + { + "epoch": 0.0537, + "grad_norm": 0.44732922315597534, + "learning_rate": 0.0004976005763971002, + "loss": 3.0820919036865235, + "step": 10740 + }, + { + "epoch": 0.05375, + "grad_norm": 0.46945932507514954, + "learning_rate": 0.0004975950907811741, + "loss": 3.1444278717041017, + "step": 10750 + }, + { + "epoch": 0.0538, + "grad_norm": 0.44444024562835693, + "learning_rate": 0.0004975895989320432, + "loss": 3.141637420654297, + "step": 10760 + }, + { + "epoch": 0.05385, + "grad_norm": 0.4521530866622925, + "learning_rate": 0.0004975841008498459, + "loss": 3.120322418212891, + "step": 10770 + }, + { + "epoch": 0.0539, + "grad_norm": 0.44179314374923706, + "learning_rate": 0.0004975785965347204, + "loss": 3.1395858764648437, + "step": 10780 + }, + { + "epoch": 0.05395, + "grad_norm": 0.44499510526657104, + "learning_rate": 0.0004975730859868054, + "loss": 3.0945198059082033, + "step": 10790 + }, + { + "epoch": 0.054, + "grad_norm": 0.431213915348053, + "learning_rate": 0.0004975675692062396, + "loss": 3.1323497772216795, + "step": 10800 + }, + { + "epoch": 0.05405, + "grad_norm": 0.46875351667404175, + "learning_rate": 0.0004975620461931619, + "loss": 3.146457481384277, + "step": 10810 + }, + { + "epoch": 0.0541, + "grad_norm": 0.4489651620388031, + "learning_rate": 0.0004975565169477114, + "loss": 3.0978736877441406, + "step": 10820 + }, + { + "epoch": 0.05415, + "grad_norm": 0.43135836720466614, + "learning_rate": 0.0004975509814700272, + "loss": 3.158289337158203, + "step": 10830 + }, + { + "epoch": 0.0542, + "grad_norm": 0.4444389343261719, + "learning_rate": 0.0004975454397602488, + "loss": 3.1519330978393554, + "step": 10840 + }, + { + "epoch": 0.05425, + "grad_norm": 0.4571950137615204, + "learning_rate": 0.0004975398918185155, + "loss": 3.0435562133789062, + "step": 10850 + }, + { + "epoch": 0.0543, + "grad_norm": 0.44448399543762207, + "learning_rate": 0.0004975343376449671, + "loss": 3.1071767807006836, + "step": 10860 + }, + { + "epoch": 0.05435, + "grad_norm": 0.4282829165458679, + "learning_rate": 0.0004975287772397434, + "loss": 3.1336496353149412, + "step": 10870 + }, + { + "epoch": 0.0544, + "grad_norm": 0.42780929803848267, + "learning_rate": 0.0004975232106029843, + "loss": 3.090181350708008, + "step": 10880 + }, + { + "epoch": 0.05445, + "grad_norm": 0.4668521583080292, + "learning_rate": 0.0004975176377348302, + "loss": 3.1423845291137695, + "step": 10890 + }, + { + "epoch": 0.0545, + "grad_norm": 0.47197920083999634, + "learning_rate": 0.0004975120586354211, + "loss": 3.15667667388916, + "step": 10900 + }, + { + "epoch": 0.05455, + "grad_norm": 0.4675155580043793, + "learning_rate": 0.0004975064733048977, + "loss": 3.1409011840820313, + "step": 10910 + }, + { + "epoch": 0.0546, + "grad_norm": 0.4414594769477844, + "learning_rate": 0.0004975008817434004, + "loss": 3.1472692489624023, + "step": 10920 + }, + { + "epoch": 0.05465, + "grad_norm": 0.468273788690567, + "learning_rate": 0.00049749528395107, + "loss": 3.1181308746337892, + "step": 10930 + }, + { + "epoch": 0.0547, + "grad_norm": 0.448397159576416, + "learning_rate": 0.0004974896799280475, + "loss": 3.1273468017578123, + "step": 10940 + }, + { + "epoch": 0.05475, + "grad_norm": 0.47403842210769653, + "learning_rate": 0.0004974840696744741, + "loss": 3.145471954345703, + "step": 10950 + }, + { + "epoch": 0.0548, + "grad_norm": 0.4314950108528137, + "learning_rate": 0.0004974784531904908, + "loss": 3.0826528549194334, + "step": 10960 + }, + { + "epoch": 0.05485, + "grad_norm": 0.4556245505809784, + "learning_rate": 0.000497472830476239, + "loss": 3.1109920501708985, + "step": 10970 + }, + { + "epoch": 0.0549, + "grad_norm": 0.4267360270023346, + "learning_rate": 0.0004974672015318605, + "loss": 3.1484264373779296, + "step": 10980 + }, + { + "epoch": 0.05495, + "grad_norm": 0.45805761218070984, + "learning_rate": 0.0004974615663574967, + "loss": 3.117898178100586, + "step": 10990 + }, + { + "epoch": 0.055, + "grad_norm": 0.4472389817237854, + "learning_rate": 0.0004974559249532898, + "loss": 3.1780839920043946, + "step": 11000 + }, + { + "epoch": 0.05505, + "grad_norm": 0.4771518409252167, + "learning_rate": 0.0004974502773193815, + "loss": 3.127610206604004, + "step": 11010 + }, + { + "epoch": 0.0551, + "grad_norm": 0.4569537937641144, + "learning_rate": 0.0004974446234559143, + "loss": 3.129824638366699, + "step": 11020 + }, + { + "epoch": 0.05515, + "grad_norm": 0.43621569871902466, + "learning_rate": 0.0004974389633630302, + "loss": 3.0492055892944334, + "step": 11030 + }, + { + "epoch": 0.0552, + "grad_norm": 0.4442897140979767, + "learning_rate": 0.000497433297040872, + "loss": 3.143612861633301, + "step": 11040 + }, + { + "epoch": 0.05525, + "grad_norm": 0.4332914650440216, + "learning_rate": 0.0004974276244895821, + "loss": 3.104469871520996, + "step": 11050 + }, + { + "epoch": 0.0553, + "grad_norm": 0.42286577820777893, + "learning_rate": 0.0004974219457093032, + "loss": 3.079830551147461, + "step": 11060 + }, + { + "epoch": 0.05535, + "grad_norm": 0.4757891297340393, + "learning_rate": 0.0004974162607001788, + "loss": 3.0957117080688477, + "step": 11070 + }, + { + "epoch": 0.0554, + "grad_norm": 0.46438732743263245, + "learning_rate": 0.0004974105694623515, + "loss": 3.0790632247924803, + "step": 11080 + }, + { + "epoch": 0.05545, + "grad_norm": 0.4160889983177185, + "learning_rate": 0.0004974048719959648, + "loss": 3.120720863342285, + "step": 11090 + }, + { + "epoch": 0.0555, + "grad_norm": 0.4786531329154968, + "learning_rate": 0.0004973991683011622, + "loss": 3.088596725463867, + "step": 11100 + }, + { + "epoch": 0.05555, + "grad_norm": 0.4379175305366516, + "learning_rate": 0.0004973934583780871, + "loss": 3.1352495193481444, + "step": 11110 + }, + { + "epoch": 0.0556, + "grad_norm": 0.44919100403785706, + "learning_rate": 0.0004973877422268833, + "loss": 3.07130126953125, + "step": 11120 + }, + { + "epoch": 0.05565, + "grad_norm": 0.424814909696579, + "learning_rate": 0.0004973820198476949, + "loss": 3.089427947998047, + "step": 11130 + }, + { + "epoch": 0.0557, + "grad_norm": 0.45324501395225525, + "learning_rate": 0.0004973762912406656, + "loss": 3.1176782608032227, + "step": 11140 + }, + { + "epoch": 0.05575, + "grad_norm": 0.4153450131416321, + "learning_rate": 0.00049737055640594, + "loss": 3.1521942138671877, + "step": 11150 + }, + { + "epoch": 0.0558, + "grad_norm": 0.41539934277534485, + "learning_rate": 0.0004973648153436622, + "loss": 3.1490337371826174, + "step": 11160 + }, + { + "epoch": 0.05585, + "grad_norm": 0.4186537563800812, + "learning_rate": 0.0004973590680539768, + "loss": 3.072358512878418, + "step": 11170 + }, + { + "epoch": 0.0559, + "grad_norm": 0.4539622962474823, + "learning_rate": 0.0004973533145370286, + "loss": 3.116567611694336, + "step": 11180 + }, + { + "epoch": 0.05595, + "grad_norm": 0.44712284207344055, + "learning_rate": 0.0004973475547929623, + "loss": 3.123878288269043, + "step": 11190 + }, + { + "epoch": 0.056, + "grad_norm": 0.4301578402519226, + "learning_rate": 0.000497341788821923, + "loss": 3.0433387756347656, + "step": 11200 + }, + { + "epoch": 0.05605, + "grad_norm": 0.46530354022979736, + "learning_rate": 0.0004973360166240558, + "loss": 3.1205005645751953, + "step": 11210 + }, + { + "epoch": 0.0561, + "grad_norm": 0.4243711531162262, + "learning_rate": 0.0004973302381995061, + "loss": 3.1203598022460937, + "step": 11220 + }, + { + "epoch": 0.05615, + "grad_norm": 0.4439530074596405, + "learning_rate": 0.0004973244535484191, + "loss": 3.063346290588379, + "step": 11230 + }, + { + "epoch": 0.0562, + "grad_norm": 0.4417268931865692, + "learning_rate": 0.0004973186626709409, + "loss": 3.094755744934082, + "step": 11240 + }, + { + "epoch": 0.05625, + "grad_norm": 0.4466833770275116, + "learning_rate": 0.0004973128655672168, + "loss": 3.127775192260742, + "step": 11250 + }, + { + "epoch": 0.0563, + "grad_norm": 0.41342031955718994, + "learning_rate": 0.0004973070622373931, + "loss": 3.1127651214599608, + "step": 11260 + }, + { + "epoch": 0.05635, + "grad_norm": 0.42843154072761536, + "learning_rate": 0.0004973012526816158, + "loss": 3.0581117630004884, + "step": 11270 + }, + { + "epoch": 0.0564, + "grad_norm": 0.45010122656822205, + "learning_rate": 0.0004972954369000309, + "loss": 3.100748825073242, + "step": 11280 + }, + { + "epoch": 0.05645, + "grad_norm": 0.4442317485809326, + "learning_rate": 0.0004972896148927853, + "loss": 3.084503746032715, + "step": 11290 + }, + { + "epoch": 0.0565, + "grad_norm": 0.457319051027298, + "learning_rate": 0.000497283786660025, + "loss": 3.1405994415283205, + "step": 11300 + }, + { + "epoch": 0.05655, + "grad_norm": 0.46403932571411133, + "learning_rate": 0.0004972779522018973, + "loss": 3.1294864654541015, + "step": 11310 + }, + { + "epoch": 0.0566, + "grad_norm": 0.4435673952102661, + "learning_rate": 0.0004972721115185487, + "loss": 3.1257503509521483, + "step": 11320 + }, + { + "epoch": 0.05665, + "grad_norm": 0.46190616488456726, + "learning_rate": 0.0004972662646101262, + "loss": 3.1140153884887694, + "step": 11330 + }, + { + "epoch": 0.0567, + "grad_norm": 0.44628575444221497, + "learning_rate": 0.0004972604114767774, + "loss": 3.12661075592041, + "step": 11340 + }, + { + "epoch": 0.05675, + "grad_norm": 0.4314859211444855, + "learning_rate": 0.0004972545521186492, + "loss": 3.047186088562012, + "step": 11350 + }, + { + "epoch": 0.0568, + "grad_norm": 0.4531655013561249, + "learning_rate": 0.0004972486865358894, + "loss": 3.0994169235229494, + "step": 11360 + }, + { + "epoch": 0.05685, + "grad_norm": 0.4922948181629181, + "learning_rate": 0.0004972428147286455, + "loss": 3.107127380371094, + "step": 11370 + }, + { + "epoch": 0.0569, + "grad_norm": 0.4709605574607849, + "learning_rate": 0.0004972369366970655, + "loss": 3.10260124206543, + "step": 11380 + }, + { + "epoch": 0.05695, + "grad_norm": 0.45326289534568787, + "learning_rate": 0.0004972310524412972, + "loss": 3.0088768005371094, + "step": 11390 + }, + { + "epoch": 0.057, + "grad_norm": 0.4321173131465912, + "learning_rate": 0.0004972251619614887, + "loss": 3.055381011962891, + "step": 11400 + }, + { + "epoch": 0.05705, + "grad_norm": 0.4606868028640747, + "learning_rate": 0.0004972192652577886, + "loss": 3.0935060501098635, + "step": 11410 + }, + { + "epoch": 0.0571, + "grad_norm": 0.45920100808143616, + "learning_rate": 0.000497213362330345, + "loss": 3.1192787170410154, + "step": 11420 + }, + { + "epoch": 0.05715, + "grad_norm": 0.44384467601776123, + "learning_rate": 0.0004972074531793067, + "loss": 3.1268178939819338, + "step": 11430 + }, + { + "epoch": 0.0572, + "grad_norm": 0.4104965925216675, + "learning_rate": 0.0004972015378048224, + "loss": 3.0661325454711914, + "step": 11440 + }, + { + "epoch": 0.05725, + "grad_norm": 0.49174484610557556, + "learning_rate": 0.0004971956162070411, + "loss": 3.1278430938720705, + "step": 11450 + }, + { + "epoch": 0.0573, + "grad_norm": 0.48326465487480164, + "learning_rate": 0.0004971896883861118, + "loss": 3.128858184814453, + "step": 11460 + }, + { + "epoch": 0.05735, + "grad_norm": 0.4307974576950073, + "learning_rate": 0.0004971837543421837, + "loss": 3.0982843399047852, + "step": 11470 + }, + { + "epoch": 0.0574, + "grad_norm": 0.4284289479255676, + "learning_rate": 0.0004971778140754063, + "loss": 3.1836509704589844, + "step": 11480 + }, + { + "epoch": 0.05745, + "grad_norm": 0.41177600622177124, + "learning_rate": 0.000497171867585929, + "loss": 3.0578861236572266, + "step": 11490 + }, + { + "epoch": 0.0575, + "grad_norm": 0.44963982701301575, + "learning_rate": 0.0004971659148739017, + "loss": 3.0789562225341798, + "step": 11500 + }, + { + "epoch": 0.05755, + "grad_norm": 0.4543377459049225, + "learning_rate": 0.000497159955939474, + "loss": 3.1187414169311523, + "step": 11510 + }, + { + "epoch": 0.0576, + "grad_norm": 0.4495135247707367, + "learning_rate": 0.0004971539907827963, + "loss": 3.136130142211914, + "step": 11520 + }, + { + "epoch": 0.05765, + "grad_norm": 0.46229344606399536, + "learning_rate": 0.0004971480194040182, + "loss": 3.054703140258789, + "step": 11530 + }, + { + "epoch": 0.0577, + "grad_norm": 0.45580559968948364, + "learning_rate": 0.0004971420418032905, + "loss": 3.140325927734375, + "step": 11540 + }, + { + "epoch": 0.05775, + "grad_norm": 0.4524634778499603, + "learning_rate": 0.0004971360579807637, + "loss": 3.0740943908691407, + "step": 11550 + }, + { + "epoch": 0.0578, + "grad_norm": 0.43880972266197205, + "learning_rate": 0.0004971300679365881, + "loss": 3.1245981216430665, + "step": 11560 + }, + { + "epoch": 0.05785, + "grad_norm": 0.4693332016468048, + "learning_rate": 0.0004971240716709148, + "loss": 3.0326465606689452, + "step": 11570 + }, + { + "epoch": 0.0579, + "grad_norm": 0.4518111050128937, + "learning_rate": 0.0004971180691838946, + "loss": 3.1315053939819335, + "step": 11580 + }, + { + "epoch": 0.05795, + "grad_norm": 0.4102519452571869, + "learning_rate": 0.0004971120604756787, + "loss": 3.094550132751465, + "step": 11590 + }, + { + "epoch": 0.058, + "grad_norm": 0.44550803303718567, + "learning_rate": 0.0004971060455464183, + "loss": 3.1077194213867188, + "step": 11600 + }, + { + "epoch": 0.05805, + "grad_norm": 0.44218048453330994, + "learning_rate": 0.0004971000243962648, + "loss": 3.1109832763671874, + "step": 11610 + }, + { + "epoch": 0.0581, + "grad_norm": 0.4592970311641693, + "learning_rate": 0.0004970939970253699, + "loss": 3.1355661392211913, + "step": 11620 + }, + { + "epoch": 0.05815, + "grad_norm": 0.422493577003479, + "learning_rate": 0.0004970879634338853, + "loss": 3.044805145263672, + "step": 11630 + }, + { + "epoch": 0.0582, + "grad_norm": 0.5107035040855408, + "learning_rate": 0.0004970819236219628, + "loss": 3.0804296493530274, + "step": 11640 + }, + { + "epoch": 0.05825, + "grad_norm": 0.41810107231140137, + "learning_rate": 0.0004970758775897546, + "loss": 3.043254852294922, + "step": 11650 + }, + { + "epoch": 0.0583, + "grad_norm": 0.42856961488723755, + "learning_rate": 0.0004970698253374127, + "loss": 3.0848926544189452, + "step": 11660 + }, + { + "epoch": 0.05835, + "grad_norm": 0.45528045296669006, + "learning_rate": 0.0004970637668650898, + "loss": 3.0564172744750975, + "step": 11670 + }, + { + "epoch": 0.0584, + "grad_norm": 0.4087652564048767, + "learning_rate": 0.0004970577021729381, + "loss": 3.1667015075683596, + "step": 11680 + }, + { + "epoch": 0.05845, + "grad_norm": 0.4470013976097107, + "learning_rate": 0.0004970516312611104, + "loss": 3.1094797134399412, + "step": 11690 + }, + { + "epoch": 0.0585, + "grad_norm": 0.43683794140815735, + "learning_rate": 0.0004970455541297595, + "loss": 3.0665843963623045, + "step": 11700 + }, + { + "epoch": 0.05855, + "grad_norm": 0.4262227416038513, + "learning_rate": 0.0004970394707790386, + "loss": 3.1093172073364257, + "step": 11710 + }, + { + "epoch": 0.0586, + "grad_norm": 0.3947440981864929, + "learning_rate": 0.0004970333812091004, + "loss": 3.08215446472168, + "step": 11720 + }, + { + "epoch": 0.05865, + "grad_norm": 0.5255452394485474, + "learning_rate": 0.0004970272854200987, + "loss": 3.053925704956055, + "step": 11730 + }, + { + "epoch": 0.0587, + "grad_norm": 0.4779245853424072, + "learning_rate": 0.0004970211834121866, + "loss": 3.1012372970581055, + "step": 11740 + }, + { + "epoch": 0.05875, + "grad_norm": 0.43874451518058777, + "learning_rate": 0.0004970150751855179, + "loss": 3.059400749206543, + "step": 11750 + }, + { + "epoch": 0.0588, + "grad_norm": 0.4329393208026886, + "learning_rate": 0.0004970089607402464, + "loss": 3.044160842895508, + "step": 11760 + }, + { + "epoch": 0.05885, + "grad_norm": 0.46670472621917725, + "learning_rate": 0.000497002840076526, + "loss": 3.0981826782226562, + "step": 11770 + }, + { + "epoch": 0.0589, + "grad_norm": 0.4327705204486847, + "learning_rate": 0.0004969967131945106, + "loss": 3.113190460205078, + "step": 11780 + }, + { + "epoch": 0.05895, + "grad_norm": 0.46607980132102966, + "learning_rate": 0.0004969905800943547, + "loss": 3.1061422348022463, + "step": 11790 + }, + { + "epoch": 0.059, + "grad_norm": 0.4682972729206085, + "learning_rate": 0.0004969844407762125, + "loss": 3.175874710083008, + "step": 11800 + }, + { + "epoch": 0.05905, + "grad_norm": 0.4719960689544678, + "learning_rate": 0.0004969782952402387, + "loss": 3.075112533569336, + "step": 11810 + }, + { + "epoch": 0.0591, + "grad_norm": 0.43895164132118225, + "learning_rate": 0.000496972143486588, + "loss": 3.069042205810547, + "step": 11820 + }, + { + "epoch": 0.05915, + "grad_norm": 0.40696412324905396, + "learning_rate": 0.0004969659855154153, + "loss": 3.0823091506958007, + "step": 11830 + }, + { + "epoch": 0.0592, + "grad_norm": 0.4299461245536804, + "learning_rate": 0.0004969598213268755, + "loss": 3.120342826843262, + "step": 11840 + }, + { + "epoch": 0.05925, + "grad_norm": 0.4452085494995117, + "learning_rate": 0.0004969536509211238, + "loss": 3.062306213378906, + "step": 11850 + }, + { + "epoch": 0.0593, + "grad_norm": 0.4445352256298065, + "learning_rate": 0.0004969474742983157, + "loss": 3.0834772109985353, + "step": 11860 + }, + { + "epoch": 0.05935, + "grad_norm": 0.4412632882595062, + "learning_rate": 0.0004969412914586065, + "loss": 3.0349857330322267, + "step": 11870 + }, + { + "epoch": 0.0594, + "grad_norm": 0.43203645944595337, + "learning_rate": 0.0004969351024021519, + "loss": 3.096047019958496, + "step": 11880 + }, + { + "epoch": 0.05945, + "grad_norm": 0.4380509853363037, + "learning_rate": 0.0004969289071291077, + "loss": 3.07810001373291, + "step": 11890 + }, + { + "epoch": 0.0595, + "grad_norm": 0.41638490557670593, + "learning_rate": 0.0004969227056396301, + "loss": 3.0903270721435545, + "step": 11900 + }, + { + "epoch": 0.05955, + "grad_norm": 0.4320518672466278, + "learning_rate": 0.0004969164979338749, + "loss": 3.1063068389892576, + "step": 11910 + }, + { + "epoch": 0.0596, + "grad_norm": 0.4880376160144806, + "learning_rate": 0.0004969102840119985, + "loss": 3.061174201965332, + "step": 11920 + }, + { + "epoch": 0.05965, + "grad_norm": 0.4328549802303314, + "learning_rate": 0.0004969040638741574, + "loss": 3.0764492034912108, + "step": 11930 + }, + { + "epoch": 0.0597, + "grad_norm": 0.45993196964263916, + "learning_rate": 0.0004968978375205082, + "loss": 3.0526498794555663, + "step": 11940 + }, + { + "epoch": 0.05975, + "grad_norm": 0.436786949634552, + "learning_rate": 0.0004968916049512076, + "loss": 3.1095481872558595, + "step": 11950 + }, + { + "epoch": 0.0598, + "grad_norm": 0.45222562551498413, + "learning_rate": 0.0004968853661664124, + "loss": 3.06588134765625, + "step": 11960 + }, + { + "epoch": 0.05985, + "grad_norm": 0.43392738699913025, + "learning_rate": 0.0004968791211662798, + "loss": 3.0787946701049806, + "step": 11970 + }, + { + "epoch": 0.0599, + "grad_norm": 0.4860612154006958, + "learning_rate": 0.0004968728699509669, + "loss": 3.095821762084961, + "step": 11980 + }, + { + "epoch": 0.05995, + "grad_norm": 0.4640517234802246, + "learning_rate": 0.0004968666125206312, + "loss": 3.0653432846069335, + "step": 11990 + }, + { + "epoch": 0.06, + "grad_norm": 0.4604586958885193, + "learning_rate": 0.0004968603488754302, + "loss": 3.07115478515625, + "step": 12000 + }, + { + "epoch": 0.06005, + "grad_norm": 0.461698442697525, + "learning_rate": 0.0004968540790155216, + "loss": 3.1184209823608398, + "step": 12010 + }, + { + "epoch": 0.0601, + "grad_norm": 0.4627610146999359, + "learning_rate": 0.0004968478029410632, + "loss": 3.0927806854248048, + "step": 12020 + }, + { + "epoch": 0.06015, + "grad_norm": 0.4640527367591858, + "learning_rate": 0.0004968415206522129, + "loss": 3.144689178466797, + "step": 12030 + }, + { + "epoch": 0.0602, + "grad_norm": 0.4458412528038025, + "learning_rate": 0.0004968352321491291, + "loss": 3.0518545150756835, + "step": 12040 + }, + { + "epoch": 0.06025, + "grad_norm": 0.4273608922958374, + "learning_rate": 0.0004968289374319699, + "loss": 3.1335847854614256, + "step": 12050 + }, + { + "epoch": 0.0603, + "grad_norm": 0.4357644021511078, + "learning_rate": 0.0004968226365008938, + "loss": 3.0753379821777345, + "step": 12060 + }, + { + "epoch": 0.06035, + "grad_norm": 0.45946410298347473, + "learning_rate": 0.0004968163293560595, + "loss": 3.1011171340942383, + "step": 12070 + }, + { + "epoch": 0.0604, + "grad_norm": 0.4340782165527344, + "learning_rate": 0.000496810015997626, + "loss": 3.0946298599243165, + "step": 12080 + }, + { + "epoch": 0.06045, + "grad_norm": 0.41644611954689026, + "learning_rate": 0.0004968036964257518, + "loss": 3.0933860778808593, + "step": 12090 + }, + { + "epoch": 0.0605, + "grad_norm": 0.42529112100601196, + "learning_rate": 0.0004967973706405961, + "loss": 3.09979133605957, + "step": 12100 + }, + { + "epoch": 0.06055, + "grad_norm": 0.4440511465072632, + "learning_rate": 0.0004967910386423183, + "loss": 3.086526107788086, + "step": 12110 + }, + { + "epoch": 0.0606, + "grad_norm": 0.3875409662723541, + "learning_rate": 0.0004967847004310778, + "loss": 3.0803552627563477, + "step": 12120 + }, + { + "epoch": 0.06065, + "grad_norm": 0.4287048578262329, + "learning_rate": 0.0004967783560070341, + "loss": 3.1448644638061523, + "step": 12130 + }, + { + "epoch": 0.0607, + "grad_norm": 0.43062517046928406, + "learning_rate": 0.000496772005370347, + "loss": 3.0767364501953125, + "step": 12140 + }, + { + "epoch": 0.06075, + "grad_norm": 0.42424097657203674, + "learning_rate": 0.0004967656485211762, + "loss": 3.060273361206055, + "step": 12150 + }, + { + "epoch": 0.0608, + "grad_norm": 0.437877357006073, + "learning_rate": 0.000496759285459682, + "loss": 3.0645572662353517, + "step": 12160 + }, + { + "epoch": 0.06085, + "grad_norm": 0.4412948191165924, + "learning_rate": 0.0004967529161860243, + "loss": 3.0988748550415037, + "step": 12170 + }, + { + "epoch": 0.0609, + "grad_norm": 0.4294641315937042, + "learning_rate": 0.0004967465407003637, + "loss": 3.0882436752319338, + "step": 12180 + }, + { + "epoch": 0.06095, + "grad_norm": 0.42651334404945374, + "learning_rate": 0.0004967401590028604, + "loss": 3.093051528930664, + "step": 12190 + }, + { + "epoch": 0.061, + "grad_norm": 0.43140408396720886, + "learning_rate": 0.0004967337710936754, + "loss": 3.12774543762207, + "step": 12200 + }, + { + "epoch": 0.06105, + "grad_norm": 0.4214419722557068, + "learning_rate": 0.0004967273769729693, + "loss": 3.064424514770508, + "step": 12210 + }, + { + "epoch": 0.0611, + "grad_norm": 0.44243958592414856, + "learning_rate": 0.0004967209766409031, + "loss": 3.077922058105469, + "step": 12220 + }, + { + "epoch": 0.06115, + "grad_norm": 0.46073150634765625, + "learning_rate": 0.0004967145700976381, + "loss": 3.0733089447021484, + "step": 12230 + }, + { + "epoch": 0.0612, + "grad_norm": 0.42104554176330566, + "learning_rate": 0.0004967081573433353, + "loss": 3.084671974182129, + "step": 12240 + }, + { + "epoch": 0.06125, + "grad_norm": 0.4329899549484253, + "learning_rate": 0.0004967017383781564, + "loss": 3.0866928100585938, + "step": 12250 + }, + { + "epoch": 0.0613, + "grad_norm": 0.4449341893196106, + "learning_rate": 0.0004966953132022628, + "loss": 3.0333723068237304, + "step": 12260 + }, + { + "epoch": 0.06135, + "grad_norm": 0.4628029465675354, + "learning_rate": 0.0004966888818158163, + "loss": 3.048601531982422, + "step": 12270 + }, + { + "epoch": 0.0614, + "grad_norm": 0.4455244541168213, + "learning_rate": 0.000496682444218979, + "loss": 3.0988893508911133, + "step": 12280 + }, + { + "epoch": 0.06145, + "grad_norm": 0.43540072441101074, + "learning_rate": 0.0004966760004119127, + "loss": 3.0892410278320312, + "step": 12290 + }, + { + "epoch": 0.0615, + "grad_norm": 0.44479528069496155, + "learning_rate": 0.0004966695503947798, + "loss": 3.0746789932250977, + "step": 12300 + }, + { + "epoch": 0.06155, + "grad_norm": 0.4398055076599121, + "learning_rate": 0.0004966630941677426, + "loss": 3.046609115600586, + "step": 12310 + }, + { + "epoch": 0.0616, + "grad_norm": 0.46076443791389465, + "learning_rate": 0.0004966566317309636, + "loss": 3.0937873840332033, + "step": 12320 + }, + { + "epoch": 0.06165, + "grad_norm": 0.44999945163726807, + "learning_rate": 0.0004966501630846057, + "loss": 3.1026506423950195, + "step": 12330 + }, + { + "epoch": 0.0617, + "grad_norm": 0.42156878113746643, + "learning_rate": 0.0004966436882288314, + "loss": 3.0777894973754885, + "step": 12340 + }, + { + "epoch": 0.06175, + "grad_norm": 0.4666925072669983, + "learning_rate": 0.000496637207163804, + "loss": 3.12052001953125, + "step": 12350 + }, + { + "epoch": 0.0618, + "grad_norm": 0.4778507351875305, + "learning_rate": 0.0004966307198896865, + "loss": 3.1371421813964844, + "step": 12360 + }, + { + "epoch": 0.06185, + "grad_norm": 0.431976318359375, + "learning_rate": 0.0004966242264066423, + "loss": 3.0366167068481444, + "step": 12370 + }, + { + "epoch": 0.0619, + "grad_norm": 0.4032762050628662, + "learning_rate": 0.0004966177267148349, + "loss": 3.1120399475097655, + "step": 12380 + }, + { + "epoch": 0.06195, + "grad_norm": 0.4329270124435425, + "learning_rate": 0.0004966112208144278, + "loss": 3.0607009887695313, + "step": 12390 + }, + { + "epoch": 0.062, + "grad_norm": 0.4243468642234802, + "learning_rate": 0.000496604708705585, + "loss": 3.0312311172485353, + "step": 12400 + }, + { + "epoch": 0.06205, + "grad_norm": 0.41873520612716675, + "learning_rate": 0.0004965981903884701, + "loss": 3.028628921508789, + "step": 12410 + }, + { + "epoch": 0.0621, + "grad_norm": 0.4434989094734192, + "learning_rate": 0.0004965916658632476, + "loss": 3.081497573852539, + "step": 12420 + }, + { + "epoch": 0.06215, + "grad_norm": 0.4676065146923065, + "learning_rate": 0.0004965851351300814, + "loss": 3.1164152145385744, + "step": 12430 + }, + { + "epoch": 0.0622, + "grad_norm": 0.4360460638999939, + "learning_rate": 0.000496578598189136, + "loss": 3.084811973571777, + "step": 12440 + }, + { + "epoch": 0.06225, + "grad_norm": 0.46777695417404175, + "learning_rate": 0.0004965720550405763, + "loss": 3.1145423889160155, + "step": 12450 + }, + { + "epoch": 0.0623, + "grad_norm": 0.39834392070770264, + "learning_rate": 0.0004965655056845666, + "loss": 3.0644210815429687, + "step": 12460 + }, + { + "epoch": 0.06235, + "grad_norm": 0.41210728883743286, + "learning_rate": 0.000496558950121272, + "loss": 3.0101625442504885, + "step": 12470 + }, + { + "epoch": 0.0624, + "grad_norm": 0.42682114243507385, + "learning_rate": 0.0004965523883508575, + "loss": 3.0912878036499025, + "step": 12480 + }, + { + "epoch": 0.06245, + "grad_norm": 0.40611159801483154, + "learning_rate": 0.0004965458203734883, + "loss": 3.0586502075195314, + "step": 12490 + }, + { + "epoch": 0.0625, + "grad_norm": 0.40737447142601013, + "learning_rate": 0.0004965392461893295, + "loss": 3.0703718185424806, + "step": 12500 + }, + { + "epoch": 0.06255, + "grad_norm": 0.43525460362434387, + "learning_rate": 0.0004965326657985471, + "loss": 3.0985620498657225, + "step": 12510 + }, + { + "epoch": 0.0626, + "grad_norm": 0.44571685791015625, + "learning_rate": 0.0004965260792013063, + "loss": 3.0649415969848635, + "step": 12520 + }, + { + "epoch": 0.06265, + "grad_norm": 0.4251353144645691, + "learning_rate": 0.0004965194863977731, + "loss": 3.067170333862305, + "step": 12530 + }, + { + "epoch": 0.0627, + "grad_norm": 0.416737824678421, + "learning_rate": 0.0004965128873881136, + "loss": 3.08145809173584, + "step": 12540 + }, + { + "epoch": 0.06275, + "grad_norm": 0.4613164961338043, + "learning_rate": 0.0004965062821724937, + "loss": 3.0679317474365235, + "step": 12550 + }, + { + "epoch": 0.0628, + "grad_norm": 0.4403628408908844, + "learning_rate": 0.0004964996707510799, + "loss": 3.044031524658203, + "step": 12560 + }, + { + "epoch": 0.06285, + "grad_norm": 0.4254177212715149, + "learning_rate": 0.0004964930531240385, + "loss": 3.1091020584106444, + "step": 12570 + }, + { + "epoch": 0.0629, + "grad_norm": 0.4298385679721832, + "learning_rate": 0.0004964864292915361, + "loss": 3.0960504531860353, + "step": 12580 + }, + { + "epoch": 0.06295, + "grad_norm": 0.43309396505355835, + "learning_rate": 0.0004964797992537394, + "loss": 3.1045766830444337, + "step": 12590 + }, + { + "epoch": 0.063, + "grad_norm": 0.42517390847206116, + "learning_rate": 0.0004964731630108156, + "loss": 3.0560550689697266, + "step": 12600 + }, + { + "epoch": 0.06305, + "grad_norm": 0.43558305501937866, + "learning_rate": 0.0004964665205629315, + "loss": 3.029275894165039, + "step": 12610 + }, + { + "epoch": 0.0631, + "grad_norm": 0.4337294101715088, + "learning_rate": 0.0004964598719102544, + "loss": 3.000980567932129, + "step": 12620 + }, + { + "epoch": 0.06315, + "grad_norm": 0.42330023646354675, + "learning_rate": 0.0004964532170529516, + "loss": 3.092368507385254, + "step": 12630 + }, + { + "epoch": 0.0632, + "grad_norm": 0.4515341520309448, + "learning_rate": 0.0004964465559911907, + "loss": 3.0179523468017577, + "step": 12640 + }, + { + "epoch": 0.06325, + "grad_norm": 0.42109692096710205, + "learning_rate": 0.0004964398887251395, + "loss": 3.1256399154663086, + "step": 12650 + }, + { + "epoch": 0.0633, + "grad_norm": 0.4251618981361389, + "learning_rate": 0.0004964332152549657, + "loss": 3.107097625732422, + "step": 12660 + }, + { + "epoch": 0.06335, + "grad_norm": 0.44023171067237854, + "learning_rate": 0.0004964265355808373, + "loss": 3.086578369140625, + "step": 12670 + }, + { + "epoch": 0.0634, + "grad_norm": 0.4044812321662903, + "learning_rate": 0.0004964198497029225, + "loss": 3.0722845077514647, + "step": 12680 + }, + { + "epoch": 0.06345, + "grad_norm": 0.42578455805778503, + "learning_rate": 0.0004964131576213897, + "loss": 3.1056888580322264, + "step": 12690 + }, + { + "epoch": 0.0635, + "grad_norm": 0.4303935766220093, + "learning_rate": 0.0004964064593364073, + "loss": 3.047984313964844, + "step": 12700 + }, + { + "epoch": 0.06355, + "grad_norm": 0.46646544337272644, + "learning_rate": 0.000496399754848144, + "loss": 3.063814163208008, + "step": 12710 + }, + { + "epoch": 0.0636, + "grad_norm": 0.3943672478199005, + "learning_rate": 0.0004963930441567683, + "loss": 3.069620132446289, + "step": 12720 + }, + { + "epoch": 0.06365, + "grad_norm": 0.4204106628894806, + "learning_rate": 0.0004963863272624496, + "loss": 2.997442436218262, + "step": 12730 + }, + { + "epoch": 0.0637, + "grad_norm": 0.4261746108531952, + "learning_rate": 0.0004963796041653565, + "loss": 3.0699777603149414, + "step": 12740 + }, + { + "epoch": 0.06375, + "grad_norm": 0.4409177601337433, + "learning_rate": 0.0004963728748656588, + "loss": 3.0831933975219727, + "step": 12750 + }, + { + "epoch": 0.0638, + "grad_norm": 0.43858516216278076, + "learning_rate": 0.0004963661393635255, + "loss": 3.0720600128173827, + "step": 12760 + }, + { + "epoch": 0.06385, + "grad_norm": 0.435518741607666, + "learning_rate": 0.0004963593976591263, + "loss": 3.119475173950195, + "step": 12770 + }, + { + "epoch": 0.0639, + "grad_norm": 0.4648002088069916, + "learning_rate": 0.0004963526497526309, + "loss": 3.071752738952637, + "step": 12780 + }, + { + "epoch": 0.06395, + "grad_norm": 0.4398486018180847, + "learning_rate": 0.0004963458956442091, + "loss": 3.0430780410766602, + "step": 12790 + }, + { + "epoch": 0.064, + "grad_norm": 0.4275256097316742, + "learning_rate": 0.000496339135334031, + "loss": 3.0892427444458006, + "step": 12800 + }, + { + "epoch": 0.06405, + "grad_norm": 0.4073059558868408, + "learning_rate": 0.000496332368822267, + "loss": 3.0533267974853517, + "step": 12810 + }, + { + "epoch": 0.0641, + "grad_norm": 0.41940388083457947, + "learning_rate": 0.0004963255961090873, + "loss": 3.070657157897949, + "step": 12820 + }, + { + "epoch": 0.06415, + "grad_norm": 0.4299185276031494, + "learning_rate": 0.0004963188171946621, + "loss": 3.0961671829223634, + "step": 12830 + }, + { + "epoch": 0.0642, + "grad_norm": 0.4494936764240265, + "learning_rate": 0.0004963120320791625, + "loss": 3.0981143951416015, + "step": 12840 + }, + { + "epoch": 0.06425, + "grad_norm": 0.39347517490386963, + "learning_rate": 0.0004963052407627591, + "loss": 3.0969696044921875, + "step": 12850 + }, + { + "epoch": 0.0643, + "grad_norm": 0.4519364535808563, + "learning_rate": 0.0004962984432456229, + "loss": 3.054274559020996, + "step": 12860 + }, + { + "epoch": 0.06435, + "grad_norm": 0.42198774218559265, + "learning_rate": 0.000496291639527925, + "loss": 3.084841728210449, + "step": 12870 + }, + { + "epoch": 0.0644, + "grad_norm": 0.40672871470451355, + "learning_rate": 0.0004962848296098368, + "loss": 3.047662925720215, + "step": 12880 + }, + { + "epoch": 0.06445, + "grad_norm": 0.4454973340034485, + "learning_rate": 0.0004962780134915297, + "loss": 3.0772159576416014, + "step": 12890 + }, + { + "epoch": 0.0645, + "grad_norm": 0.4159931242465973, + "learning_rate": 0.0004962711911731751, + "loss": 3.0430816650390624, + "step": 12900 + }, + { + "epoch": 0.06455, + "grad_norm": 0.42769575119018555, + "learning_rate": 0.000496264362654945, + "loss": 3.0759033203125, + "step": 12910 + }, + { + "epoch": 0.0646, + "grad_norm": 0.45706847310066223, + "learning_rate": 0.0004962575279370112, + "loss": 3.1311113357543947, + "step": 12920 + }, + { + "epoch": 0.06465, + "grad_norm": 0.4372093081474304, + "learning_rate": 0.0004962506870195458, + "loss": 3.0485519409179687, + "step": 12930 + }, + { + "epoch": 0.0647, + "grad_norm": 0.41135138273239136, + "learning_rate": 0.000496243839902721, + "loss": 3.066899299621582, + "step": 12940 + }, + { + "epoch": 0.06475, + "grad_norm": 0.46954724192619324, + "learning_rate": 0.0004962369865867091, + "loss": 3.0878908157348635, + "step": 12950 + }, + { + "epoch": 0.0648, + "grad_norm": 0.4438859522342682, + "learning_rate": 0.0004962301270716828, + "loss": 3.0730493545532225, + "step": 12960 + }, + { + "epoch": 0.06485, + "grad_norm": 0.44842100143432617, + "learning_rate": 0.0004962232613578147, + "loss": 3.1094865798950195, + "step": 12970 + }, + { + "epoch": 0.0649, + "grad_norm": 0.42005571722984314, + "learning_rate": 0.0004962163894452776, + "loss": 3.0667959213256837, + "step": 12980 + }, + { + "epoch": 0.06495, + "grad_norm": 0.42673608660697937, + "learning_rate": 0.0004962095113342445, + "loss": 3.048061180114746, + "step": 12990 + }, + { + "epoch": 0.065, + "grad_norm": 0.4259413182735443, + "learning_rate": 0.0004962026270248887, + "loss": 3.039993667602539, + "step": 13000 + }, + { + "epoch": 0.06505, + "grad_norm": 0.43681538105010986, + "learning_rate": 0.0004961957365173834, + "loss": 3.1092639923095704, + "step": 13010 + }, + { + "epoch": 0.0651, + "grad_norm": 0.43655750155448914, + "learning_rate": 0.000496188839811902, + "loss": 3.054853820800781, + "step": 13020 + }, + { + "epoch": 0.06515, + "grad_norm": 0.40653282403945923, + "learning_rate": 0.0004961819369086182, + "loss": 3.024944305419922, + "step": 13030 + }, + { + "epoch": 0.0652, + "grad_norm": 0.42998287081718445, + "learning_rate": 0.0004961750278077059, + "loss": 3.020181655883789, + "step": 13040 + }, + { + "epoch": 0.06525, + "grad_norm": 0.4043445587158203, + "learning_rate": 0.0004961681125093388, + "loss": 3.039630126953125, + "step": 13050 + }, + { + "epoch": 0.0653, + "grad_norm": 0.4102916717529297, + "learning_rate": 0.0004961611910136911, + "loss": 3.0329830169677736, + "step": 13060 + }, + { + "epoch": 0.06535, + "grad_norm": 0.42298391461372375, + "learning_rate": 0.0004961542633209372, + "loss": 3.0262718200683594, + "step": 13070 + }, + { + "epoch": 0.0654, + "grad_norm": 0.46637362241744995, + "learning_rate": 0.0004961473294312512, + "loss": 3.0498531341552733, + "step": 13080 + }, + { + "epoch": 0.06545, + "grad_norm": 0.38563141226768494, + "learning_rate": 0.000496140389344808, + "loss": 3.084819030761719, + "step": 13090 + }, + { + "epoch": 0.0655, + "grad_norm": 0.4165147542953491, + "learning_rate": 0.000496133443061782, + "loss": 3.0617265701293945, + "step": 13100 + }, + { + "epoch": 0.06555, + "grad_norm": 0.4005628526210785, + "learning_rate": 0.0004961264905823482, + "loss": 3.056559181213379, + "step": 13110 + }, + { + "epoch": 0.0656, + "grad_norm": 0.42206570506095886, + "learning_rate": 0.0004961195319066817, + "loss": 3.061668586730957, + "step": 13120 + }, + { + "epoch": 0.06565, + "grad_norm": 0.4484948217868805, + "learning_rate": 0.0004961125670349577, + "loss": 3.0642635345458986, + "step": 13130 + }, + { + "epoch": 0.0657, + "grad_norm": 0.4593627154827118, + "learning_rate": 0.0004961055959673514, + "loss": 3.0886993408203125, + "step": 13140 + }, + { + "epoch": 0.06575, + "grad_norm": 0.40536049008369446, + "learning_rate": 0.0004960986187040384, + "loss": 3.054697608947754, + "step": 13150 + }, + { + "epoch": 0.0658, + "grad_norm": 0.4241085648536682, + "learning_rate": 0.0004960916352451943, + "loss": 3.0130767822265625, + "step": 13160 + }, + { + "epoch": 0.06585, + "grad_norm": 0.42358672618865967, + "learning_rate": 0.0004960846455909949, + "loss": 3.074187469482422, + "step": 13170 + }, + { + "epoch": 0.0659, + "grad_norm": 0.44287270307540894, + "learning_rate": 0.0004960776497416161, + "loss": 3.0324588775634767, + "step": 13180 + }, + { + "epoch": 0.06595, + "grad_norm": 0.4137684106826782, + "learning_rate": 0.0004960706476972343, + "loss": 3.0051416397094726, + "step": 13190 + }, + { + "epoch": 0.066, + "grad_norm": 0.408628910779953, + "learning_rate": 0.0004960636394580254, + "loss": 3.0680124282836916, + "step": 13200 + }, + { + "epoch": 0.06605, + "grad_norm": 0.4451737701892853, + "learning_rate": 0.0004960566250241662, + "loss": 3.0856391906738283, + "step": 13210 + }, + { + "epoch": 0.0661, + "grad_norm": 0.3966505825519562, + "learning_rate": 0.0004960496043958331, + "loss": 3.068583869934082, + "step": 13220 + }, + { + "epoch": 0.06615, + "grad_norm": 0.39775320887565613, + "learning_rate": 0.0004960425775732027, + "loss": 3.1176425933837892, + "step": 13230 + }, + { + "epoch": 0.0662, + "grad_norm": 0.43097108602523804, + "learning_rate": 0.0004960355445564523, + "loss": 3.0790922164916994, + "step": 13240 + }, + { + "epoch": 0.06625, + "grad_norm": 0.4682798683643341, + "learning_rate": 0.0004960285053457585, + "loss": 3.030688667297363, + "step": 13250 + }, + { + "epoch": 0.0663, + "grad_norm": 0.43764975666999817, + "learning_rate": 0.0004960214599412987, + "loss": 3.0811805725097656, + "step": 13260 + }, + { + "epoch": 0.06635, + "grad_norm": 0.4196675717830658, + "learning_rate": 0.0004960144083432504, + "loss": 3.0694129943847654, + "step": 13270 + }, + { + "epoch": 0.0664, + "grad_norm": 0.4495108723640442, + "learning_rate": 0.0004960073505517911, + "loss": 3.0427200317382814, + "step": 13280 + }, + { + "epoch": 0.06645, + "grad_norm": 0.41927406191825867, + "learning_rate": 0.0004960002865670982, + "loss": 3.0375852584838867, + "step": 13290 + }, + { + "epoch": 0.0665, + "grad_norm": 0.4418603777885437, + "learning_rate": 0.0004959932163893498, + "loss": 2.964359664916992, + "step": 13300 + }, + { + "epoch": 0.06655, + "grad_norm": 0.4069922864437103, + "learning_rate": 0.0004959861400187237, + "loss": 3.0728763580322265, + "step": 13310 + }, + { + "epoch": 0.0666, + "grad_norm": 0.40913358330726624, + "learning_rate": 0.0004959790574553984, + "loss": 3.0827348709106444, + "step": 13320 + }, + { + "epoch": 0.06665, + "grad_norm": 0.4758280813694, + "learning_rate": 0.0004959719686995519, + "loss": 3.0613563537597654, + "step": 13330 + }, + { + "epoch": 0.0667, + "grad_norm": 0.45922207832336426, + "learning_rate": 0.0004959648737513627, + "loss": 3.107474708557129, + "step": 13340 + }, + { + "epoch": 0.06675, + "grad_norm": 0.43403008580207825, + "learning_rate": 0.0004959577726110093, + "loss": 3.076357841491699, + "step": 13350 + }, + { + "epoch": 0.0668, + "grad_norm": 0.4216567575931549, + "learning_rate": 0.0004959506652786708, + "loss": 3.0198015213012694, + "step": 13360 + }, + { + "epoch": 0.06685, + "grad_norm": 0.3921400010585785, + "learning_rate": 0.000495943551754526, + "loss": 3.0764623641967774, + "step": 13370 + }, + { + "epoch": 0.0669, + "grad_norm": 0.4183896780014038, + "learning_rate": 0.0004959364320387538, + "loss": 3.074908447265625, + "step": 13380 + }, + { + "epoch": 0.06695, + "grad_norm": 0.3848344683647156, + "learning_rate": 0.0004959293061315336, + "loss": 3.068827819824219, + "step": 13390 + }, + { + "epoch": 0.067, + "grad_norm": 0.3933485448360443, + "learning_rate": 0.0004959221740330448, + "loss": 3.0837787628173827, + "step": 13400 + }, + { + "epoch": 0.06705, + "grad_norm": 0.43187573552131653, + "learning_rate": 0.0004959150357434669, + "loss": 3.029253387451172, + "step": 13410 + }, + { + "epoch": 0.0671, + "grad_norm": 0.4535958766937256, + "learning_rate": 0.0004959078912629795, + "loss": 3.0557483673095702, + "step": 13420 + }, + { + "epoch": 0.06715, + "grad_norm": 0.4150027930736542, + "learning_rate": 0.0004959007405917627, + "loss": 3.0515743255615235, + "step": 13430 + }, + { + "epoch": 0.0672, + "grad_norm": 0.5156484842300415, + "learning_rate": 0.0004958935837299964, + "loss": 3.0545581817626952, + "step": 13440 + }, + { + "epoch": 0.06725, + "grad_norm": 0.4644143879413605, + "learning_rate": 0.0004958864206778607, + "loss": 3.0592977523803713, + "step": 13450 + }, + { + "epoch": 0.0673, + "grad_norm": 0.4512185752391815, + "learning_rate": 0.0004958792514355361, + "loss": 3.056985092163086, + "step": 13460 + }, + { + "epoch": 0.06735, + "grad_norm": 0.3998720347881317, + "learning_rate": 0.0004958720760032029, + "loss": 3.0489158630371094, + "step": 13470 + }, + { + "epoch": 0.0674, + "grad_norm": 0.44172775745391846, + "learning_rate": 0.0004958648943810418, + "loss": 3.057107925415039, + "step": 13480 + }, + { + "epoch": 0.06745, + "grad_norm": 0.4343692362308502, + "learning_rate": 0.0004958577065692338, + "loss": 3.0350465774536133, + "step": 13490 + }, + { + "epoch": 0.0675, + "grad_norm": 0.3983580470085144, + "learning_rate": 0.0004958505125679596, + "loss": 3.0317499160766603, + "step": 13500 + }, + { + "epoch": 0.06755, + "grad_norm": 0.4575647711753845, + "learning_rate": 0.0004958433123774004, + "loss": 3.0480283737182616, + "step": 13510 + }, + { + "epoch": 0.0676, + "grad_norm": 0.4236801266670227, + "learning_rate": 0.0004958361059977373, + "loss": 3.064761161804199, + "step": 13520 + }, + { + "epoch": 0.06765, + "grad_norm": 0.45026737451553345, + "learning_rate": 0.0004958288934291521, + "loss": 3.030267906188965, + "step": 13530 + }, + { + "epoch": 0.0677, + "grad_norm": 0.49033811688423157, + "learning_rate": 0.000495821674671826, + "loss": 3.05400390625, + "step": 13540 + }, + { + "epoch": 0.06775, + "grad_norm": 0.41808491945266724, + "learning_rate": 0.000495814449725941, + "loss": 3.0657394409179686, + "step": 13550 + }, + { + "epoch": 0.0678, + "grad_norm": 0.44676774740219116, + "learning_rate": 0.0004958072185916787, + "loss": 3.0624601364135744, + "step": 13560 + }, + { + "epoch": 0.06785, + "grad_norm": 0.4425623416900635, + "learning_rate": 0.0004957999812692215, + "loss": 3.0684131622314452, + "step": 13570 + }, + { + "epoch": 0.0679, + "grad_norm": 0.43752047419548035, + "learning_rate": 0.0004957927377587512, + "loss": 3.06956787109375, + "step": 13580 + }, + { + "epoch": 0.06795, + "grad_norm": 0.47310513257980347, + "learning_rate": 0.0004957854880604506, + "loss": 2.9847490310668947, + "step": 13590 + }, + { + "epoch": 0.068, + "grad_norm": 0.474104642868042, + "learning_rate": 0.0004957782321745018, + "loss": 3.0415035247802735, + "step": 13600 + }, + { + "epoch": 0.06805, + "grad_norm": 0.407358318567276, + "learning_rate": 0.0004957709701010878, + "loss": 3.023394966125488, + "step": 13610 + }, + { + "epoch": 0.0681, + "grad_norm": Infinity, + "learning_rate": 0.0004957637018403911, + "loss": 3.0967655181884766, + "step": 13620 + }, + { + "epoch": 0.06815, + "grad_norm": 0.41001370549201965, + "learning_rate": 0.0004957564273925949, + "loss": 3.0575170516967773, + "step": 13630 + }, + { + "epoch": 0.0682, + "grad_norm": 0.39885449409484863, + "learning_rate": 0.0004957491467578824, + "loss": 3.0892404556274413, + "step": 13640 + }, + { + "epoch": 0.06825, + "grad_norm": 0.46730849146842957, + "learning_rate": 0.0004957418599364367, + "loss": 3.0483850479125976, + "step": 13650 + }, + { + "epoch": 0.0683, + "grad_norm": 0.44512632489204407, + "learning_rate": 0.0004957345669284413, + "loss": 3.0486120223999023, + "step": 13660 + }, + { + "epoch": 0.06835, + "grad_norm": 0.44009900093078613, + "learning_rate": 0.0004957272677340799, + "loss": 3.0199108123779297, + "step": 13670 + }, + { + "epoch": 0.0684, + "grad_norm": 0.43925291299819946, + "learning_rate": 0.000495719962353536, + "loss": 3.040329360961914, + "step": 13680 + }, + { + "epoch": 0.06845, + "grad_norm": 0.41885459423065186, + "learning_rate": 0.0004957126507869938, + "loss": 3.100498390197754, + "step": 13690 + }, + { + "epoch": 0.0685, + "grad_norm": 0.4566747546195984, + "learning_rate": 0.0004957053330346372, + "loss": 3.0620933532714845, + "step": 13700 + }, + { + "epoch": 0.06855, + "grad_norm": 0.4159857928752899, + "learning_rate": 0.0004956980090966507, + "loss": 3.0453922271728517, + "step": 13710 + }, + { + "epoch": 0.0686, + "grad_norm": 0.44535091519355774, + "learning_rate": 0.0004956906789732182, + "loss": 3.044849395751953, + "step": 13720 + }, + { + "epoch": 0.06865, + "grad_norm": 0.42268291115760803, + "learning_rate": 0.0004956833426645246, + "loss": 3.0922595977783205, + "step": 13730 + }, + { + "epoch": 0.0687, + "grad_norm": 0.43735113739967346, + "learning_rate": 0.0004956760001707545, + "loss": 3.015547752380371, + "step": 13740 + }, + { + "epoch": 0.06875, + "grad_norm": 0.411480575799942, + "learning_rate": 0.0004956686514920927, + "loss": 3.0416345596313477, + "step": 13750 + }, + { + "epoch": 0.0688, + "grad_norm": 0.4191260039806366, + "learning_rate": 0.0004956612966287243, + "loss": 3.0467256546020507, + "step": 13760 + }, + { + "epoch": 0.06885, + "grad_norm": 0.39067938923835754, + "learning_rate": 0.0004956539355808344, + "loss": 3.1227474212646484, + "step": 13770 + }, + { + "epoch": 0.0689, + "grad_norm": 0.45102792978286743, + "learning_rate": 0.0004956465683486082, + "loss": 3.0380462646484374, + "step": 13780 + }, + { + "epoch": 0.06895, + "grad_norm": 0.412813276052475, + "learning_rate": 0.0004956391949322315, + "loss": 3.039482307434082, + "step": 13790 + }, + { + "epoch": 0.069, + "grad_norm": 0.4328019320964813, + "learning_rate": 0.0004956318153318895, + "loss": 3.0902366638183594, + "step": 13800 + }, + { + "epoch": 0.06905, + "grad_norm": 0.4204021394252777, + "learning_rate": 0.0004956244295477682, + "loss": 3.083039093017578, + "step": 13810 + }, + { + "epoch": 0.0691, + "grad_norm": 0.4725804924964905, + "learning_rate": 0.0004956170375800536, + "loss": 3.0504119873046873, + "step": 13820 + }, + { + "epoch": 0.06915, + "grad_norm": 0.4581315219402313, + "learning_rate": 0.0004956096394289317, + "loss": 3.064915084838867, + "step": 13830 + }, + { + "epoch": 0.0692, + "grad_norm": 0.41387972235679626, + "learning_rate": 0.0004956022350945888, + "loss": 3.023785209655762, + "step": 13840 + }, + { + "epoch": 0.06925, + "grad_norm": 0.4439130425453186, + "learning_rate": 0.0004955948245772112, + "loss": 3.0693410873413085, + "step": 13850 + }, + { + "epoch": 0.0693, + "grad_norm": 0.41072896122932434, + "learning_rate": 0.0004955874078769857, + "loss": 3.0425281524658203, + "step": 13860 + }, + { + "epoch": 0.06935, + "grad_norm": 0.43224000930786133, + "learning_rate": 0.0004955799849940987, + "loss": 3.0803417205810546, + "step": 13870 + }, + { + "epoch": 0.0694, + "grad_norm": 0.4159993827342987, + "learning_rate": 0.0004955725559287372, + "loss": 3.031279945373535, + "step": 13880 + }, + { + "epoch": 0.06945, + "grad_norm": 0.40543249249458313, + "learning_rate": 0.0004955651206810883, + "loss": 3.0478404998779296, + "step": 13890 + }, + { + "epoch": 0.0695, + "grad_norm": 0.4262925982475281, + "learning_rate": 0.0004955576792513391, + "loss": 3.056463050842285, + "step": 13900 + }, + { + "epoch": 0.06955, + "grad_norm": 0.4395013153553009, + "learning_rate": 0.000495550231639677, + "loss": 3.049996185302734, + "step": 13910 + }, + { + "epoch": 0.0696, + "grad_norm": 0.4323171377182007, + "learning_rate": 0.0004955427778462895, + "loss": 2.985594940185547, + "step": 13920 + }, + { + "epoch": 0.06965, + "grad_norm": 0.4359630346298218, + "learning_rate": 0.0004955353178713642, + "loss": 3.0153345108032226, + "step": 13930 + }, + { + "epoch": 0.0697, + "grad_norm": 0.41032522916793823, + "learning_rate": 0.0004955278517150888, + "loss": 3.044835090637207, + "step": 13940 + }, + { + "epoch": 0.06975, + "grad_norm": 0.42837467789649963, + "learning_rate": 0.0004955203793776514, + "loss": 3.056684875488281, + "step": 13950 + }, + { + "epoch": 0.0698, + "grad_norm": 0.455090194940567, + "learning_rate": 0.0004955129008592402, + "loss": 3.0634700775146486, + "step": 13960 + }, + { + "epoch": 0.06985, + "grad_norm": 0.3931862711906433, + "learning_rate": 0.0004955054161600433, + "loss": 3.0810726165771483, + "step": 13970 + }, + { + "epoch": 0.0699, + "grad_norm": 0.45165038108825684, + "learning_rate": 0.0004954979252802491, + "loss": 3.0796512603759765, + "step": 13980 + }, + { + "epoch": 0.06995, + "grad_norm": 0.3868260085582733, + "learning_rate": 0.0004954904282200465, + "loss": 3.012904739379883, + "step": 13990 + }, + { + "epoch": 0.07, + "grad_norm": 0.4215471148490906, + "learning_rate": 0.0004954829249796238, + "loss": 2.997014617919922, + "step": 14000 + }, + { + "epoch": 0.07005, + "grad_norm": 0.4310068190097809, + "learning_rate": 0.0004954754155591702, + "loss": 3.0785400390625, + "step": 14010 + }, + { + "epoch": 0.0701, + "grad_norm": 0.4491886496543884, + "learning_rate": 0.0004954678999588745, + "loss": 3.0604406356811524, + "step": 14020 + }, + { + "epoch": 0.07015, + "grad_norm": 0.39590463042259216, + "learning_rate": 0.0004954603781789263, + "loss": 3.0131013870239256, + "step": 14030 + }, + { + "epoch": 0.0702, + "grad_norm": 0.42315778136253357, + "learning_rate": 0.0004954528502195145, + "loss": 3.0353607177734374, + "step": 14040 + }, + { + "epoch": 0.07025, + "grad_norm": 0.4360930919647217, + "learning_rate": 0.000495445316080829, + "loss": 3.0252779006958006, + "step": 14050 + }, + { + "epoch": 0.0703, + "grad_norm": 0.4322260022163391, + "learning_rate": 0.0004954377757630593, + "loss": 3.0598237991333006, + "step": 14060 + }, + { + "epoch": 0.07035, + "grad_norm": 0.43572115898132324, + "learning_rate": 0.0004954302292663951, + "loss": 3.053603172302246, + "step": 14070 + }, + { + "epoch": 0.0704, + "grad_norm": 0.4054493308067322, + "learning_rate": 0.0004954226765910266, + "loss": 2.9940362930297852, + "step": 14080 + }, + { + "epoch": 0.07045, + "grad_norm": 0.416750431060791, + "learning_rate": 0.0004954151177371438, + "loss": 3.052347755432129, + "step": 14090 + }, + { + "epoch": 0.0705, + "grad_norm": 0.4071808159351349, + "learning_rate": 0.0004954075527049372, + "loss": 3.0601900100708006, + "step": 14100 + }, + { + "epoch": 0.07055, + "grad_norm": 0.44921445846557617, + "learning_rate": 0.000495399981494597, + "loss": 3.0524709701538084, + "step": 14110 + }, + { + "epoch": 0.0706, + "grad_norm": 0.4071250557899475, + "learning_rate": 0.0004953924041063138, + "loss": 3.0477027893066406, + "step": 14120 + }, + { + "epoch": 0.07065, + "grad_norm": 0.44183260202407837, + "learning_rate": 0.0004953848205402787, + "loss": 3.072732162475586, + "step": 14130 + }, + { + "epoch": 0.0707, + "grad_norm": 0.435482919216156, + "learning_rate": 0.0004953772307966822, + "loss": 3.0654884338378907, + "step": 14140 + }, + { + "epoch": 0.07075, + "grad_norm": 0.4094671607017517, + "learning_rate": 0.0004953696348757156, + "loss": 2.9849170684814452, + "step": 14150 + }, + { + "epoch": 0.0708, + "grad_norm": 0.4346947968006134, + "learning_rate": 0.0004953620327775701, + "loss": 3.0429378509521485, + "step": 14160 + }, + { + "epoch": 0.07085, + "grad_norm": 0.4008183181285858, + "learning_rate": 0.000495354424502437, + "loss": 3.0788972854614256, + "step": 14170 + }, + { + "epoch": 0.0709, + "grad_norm": 0.42749330401420593, + "learning_rate": 0.000495346810050508, + "loss": 3.0733192443847654, + "step": 14180 + }, + { + "epoch": 0.07095, + "grad_norm": 0.38859328627586365, + "learning_rate": 0.0004953391894219747, + "loss": 3.027678298950195, + "step": 14190 + }, + { + "epoch": 0.071, + "grad_norm": 0.43580374121665955, + "learning_rate": 0.0004953315626170289, + "loss": 3.114019775390625, + "step": 14200 + }, + { + "epoch": 0.07105, + "grad_norm": 0.4269464910030365, + "learning_rate": 0.0004953239296358626, + "loss": 3.017792892456055, + "step": 14210 + }, + { + "epoch": 0.0711, + "grad_norm": 0.41402947902679443, + "learning_rate": 0.0004953162904786681, + "loss": 3.025689697265625, + "step": 14220 + }, + { + "epoch": 0.07115, + "grad_norm": 0.4048073887825012, + "learning_rate": 0.0004953086451456376, + "loss": 3.021601104736328, + "step": 14230 + }, + { + "epoch": 0.0712, + "grad_norm": 0.401803582906723, + "learning_rate": 0.0004953009936369636, + "loss": 3.0076366424560548, + "step": 14240 + }, + { + "epoch": 0.07125, + "grad_norm": 0.4070033133029938, + "learning_rate": 0.0004952933359528387, + "loss": 3.0604631423950197, + "step": 14250 + }, + { + "epoch": 0.0713, + "grad_norm": 0.4213806688785553, + "learning_rate": 0.0004952856720934558, + "loss": 3.0531105041503905, + "step": 14260 + }, + { + "epoch": 0.07135, + "grad_norm": 0.4445253610610962, + "learning_rate": 0.0004952780020590076, + "loss": 3.069401168823242, + "step": 14270 + }, + { + "epoch": 0.0714, + "grad_norm": 0.4177786409854889, + "learning_rate": 0.0004952703258496874, + "loss": 3.084978485107422, + "step": 14280 + }, + { + "epoch": 0.07145, + "grad_norm": 0.4253176748752594, + "learning_rate": 0.0004952626434656884, + "loss": 3.02628231048584, + "step": 14290 + }, + { + "epoch": 0.0715, + "grad_norm": 0.398554265499115, + "learning_rate": 0.000495254954907204, + "loss": 3.0452856063842773, + "step": 14300 + }, + { + "epoch": 0.07155, + "grad_norm": 0.4095679223537445, + "learning_rate": 0.0004952472601744277, + "loss": 3.04653377532959, + "step": 14310 + }, + { + "epoch": 0.0716, + "grad_norm": 0.41034841537475586, + "learning_rate": 0.0004952395592675532, + "loss": 3.0917379379272463, + "step": 14320 + }, + { + "epoch": 0.07165, + "grad_norm": 0.4052357077598572, + "learning_rate": 0.0004952318521867745, + "loss": 2.970050239562988, + "step": 14330 + }, + { + "epoch": 0.0717, + "grad_norm": 0.4163179397583008, + "learning_rate": 0.0004952241389322857, + "loss": 3.1234582901000976, + "step": 14340 + }, + { + "epoch": 0.07175, + "grad_norm": 0.4375966787338257, + "learning_rate": 0.0004952164195042807, + "loss": 3.081032371520996, + "step": 14350 + }, + { + "epoch": 0.0718, + "grad_norm": 0.4549047648906708, + "learning_rate": 0.000495208693902954, + "loss": 3.0053888320922852, + "step": 14360 + }, + { + "epoch": 0.07185, + "grad_norm": 0.44842204451560974, + "learning_rate": 0.0004952009621285, + "loss": 3.005378532409668, + "step": 14370 + }, + { + "epoch": 0.0719, + "grad_norm": 0.44110289216041565, + "learning_rate": 0.0004951932241811136, + "loss": 3.05487117767334, + "step": 14380 + }, + { + "epoch": 0.07195, + "grad_norm": 0.41598138213157654, + "learning_rate": 0.0004951854800609893, + "loss": 3.063400459289551, + "step": 14390 + }, + { + "epoch": 0.072, + "grad_norm": 0.4587278664112091, + "learning_rate": 0.0004951777297683222, + "loss": 2.99243106842041, + "step": 14400 + }, + { + "epoch": 0.07205, + "grad_norm": 0.40660879015922546, + "learning_rate": 0.0004951699733033074, + "loss": 3.0594390869140624, + "step": 14410 + }, + { + "epoch": 0.0721, + "grad_norm": 0.4190792143344879, + "learning_rate": 0.0004951622106661401, + "loss": 3.0662994384765625, + "step": 14420 + }, + { + "epoch": 0.07215, + "grad_norm": 0.4463399648666382, + "learning_rate": 0.0004951544418570159, + "loss": 3.0860157012939453, + "step": 14430 + }, + { + "epoch": 0.0722, + "grad_norm": 0.4607914984226227, + "learning_rate": 0.0004951466668761301, + "loss": 3.0297698974609375, + "step": 14440 + }, + { + "epoch": 0.07225, + "grad_norm": 0.4298783540725708, + "learning_rate": 0.0004951388857236788, + "loss": 3.0540924072265625, + "step": 14450 + }, + { + "epoch": 0.0723, + "grad_norm": 0.41715288162231445, + "learning_rate": 0.0004951310983998575, + "loss": 3.03688907623291, + "step": 14460 + }, + { + "epoch": 0.07235, + "grad_norm": 0.44714051485061646, + "learning_rate": 0.0004951233049048624, + "loss": 3.072092819213867, + "step": 14470 + }, + { + "epoch": 0.0724, + "grad_norm": 0.4517274498939514, + "learning_rate": 0.00049511550523889, + "loss": 3.0159027099609377, + "step": 14480 + }, + { + "epoch": 0.07245, + "grad_norm": 0.417514830827713, + "learning_rate": 0.0004951076994021362, + "loss": 3.078214645385742, + "step": 14490 + }, + { + "epoch": 0.0725, + "grad_norm": 0.4170454144477844, + "learning_rate": 0.0004950998873947978, + "loss": 3.0436716079711914, + "step": 14500 + }, + { + "epoch": 0.07255, + "grad_norm": 0.4666653573513031, + "learning_rate": 0.0004950920692170712, + "loss": 3.036488914489746, + "step": 14510 + }, + { + "epoch": 0.0726, + "grad_norm": 0.4736715257167816, + "learning_rate": 0.0004950842448691535, + "loss": 3.0795412063598633, + "step": 14520 + }, + { + "epoch": 0.07265, + "grad_norm": 0.42345431447029114, + "learning_rate": 0.0004950764143512416, + "loss": 3.0593048095703126, + "step": 14530 + }, + { + "epoch": 0.0727, + "grad_norm": 0.426814466714859, + "learning_rate": 0.0004950685776635327, + "loss": 3.0710336685180666, + "step": 14540 + }, + { + "epoch": 0.07275, + "grad_norm": 0.4359742999076843, + "learning_rate": 0.0004950607348062239, + "loss": 3.0226369857788087, + "step": 14550 + }, + { + "epoch": 0.0728, + "grad_norm": 0.417744517326355, + "learning_rate": 0.0004950528857795126, + "loss": 3.0550174713134766, + "step": 14560 + }, + { + "epoch": 0.07285, + "grad_norm": 0.3928349018096924, + "learning_rate": 0.0004950450305835967, + "loss": 3.026188850402832, + "step": 14570 + }, + { + "epoch": 0.0729, + "grad_norm": 0.39700549840927124, + "learning_rate": 0.0004950371692186737, + "loss": 3.0028194427490233, + "step": 14580 + }, + { + "epoch": 0.07295, + "grad_norm": 0.43120020627975464, + "learning_rate": 0.0004950293016849418, + "loss": 3.0456228256225586, + "step": 14590 + }, + { + "epoch": 0.073, + "grad_norm": 0.42133063077926636, + "learning_rate": 0.0004950214279825987, + "loss": 3.0589826583862303, + "step": 14600 + }, + { + "epoch": 0.07305, + "grad_norm": 0.4256994426250458, + "learning_rate": 0.0004950135481118428, + "loss": 3.016509437561035, + "step": 14610 + }, + { + "epoch": 0.0731, + "grad_norm": 0.4473554790019989, + "learning_rate": 0.0004950056620728724, + "loss": 3.0570693969726563, + "step": 14620 + }, + { + "epoch": 0.07315, + "grad_norm": 0.44516581296920776, + "learning_rate": 0.0004949977698658862, + "loss": 3.071784019470215, + "step": 14630 + }, + { + "epoch": 0.0732, + "grad_norm": 0.4134456217288971, + "learning_rate": 0.0004949898714910828, + "loss": 3.0104108810424806, + "step": 14640 + }, + { + "epoch": 0.07325, + "grad_norm": 0.397262305021286, + "learning_rate": 0.0004949819669486609, + "loss": 3.0050783157348633, + "step": 14650 + }, + { + "epoch": 0.0733, + "grad_norm": 0.42361247539520264, + "learning_rate": 0.0004949740562388197, + "loss": 3.067057418823242, + "step": 14660 + }, + { + "epoch": 0.07335, + "grad_norm": 0.4103213846683502, + "learning_rate": 0.0004949661393617582, + "loss": 3.086886405944824, + "step": 14670 + }, + { + "epoch": 0.0734, + "grad_norm": 0.45061561465263367, + "learning_rate": 0.000494958216317676, + "loss": 3.053594779968262, + "step": 14680 + }, + { + "epoch": 0.07345, + "grad_norm": 0.4199714660644531, + "learning_rate": 0.0004949502871067721, + "loss": 3.036746597290039, + "step": 14690 + }, + { + "epoch": 0.0735, + "grad_norm": 0.4679690897464752, + "learning_rate": 0.0004949423517292465, + "loss": 3.052125358581543, + "step": 14700 + }, + { + "epoch": 0.07355, + "grad_norm": 0.4157259166240692, + "learning_rate": 0.0004949344101852987, + "loss": 3.040884590148926, + "step": 14710 + }, + { + "epoch": 0.0736, + "grad_norm": 0.4049403667449951, + "learning_rate": 0.0004949264624751289, + "loss": 3.002482604980469, + "step": 14720 + }, + { + "epoch": 0.07365, + "grad_norm": 0.4363904893398285, + "learning_rate": 0.000494918508598937, + "loss": 3.0583377838134767, + "step": 14730 + }, + { + "epoch": 0.0737, + "grad_norm": 0.4160439074039459, + "learning_rate": 0.0004949105485569233, + "loss": 3.0266529083251954, + "step": 14740 + }, + { + "epoch": 0.07375, + "grad_norm": 0.4420735538005829, + "learning_rate": 0.0004949025823492881, + "loss": 3.0204465866088865, + "step": 14750 + }, + { + "epoch": 0.0738, + "grad_norm": 0.40060293674468994, + "learning_rate": 0.000494894609976232, + "loss": 3.045253372192383, + "step": 14760 + }, + { + "epoch": 0.07385, + "grad_norm": 0.40863099694252014, + "learning_rate": 0.0004948866314379559, + "loss": 3.046921730041504, + "step": 14770 + }, + { + "epoch": 0.0739, + "grad_norm": 0.4431397020816803, + "learning_rate": 0.0004948786467346603, + "loss": 3.0493425369262694, + "step": 14780 + }, + { + "epoch": 0.07395, + "grad_norm": 0.4319309592247009, + "learning_rate": 0.0004948706558665465, + "loss": 3.0201263427734375, + "step": 14790 + }, + { + "epoch": 0.074, + "grad_norm": 0.39743417501449585, + "learning_rate": 0.0004948626588338156, + "loss": 3.0179458618164063, + "step": 14800 + }, + { + "epoch": 0.07405, + "grad_norm": 0.42042291164398193, + "learning_rate": 0.0004948546556366688, + "loss": 3.055504035949707, + "step": 14810 + }, + { + "epoch": 0.0741, + "grad_norm": 0.4455428421497345, + "learning_rate": 0.0004948466462753079, + "loss": 3.040127182006836, + "step": 14820 + }, + { + "epoch": 0.07415, + "grad_norm": 0.424570232629776, + "learning_rate": 0.0004948386307499341, + "loss": 3.0315185546875, + "step": 14830 + }, + { + "epoch": 0.0742, + "grad_norm": 0.43550926446914673, + "learning_rate": 0.0004948306090607495, + "loss": 3.0151220321655274, + "step": 14840 + }, + { + "epoch": 0.07425, + "grad_norm": 0.4377257227897644, + "learning_rate": 0.0004948225812079559, + "loss": 2.9626670837402345, + "step": 14850 + }, + { + "epoch": 0.0743, + "grad_norm": 0.42741337418556213, + "learning_rate": 0.0004948145471917554, + "loss": 3.076620864868164, + "step": 14860 + }, + { + "epoch": 0.07435, + "grad_norm": 0.4463099539279938, + "learning_rate": 0.0004948065070123504, + "loss": 3.028512191772461, + "step": 14870 + }, + { + "epoch": 0.0744, + "grad_norm": 0.3998011648654938, + "learning_rate": 0.0004947984606699432, + "loss": 3.041858100891113, + "step": 14880 + }, + { + "epoch": 0.07445, + "grad_norm": 0.4223368167877197, + "learning_rate": 0.0004947904081647364, + "loss": 3.0040191650390624, + "step": 14890 + }, + { + "epoch": 0.0745, + "grad_norm": 0.3911333680152893, + "learning_rate": 0.0004947823494969327, + "loss": 3.0630041122436524, + "step": 14900 + }, + { + "epoch": 0.07455, + "grad_norm": 0.4357396960258484, + "learning_rate": 0.000494774284666735, + "loss": 3.050481414794922, + "step": 14910 + }, + { + "epoch": 0.0746, + "grad_norm": 0.3790024220943451, + "learning_rate": 0.0004947662136743463, + "loss": 3.003383255004883, + "step": 14920 + }, + { + "epoch": 0.07465, + "grad_norm": 0.4218028783798218, + "learning_rate": 0.0004947581365199698, + "loss": 3.0250484466552736, + "step": 14930 + }, + { + "epoch": 0.0747, + "grad_norm": 0.42349016666412354, + "learning_rate": 0.0004947500532038089, + "loss": 3.032253074645996, + "step": 14940 + }, + { + "epoch": 0.07475, + "grad_norm": 0.4140298664569855, + "learning_rate": 0.0004947419637260671, + "loss": 3.03476505279541, + "step": 14950 + }, + { + "epoch": 0.0748, + "grad_norm": 0.4385324716567993, + "learning_rate": 0.0004947338680869478, + "loss": 3.0342405319213865, + "step": 14960 + }, + { + "epoch": 0.07485, + "grad_norm": 0.45654842257499695, + "learning_rate": 0.0004947257662866551, + "loss": 3.020151901245117, + "step": 14970 + }, + { + "epoch": 0.0749, + "grad_norm": 0.4099375605583191, + "learning_rate": 0.000494717658325393, + "loss": 3.0317527770996096, + "step": 14980 + }, + { + "epoch": 0.07495, + "grad_norm": 0.4464881420135498, + "learning_rate": 0.0004947095442033653, + "loss": 3.0880722045898437, + "step": 14990 + }, + { + "epoch": 0.075, + "grad_norm": 0.441891610622406, + "learning_rate": 0.0004947014239207766, + "loss": 3.0220977783203127, + "step": 15000 + }, + { + "epoch": 0.07505, + "grad_norm": 0.4136958420276642, + "learning_rate": 0.0004946932974778311, + "loss": 2.9815036773681642, + "step": 15010 + }, + { + "epoch": 0.0751, + "grad_norm": 0.420322448015213, + "learning_rate": 0.0004946851648747335, + "loss": 3.044084167480469, + "step": 15020 + }, + { + "epoch": 0.07515, + "grad_norm": 0.42696380615234375, + "learning_rate": 0.0004946770261116884, + "loss": 3.008147048950195, + "step": 15030 + }, + { + "epoch": 0.0752, + "grad_norm": 0.41254329681396484, + "learning_rate": 0.000494668881188901, + "loss": 3.029206466674805, + "step": 15040 + }, + { + "epoch": 0.07525, + "grad_norm": 0.41019389033317566, + "learning_rate": 0.0004946607301065761, + "loss": 3.0353115081787108, + "step": 15050 + }, + { + "epoch": 0.0753, + "grad_norm": 0.4046579897403717, + "learning_rate": 0.0004946525728649189, + "loss": 3.0178272247314455, + "step": 15060 + }, + { + "epoch": 0.07535, + "grad_norm": 0.3974760174751282, + "learning_rate": 0.0004946444094641348, + "loss": 2.9889122009277345, + "step": 15070 + }, + { + "epoch": 0.0754, + "grad_norm": 0.40915051102638245, + "learning_rate": 0.0004946362399044293, + "loss": 3.0426544189453124, + "step": 15080 + }, + { + "epoch": 0.07545, + "grad_norm": 0.4259672462940216, + "learning_rate": 0.0004946280641860082, + "loss": 2.9978145599365233, + "step": 15090 + }, + { + "epoch": 0.0755, + "grad_norm": 0.4192453622817993, + "learning_rate": 0.0004946198823090771, + "loss": 3.06632080078125, + "step": 15100 + }, + { + "epoch": 0.07555, + "grad_norm": 0.4059858024120331, + "learning_rate": 0.0004946116942738422, + "loss": 3.029409408569336, + "step": 15110 + }, + { + "epoch": 0.0756, + "grad_norm": 0.48828163743019104, + "learning_rate": 0.0004946035000805095, + "loss": 3.035860061645508, + "step": 15120 + }, + { + "epoch": 0.07565, + "grad_norm": 0.4310038685798645, + "learning_rate": 0.0004945952997292852, + "loss": 2.999332809448242, + "step": 15130 + }, + { + "epoch": 0.0757, + "grad_norm": 0.5463998317718506, + "learning_rate": 0.0004945870932203761, + "loss": 3.0294782638549806, + "step": 15140 + }, + { + "epoch": 0.07575, + "grad_norm": 0.41085031628608704, + "learning_rate": 0.0004945788805539884, + "loss": 3.0490779876708984, + "step": 15150 + }, + { + "epoch": 0.0758, + "grad_norm": 0.41146039962768555, + "learning_rate": 0.000494570661730329, + "loss": 3.0853240966796873, + "step": 15160 + }, + { + "epoch": 0.07585, + "grad_norm": 0.41574349999427795, + "learning_rate": 0.000494562436749605, + "loss": 3.0029788970947267, + "step": 15170 + }, + { + "epoch": 0.0759, + "grad_norm": 0.4222142696380615, + "learning_rate": 0.0004945542056120231, + "loss": 3.0169275283813475, + "step": 15180 + }, + { + "epoch": 0.07595, + "grad_norm": 0.4162946343421936, + "learning_rate": 0.0004945459683177907, + "loss": 2.972304916381836, + "step": 15190 + }, + { + "epoch": 0.076, + "grad_norm": 0.42350664734840393, + "learning_rate": 0.0004945377248671153, + "loss": 3.0511598587036133, + "step": 15200 + }, + { + "epoch": 0.07605, + "grad_norm": 0.41618606448173523, + "learning_rate": 0.0004945294752602043, + "loss": 3.0345659255981445, + "step": 15210 + }, + { + "epoch": 0.0761, + "grad_norm": 0.40396299958229065, + "learning_rate": 0.0004945212194972654, + "loss": 3.016426658630371, + "step": 15220 + }, + { + "epoch": 0.07615, + "grad_norm": 0.4166755676269531, + "learning_rate": 0.0004945129575785063, + "loss": 3.0199813842773438, + "step": 15230 + }, + { + "epoch": 0.0762, + "grad_norm": 0.4646637439727783, + "learning_rate": 0.0004945046895041353, + "loss": 3.035883140563965, + "step": 15240 + }, + { + "epoch": 0.07625, + "grad_norm": 0.43878182768821716, + "learning_rate": 0.0004944964152743603, + "loss": 3.025874137878418, + "step": 15250 + }, + { + "epoch": 0.0763, + "grad_norm": 0.43151336908340454, + "learning_rate": 0.0004944881348893897, + "loss": 3.057098388671875, + "step": 15260 + }, + { + "epoch": 0.07635, + "grad_norm": 0.4237331748008728, + "learning_rate": 0.0004944798483494319, + "loss": 3.0365501403808595, + "step": 15270 + }, + { + "epoch": 0.0764, + "grad_norm": 0.40569546818733215, + "learning_rate": 0.0004944715556546956, + "loss": 3.0207820892333985, + "step": 15280 + }, + { + "epoch": 0.07645, + "grad_norm": 0.418804407119751, + "learning_rate": 0.0004944632568053895, + "loss": 3.0654834747314452, + "step": 15290 + }, + { + "epoch": 0.0765, + "grad_norm": 0.4376133978366852, + "learning_rate": 0.0004944549518017225, + "loss": 3.0015584945678713, + "step": 15300 + }, + { + "epoch": 0.07655, + "grad_norm": 0.4002218544483185, + "learning_rate": 0.0004944466406439038, + "loss": 3.0127262115478515, + "step": 15310 + }, + { + "epoch": 0.0766, + "grad_norm": 0.4170820116996765, + "learning_rate": 0.0004944383233321425, + "loss": 3.055810737609863, + "step": 15320 + }, + { + "epoch": 0.07665, + "grad_norm": 0.3868899643421173, + "learning_rate": 0.0004944299998666479, + "loss": 2.9739154815673827, + "step": 15330 + }, + { + "epoch": 0.0767, + "grad_norm": 0.4745780825614929, + "learning_rate": 0.00049442167024763, + "loss": 3.040243911743164, + "step": 15340 + }, + { + "epoch": 0.07675, + "grad_norm": 0.41840097308158875, + "learning_rate": 0.0004944133344752979, + "loss": 3.036795997619629, + "step": 15350 + }, + { + "epoch": 0.0768, + "grad_norm": 0.4274292290210724, + "learning_rate": 0.0004944049925498619, + "loss": 3.0502593994140623, + "step": 15360 + }, + { + "epoch": 0.07685, + "grad_norm": 0.3790803551673889, + "learning_rate": 0.0004943966444715317, + "loss": 3.051592254638672, + "step": 15370 + }, + { + "epoch": 0.0769, + "grad_norm": 0.4218064248561859, + "learning_rate": 0.0004943882902405178, + "loss": 3.0085887908935547, + "step": 15380 + }, + { + "epoch": 0.07695, + "grad_norm": 0.407588928937912, + "learning_rate": 0.0004943799298570303, + "loss": 3.018474006652832, + "step": 15390 + }, + { + "epoch": 0.077, + "grad_norm": 0.4012243449687958, + "learning_rate": 0.0004943715633212797, + "loss": 3.040899085998535, + "step": 15400 + }, + { + "epoch": 0.07705, + "grad_norm": 0.39953622221946716, + "learning_rate": 0.0004943631906334765, + "loss": 3.048007011413574, + "step": 15410 + }, + { + "epoch": 0.0771, + "grad_norm": 0.4310099184513092, + "learning_rate": 0.0004943548117938316, + "loss": 3.0128854751586913, + "step": 15420 + }, + { + "epoch": 0.07715, + "grad_norm": 0.4125700891017914, + "learning_rate": 0.0004943464268025561, + "loss": 3.023000144958496, + "step": 15430 + }, + { + "epoch": 0.0772, + "grad_norm": 0.43228572607040405, + "learning_rate": 0.0004943380356598608, + "loss": 3.0453975677490233, + "step": 15440 + }, + { + "epoch": 0.07725, + "grad_norm": 0.46034032106399536, + "learning_rate": 0.0004943296383659572, + "loss": 3.082236480712891, + "step": 15450 + }, + { + "epoch": 0.0773, + "grad_norm": 0.40177884697914124, + "learning_rate": 0.0004943212349210565, + "loss": 3.009682846069336, + "step": 15460 + }, + { + "epoch": 0.07735, + "grad_norm": 0.4239720106124878, + "learning_rate": 0.0004943128253253703, + "loss": 3.0431360244750976, + "step": 15470 + }, + { + "epoch": 0.0774, + "grad_norm": 0.38665422797203064, + "learning_rate": 0.0004943044095791105, + "loss": 3.0014976501464843, + "step": 15480 + }, + { + "epoch": 0.07745, + "grad_norm": 0.4157944917678833, + "learning_rate": 0.0004942959876824887, + "loss": 3.0052492141723635, + "step": 15490 + }, + { + "epoch": 0.0775, + "grad_norm": 0.40721017122268677, + "learning_rate": 0.0004942875596357172, + "loss": 2.9916847229003904, + "step": 15500 + }, + { + "epoch": 0.07755, + "grad_norm": 0.4126449227333069, + "learning_rate": 0.0004942791254390078, + "loss": 2.9743717193603514, + "step": 15510 + }, + { + "epoch": 0.0776, + "grad_norm": 0.432405024766922, + "learning_rate": 0.0004942706850925733, + "loss": 3.1199457168579103, + "step": 15520 + }, + { + "epoch": 0.07765, + "grad_norm": 0.3919544517993927, + "learning_rate": 0.0004942622385966256, + "loss": 3.0502904891967773, + "step": 15530 + }, + { + "epoch": 0.0777, + "grad_norm": 0.46674874424934387, + "learning_rate": 0.0004942537859513779, + "loss": 3.0506708145141603, + "step": 15540 + }, + { + "epoch": 0.07775, + "grad_norm": 0.4476867616176605, + "learning_rate": 0.0004942453271570428, + "loss": 2.9856950759887697, + "step": 15550 + }, + { + "epoch": 0.0778, + "grad_norm": 0.4190482199192047, + "learning_rate": 0.0004942368622138331, + "loss": 3.005096435546875, + "step": 15560 + }, + { + "epoch": 0.07785, + "grad_norm": 0.415885329246521, + "learning_rate": 0.0004942283911219622, + "loss": 3.0152376174926756, + "step": 15570 + }, + { + "epoch": 0.0779, + "grad_norm": 0.39139848947525024, + "learning_rate": 0.0004942199138816431, + "loss": 3.0557058334350584, + "step": 15580 + }, + { + "epoch": 0.07795, + "grad_norm": 0.4102530777454376, + "learning_rate": 0.0004942114304930892, + "loss": 3.046027183532715, + "step": 15590 + }, + { + "epoch": 0.078, + "grad_norm": 0.40563368797302246, + "learning_rate": 0.0004942029409565142, + "loss": 3.0338916778564453, + "step": 15600 + }, + { + "epoch": 0.07805, + "grad_norm": 0.4070001244544983, + "learning_rate": 0.0004941944452721319, + "loss": 3.0510974884033204, + "step": 15610 + }, + { + "epoch": 0.0781, + "grad_norm": 0.3999904692173004, + "learning_rate": 0.000494185943440156, + "loss": 3.023247718811035, + "step": 15620 + }, + { + "epoch": 0.07815, + "grad_norm": 0.4079655706882477, + "learning_rate": 0.0004941774354608006, + "loss": 3.033884620666504, + "step": 15630 + }, + { + "epoch": 0.0782, + "grad_norm": 0.4317818582057953, + "learning_rate": 0.0004941689213342799, + "loss": 3.03594970703125, + "step": 15640 + }, + { + "epoch": 0.07825, + "grad_norm": 0.4103674590587616, + "learning_rate": 0.0004941604010608083, + "loss": 3.0427181243896486, + "step": 15650 + }, + { + "epoch": 0.0783, + "grad_norm": 0.42370083928108215, + "learning_rate": 0.0004941518746406001, + "loss": 3.0357107162475585, + "step": 15660 + }, + { + "epoch": 0.07835, + "grad_norm": 0.42844510078430176, + "learning_rate": 0.0004941433420738702, + "loss": 2.9782712936401365, + "step": 15670 + }, + { + "epoch": 0.0784, + "grad_norm": 0.42992016673088074, + "learning_rate": 0.0004941348033608333, + "loss": 3.030440330505371, + "step": 15680 + }, + { + "epoch": 0.07845, + "grad_norm": 0.40208572149276733, + "learning_rate": 0.0004941262585017042, + "loss": 3.0682470321655275, + "step": 15690 + }, + { + "epoch": 0.0785, + "grad_norm": 0.41560444235801697, + "learning_rate": 0.0004941177074966984, + "loss": 3.035159683227539, + "step": 15700 + }, + { + "epoch": 0.07855, + "grad_norm": 0.4042159616947174, + "learning_rate": 0.0004941091503460307, + "loss": 3.042405891418457, + "step": 15710 + }, + { + "epoch": 0.0786, + "grad_norm": 0.4276389181613922, + "learning_rate": 0.000494100587049917, + "loss": 2.9607845306396485, + "step": 15720 + }, + { + "epoch": 0.07865, + "grad_norm": 0.4670949876308441, + "learning_rate": 0.0004940920176085726, + "loss": 3.024188423156738, + "step": 15730 + }, + { + "epoch": 0.0787, + "grad_norm": 0.40999481081962585, + "learning_rate": 0.0004940834420222132, + "loss": 3.045955848693848, + "step": 15740 + }, + { + "epoch": 0.07875, + "grad_norm": 0.4095138907432556, + "learning_rate": 0.0004940748602910548, + "loss": 3.0176820755004883, + "step": 15750 + }, + { + "epoch": 0.0788, + "grad_norm": 0.4186955988407135, + "learning_rate": 0.0004940662724153134, + "loss": 3.014872169494629, + "step": 15760 + }, + { + "epoch": 0.07885, + "grad_norm": 0.4270552694797516, + "learning_rate": 0.0004940576783952054, + "loss": 3.0409141540527345, + "step": 15770 + }, + { + "epoch": 0.0789, + "grad_norm": 0.4293685853481293, + "learning_rate": 0.0004940490782309468, + "loss": 3.027055358886719, + "step": 15780 + }, + { + "epoch": 0.07895, + "grad_norm": 0.456049382686615, + "learning_rate": 0.0004940404719227543, + "loss": 3.0010168075561525, + "step": 15790 + }, + { + "epoch": 0.079, + "grad_norm": 0.42186132073402405, + "learning_rate": 0.0004940318594708446, + "loss": 3.010273742675781, + "step": 15800 + }, + { + "epoch": 0.07905, + "grad_norm": 0.40046226978302, + "learning_rate": 0.0004940232408754345, + "loss": 2.969969367980957, + "step": 15810 + }, + { + "epoch": 0.0791, + "grad_norm": 0.39585933089256287, + "learning_rate": 0.0004940146161367408, + "loss": 3.06337833404541, + "step": 15820 + }, + { + "epoch": 0.07915, + "grad_norm": 0.4315457046031952, + "learning_rate": 0.000494005985254981, + "loss": 3.040847969055176, + "step": 15830 + }, + { + "epoch": 0.0792, + "grad_norm": 0.49437007308006287, + "learning_rate": 0.0004939973482303719, + "loss": 3.0027137756347657, + "step": 15840 + }, + { + "epoch": 0.07925, + "grad_norm": 0.43426021933555603, + "learning_rate": 0.0004939887050631314, + "loss": 2.9905397415161135, + "step": 15850 + }, + { + "epoch": 0.0793, + "grad_norm": 0.3832433521747589, + "learning_rate": 0.0004939800557534767, + "loss": 2.9981452941894533, + "step": 15860 + }, + { + "epoch": 0.07935, + "grad_norm": 0.4148855209350586, + "learning_rate": 0.0004939714003016258, + "loss": 3.023213577270508, + "step": 15870 + }, + { + "epoch": 0.0794, + "grad_norm": 0.3952372968196869, + "learning_rate": 0.0004939627387077966, + "loss": 3.009656524658203, + "step": 15880 + }, + { + "epoch": 0.07945, + "grad_norm": 0.3981904983520508, + "learning_rate": 0.0004939540709722071, + "loss": 3.05117244720459, + "step": 15890 + }, + { + "epoch": 0.0795, + "grad_norm": 0.4229997992515564, + "learning_rate": 0.0004939453970950753, + "loss": 3.049421691894531, + "step": 15900 + }, + { + "epoch": 0.07955, + "grad_norm": 0.41993799805641174, + "learning_rate": 0.0004939367170766198, + "loss": 3.0530269622802733, + "step": 15910 + }, + { + "epoch": 0.0796, + "grad_norm": 0.38749998807907104, + "learning_rate": 0.0004939280309170592, + "loss": 2.994222068786621, + "step": 15920 + }, + { + "epoch": 0.07965, + "grad_norm": 0.4196079671382904, + "learning_rate": 0.000493919338616612, + "loss": 2.9984846115112305, + "step": 15930 + }, + { + "epoch": 0.0797, + "grad_norm": 0.3915437161922455, + "learning_rate": 0.0004939106401754971, + "loss": 3.0278448104858398, + "step": 15940 + }, + { + "epoch": 0.07975, + "grad_norm": 0.45901378989219666, + "learning_rate": 0.0004939019355939335, + "loss": 3.033346939086914, + "step": 15950 + }, + { + "epoch": 0.0798, + "grad_norm": 0.40597325563430786, + "learning_rate": 0.0004938932248721401, + "loss": 2.971295166015625, + "step": 15960 + }, + { + "epoch": 0.07985, + "grad_norm": 0.40412160754203796, + "learning_rate": 0.0004938845080103366, + "loss": 3.0150901794433596, + "step": 15970 + }, + { + "epoch": 0.0799, + "grad_norm": 0.37738466262817383, + "learning_rate": 0.0004938757850087422, + "loss": 3.0198696136474608, + "step": 15980 + }, + { + "epoch": 0.07995, + "grad_norm": 0.3915244936943054, + "learning_rate": 0.0004938670558675766, + "loss": 3.0286592483520507, + "step": 15990 + }, + { + "epoch": 0.08, + "grad_norm": 0.4346381425857544, + "learning_rate": 0.0004938583205870594, + "loss": 2.984975814819336, + "step": 16000 + }, + { + "epoch": 0.08005, + "grad_norm": 0.4381770193576813, + "learning_rate": 0.0004938495791674106, + "loss": 3.010390853881836, + "step": 16010 + }, + { + "epoch": 0.0801, + "grad_norm": 0.4325791299343109, + "learning_rate": 0.0004938408316088503, + "loss": 3.0067333221435546, + "step": 16020 + }, + { + "epoch": 0.08015, + "grad_norm": 0.42161616683006287, + "learning_rate": 0.0004938320779115987, + "loss": 3.0132516860961913, + "step": 16030 + }, + { + "epoch": 0.0802, + "grad_norm": 0.46755197644233704, + "learning_rate": 0.0004938233180758762, + "loss": 3.0077991485595703, + "step": 16040 + }, + { + "epoch": 0.08025, + "grad_norm": 0.43869516253471375, + "learning_rate": 0.0004938145521019032, + "loss": 3.0380022048950197, + "step": 16050 + }, + { + "epoch": 0.0803, + "grad_norm": 0.4321464002132416, + "learning_rate": 0.0004938057799899005, + "loss": 3.035447120666504, + "step": 16060 + }, + { + "epoch": 0.08035, + "grad_norm": 0.4282110035419464, + "learning_rate": 0.000493797001740089, + "loss": 3.003855514526367, + "step": 16070 + }, + { + "epoch": 0.0804, + "grad_norm": 0.395384818315506, + "learning_rate": 0.0004937882173526896, + "loss": 2.990129089355469, + "step": 16080 + }, + { + "epoch": 0.08045, + "grad_norm": 0.45007553696632385, + "learning_rate": 0.0004937794268279234, + "loss": 3.0123531341552736, + "step": 16090 + }, + { + "epoch": 0.0805, + "grad_norm": 0.41222935914993286, + "learning_rate": 0.0004937706301660117, + "loss": 3.0326547622680664, + "step": 16100 + }, + { + "epoch": 0.08055, + "grad_norm": 0.42049330472946167, + "learning_rate": 0.0004937618273671762, + "loss": 3.0699968338012695, + "step": 16110 + }, + { + "epoch": 0.0806, + "grad_norm": 0.4265180826187134, + "learning_rate": 0.0004937530184316383, + "loss": 3.02563533782959, + "step": 16120 + }, + { + "epoch": 0.08065, + "grad_norm": 0.4546215534210205, + "learning_rate": 0.0004937442033596197, + "loss": 2.967991256713867, + "step": 16130 + }, + { + "epoch": 0.0807, + "grad_norm": 0.40302103757858276, + "learning_rate": 0.0004937353821513424, + "loss": 3.0401132583618162, + "step": 16140 + }, + { + "epoch": 0.08075, + "grad_norm": 0.41792747378349304, + "learning_rate": 0.0004937265548070284, + "loss": 2.965079116821289, + "step": 16150 + }, + { + "epoch": 0.0808, + "grad_norm": 0.4012744128704071, + "learning_rate": 0.0004937177213269002, + "loss": 3.043223190307617, + "step": 16160 + }, + { + "epoch": 0.08085, + "grad_norm": 0.4195712208747864, + "learning_rate": 0.0004937088817111799, + "loss": 3.0428131103515623, + "step": 16170 + }, + { + "epoch": 0.0809, + "grad_norm": 0.42785483598709106, + "learning_rate": 0.0004937000359600902, + "loss": 2.9972002029418947, + "step": 16180 + }, + { + "epoch": 0.08095, + "grad_norm": 0.3999212980270386, + "learning_rate": 0.0004936911840738536, + "loss": 2.9583263397216797, + "step": 16190 + }, + { + "epoch": 0.081, + "grad_norm": 0.4024742841720581, + "learning_rate": 0.0004936823260526932, + "loss": 3.0959423065185545, + "step": 16200 + }, + { + "epoch": 0.08105, + "grad_norm": 0.4020160734653473, + "learning_rate": 0.0004936734618968317, + "loss": 3.0087312698364257, + "step": 16210 + }, + { + "epoch": 0.0811, + "grad_norm": 0.41929057240486145, + "learning_rate": 0.0004936645916064925, + "loss": 2.9716363906860352, + "step": 16220 + }, + { + "epoch": 0.08115, + "grad_norm": 0.4329576790332794, + "learning_rate": 0.0004936557151818989, + "loss": 3.0107952117919923, + "step": 16230 + }, + { + "epoch": 0.0812, + "grad_norm": 0.3996362090110779, + "learning_rate": 0.0004936468326232743, + "loss": 3.0199605941772463, + "step": 16240 + }, + { + "epoch": 0.08125, + "grad_norm": 0.39715269207954407, + "learning_rate": 0.0004936379439308422, + "loss": 3.0336109161376954, + "step": 16250 + }, + { + "epoch": 0.0813, + "grad_norm": 0.41527289152145386, + "learning_rate": 0.0004936290491048266, + "loss": 3.0209623336791993, + "step": 16260 + }, + { + "epoch": 0.08135, + "grad_norm": 0.43068552017211914, + "learning_rate": 0.0004936201481454513, + "loss": 2.9864025115966797, + "step": 16270 + }, + { + "epoch": 0.0814, + "grad_norm": 0.4180382192134857, + "learning_rate": 0.0004936112410529404, + "loss": 2.9835086822509767, + "step": 16280 + }, + { + "epoch": 0.08145, + "grad_norm": 0.4120452105998993, + "learning_rate": 0.0004936023278275182, + "loss": 3.050226402282715, + "step": 16290 + }, + { + "epoch": 0.0815, + "grad_norm": 0.4436265528202057, + "learning_rate": 0.000493593408469409, + "loss": 3.027474594116211, + "step": 16300 + }, + { + "epoch": 0.08155, + "grad_norm": 0.39533382654190063, + "learning_rate": 0.0004935844829788372, + "loss": 3.036281967163086, + "step": 16310 + }, + { + "epoch": 0.0816, + "grad_norm": 0.40083983540534973, + "learning_rate": 0.0004935755513560278, + "loss": 2.9857366561889647, + "step": 16320 + }, + { + "epoch": 0.08165, + "grad_norm": 0.4030260145664215, + "learning_rate": 0.0004935666136012055, + "loss": 2.9935346603393556, + "step": 16330 + }, + { + "epoch": 0.0817, + "grad_norm": 0.4494929909706116, + "learning_rate": 0.0004935576697145953, + "loss": 3.010569763183594, + "step": 16340 + }, + { + "epoch": 0.08175, + "grad_norm": 0.4400009214878082, + "learning_rate": 0.0004935487196964225, + "loss": 2.9673206329345705, + "step": 16350 + }, + { + "epoch": 0.0818, + "grad_norm": 0.43894970417022705, + "learning_rate": 0.0004935397635469122, + "loss": 3.018828582763672, + "step": 16360 + }, + { + "epoch": 0.08185, + "grad_norm": 0.43460342288017273, + "learning_rate": 0.00049353080126629, + "loss": 3.0074331283569338, + "step": 16370 + }, + { + "epoch": 0.0819, + "grad_norm": 0.37929776310920715, + "learning_rate": 0.0004935218328547815, + "loss": 2.9867956161499025, + "step": 16380 + }, + { + "epoch": 0.08195, + "grad_norm": 0.44037896394729614, + "learning_rate": 0.0004935128583126125, + "loss": 2.9975425720214846, + "step": 16390 + }, + { + "epoch": 0.082, + "grad_norm": 0.41553136706352234, + "learning_rate": 0.0004935038776400089, + "loss": 3.048304557800293, + "step": 16400 + }, + { + "epoch": 0.08205, + "grad_norm": 0.42872098088264465, + "learning_rate": 0.0004934948908371969, + "loss": 3.010783004760742, + "step": 16410 + }, + { + "epoch": 0.0821, + "grad_norm": 0.45513424277305603, + "learning_rate": 0.0004934858979044025, + "loss": 2.9912113189697265, + "step": 16420 + }, + { + "epoch": 0.08215, + "grad_norm": 0.4262707829475403, + "learning_rate": 0.0004934768988418525, + "loss": 3.063294219970703, + "step": 16430 + }, + { + "epoch": 0.0822, + "grad_norm": 0.40714019536972046, + "learning_rate": 0.0004934678936497729, + "loss": 3.026559257507324, + "step": 16440 + }, + { + "epoch": 0.08225, + "grad_norm": 0.44140079617500305, + "learning_rate": 0.0004934588823283909, + "loss": 3.0218427658081053, + "step": 16450 + }, + { + "epoch": 0.0823, + "grad_norm": 0.4251779019832611, + "learning_rate": 0.0004934498648779331, + "loss": 2.9946861267089844, + "step": 16460 + }, + { + "epoch": 0.08235, + "grad_norm": 0.41769903898239136, + "learning_rate": 0.0004934408412986265, + "loss": 3.0016429901123045, + "step": 16470 + }, + { + "epoch": 0.0824, + "grad_norm": 0.40858983993530273, + "learning_rate": 0.0004934318115906984, + "loss": 2.943284034729004, + "step": 16480 + }, + { + "epoch": 0.08245, + "grad_norm": 0.4180353283882141, + "learning_rate": 0.0004934227757543761, + "loss": 2.9940858840942384, + "step": 16490 + }, + { + "epoch": 0.0825, + "grad_norm": 0.40011829137802124, + "learning_rate": 0.0004934137337898871, + "loss": 3.0082279205322267, + "step": 16500 + }, + { + "epoch": 0.08255, + "grad_norm": 0.4016377031803131, + "learning_rate": 0.0004934046856974589, + "loss": 2.9812044143676757, + "step": 16510 + }, + { + "epoch": 0.0826, + "grad_norm": 0.4047131836414337, + "learning_rate": 0.0004933956314773193, + "loss": 3.032417869567871, + "step": 16520 + }, + { + "epoch": 0.08265, + "grad_norm": 0.403673380613327, + "learning_rate": 0.0004933865711296964, + "loss": 3.0419898986816407, + "step": 16530 + }, + { + "epoch": 0.0827, + "grad_norm": 0.3682062029838562, + "learning_rate": 0.0004933775046548182, + "loss": 3.032960319519043, + "step": 16540 + }, + { + "epoch": 0.08275, + "grad_norm": 0.40666288137435913, + "learning_rate": 0.000493368432052913, + "loss": 2.9818950653076173, + "step": 16550 + }, + { + "epoch": 0.0828, + "grad_norm": 0.37660762667655945, + "learning_rate": 0.000493359353324209, + "loss": 2.9934425354003906, + "step": 16560 + }, + { + "epoch": 0.08285, + "grad_norm": 0.4196010231971741, + "learning_rate": 0.0004933502684689351, + "loss": 3.0153261184692384, + "step": 16570 + }, + { + "epoch": 0.0829, + "grad_norm": 0.383253276348114, + "learning_rate": 0.0004933411774873197, + "loss": 3.013417434692383, + "step": 16580 + }, + { + "epoch": 0.08295, + "grad_norm": 0.4191989302635193, + "learning_rate": 0.0004933320803795917, + "loss": 3.013052749633789, + "step": 16590 + }, + { + "epoch": 0.083, + "grad_norm": 0.41527479887008667, + "learning_rate": 0.0004933229771459803, + "loss": 3.0130212783813475, + "step": 16600 + }, + { + "epoch": 0.08305, + "grad_norm": 0.4459558129310608, + "learning_rate": 0.0004933138677867147, + "loss": 3.0487133026123048, + "step": 16610 + }, + { + "epoch": 0.0831, + "grad_norm": 0.3924042284488678, + "learning_rate": 0.0004933047523020239, + "loss": 2.9990524291992187, + "step": 16620 + }, + { + "epoch": 0.08315, + "grad_norm": 0.4030965268611908, + "learning_rate": 0.0004932956306921377, + "loss": 3.006658363342285, + "step": 16630 + }, + { + "epoch": 0.0832, + "grad_norm": 0.3876161575317383, + "learning_rate": 0.0004932865029572855, + "loss": 3.007305717468262, + "step": 16640 + }, + { + "epoch": 0.08325, + "grad_norm": 0.40160447359085083, + "learning_rate": 0.0004932773690976973, + "loss": 2.9770181655883787, + "step": 16650 + }, + { + "epoch": 0.0833, + "grad_norm": 0.38483142852783203, + "learning_rate": 0.0004932682291136029, + "loss": 3.0307680130004884, + "step": 16660 + }, + { + "epoch": 0.08335, + "grad_norm": 0.3801630437374115, + "learning_rate": 0.0004932590830052326, + "loss": 3.021532440185547, + "step": 16670 + }, + { + "epoch": 0.0834, + "grad_norm": 0.4537679851055145, + "learning_rate": 0.0004932499307728163, + "loss": 2.989686965942383, + "step": 16680 + }, + { + "epoch": 0.08345, + "grad_norm": 0.4299713671207428, + "learning_rate": 0.0004932407724165848, + "loss": 3.015716552734375, + "step": 16690 + }, + { + "epoch": 0.0835, + "grad_norm": 0.4123026728630066, + "learning_rate": 0.0004932316079367684, + "loss": 3.0031387329101564, + "step": 16700 + }, + { + "epoch": 0.08355, + "grad_norm": 0.41168275475502014, + "learning_rate": 0.0004932224373335979, + "loss": 2.982390785217285, + "step": 16710 + }, + { + "epoch": 0.0836, + "grad_norm": 0.4063844382762909, + "learning_rate": 0.0004932132606073042, + "loss": 3.021174430847168, + "step": 16720 + }, + { + "epoch": 0.08365, + "grad_norm": 0.41638314723968506, + "learning_rate": 0.0004932040777581182, + "loss": 2.962446594238281, + "step": 16730 + }, + { + "epoch": 0.0837, + "grad_norm": 0.42228591442108154, + "learning_rate": 0.0004931948887862713, + "loss": 3.0034015655517576, + "step": 16740 + }, + { + "epoch": 0.08375, + "grad_norm": 0.4264296293258667, + "learning_rate": 0.0004931856936919946, + "loss": 3.0277694702148437, + "step": 16750 + }, + { + "epoch": 0.0838, + "grad_norm": 0.40793943405151367, + "learning_rate": 0.0004931764924755198, + "loss": 3.024114227294922, + "step": 16760 + }, + { + "epoch": 0.08385, + "grad_norm": 0.37891241908073425, + "learning_rate": 0.0004931672851370783, + "loss": 2.9812429428100584, + "step": 16770 + }, + { + "epoch": 0.0839, + "grad_norm": 0.4109377861022949, + "learning_rate": 0.0004931580716769021, + "loss": 3.0002317428588867, + "step": 16780 + }, + { + "epoch": 0.08395, + "grad_norm": 0.38988766074180603, + "learning_rate": 0.0004931488520952229, + "loss": 3.003449630737305, + "step": 16790 + }, + { + "epoch": 0.084, + "grad_norm": 0.43937286734580994, + "learning_rate": 0.0004931396263922732, + "loss": 2.997350883483887, + "step": 16800 + }, + { + "epoch": 0.08405, + "grad_norm": 0.3996603190898895, + "learning_rate": 0.000493130394568285, + "loss": 3.006118392944336, + "step": 16810 + }, + { + "epoch": 0.0841, + "grad_norm": 0.3988634943962097, + "learning_rate": 0.0004931211566234907, + "loss": 3.0105772018432617, + "step": 16820 + }, + { + "epoch": 0.08415, + "grad_norm": 0.4089464843273163, + "learning_rate": 0.0004931119125581228, + "loss": 3.0182571411132812, + "step": 16830 + }, + { + "epoch": 0.0842, + "grad_norm": 0.4601038992404938, + "learning_rate": 0.0004931026623724142, + "loss": 3.0478816986083985, + "step": 16840 + }, + { + "epoch": 0.08425, + "grad_norm": 0.42933860421180725, + "learning_rate": 0.0004930934060665978, + "loss": 3.012487602233887, + "step": 16850 + }, + { + "epoch": 0.0843, + "grad_norm": 0.43135005235671997, + "learning_rate": 0.0004930841436409064, + "loss": 3.038133239746094, + "step": 16860 + }, + { + "epoch": 0.08435, + "grad_norm": 0.3944123685359955, + "learning_rate": 0.0004930748750955734, + "loss": 3.017244338989258, + "step": 16870 + }, + { + "epoch": 0.0844, + "grad_norm": 0.3934997320175171, + "learning_rate": 0.0004930656004308319, + "loss": 3.05208683013916, + "step": 16880 + }, + { + "epoch": 0.08445, + "grad_norm": 0.4046505093574524, + "learning_rate": 0.0004930563196469157, + "loss": 3.0352298736572267, + "step": 16890 + }, + { + "epoch": 0.0845, + "grad_norm": 0.3931903839111328, + "learning_rate": 0.0004930470327440582, + "loss": 2.987870216369629, + "step": 16900 + }, + { + "epoch": 0.08455, + "grad_norm": 0.4182949662208557, + "learning_rate": 0.0004930377397224933, + "loss": 2.9846614837646483, + "step": 16910 + }, + { + "epoch": 0.0846, + "grad_norm": 0.42151540517807007, + "learning_rate": 0.0004930284405824549, + "loss": 2.99737606048584, + "step": 16920 + }, + { + "epoch": 0.08465, + "grad_norm": 0.4217252731323242, + "learning_rate": 0.0004930191353241771, + "loss": 2.9877853393554688, + "step": 16930 + }, + { + "epoch": 0.0847, + "grad_norm": 0.39646390080451965, + "learning_rate": 0.0004930098239478943, + "loss": 2.9519441604614256, + "step": 16940 + }, + { + "epoch": 0.08475, + "grad_norm": 0.39371734857559204, + "learning_rate": 0.0004930005064538406, + "loss": 2.9866552352905273, + "step": 16950 + }, + { + "epoch": 0.0848, + "grad_norm": 0.40942394733428955, + "learning_rate": 0.000492991182842251, + "loss": 3.034608840942383, + "step": 16960 + }, + { + "epoch": 0.08485, + "grad_norm": 0.39126285910606384, + "learning_rate": 0.0004929818531133599, + "loss": 3.030436706542969, + "step": 16970 + }, + { + "epoch": 0.0849, + "grad_norm": 0.4224851727485657, + "learning_rate": 0.0004929725172674022, + "loss": 3.014138412475586, + "step": 16980 + }, + { + "epoch": 0.08495, + "grad_norm": 0.4108598530292511, + "learning_rate": 0.0004929631753046132, + "loss": 2.9737569808959963, + "step": 16990 + }, + { + "epoch": 0.085, + "grad_norm": 0.42497941851615906, + "learning_rate": 0.0004929538272252277, + "loss": 2.9872474670410156, + "step": 17000 + }, + { + "epoch": 0.08505, + "grad_norm": 0.3986911475658417, + "learning_rate": 0.0004929444730294813, + "loss": 2.9395162582397463, + "step": 17010 + }, + { + "epoch": 0.0851, + "grad_norm": 0.42162349820137024, + "learning_rate": 0.0004929351127176093, + "loss": 3.01477165222168, + "step": 17020 + }, + { + "epoch": 0.08515, + "grad_norm": 0.3920750916004181, + "learning_rate": 0.0004929257462898477, + "loss": 2.993110466003418, + "step": 17030 + }, + { + "epoch": 0.0852, + "grad_norm": 0.38869237899780273, + "learning_rate": 0.0004929163737464318, + "loss": 3.0515056610107423, + "step": 17040 + }, + { + "epoch": 0.08525, + "grad_norm": 0.4019846022129059, + "learning_rate": 0.000492906995087598, + "loss": 2.982625198364258, + "step": 17050 + }, + { + "epoch": 0.0853, + "grad_norm": 0.3902756869792938, + "learning_rate": 0.0004928976103135822, + "loss": 3.016064453125, + "step": 17060 + }, + { + "epoch": 0.08535, + "grad_norm": 0.4064615070819855, + "learning_rate": 0.0004928882194246205, + "loss": 3.012543869018555, + "step": 17070 + }, + { + "epoch": 0.0854, + "grad_norm": 0.3905491828918457, + "learning_rate": 0.0004928788224209497, + "loss": 2.975779914855957, + "step": 17080 + }, + { + "epoch": 0.08545, + "grad_norm": 0.3932758867740631, + "learning_rate": 0.0004928694193028062, + "loss": 2.9724990844726564, + "step": 17090 + }, + { + "epoch": 0.0855, + "grad_norm": 0.3800906538963318, + "learning_rate": 0.0004928600100704266, + "loss": 2.9835941314697267, + "step": 17100 + }, + { + "epoch": 0.08555, + "grad_norm": 0.41058167815208435, + "learning_rate": 0.0004928505947240478, + "loss": 2.9919198989868163, + "step": 17110 + }, + { + "epoch": 0.0856, + "grad_norm": 0.4087032079696655, + "learning_rate": 0.000492841173263907, + "loss": 2.9493921279907225, + "step": 17120 + }, + { + "epoch": 0.08565, + "grad_norm": 0.41807740926742554, + "learning_rate": 0.0004928317456902412, + "loss": 3.0035457611083984, + "step": 17130 + }, + { + "epoch": 0.0857, + "grad_norm": 0.42784151434898376, + "learning_rate": 0.000492822312003288, + "loss": 2.98780517578125, + "step": 17140 + }, + { + "epoch": 0.08575, + "grad_norm": 0.3930789828300476, + "learning_rate": 0.0004928128722032846, + "loss": 2.969643974304199, + "step": 17150 + }, + { + "epoch": 0.0858, + "grad_norm": 0.4237373173236847, + "learning_rate": 0.0004928034262904687, + "loss": 3.0148294448852537, + "step": 17160 + }, + { + "epoch": 0.08585, + "grad_norm": 0.4123910367488861, + "learning_rate": 0.0004927939742650782, + "loss": 2.9749982833862303, + "step": 17170 + }, + { + "epoch": 0.0859, + "grad_norm": 0.4068756103515625, + "learning_rate": 0.000492784516127351, + "loss": 3.044289970397949, + "step": 17180 + }, + { + "epoch": 0.08595, + "grad_norm": 0.5067603588104248, + "learning_rate": 0.0004927750518775254, + "loss": 3.030312919616699, + "step": 17190 + }, + { + "epoch": 0.086, + "grad_norm": 0.42751988768577576, + "learning_rate": 0.0004927655815158393, + "loss": 3.0210588455200194, + "step": 17200 + }, + { + "epoch": 0.08605, + "grad_norm": 0.40351900458335876, + "learning_rate": 0.0004927561050425314, + "loss": 2.9825557708740233, + "step": 17210 + }, + { + "epoch": 0.0861, + "grad_norm": 0.4107872545719147, + "learning_rate": 0.0004927466224578402, + "loss": 3.0564807891845702, + "step": 17220 + }, + { + "epoch": 0.08615, + "grad_norm": 0.3824727535247803, + "learning_rate": 0.0004927371337620043, + "loss": 2.960910415649414, + "step": 17230 + }, + { + "epoch": 0.0862, + "grad_norm": 0.428481787443161, + "learning_rate": 0.0004927276389552627, + "loss": 3.043750190734863, + "step": 17240 + }, + { + "epoch": 0.08625, + "grad_norm": 0.4221065938472748, + "learning_rate": 0.0004927181380378544, + "loss": 2.9772539138793945, + "step": 17250 + }, + { + "epoch": 0.0863, + "grad_norm": 0.4316578805446625, + "learning_rate": 0.0004927086310100188, + "loss": 2.988338088989258, + "step": 17260 + }, + { + "epoch": 0.08635, + "grad_norm": 0.3734188377857208, + "learning_rate": 0.0004926991178719948, + "loss": 2.9860368728637696, + "step": 17270 + }, + { + "epoch": 0.0864, + "grad_norm": 0.3738458752632141, + "learning_rate": 0.0004926895986240222, + "loss": 3.0049415588378907, + "step": 17280 + }, + { + "epoch": 0.08645, + "grad_norm": 0.4638879597187042, + "learning_rate": 0.0004926800732663406, + "loss": 2.980580520629883, + "step": 17290 + }, + { + "epoch": 0.0865, + "grad_norm": 0.3773985207080841, + "learning_rate": 0.0004926705417991898, + "loss": 3.0081808090209963, + "step": 17300 + }, + { + "epoch": 0.08655, + "grad_norm": 0.4355092942714691, + "learning_rate": 0.0004926610042228096, + "loss": 2.941610336303711, + "step": 17310 + }, + { + "epoch": 0.0866, + "grad_norm": 0.39183536171913147, + "learning_rate": 0.0004926514605374404, + "loss": 3.0094675064086913, + "step": 17320 + }, + { + "epoch": 0.08665, + "grad_norm": 0.45733004808425903, + "learning_rate": 0.0004926419107433223, + "loss": 3.038616180419922, + "step": 17330 + }, + { + "epoch": 0.0867, + "grad_norm": 0.4076864719390869, + "learning_rate": 0.0004926323548406956, + "loss": 3.0203866958618164, + "step": 17340 + }, + { + "epoch": 0.08675, + "grad_norm": 0.3865482211112976, + "learning_rate": 0.000492622792829801, + "loss": 3.0035964965820314, + "step": 17350 + }, + { + "epoch": 0.0868, + "grad_norm": 0.4124213457107544, + "learning_rate": 0.0004926132247108793, + "loss": 3.0204648971557617, + "step": 17360 + }, + { + "epoch": 0.08685, + "grad_norm": 0.41598156094551086, + "learning_rate": 0.0004926036504841712, + "loss": 2.98094482421875, + "step": 17370 + }, + { + "epoch": 0.0869, + "grad_norm": 0.40575432777404785, + "learning_rate": 0.0004925940701499179, + "loss": 2.9946514129638673, + "step": 17380 + }, + { + "epoch": 0.08695, + "grad_norm": 0.42477890849113464, + "learning_rate": 0.0004925844837083605, + "loss": 3.0268035888671876, + "step": 17390 + }, + { + "epoch": 0.087, + "grad_norm": 0.4698100686073303, + "learning_rate": 0.0004925748911597403, + "loss": 3.031271743774414, + "step": 17400 + }, + { + "epoch": 0.08705, + "grad_norm": 0.43398401141166687, + "learning_rate": 0.0004925652925042988, + "loss": 2.9769147872924804, + "step": 17410 + }, + { + "epoch": 0.0871, + "grad_norm": 0.41419973969459534, + "learning_rate": 0.0004925556877422777, + "loss": 2.9908552169799805, + "step": 17420 + }, + { + "epoch": 0.08715, + "grad_norm": 0.38204243779182434, + "learning_rate": 0.0004925460768739189, + "loss": 2.9585245132446287, + "step": 17430 + }, + { + "epoch": 0.0872, + "grad_norm": 0.3965616226196289, + "learning_rate": 0.0004925364598994641, + "loss": 2.938535690307617, + "step": 17440 + }, + { + "epoch": 0.08725, + "grad_norm": 0.4577621817588806, + "learning_rate": 0.0004925268368191557, + "loss": 2.987740135192871, + "step": 17450 + }, + { + "epoch": 0.0873, + "grad_norm": 0.3953305184841156, + "learning_rate": 0.0004925172076332356, + "loss": 2.99172420501709, + "step": 17460 + }, + { + "epoch": 0.08735, + "grad_norm": 0.3966676890850067, + "learning_rate": 0.0004925075723419465, + "loss": 3.0148235321044923, + "step": 17470 + }, + { + "epoch": 0.0874, + "grad_norm": 0.3715584874153137, + "learning_rate": 0.000492497930945531, + "loss": 2.975336456298828, + "step": 17480 + }, + { + "epoch": 0.08745, + "grad_norm": 0.4323531985282898, + "learning_rate": 0.0004924882834442315, + "loss": 2.9948333740234374, + "step": 17490 + }, + { + "epoch": 0.0875, + "grad_norm": 0.3979775607585907, + "learning_rate": 0.0004924786298382913, + "loss": 2.9859235763549803, + "step": 17500 + }, + { + "epoch": 0.08755, + "grad_norm": 0.3850301504135132, + "learning_rate": 0.000492468970127953, + "loss": 3.0310834884643554, + "step": 17510 + }, + { + "epoch": 0.0876, + "grad_norm": 0.41494014859199524, + "learning_rate": 0.0004924593043134602, + "loss": 2.92403564453125, + "step": 17520 + }, + { + "epoch": 0.08765, + "grad_norm": 0.4225090444087982, + "learning_rate": 0.0004924496323950559, + "loss": 3.0617284774780273, + "step": 17530 + }, + { + "epoch": 0.0877, + "grad_norm": 0.4201820194721222, + "learning_rate": 0.0004924399543729838, + "loss": 3.0299983978271485, + "step": 17540 + }, + { + "epoch": 0.08775, + "grad_norm": 0.3822593688964844, + "learning_rate": 0.0004924302702474873, + "loss": 3.0237443923950194, + "step": 17550 + }, + { + "epoch": 0.0878, + "grad_norm": 0.42000189423561096, + "learning_rate": 0.0004924205800188106, + "loss": 3.0011774063110352, + "step": 17560 + }, + { + "epoch": 0.08785, + "grad_norm": 0.3943185806274414, + "learning_rate": 0.0004924108836871973, + "loss": 2.96072940826416, + "step": 17570 + }, + { + "epoch": 0.0879, + "grad_norm": 0.42101380228996277, + "learning_rate": 0.0004924011812528917, + "loss": 2.967290496826172, + "step": 17580 + }, + { + "epoch": 0.08795, + "grad_norm": 0.3846522867679596, + "learning_rate": 0.0004923914727161378, + "loss": 3.017996978759766, + "step": 17590 + }, + { + "epoch": 0.088, + "grad_norm": 0.4257715344429016, + "learning_rate": 0.0004923817580771804, + "loss": 2.9584531784057617, + "step": 17600 + }, + { + "epoch": 0.08805, + "grad_norm": 0.44196781516075134, + "learning_rate": 0.0004923720373362638, + "loss": 2.979490852355957, + "step": 17610 + }, + { + "epoch": 0.0881, + "grad_norm": 0.41167670488357544, + "learning_rate": 0.0004923623104936326, + "loss": 3.070339393615723, + "step": 17620 + }, + { + "epoch": 0.08815, + "grad_norm": 0.4211384356021881, + "learning_rate": 0.0004923525775495321, + "loss": 2.987842559814453, + "step": 17630 + }, + { + "epoch": 0.0882, + "grad_norm": 0.40066713094711304, + "learning_rate": 0.000492342838504207, + "loss": 2.9889583587646484, + "step": 17640 + }, + { + "epoch": 0.08825, + "grad_norm": 0.4206503629684448, + "learning_rate": 0.0004923330933579025, + "loss": 2.975999641418457, + "step": 17650 + }, + { + "epoch": 0.0883, + "grad_norm": 0.42546308040618896, + "learning_rate": 0.0004923233421108641, + "loss": 3.0239078521728517, + "step": 17660 + }, + { + "epoch": 0.08835, + "grad_norm": 0.3984076976776123, + "learning_rate": 0.0004923135847633371, + "loss": 2.974102783203125, + "step": 17670 + }, + { + "epoch": 0.0884, + "grad_norm": 0.38649871945381165, + "learning_rate": 0.0004923038213155672, + "loss": 3.023240852355957, + "step": 17680 + }, + { + "epoch": 0.08845, + "grad_norm": 0.3946269750595093, + "learning_rate": 0.0004922940517678003, + "loss": 3.0438411712646483, + "step": 17690 + }, + { + "epoch": 0.0885, + "grad_norm": 0.421010285615921, + "learning_rate": 0.0004922842761202823, + "loss": 2.989790916442871, + "step": 17700 + }, + { + "epoch": 0.08855, + "grad_norm": 0.3872404098510742, + "learning_rate": 0.0004922744943732591, + "loss": 2.972174072265625, + "step": 17710 + }, + { + "epoch": 0.0886, + "grad_norm": 0.41067832708358765, + "learning_rate": 0.0004922647065269773, + "loss": 3.028604507446289, + "step": 17720 + }, + { + "epoch": 0.08865, + "grad_norm": 0.39123010635375977, + "learning_rate": 0.000492254912581683, + "loss": 2.939625930786133, + "step": 17730 + }, + { + "epoch": 0.0887, + "grad_norm": 0.39841529726982117, + "learning_rate": 0.000492245112537623, + "loss": 2.988880157470703, + "step": 17740 + }, + { + "epoch": 0.08875, + "grad_norm": 0.37499192357063293, + "learning_rate": 0.0004922353063950438, + "loss": 2.983123207092285, + "step": 17750 + }, + { + "epoch": 0.0888, + "grad_norm": 0.405704528093338, + "learning_rate": 0.0004922254941541924, + "loss": 2.990595245361328, + "step": 17760 + }, + { + "epoch": 0.08885, + "grad_norm": 0.38841286301612854, + "learning_rate": 0.0004922156758153159, + "loss": 3.0095653533935547, + "step": 17770 + }, + { + "epoch": 0.0889, + "grad_norm": 0.4298284947872162, + "learning_rate": 0.0004922058513786613, + "loss": 2.9548786163330076, + "step": 17780 + }, + { + "epoch": 0.08895, + "grad_norm": 0.3870551586151123, + "learning_rate": 0.000492196020844476, + "loss": 3.004525375366211, + "step": 17790 + }, + { + "epoch": 0.089, + "grad_norm": 0.41583871841430664, + "learning_rate": 0.0004921861842130075, + "loss": 2.9970287322998046, + "step": 17800 + }, + { + "epoch": 0.08905, + "grad_norm": 0.42432618141174316, + "learning_rate": 0.0004921763414845035, + "loss": 2.9944084167480467, + "step": 17810 + }, + { + "epoch": 0.0891, + "grad_norm": 0.3945760726928711, + "learning_rate": 0.0004921664926592116, + "loss": 3.0533899307250976, + "step": 17820 + }, + { + "epoch": 0.08915, + "grad_norm": 0.41853660345077515, + "learning_rate": 0.00049215663773738, + "loss": 2.9467374801635744, + "step": 17830 + }, + { + "epoch": 0.0892, + "grad_norm": 0.42245805263519287, + "learning_rate": 0.0004921467767192566, + "loss": 3.0182563781738283, + "step": 17840 + }, + { + "epoch": 0.08925, + "grad_norm": 0.46266183257102966, + "learning_rate": 0.0004921369096050897, + "loss": 2.987392234802246, + "step": 17850 + }, + { + "epoch": 0.0893, + "grad_norm": 0.38459092378616333, + "learning_rate": 0.0004921270363951278, + "loss": 2.967350196838379, + "step": 17860 + }, + { + "epoch": 0.08935, + "grad_norm": 0.4110516905784607, + "learning_rate": 0.0004921171570896193, + "loss": 3.016069030761719, + "step": 17870 + }, + { + "epoch": 0.0894, + "grad_norm": 0.40413907170295715, + "learning_rate": 0.0004921072716888129, + "loss": 2.9457868576049804, + "step": 17880 + }, + { + "epoch": 0.08945, + "grad_norm": 0.43131473660469055, + "learning_rate": 0.0004920973801929578, + "loss": 2.981226348876953, + "step": 17890 + }, + { + "epoch": 0.0895, + "grad_norm": 0.3865971267223358, + "learning_rate": 0.0004920874826023026, + "loss": 3.029834175109863, + "step": 17900 + }, + { + "epoch": 0.08955, + "grad_norm": 0.41990500688552856, + "learning_rate": 0.0004920775789170966, + "loss": 3.0425142288208007, + "step": 17910 + }, + { + "epoch": 0.0896, + "grad_norm": 0.40893077850341797, + "learning_rate": 0.0004920676691375893, + "loss": 3.0135936737060547, + "step": 17920 + }, + { + "epoch": 0.08965, + "grad_norm": 0.4683728814125061, + "learning_rate": 0.00049205775326403, + "loss": 3.008383941650391, + "step": 17930 + }, + { + "epoch": 0.0897, + "grad_norm": 0.37682345509529114, + "learning_rate": 0.0004920478312966683, + "loss": 2.9903472900390624, + "step": 17940 + }, + { + "epoch": 0.08975, + "grad_norm": 0.3855457901954651, + "learning_rate": 0.0004920379032357542, + "loss": 3.015820121765137, + "step": 17950 + }, + { + "epoch": 0.0898, + "grad_norm": 0.3805607259273529, + "learning_rate": 0.0004920279690815374, + "loss": 2.9646188735961916, + "step": 17960 + }, + { + "epoch": 0.08985, + "grad_norm": 0.40327686071395874, + "learning_rate": 0.0004920180288342681, + "loss": 2.9738250732421876, + "step": 17970 + }, + { + "epoch": 0.0899, + "grad_norm": 0.4261651337146759, + "learning_rate": 0.0004920080824941966, + "loss": 2.9936969757080076, + "step": 17980 + }, + { + "epoch": 0.08995, + "grad_norm": 0.39965248107910156, + "learning_rate": 0.0004919981300615732, + "loss": 2.9980243682861327, + "step": 17990 + }, + { + "epoch": 0.09, + "grad_norm": 0.37291571497917175, + "learning_rate": 0.0004919881715366484, + "loss": 3.048062324523926, + "step": 18000 + }, + { + "epoch": 0.09005, + "grad_norm": 0.38934391736984253, + "learning_rate": 0.0004919782069196731, + "loss": 3.0418373107910157, + "step": 18010 + }, + { + "epoch": 0.0901, + "grad_norm": 0.39115452766418457, + "learning_rate": 0.0004919682362108981, + "loss": 2.996799659729004, + "step": 18020 + }, + { + "epoch": 0.09015, + "grad_norm": 0.404714435338974, + "learning_rate": 0.0004919582594105744, + "loss": 2.9656393051147463, + "step": 18030 + }, + { + "epoch": 0.0902, + "grad_norm": 0.44513630867004395, + "learning_rate": 0.000491948276518953, + "loss": 3.015511894226074, + "step": 18040 + }, + { + "epoch": 0.09025, + "grad_norm": 0.47008761763572693, + "learning_rate": 0.0004919382875362854, + "loss": 2.966638374328613, + "step": 18050 + }, + { + "epoch": 0.0903, + "grad_norm": 0.40414896607398987, + "learning_rate": 0.0004919282924628231, + "loss": 3.012514114379883, + "step": 18060 + }, + { + "epoch": 0.09035, + "grad_norm": 0.415902316570282, + "learning_rate": 0.0004919182912988176, + "loss": 3.0080854415893556, + "step": 18070 + }, + { + "epoch": 0.0904, + "grad_norm": 0.42996370792388916, + "learning_rate": 0.0004919082840445208, + "loss": 3.0051286697387694, + "step": 18080 + }, + { + "epoch": 0.09045, + "grad_norm": 0.4032011926174164, + "learning_rate": 0.0004918982707001844, + "loss": 3.000331497192383, + "step": 18090 + }, + { + "epoch": 0.0905, + "grad_norm": 0.40351954102516174, + "learning_rate": 0.0004918882512660607, + "loss": 2.9545690536499025, + "step": 18100 + }, + { + "epoch": 0.09055, + "grad_norm": 0.3843930661678314, + "learning_rate": 0.0004918782257424019, + "loss": 3.0093151092529298, + "step": 18110 + }, + { + "epoch": 0.0906, + "grad_norm": 0.39468011260032654, + "learning_rate": 0.0004918681941294605, + "loss": 2.997275543212891, + "step": 18120 + }, + { + "epoch": 0.09065, + "grad_norm": 0.41943347454071045, + "learning_rate": 0.0004918581564274889, + "loss": 2.9634376525878907, + "step": 18130 + }, + { + "epoch": 0.0907, + "grad_norm": 0.43022438883781433, + "learning_rate": 0.0004918481126367398, + "loss": 2.9822837829589846, + "step": 18140 + }, + { + "epoch": 0.09075, + "grad_norm": 0.40457257628440857, + "learning_rate": 0.000491838062757466, + "loss": 2.9822866439819338, + "step": 18150 + }, + { + "epoch": 0.0908, + "grad_norm": 0.3885253071784973, + "learning_rate": 0.0004918280067899207, + "loss": 2.971511459350586, + "step": 18160 + }, + { + "epoch": 0.09085, + "grad_norm": 0.3854003846645355, + "learning_rate": 0.0004918179447343569, + "loss": 3.0337551116943358, + "step": 18170 + }, + { + "epoch": 0.0909, + "grad_norm": 0.40261343121528625, + "learning_rate": 0.0004918078765910279, + "loss": 3.001726531982422, + "step": 18180 + }, + { + "epoch": 0.09095, + "grad_norm": 0.4398430585861206, + "learning_rate": 0.0004917978023601873, + "loss": 2.968094253540039, + "step": 18190 + }, + { + "epoch": 0.091, + "grad_norm": 0.39529842138290405, + "learning_rate": 0.0004917877220420886, + "loss": 3.0040868759155273, + "step": 18200 + }, + { + "epoch": 0.09105, + "grad_norm": 0.41797688603401184, + "learning_rate": 0.0004917776356369856, + "loss": 3.038773536682129, + "step": 18210 + }, + { + "epoch": 0.0911, + "grad_norm": 0.368235319852829, + "learning_rate": 0.0004917675431451323, + "loss": 3.0480697631835936, + "step": 18220 + }, + { + "epoch": 0.09115, + "grad_norm": 0.4428688585758209, + "learning_rate": 0.0004917574445667827, + "loss": 2.9280349731445314, + "step": 18230 + }, + { + "epoch": 0.0912, + "grad_norm": 0.38213610649108887, + "learning_rate": 0.000491747339902191, + "loss": 3.038812255859375, + "step": 18240 + }, + { + "epoch": 0.09125, + "grad_norm": 0.39291203022003174, + "learning_rate": 0.0004917372291516117, + "loss": 2.966519355773926, + "step": 18250 + }, + { + "epoch": 0.0913, + "grad_norm": 0.40589725971221924, + "learning_rate": 0.0004917271123152992, + "loss": 2.991139221191406, + "step": 18260 + }, + { + "epoch": 0.09135, + "grad_norm": 0.3944251239299774, + "learning_rate": 0.0004917169893935083, + "loss": 2.984934616088867, + "step": 18270 + }, + { + "epoch": 0.0914, + "grad_norm": 0.3897981643676758, + "learning_rate": 0.0004917068603864938, + "loss": 2.987991142272949, + "step": 18280 + }, + { + "epoch": 0.09145, + "grad_norm": 0.37574049830436707, + "learning_rate": 0.0004916967252945107, + "loss": 2.9791627883911134, + "step": 18290 + }, + { + "epoch": 0.0915, + "grad_norm": 0.3890710771083832, + "learning_rate": 0.0004916865841178142, + "loss": 3.019578552246094, + "step": 18300 + }, + { + "epoch": 0.09155, + "grad_norm": 0.36661022901535034, + "learning_rate": 0.0004916764368566595, + "loss": 2.991841125488281, + "step": 18310 + }, + { + "epoch": 0.0916, + "grad_norm": 0.4084882140159607, + "learning_rate": 0.0004916662835113022, + "loss": 2.941826057434082, + "step": 18320 + }, + { + "epoch": 0.09165, + "grad_norm": 0.3709961473941803, + "learning_rate": 0.0004916561240819976, + "loss": 2.999320411682129, + "step": 18330 + }, + { + "epoch": 0.0917, + "grad_norm": 0.4036893844604492, + "learning_rate": 0.0004916459585690019, + "loss": 2.9645723342895507, + "step": 18340 + }, + { + "epoch": 0.09175, + "grad_norm": 0.3924657106399536, + "learning_rate": 0.0004916357869725708, + "loss": 2.9956165313720704, + "step": 18350 + }, + { + "epoch": 0.0918, + "grad_norm": 0.41972780227661133, + "learning_rate": 0.0004916256092929604, + "loss": 3.0226003646850588, + "step": 18360 + }, + { + "epoch": 0.09185, + "grad_norm": 0.4337574541568756, + "learning_rate": 0.0004916154255304268, + "loss": 3.009743499755859, + "step": 18370 + }, + { + "epoch": 0.0919, + "grad_norm": 0.3900679349899292, + "learning_rate": 0.0004916052356852265, + "loss": 3.0439102172851564, + "step": 18380 + }, + { + "epoch": 0.09195, + "grad_norm": 0.4081436097621918, + "learning_rate": 0.0004915950397576161, + "loss": 2.958689308166504, + "step": 18390 + }, + { + "epoch": 0.092, + "grad_norm": 0.39790183305740356, + "learning_rate": 0.0004915848377478521, + "loss": 2.9773317337036134, + "step": 18400 + }, + { + "epoch": 0.09205, + "grad_norm": 0.4190053939819336, + "learning_rate": 0.0004915746296561914, + "loss": 2.9791927337646484, + "step": 18410 + }, + { + "epoch": 0.0921, + "grad_norm": 0.41204601526260376, + "learning_rate": 0.0004915644154828911, + "loss": 2.9643911361694335, + "step": 18420 + }, + { + "epoch": 0.09215, + "grad_norm": 0.5103815197944641, + "learning_rate": 0.0004915541952282082, + "loss": 2.963954544067383, + "step": 18430 + }, + { + "epoch": 0.0922, + "grad_norm": 0.3838765621185303, + "learning_rate": 0.0004915439688924, + "loss": 2.9774723052978516, + "step": 18440 + }, + { + "epoch": 0.09225, + "grad_norm": 0.4183393716812134, + "learning_rate": 0.0004915337364757241, + "loss": 3.0017934799194337, + "step": 18450 + }, + { + "epoch": 0.0923, + "grad_norm": 0.3899684548377991, + "learning_rate": 0.000491523497978438, + "loss": 3.0249990463256835, + "step": 18460 + }, + { + "epoch": 0.09235, + "grad_norm": 0.4032674729824066, + "learning_rate": 0.0004915132534007993, + "loss": 2.9727718353271486, + "step": 18470 + }, + { + "epoch": 0.0924, + "grad_norm": 0.38946646451950073, + "learning_rate": 0.0004915030027430662, + "loss": 2.952444648742676, + "step": 18480 + }, + { + "epoch": 0.09245, + "grad_norm": 0.42766717076301575, + "learning_rate": 0.0004914927460054967, + "loss": 2.9869878768920897, + "step": 18490 + }, + { + "epoch": 0.0925, + "grad_norm": 0.4101198613643646, + "learning_rate": 0.0004914824831883487, + "loss": 2.967674446105957, + "step": 18500 + }, + { + "epoch": 0.09255, + "grad_norm": 0.43074607849121094, + "learning_rate": 0.000491472214291881, + "loss": 3.0085756301879885, + "step": 18510 + }, + { + "epoch": 0.0926, + "grad_norm": 0.40602612495422363, + "learning_rate": 0.0004914619393163517, + "loss": 3.0136129379272463, + "step": 18520 + }, + { + "epoch": 0.09265, + "grad_norm": 0.3938225209712982, + "learning_rate": 0.0004914516582620198, + "loss": 2.9691225051879884, + "step": 18530 + }, + { + "epoch": 0.0927, + "grad_norm": 0.3827350437641144, + "learning_rate": 0.0004914413711291441, + "loss": 2.938970756530762, + "step": 18540 + }, + { + "epoch": 0.09275, + "grad_norm": 0.38338005542755127, + "learning_rate": 0.0004914310779179834, + "loss": 3.0242143630981446, + "step": 18550 + }, + { + "epoch": 0.0928, + "grad_norm": 0.399312287569046, + "learning_rate": 0.0004914207786287971, + "loss": 3.0008867263793944, + "step": 18560 + }, + { + "epoch": 0.09285, + "grad_norm": 0.3935619294643402, + "learning_rate": 0.0004914104732618442, + "loss": 2.94429931640625, + "step": 18570 + }, + { + "epoch": 0.0929, + "grad_norm": 0.3991539180278778, + "learning_rate": 0.0004914001618173842, + "loss": 2.9652395248413086, + "step": 18580 + }, + { + "epoch": 0.09295, + "grad_norm": 0.39422252774238586, + "learning_rate": 0.0004913898442956767, + "loss": 3.0603654861450194, + "step": 18590 + }, + { + "epoch": 0.093, + "grad_norm": 0.39744332432746887, + "learning_rate": 0.0004913795206969815, + "loss": 2.981831741333008, + "step": 18600 + }, + { + "epoch": 0.09305, + "grad_norm": 0.377719521522522, + "learning_rate": 0.0004913691910215586, + "loss": 3.004331588745117, + "step": 18610 + }, + { + "epoch": 0.0931, + "grad_norm": 0.3813129663467407, + "learning_rate": 0.0004913588552696678, + "loss": 2.975150299072266, + "step": 18620 + }, + { + "epoch": 0.09315, + "grad_norm": 0.43309611082077026, + "learning_rate": 0.0004913485134415695, + "loss": 2.9659906387329102, + "step": 18630 + }, + { + "epoch": 0.0932, + "grad_norm": 0.410806804895401, + "learning_rate": 0.000491338165537524, + "loss": 2.96295223236084, + "step": 18640 + }, + { + "epoch": 0.09325, + "grad_norm": 0.4079122543334961, + "learning_rate": 0.0004913278115577917, + "loss": 2.9726701736450196, + "step": 18650 + }, + { + "epoch": 0.0933, + "grad_norm": 0.4110164940357208, + "learning_rate": 0.0004913174515026335, + "loss": 3.0182479858398437, + "step": 18660 + }, + { + "epoch": 0.09335, + "grad_norm": 0.3746974766254425, + "learning_rate": 0.00049130708537231, + "loss": 3.007017135620117, + "step": 18670 + }, + { + "epoch": 0.0934, + "grad_norm": 0.38734012842178345, + "learning_rate": 0.0004912967131670823, + "loss": 3.0476789474487305, + "step": 18680 + }, + { + "epoch": 0.09345, + "grad_norm": 0.407429039478302, + "learning_rate": 0.0004912863348872113, + "loss": 2.9588764190673826, + "step": 18690 + }, + { + "epoch": 0.0935, + "grad_norm": 0.3694862723350525, + "learning_rate": 0.0004912759505329586, + "loss": 2.983528709411621, + "step": 18700 + }, + { + "epoch": 0.09355, + "grad_norm": 0.39440470933914185, + "learning_rate": 0.0004912655601045854, + "loss": 3.0086294174194337, + "step": 18710 + }, + { + "epoch": 0.0936, + "grad_norm": 0.39752596616744995, + "learning_rate": 0.0004912551636023532, + "loss": 2.9710350036621094, + "step": 18720 + }, + { + "epoch": 0.09365, + "grad_norm": 0.4439707398414612, + "learning_rate": 0.0004912447610265241, + "loss": 2.9757854461669924, + "step": 18730 + }, + { + "epoch": 0.0937, + "grad_norm": 0.3853877782821655, + "learning_rate": 0.0004912343523773596, + "loss": 2.9856185913085938, + "step": 18740 + }, + { + "epoch": 0.09375, + "grad_norm": 0.41019850969314575, + "learning_rate": 0.0004912239376551219, + "loss": 2.972187805175781, + "step": 18750 + }, + { + "epoch": 0.0938, + "grad_norm": 0.4183914363384247, + "learning_rate": 0.0004912135168600732, + "loss": 2.986387825012207, + "step": 18760 + }, + { + "epoch": 0.09385, + "grad_norm": 0.39281225204467773, + "learning_rate": 0.0004912030899924759, + "loss": 2.9675518035888673, + "step": 18770 + }, + { + "epoch": 0.0939, + "grad_norm": 0.3819749057292938, + "learning_rate": 0.0004911926570525923, + "loss": 2.960123634338379, + "step": 18780 + }, + { + "epoch": 0.09395, + "grad_norm": 0.40241071581840515, + "learning_rate": 0.0004911822180406853, + "loss": 2.9595239639282225, + "step": 18790 + }, + { + "epoch": 0.094, + "grad_norm": 0.37682175636291504, + "learning_rate": 0.0004911717729570174, + "loss": 2.998984146118164, + "step": 18800 + }, + { + "epoch": 0.09405, + "grad_norm": 0.40527254343032837, + "learning_rate": 0.0004911613218018518, + "loss": 2.9891357421875, + "step": 18810 + }, + { + "epoch": 0.0941, + "grad_norm": 0.3959904611110687, + "learning_rate": 0.0004911508645754517, + "loss": 3.0157953262329102, + "step": 18820 + }, + { + "epoch": 0.09415, + "grad_norm": 0.4172649085521698, + "learning_rate": 0.00049114040127808, + "loss": 2.9116899490356447, + "step": 18830 + }, + { + "epoch": 0.0942, + "grad_norm": 0.3793926537036896, + "learning_rate": 0.0004911299319100003, + "loss": 2.977753829956055, + "step": 18840 + }, + { + "epoch": 0.09425, + "grad_norm": 0.3747255802154541, + "learning_rate": 0.0004911194564714762, + "loss": 3.009552574157715, + "step": 18850 + }, + { + "epoch": 0.0943, + "grad_norm": 0.4023728668689728, + "learning_rate": 0.0004911089749627715, + "loss": 2.9834741592407226, + "step": 18860 + }, + { + "epoch": 0.09435, + "grad_norm": 0.38110074400901794, + "learning_rate": 0.00049109848738415, + "loss": 3.017152786254883, + "step": 18870 + }, + { + "epoch": 0.0944, + "grad_norm": 0.42476993799209595, + "learning_rate": 0.0004910879937358756, + "loss": 2.9816707611083983, + "step": 18880 + }, + { + "epoch": 0.09445, + "grad_norm": 0.4043543338775635, + "learning_rate": 0.0004910774940182125, + "loss": 3.0149526596069336, + "step": 18890 + }, + { + "epoch": 0.0945, + "grad_norm": 0.43235352635383606, + "learning_rate": 0.0004910669882314251, + "loss": 2.973556900024414, + "step": 18900 + }, + { + "epoch": 0.09455, + "grad_norm": 0.3916379511356354, + "learning_rate": 0.000491056476375778, + "loss": 2.953691291809082, + "step": 18910 + }, + { + "epoch": 0.0946, + "grad_norm": 0.3803953230381012, + "learning_rate": 0.0004910459584515357, + "loss": 2.9836523056030275, + "step": 18920 + }, + { + "epoch": 0.09465, + "grad_norm": 0.3943098187446594, + "learning_rate": 0.000491035434458963, + "loss": 3.0307125091552733, + "step": 18930 + }, + { + "epoch": 0.0947, + "grad_norm": 0.38199570775032043, + "learning_rate": 0.0004910249043983247, + "loss": 3.0256635665893556, + "step": 18940 + }, + { + "epoch": 0.09475, + "grad_norm": 0.42568153142929077, + "learning_rate": 0.0004910143682698862, + "loss": 3.0262908935546875, + "step": 18950 + }, + { + "epoch": 0.0948, + "grad_norm": 0.40815219283103943, + "learning_rate": 0.0004910038260739126, + "loss": 3.0066089630126953, + "step": 18960 + }, + { + "epoch": 0.09485, + "grad_norm": 0.3755906820297241, + "learning_rate": 0.0004909932778106692, + "loss": 2.9631044387817385, + "step": 18970 + }, + { + "epoch": 0.0949, + "grad_norm": 0.4077896475791931, + "learning_rate": 0.0004909827234804217, + "loss": 2.910666656494141, + "step": 18980 + }, + { + "epoch": 0.09495, + "grad_norm": 0.3936624825000763, + "learning_rate": 0.0004909721630834358, + "loss": 3.0245832443237304, + "step": 18990 + }, + { + "epoch": 0.095, + "grad_norm": 0.391925185918808, + "learning_rate": 0.0004909615966199771, + "loss": 3.0177118301391603, + "step": 19000 + }, + { + "epoch": 0.09505, + "grad_norm": 0.4142904579639435, + "learning_rate": 0.000490951024090312, + "loss": 2.976351737976074, + "step": 19010 + }, + { + "epoch": 0.0951, + "grad_norm": 0.4174162745475769, + "learning_rate": 0.0004909404454947064, + "loss": 3.0077146530151366, + "step": 19020 + }, + { + "epoch": 0.09515, + "grad_norm": 0.44926539063453674, + "learning_rate": 0.0004909298608334268, + "loss": 3.0333045959472655, + "step": 19030 + }, + { + "epoch": 0.0952, + "grad_norm": 0.38256359100341797, + "learning_rate": 0.0004909192701067394, + "loss": 2.957597351074219, + "step": 19040 + }, + { + "epoch": 0.09525, + "grad_norm": 0.3904925286769867, + "learning_rate": 0.0004909086733149111, + "loss": 2.963376045227051, + "step": 19050 + }, + { + "epoch": 0.0953, + "grad_norm": 0.3996763527393341, + "learning_rate": 0.0004908980704582085, + "loss": 2.97432918548584, + "step": 19060 + }, + { + "epoch": 0.09535, + "grad_norm": 0.37045201659202576, + "learning_rate": 0.0004908874615368986, + "loss": 3.039846420288086, + "step": 19070 + }, + { + "epoch": 0.0954, + "grad_norm": 0.39749854803085327, + "learning_rate": 0.0004908768465512485, + "loss": 3.0327983856201173, + "step": 19080 + }, + { + "epoch": 0.09545, + "grad_norm": 0.3814409673213959, + "learning_rate": 0.0004908662255015254, + "loss": 2.973048973083496, + "step": 19090 + }, + { + "epoch": 0.0955, + "grad_norm": 0.40760335326194763, + "learning_rate": 0.0004908555983879966, + "loss": 2.9828868865966798, + "step": 19100 + }, + { + "epoch": 0.09555, + "grad_norm": 0.414949506521225, + "learning_rate": 0.0004908449652109298, + "loss": 2.975277900695801, + "step": 19110 + }, + { + "epoch": 0.0956, + "grad_norm": 0.3996283710002899, + "learning_rate": 0.0004908343259705926, + "loss": 2.9586788177490235, + "step": 19120 + }, + { + "epoch": 0.09565, + "grad_norm": 0.36682581901550293, + "learning_rate": 0.0004908236806672529, + "loss": 2.926822280883789, + "step": 19130 + }, + { + "epoch": 0.0957, + "grad_norm": 0.40916314721107483, + "learning_rate": 0.0004908130293011786, + "loss": 2.976786994934082, + "step": 19140 + }, + { + "epoch": 0.09575, + "grad_norm": 0.3930196166038513, + "learning_rate": 0.0004908023718726379, + "loss": 2.9508541107177733, + "step": 19150 + }, + { + "epoch": 0.0958, + "grad_norm": 0.3967970311641693, + "learning_rate": 0.000490791708381899, + "loss": 2.9449539184570312, + "step": 19160 + }, + { + "epoch": 0.09585, + "grad_norm": 0.3835717737674713, + "learning_rate": 0.0004907810388292306, + "loss": 3.0133771896362305, + "step": 19170 + }, + { + "epoch": 0.0959, + "grad_norm": 0.39216122031211853, + "learning_rate": 0.0004907703632149009, + "loss": 2.937163162231445, + "step": 19180 + }, + { + "epoch": 0.09595, + "grad_norm": 0.38923653960227966, + "learning_rate": 0.0004907596815391793, + "loss": 2.9490634918212892, + "step": 19190 + }, + { + "epoch": 0.096, + "grad_norm": 0.3991347551345825, + "learning_rate": 0.0004907489938023341, + "loss": 2.958162307739258, + "step": 19200 + }, + { + "epoch": 0.09605, + "grad_norm": 0.3898104131221771, + "learning_rate": 0.0004907383000046346, + "loss": 2.9918180465698243, + "step": 19210 + }, + { + "epoch": 0.0961, + "grad_norm": 0.38518938422203064, + "learning_rate": 0.00049072760014635, + "loss": 2.948746109008789, + "step": 19220 + }, + { + "epoch": 0.09615, + "grad_norm": 0.3697531223297119, + "learning_rate": 0.0004907168942277497, + "loss": 2.959552764892578, + "step": 19230 + }, + { + "epoch": 0.0962, + "grad_norm": 0.39099472761154175, + "learning_rate": 0.0004907061822491032, + "loss": 2.981371307373047, + "step": 19240 + }, + { + "epoch": 0.09625, + "grad_norm": 0.3856882154941559, + "learning_rate": 0.0004906954642106802, + "loss": 2.9881965637207033, + "step": 19250 + }, + { + "epoch": 0.0963, + "grad_norm": 0.38984906673431396, + "learning_rate": 0.0004906847401127504, + "loss": 2.9668521881103516, + "step": 19260 + }, + { + "epoch": 0.09635, + "grad_norm": 0.38104692101478577, + "learning_rate": 0.000490674009955584, + "loss": 3.000883102416992, + "step": 19270 + }, + { + "epoch": 0.0964, + "grad_norm": 0.397260457277298, + "learning_rate": 0.0004906632737394509, + "loss": 2.992604446411133, + "step": 19280 + }, + { + "epoch": 0.09645, + "grad_norm": 0.40780889987945557, + "learning_rate": 0.0004906525314646215, + "loss": 2.9791034698486327, + "step": 19290 + }, + { + "epoch": 0.0965, + "grad_norm": 0.39408230781555176, + "learning_rate": 0.0004906417831313663, + "loss": 2.951760101318359, + "step": 19300 + }, + { + "epoch": 0.09655, + "grad_norm": 0.37632763385772705, + "learning_rate": 0.0004906310287399558, + "loss": 2.9673440933227537, + "step": 19310 + }, + { + "epoch": 0.0966, + "grad_norm": 0.37579309940338135, + "learning_rate": 0.0004906202682906607, + "loss": 2.9782684326171873, + "step": 19320 + }, + { + "epoch": 0.09665, + "grad_norm": 0.4097176790237427, + "learning_rate": 0.0004906095017837521, + "loss": 2.9465915679931642, + "step": 19330 + }, + { + "epoch": 0.0967, + "grad_norm": 0.422818660736084, + "learning_rate": 0.0004905987292195007, + "loss": 2.9591306686401366, + "step": 19340 + }, + { + "epoch": 0.09675, + "grad_norm": 0.4140697121620178, + "learning_rate": 0.0004905879505981781, + "loss": 2.9982009887695313, + "step": 19350 + }, + { + "epoch": 0.0968, + "grad_norm": 0.39838728308677673, + "learning_rate": 0.0004905771659200553, + "loss": 3.006623458862305, + "step": 19360 + }, + { + "epoch": 0.09685, + "grad_norm": 0.3867175877094269, + "learning_rate": 0.000490566375185404, + "loss": 2.988350677490234, + "step": 19370 + }, + { + "epoch": 0.0969, + "grad_norm": 0.382048636674881, + "learning_rate": 0.0004905555783944959, + "loss": 2.9400619506835937, + "step": 19380 + }, + { + "epoch": 0.09695, + "grad_norm": 0.39138567447662354, + "learning_rate": 0.0004905447755476026, + "loss": 2.9332738876342774, + "step": 19390 + }, + { + "epoch": 0.097, + "grad_norm": 0.3962591588497162, + "learning_rate": 0.0004905339666449962, + "loss": 3.0045372009277345, + "step": 19400 + }, + { + "epoch": 0.09705, + "grad_norm": 0.4030730724334717, + "learning_rate": 0.0004905231516869488, + "loss": 3.0025087356567384, + "step": 19410 + }, + { + "epoch": 0.0971, + "grad_norm": 0.40323394536972046, + "learning_rate": 0.0004905123306737327, + "loss": 2.9731267929077148, + "step": 19420 + }, + { + "epoch": 0.09715, + "grad_norm": 0.3881135582923889, + "learning_rate": 0.0004905015036056203, + "loss": 2.950571823120117, + "step": 19430 + }, + { + "epoch": 0.0972, + "grad_norm": 0.38832414150238037, + "learning_rate": 0.0004904906704828841, + "loss": 2.993998908996582, + "step": 19440 + }, + { + "epoch": 0.09725, + "grad_norm": 0.40437671542167664, + "learning_rate": 0.0004904798313057969, + "loss": 2.993727111816406, + "step": 19450 + }, + { + "epoch": 0.0973, + "grad_norm": 0.39097392559051514, + "learning_rate": 0.0004904689860746315, + "loss": 2.935964012145996, + "step": 19460 + }, + { + "epoch": 0.09735, + "grad_norm": 0.4031628668308258, + "learning_rate": 0.000490458134789661, + "loss": 2.981618309020996, + "step": 19470 + }, + { + "epoch": 0.0974, + "grad_norm": 0.40032529830932617, + "learning_rate": 0.0004904472774511587, + "loss": 2.9553300857543947, + "step": 19480 + }, + { + "epoch": 0.09745, + "grad_norm": 0.396975576877594, + "learning_rate": 0.0004904364140593977, + "loss": 2.965633773803711, + "step": 19490 + }, + { + "epoch": 0.0975, + "grad_norm": 0.40496182441711426, + "learning_rate": 0.0004904255446146515, + "loss": 2.9372879028320313, + "step": 19500 + }, + { + "epoch": 0.09755, + "grad_norm": 0.3880067765712738, + "learning_rate": 0.0004904146691171939, + "loss": 2.9094032287597655, + "step": 19510 + }, + { + "epoch": 0.0976, + "grad_norm": 0.41682863235473633, + "learning_rate": 0.0004904037875672986, + "loss": 2.9920440673828126, + "step": 19520 + }, + { + "epoch": 0.09765, + "grad_norm": 0.37528547644615173, + "learning_rate": 0.0004903928999652396, + "loss": 2.9809059143066405, + "step": 19530 + }, + { + "epoch": 0.0977, + "grad_norm": 0.3911088705062866, + "learning_rate": 0.0004903820063112909, + "loss": 2.996359443664551, + "step": 19540 + }, + { + "epoch": 0.09775, + "grad_norm": 0.3862791061401367, + "learning_rate": 0.0004903711066057268, + "loss": 3.021472930908203, + "step": 19550 + }, + { + "epoch": 0.0978, + "grad_norm": 0.4281257092952728, + "learning_rate": 0.0004903602008488217, + "loss": 2.9633403778076173, + "step": 19560 + }, + { + "epoch": 0.09785, + "grad_norm": 0.3906274437904358, + "learning_rate": 0.0004903492890408502, + "loss": 2.9495540618896485, + "step": 19570 + }, + { + "epoch": 0.0979, + "grad_norm": 0.40768060088157654, + "learning_rate": 0.0004903383711820869, + "loss": 2.988279342651367, + "step": 19580 + }, + { + "epoch": 0.09795, + "grad_norm": 0.3895949423313141, + "learning_rate": 0.0004903274472728067, + "loss": 2.966048240661621, + "step": 19590 + }, + { + "epoch": 0.098, + "grad_norm": 0.425809383392334, + "learning_rate": 0.0004903165173132846, + "loss": 2.9570075988769533, + "step": 19600 + }, + { + "epoch": 0.09805, + "grad_norm": 0.4019777476787567, + "learning_rate": 0.0004903055813037959, + "loss": 2.949747085571289, + "step": 19610 + }, + { + "epoch": 0.0981, + "grad_norm": 0.45198366045951843, + "learning_rate": 0.0004902946392446157, + "loss": 2.974771499633789, + "step": 19620 + }, + { + "epoch": 0.09815, + "grad_norm": 0.4165000915527344, + "learning_rate": 0.0004902836911360196, + "loss": 2.9366188049316406, + "step": 19630 + }, + { + "epoch": 0.0982, + "grad_norm": 0.4066777229309082, + "learning_rate": 0.0004902727369782831, + "loss": 3.0289447784423826, + "step": 19640 + }, + { + "epoch": 0.09825, + "grad_norm": 0.3914802372455597, + "learning_rate": 0.0004902617767716821, + "loss": 2.9780706405639648, + "step": 19650 + }, + { + "epoch": 0.0983, + "grad_norm": 0.3934268653392792, + "learning_rate": 0.0004902508105164926, + "loss": 2.9779333114624023, + "step": 19660 + }, + { + "epoch": 0.09835, + "grad_norm": 0.38340210914611816, + "learning_rate": 0.0004902398382129904, + "loss": 2.978034019470215, + "step": 19670 + }, + { + "epoch": 0.0984, + "grad_norm": 0.4126887917518616, + "learning_rate": 0.0004902288598614519, + "loss": 2.9518320083618166, + "step": 19680 + }, + { + "epoch": 0.09845, + "grad_norm": 0.4056738615036011, + "learning_rate": 0.0004902178754621534, + "loss": 3.0234336853027344, + "step": 19690 + }, + { + "epoch": 0.0985, + "grad_norm": 0.4025857150554657, + "learning_rate": 0.0004902068850153717, + "loss": 3.0067588806152346, + "step": 19700 + }, + { + "epoch": 0.09855, + "grad_norm": 0.3746219873428345, + "learning_rate": 0.0004901958885213831, + "loss": 2.979646110534668, + "step": 19710 + }, + { + "epoch": 0.0986, + "grad_norm": 0.4186365604400635, + "learning_rate": 0.0004901848859804646, + "loss": 2.9659618377685546, + "step": 19720 + }, + { + "epoch": 0.09865, + "grad_norm": 0.3995031416416168, + "learning_rate": 0.0004901738773928934, + "loss": 2.959844970703125, + "step": 19730 + }, + { + "epoch": 0.0987, + "grad_norm": 0.3914014995098114, + "learning_rate": 0.0004901628627589462, + "loss": 2.957374382019043, + "step": 19740 + }, + { + "epoch": 0.09875, + "grad_norm": 0.39196091890335083, + "learning_rate": 0.0004901518420789006, + "loss": 3.0103128433227537, + "step": 19750 + }, + { + "epoch": 0.0988, + "grad_norm": 0.4024708867073059, + "learning_rate": 0.000490140815353034, + "loss": 2.9846357345581054, + "step": 19760 + }, + { + "epoch": 0.09885, + "grad_norm": 0.4081383943557739, + "learning_rate": 0.000490129782581624, + "loss": 2.95188045501709, + "step": 19770 + }, + { + "epoch": 0.0989, + "grad_norm": 0.3541073799133301, + "learning_rate": 0.0004901187437649484, + "loss": 2.9810501098632813, + "step": 19780 + }, + { + "epoch": 0.09895, + "grad_norm": 0.44275572896003723, + "learning_rate": 0.0004901076989032849, + "loss": 2.947813034057617, + "step": 19790 + }, + { + "epoch": 0.099, + "grad_norm": 0.3689028322696686, + "learning_rate": 0.0004900966479969117, + "loss": 2.9474563598632812, + "step": 19800 + }, + { + "epoch": 0.09905, + "grad_norm": 0.3874076008796692, + "learning_rate": 0.000490085591046107, + "loss": 2.991294097900391, + "step": 19810 + }, + { + "epoch": 0.0991, + "grad_norm": 0.39967960119247437, + "learning_rate": 0.0004900745280511493, + "loss": 2.9413663864135744, + "step": 19820 + }, + { + "epoch": 0.09915, + "grad_norm": 0.41277262568473816, + "learning_rate": 0.0004900634590123168, + "loss": 2.9940723419189452, + "step": 19830 + }, + { + "epoch": 0.0992, + "grad_norm": 0.4144029915332794, + "learning_rate": 0.0004900523839298884, + "loss": 2.9978260040283202, + "step": 19840 + }, + { + "epoch": 0.09925, + "grad_norm": 0.44369032979011536, + "learning_rate": 0.0004900413028041427, + "loss": 2.9950075149536133, + "step": 19850 + }, + { + "epoch": 0.0993, + "grad_norm": 0.41391798853874207, + "learning_rate": 0.000490030215635359, + "loss": 2.9834800720214845, + "step": 19860 + }, + { + "epoch": 0.09935, + "grad_norm": 0.4419083893299103, + "learning_rate": 0.000490019122423816, + "loss": 2.983339309692383, + "step": 19870 + }, + { + "epoch": 0.0994, + "grad_norm": 0.39569276571273804, + "learning_rate": 0.0004900080231697935, + "loss": 2.961771774291992, + "step": 19880 + }, + { + "epoch": 0.09945, + "grad_norm": 0.4047808349132538, + "learning_rate": 0.0004899969178735704, + "loss": 2.9922624588012696, + "step": 19890 + }, + { + "epoch": 0.0995, + "grad_norm": 0.3924248516559601, + "learning_rate": 0.0004899858065354265, + "loss": 2.9127595901489256, + "step": 19900 + }, + { + "epoch": 0.09955, + "grad_norm": 0.36505556106567383, + "learning_rate": 0.0004899746891556416, + "loss": 2.969791030883789, + "step": 19910 + }, + { + "epoch": 0.0996, + "grad_norm": 0.4115752875804901, + "learning_rate": 0.0004899635657344954, + "loss": 3.0040660858154298, + "step": 19920 + }, + { + "epoch": 0.09965, + "grad_norm": 0.40807127952575684, + "learning_rate": 0.0004899524362722682, + "loss": 2.889766311645508, + "step": 19930 + }, + { + "epoch": 0.0997, + "grad_norm": 0.3604501187801361, + "learning_rate": 0.0004899413007692398, + "loss": 2.9416473388671873, + "step": 19940 + }, + { + "epoch": 0.09975, + "grad_norm": 0.37806475162506104, + "learning_rate": 0.000489930159225691, + "loss": 2.959506034851074, + "step": 19950 + }, + { + "epoch": 0.0998, + "grad_norm": 0.38627564907073975, + "learning_rate": 0.0004899190116419019, + "loss": 2.9557071685791017, + "step": 19960 + }, + { + "epoch": 0.09985, + "grad_norm": 0.3712664246559143, + "learning_rate": 0.0004899078580181533, + "loss": 3.0005107879638673, + "step": 19970 + }, + { + "epoch": 0.0999, + "grad_norm": 0.3781311810016632, + "learning_rate": 0.0004898966983547259, + "loss": 2.976605224609375, + "step": 19980 + }, + { + "epoch": 0.09995, + "grad_norm": 0.4015108644962311, + "learning_rate": 0.0004898855326519009, + "loss": 2.968572425842285, + "step": 19990 + }, + { + "epoch": 0.1, + "grad_norm": 0.42435985803604126, + "learning_rate": 0.0004898743609099591, + "loss": 2.9779685974121093, + "step": 20000 + }, + { + "epoch": 0.10005, + "grad_norm": 0.39682188630104065, + "learning_rate": 0.000489863183129182, + "loss": 2.9327795028686525, + "step": 20010 + }, + { + "epoch": 0.1001, + "grad_norm": 0.39524394273757935, + "learning_rate": 0.0004898519993098507, + "loss": 3.0287891387939454, + "step": 20020 + }, + { + "epoch": 0.10015, + "grad_norm": 0.3983134627342224, + "learning_rate": 0.0004898408094522469, + "loss": 2.959438133239746, + "step": 20030 + }, + { + "epoch": 0.1002, + "grad_norm": 0.43169671297073364, + "learning_rate": 0.0004898296135566524, + "loss": 2.9471006393432617, + "step": 20040 + }, + { + "epoch": 0.10025, + "grad_norm": 0.37834474444389343, + "learning_rate": 0.0004898184116233491, + "loss": 2.9703922271728516, + "step": 20050 + }, + { + "epoch": 0.1003, + "grad_norm": 0.38258305191993713, + "learning_rate": 0.0004898072036526187, + "loss": 2.943067741394043, + "step": 20060 + }, + { + "epoch": 0.10035, + "grad_norm": 0.3812789022922516, + "learning_rate": 0.0004897959896447437, + "loss": 2.9764387130737306, + "step": 20070 + }, + { + "epoch": 0.1004, + "grad_norm": 0.374642014503479, + "learning_rate": 0.0004897847696000062, + "loss": 2.956574058532715, + "step": 20080 + }, + { + "epoch": 0.10045, + "grad_norm": 0.38133659958839417, + "learning_rate": 0.0004897735435186887, + "loss": 2.9550592422485353, + "step": 20090 + }, + { + "epoch": 0.1005, + "grad_norm": 0.390699565410614, + "learning_rate": 0.0004897623114010739, + "loss": 2.9822641372680665, + "step": 20100 + }, + { + "epoch": 0.10055, + "grad_norm": 0.39921697974205017, + "learning_rate": 0.0004897510732474444, + "loss": 3.029366302490234, + "step": 20110 + }, + { + "epoch": 0.1006, + "grad_norm": 0.4191201329231262, + "learning_rate": 0.0004897398290580834, + "loss": 2.982554626464844, + "step": 20120 + }, + { + "epoch": 0.10065, + "grad_norm": 0.41726958751678467, + "learning_rate": 0.0004897285788332737, + "loss": 2.9553497314453123, + "step": 20130 + }, + { + "epoch": 0.1007, + "grad_norm": 0.3848062753677368, + "learning_rate": 0.0004897173225732986, + "loss": 2.9639907836914063, + "step": 20140 + }, + { + "epoch": 0.10075, + "grad_norm": 0.41073381900787354, + "learning_rate": 0.0004897060602784415, + "loss": 2.991826057434082, + "step": 20150 + }, + { + "epoch": 0.1008, + "grad_norm": 0.38719645142555237, + "learning_rate": 0.000489694791948986, + "loss": 2.9622350692749024, + "step": 20160 + }, + { + "epoch": 0.10085, + "grad_norm": 0.39311179518699646, + "learning_rate": 0.0004896835175852158, + "loss": 2.992462158203125, + "step": 20170 + }, + { + "epoch": 0.1009, + "grad_norm": 0.35227981209754944, + "learning_rate": 0.0004896722371874145, + "loss": 2.9858070373535157, + "step": 20180 + }, + { + "epoch": 0.10095, + "grad_norm": 0.4083103835582733, + "learning_rate": 0.0004896609507558662, + "loss": 3.0159233093261717, + "step": 20190 + }, + { + "epoch": 0.101, + "grad_norm": 0.3757365942001343, + "learning_rate": 0.0004896496582908552, + "loss": 2.9733808517456053, + "step": 20200 + }, + { + "epoch": 0.10105, + "grad_norm": 0.4185107946395874, + "learning_rate": 0.0004896383597926656, + "loss": 2.953194427490234, + "step": 20210 + }, + { + "epoch": 0.1011, + "grad_norm": 0.39406901597976685, + "learning_rate": 0.000489627055261582, + "loss": 2.9097986221313477, + "step": 20220 + }, + { + "epoch": 0.10115, + "grad_norm": 0.3987748324871063, + "learning_rate": 0.0004896157446978887, + "loss": 2.923099899291992, + "step": 20230 + }, + { + "epoch": 0.1012, + "grad_norm": 0.42074844241142273, + "learning_rate": 0.0004896044281018707, + "loss": 2.9660953521728515, + "step": 20240 + }, + { + "epoch": 0.10125, + "grad_norm": 0.3616711497306824, + "learning_rate": 0.0004895931054738128, + "loss": 2.9509143829345703, + "step": 20250 + }, + { + "epoch": 0.1013, + "grad_norm": 0.4172412157058716, + "learning_rate": 0.0004895817768140002, + "loss": 2.964451026916504, + "step": 20260 + }, + { + "epoch": 0.10135, + "grad_norm": 0.38361048698425293, + "learning_rate": 0.0004895704421227178, + "loss": 2.9665393829345703, + "step": 20270 + }, + { + "epoch": 0.1014, + "grad_norm": 0.41213080286979675, + "learning_rate": 0.0004895591014002511, + "loss": 2.942448043823242, + "step": 20280 + }, + { + "epoch": 0.10145, + "grad_norm": 0.41476377844810486, + "learning_rate": 0.0004895477546468858, + "loss": 2.9384521484375, + "step": 20290 + }, + { + "epoch": 0.1015, + "grad_norm": 0.4072470963001251, + "learning_rate": 0.0004895364018629073, + "loss": 3.002876091003418, + "step": 20300 + }, + { + "epoch": 0.10155, + "grad_norm": 0.39756664633750916, + "learning_rate": 0.0004895250430486015, + "loss": 3.0363426208496094, + "step": 20310 + }, + { + "epoch": 0.1016, + "grad_norm": 0.42899423837661743, + "learning_rate": 0.0004895136782042543, + "loss": 3.003566932678223, + "step": 20320 + }, + { + "epoch": 0.10165, + "grad_norm": 0.40758395195007324, + "learning_rate": 0.0004895023073301518, + "loss": 2.9608524322509764, + "step": 20330 + }, + { + "epoch": 0.1017, + "grad_norm": 0.35966193675994873, + "learning_rate": 0.0004894909304265805, + "loss": 2.9731550216674805, + "step": 20340 + }, + { + "epoch": 0.10175, + "grad_norm": 0.3729173243045807, + "learning_rate": 0.0004894795474938264, + "loss": 3.0460430145263673, + "step": 20350 + }, + { + "epoch": 0.1018, + "grad_norm": 0.371165007352829, + "learning_rate": 0.0004894681585321764, + "loss": 2.9637290954589846, + "step": 20360 + }, + { + "epoch": 0.10185, + "grad_norm": 0.3857811689376831, + "learning_rate": 0.0004894567635419172, + "loss": 2.950530433654785, + "step": 20370 + }, + { + "epoch": 0.1019, + "grad_norm": 0.4027003347873688, + "learning_rate": 0.0004894453625233354, + "loss": 3.0137176513671875, + "step": 20380 + }, + { + "epoch": 0.10195, + "grad_norm": 0.40145841240882874, + "learning_rate": 0.0004894339554767183, + "loss": 2.987253761291504, + "step": 20390 + }, + { + "epoch": 0.102, + "grad_norm": 0.48178330063819885, + "learning_rate": 0.000489422542402353, + "loss": 2.948137664794922, + "step": 20400 + }, + { + "epoch": 0.10205, + "grad_norm": 0.4037582576274872, + "learning_rate": 0.0004894111233005267, + "loss": 2.9978107452392577, + "step": 20410 + }, + { + "epoch": 0.1021, + "grad_norm": 0.39000728726387024, + "learning_rate": 0.0004893996981715271, + "loss": 2.984455871582031, + "step": 20420 + }, + { + "epoch": 0.10215, + "grad_norm": 0.422165185213089, + "learning_rate": 0.0004893882670156416, + "loss": 2.937836456298828, + "step": 20430 + }, + { + "epoch": 0.1022, + "grad_norm": 0.4187389016151428, + "learning_rate": 0.000489376829833158, + "loss": 2.9458118438720704, + "step": 20440 + }, + { + "epoch": 0.10225, + "grad_norm": 0.385634183883667, + "learning_rate": 0.0004893653866243644, + "loss": 3.019972801208496, + "step": 20450 + }, + { + "epoch": 0.1023, + "grad_norm": 0.39040181040763855, + "learning_rate": 0.0004893539373895488, + "loss": 2.9124135971069336, + "step": 20460 + }, + { + "epoch": 0.10235, + "grad_norm": 0.3889082074165344, + "learning_rate": 0.0004893424821289994, + "loss": 2.9342844009399416, + "step": 20470 + }, + { + "epoch": 0.1024, + "grad_norm": 0.3914584517478943, + "learning_rate": 0.0004893310208430047, + "loss": 2.923242950439453, + "step": 20480 + }, + { + "epoch": 0.10245, + "grad_norm": 0.4013681411743164, + "learning_rate": 0.000489319553531853, + "loss": 2.960580253601074, + "step": 20490 + }, + { + "epoch": 0.1025, + "grad_norm": 0.384805828332901, + "learning_rate": 0.0004893080801958333, + "loss": 2.9432722091674806, + "step": 20500 + }, + { + "epoch": 0.10255, + "grad_norm": 0.35287225246429443, + "learning_rate": 0.0004892966008352342, + "loss": 2.962895965576172, + "step": 20510 + }, + { + "epoch": 0.1026, + "grad_norm": 0.380937784910202, + "learning_rate": 0.0004892851154503448, + "loss": 2.909294319152832, + "step": 20520 + }, + { + "epoch": 0.10265, + "grad_norm": 0.3719380795955658, + "learning_rate": 0.0004892736240414542, + "loss": 2.9623008728027345, + "step": 20530 + }, + { + "epoch": 0.1027, + "grad_norm": 0.4022459387779236, + "learning_rate": 0.0004892621266088516, + "loss": 2.991512489318848, + "step": 20540 + }, + { + "epoch": 0.10275, + "grad_norm": 0.4024989604949951, + "learning_rate": 0.0004892506231528268, + "loss": 2.944444274902344, + "step": 20550 + }, + { + "epoch": 0.1028, + "grad_norm": 0.3824109435081482, + "learning_rate": 0.0004892391136736689, + "loss": 2.954618453979492, + "step": 20560 + }, + { + "epoch": 0.10285, + "grad_norm": 0.41220739483833313, + "learning_rate": 0.0004892275981716681, + "loss": 2.9455032348632812, + "step": 20570 + }, + { + "epoch": 0.1029, + "grad_norm": 0.4314872920513153, + "learning_rate": 0.0004892160766471141, + "loss": 2.973681831359863, + "step": 20580 + }, + { + "epoch": 0.10295, + "grad_norm": 0.3885858654975891, + "learning_rate": 0.0004892045491002968, + "loss": 2.9467016220092774, + "step": 20590 + }, + { + "epoch": 0.103, + "grad_norm": 0.3951936960220337, + "learning_rate": 0.0004891930155315067, + "loss": 2.9564619064331055, + "step": 20600 + }, + { + "epoch": 0.10305, + "grad_norm": 0.3978336751461029, + "learning_rate": 0.0004891814759410341, + "loss": 3.0157981872558595, + "step": 20610 + }, + { + "epoch": 0.1031, + "grad_norm": 0.3826766312122345, + "learning_rate": 0.0004891699303291693, + "loss": 2.9833093643188477, + "step": 20620 + }, + { + "epoch": 0.10315, + "grad_norm": 0.388771116733551, + "learning_rate": 0.0004891583786962033, + "loss": 2.9781730651855467, + "step": 20630 + }, + { + "epoch": 0.1032, + "grad_norm": 0.4006737172603607, + "learning_rate": 0.0004891468210424265, + "loss": 2.977894592285156, + "step": 20640 + }, + { + "epoch": 0.10325, + "grad_norm": 0.43237870931625366, + "learning_rate": 0.0004891352573681302, + "loss": 2.8839847564697267, + "step": 20650 + }, + { + "epoch": 0.1033, + "grad_norm": 0.38946038484573364, + "learning_rate": 0.0004891236876736054, + "loss": 3.003225326538086, + "step": 20660 + }, + { + "epoch": 0.10335, + "grad_norm": 0.3754189908504486, + "learning_rate": 0.0004891121119591433, + "loss": 2.9886152267456056, + "step": 20670 + }, + { + "epoch": 0.1034, + "grad_norm": 0.4202742576599121, + "learning_rate": 0.0004891005302250354, + "loss": 2.993182373046875, + "step": 20680 + }, + { + "epoch": 0.10345, + "grad_norm": 0.41362929344177246, + "learning_rate": 0.0004890889424715732, + "loss": 2.986987495422363, + "step": 20690 + }, + { + "epoch": 0.1035, + "grad_norm": 0.39901265501976013, + "learning_rate": 0.0004890773486990486, + "loss": 2.955605316162109, + "step": 20700 + }, + { + "epoch": 0.10355, + "grad_norm": 0.42339152097702026, + "learning_rate": 0.0004890657489077532, + "loss": 2.965118408203125, + "step": 20710 + }, + { + "epoch": 0.1036, + "grad_norm": 0.3959466516971588, + "learning_rate": 0.0004890541430979793, + "loss": 2.975650596618652, + "step": 20720 + }, + { + "epoch": 0.10365, + "grad_norm": 0.3771475553512573, + "learning_rate": 0.0004890425312700188, + "loss": 2.9265432357788086, + "step": 20730 + }, + { + "epoch": 0.1037, + "grad_norm": 0.38924840092658997, + "learning_rate": 0.0004890309134241643, + "loss": 2.992650032043457, + "step": 20740 + }, + { + "epoch": 0.10375, + "grad_norm": 0.38611963391304016, + "learning_rate": 0.000489019289560708, + "loss": 3.0349052429199217, + "step": 20750 + }, + { + "epoch": 0.1038, + "grad_norm": 0.3922240138053894, + "learning_rate": 0.0004890076596799429, + "loss": 2.947718048095703, + "step": 20760 + }, + { + "epoch": 0.10385, + "grad_norm": 0.3893890976905823, + "learning_rate": 0.0004889960237821614, + "loss": 2.9464031219482423, + "step": 20770 + }, + { + "epoch": 0.1039, + "grad_norm": 0.3605366349220276, + "learning_rate": 0.0004889843818676567, + "loss": 2.959461212158203, + "step": 20780 + }, + { + "epoch": 0.10395, + "grad_norm": 0.37678974866867065, + "learning_rate": 0.0004889727339367217, + "loss": 2.944121551513672, + "step": 20790 + }, + { + "epoch": 0.104, + "grad_norm": 0.3674085736274719, + "learning_rate": 0.0004889610799896498, + "loss": 2.9310634613037108, + "step": 20800 + }, + { + "epoch": 0.10405, + "grad_norm": 0.3860985040664673, + "learning_rate": 0.0004889494200267342, + "loss": 2.984465789794922, + "step": 20810 + }, + { + "epoch": 0.1041, + "grad_norm": 0.4493478238582611, + "learning_rate": 0.0004889377540482685, + "loss": 2.9944284439086912, + "step": 20820 + }, + { + "epoch": 0.10415, + "grad_norm": 0.38786405324935913, + "learning_rate": 0.0004889260820545466, + "loss": 2.9815210342407226, + "step": 20830 + }, + { + "epoch": 0.1042, + "grad_norm": 0.399197518825531, + "learning_rate": 0.0004889144040458622, + "loss": 2.9951858520507812, + "step": 20840 + }, + { + "epoch": 0.10425, + "grad_norm": 0.3915381133556366, + "learning_rate": 0.0004889027200225091, + "loss": 3.0038633346557617, + "step": 20850 + }, + { + "epoch": 0.1043, + "grad_norm": 0.3994583487510681, + "learning_rate": 0.0004888910299847818, + "loss": 2.996342086791992, + "step": 20860 + }, + { + "epoch": 0.10435, + "grad_norm": 0.3946007490158081, + "learning_rate": 0.0004888793339329744, + "loss": 2.9958894729614256, + "step": 20870 + }, + { + "epoch": 0.1044, + "grad_norm": 0.4150010347366333, + "learning_rate": 0.0004888676318673813, + "loss": 2.999541091918945, + "step": 20880 + }, + { + "epoch": 0.10445, + "grad_norm": 0.3786386549472809, + "learning_rate": 0.0004888559237882971, + "loss": 2.954795265197754, + "step": 20890 + }, + { + "epoch": 0.1045, + "grad_norm": 0.42667296528816223, + "learning_rate": 0.0004888442096960166, + "loss": 2.9535240173339843, + "step": 20900 + }, + { + "epoch": 0.10455, + "grad_norm": 0.39785662293434143, + "learning_rate": 0.0004888324895908349, + "loss": 2.957077407836914, + "step": 20910 + }, + { + "epoch": 0.1046, + "grad_norm": 0.38371527194976807, + "learning_rate": 0.0004888207634730467, + "loss": 2.98031063079834, + "step": 20920 + }, + { + "epoch": 0.10465, + "grad_norm": 0.3893054723739624, + "learning_rate": 0.0004888090313429474, + "loss": 2.9707273483276366, + "step": 20930 + }, + { + "epoch": 0.1047, + "grad_norm": 0.3922124207019806, + "learning_rate": 0.0004887972932008324, + "loss": 2.9364946365356444, + "step": 20940 + }, + { + "epoch": 0.10475, + "grad_norm": 0.40581828355789185, + "learning_rate": 0.0004887855490469971, + "loss": 2.939897346496582, + "step": 20950 + }, + { + "epoch": 0.1048, + "grad_norm": 0.411964476108551, + "learning_rate": 0.0004887737988817372, + "loss": 2.994437408447266, + "step": 20960 + }, + { + "epoch": 0.10485, + "grad_norm": 0.41957199573516846, + "learning_rate": 0.0004887620427053486, + "loss": 2.9818111419677735, + "step": 20970 + }, + { + "epoch": 0.1049, + "grad_norm": 0.39872968196868896, + "learning_rate": 0.0004887502805181271, + "loss": 2.973933792114258, + "step": 20980 + }, + { + "epoch": 0.10495, + "grad_norm": 0.3733173906803131, + "learning_rate": 0.0004887385123203688, + "loss": 2.9714788436889648, + "step": 20990 + }, + { + "epoch": 0.105, + "grad_norm": 0.40305814146995544, + "learning_rate": 0.0004887267381123701, + "loss": 3.025537300109863, + "step": 21000 + }, + { + "epoch": 0.10505, + "grad_norm": 0.3716965913772583, + "learning_rate": 0.0004887149578944274, + "loss": 2.9487762451171875, + "step": 21010 + }, + { + "epoch": 0.1051, + "grad_norm": 0.3723551034927368, + "learning_rate": 0.0004887031716668373, + "loss": 2.9825460433959963, + "step": 21020 + }, + { + "epoch": 0.10515, + "grad_norm": 0.4108482003211975, + "learning_rate": 0.0004886913794298963, + "loss": 3.0087749481201174, + "step": 21030 + }, + { + "epoch": 0.1052, + "grad_norm": 0.42818108201026917, + "learning_rate": 0.0004886795811839016, + "loss": 3.0135900497436525, + "step": 21040 + }, + { + "epoch": 0.10525, + "grad_norm": 0.4110470414161682, + "learning_rate": 0.0004886677769291498, + "loss": 3.018907165527344, + "step": 21050 + }, + { + "epoch": 0.1053, + "grad_norm": 0.3972264230251312, + "learning_rate": 0.0004886559666659384, + "loss": 2.967942810058594, + "step": 21060 + }, + { + "epoch": 0.10535, + "grad_norm": 0.4051559865474701, + "learning_rate": 0.0004886441503945647, + "loss": 2.9748149871826173, + "step": 21070 + }, + { + "epoch": 0.1054, + "grad_norm": 0.39241841435432434, + "learning_rate": 0.000488632328115326, + "loss": 3.014853096008301, + "step": 21080 + }, + { + "epoch": 0.10545, + "grad_norm": 0.36562609672546387, + "learning_rate": 0.0004886204998285202, + "loss": 2.9841379165649413, + "step": 21090 + }, + { + "epoch": 0.1055, + "grad_norm": 0.401610404253006, + "learning_rate": 0.0004886086655344448, + "loss": 2.9908683776855467, + "step": 21100 + }, + { + "epoch": 0.10555, + "grad_norm": 0.38145124912261963, + "learning_rate": 0.0004885968252333979, + "loss": 2.9305046081542967, + "step": 21110 + }, + { + "epoch": 0.1056, + "grad_norm": 0.39042288064956665, + "learning_rate": 0.0004885849789256774, + "loss": 2.930769920349121, + "step": 21120 + }, + { + "epoch": 0.10565, + "grad_norm": 0.39685070514678955, + "learning_rate": 0.0004885731266115817, + "loss": 3.000294303894043, + "step": 21130 + }, + { + "epoch": 0.1057, + "grad_norm": 0.40303993225097656, + "learning_rate": 0.0004885612682914093, + "loss": 2.8775169372558596, + "step": 21140 + }, + { + "epoch": 0.10575, + "grad_norm": 0.36372116208076477, + "learning_rate": 0.0004885494039654583, + "loss": 2.9807903289794924, + "step": 21150 + }, + { + "epoch": 0.1058, + "grad_norm": 0.37611624598503113, + "learning_rate": 0.0004885375336340279, + "loss": 2.916712760925293, + "step": 21160 + }, + { + "epoch": 0.10585, + "grad_norm": 0.37016594409942627, + "learning_rate": 0.0004885256572974165, + "loss": 2.96059627532959, + "step": 21170 + }, + { + "epoch": 0.1059, + "grad_norm": 0.4070974588394165, + "learning_rate": 0.0004885137749559233, + "loss": 2.9176555633544923, + "step": 21180 + }, + { + "epoch": 0.10595, + "grad_norm": 0.3740384578704834, + "learning_rate": 0.0004885018866098474, + "loss": 2.965140151977539, + "step": 21190 + }, + { + "epoch": 0.106, + "grad_norm": 0.39551809430122375, + "learning_rate": 0.0004884899922594882, + "loss": 2.9547500610351562, + "step": 21200 + }, + { + "epoch": 0.10605, + "grad_norm": 0.3712783455848694, + "learning_rate": 0.0004884780919051449, + "loss": 2.994594192504883, + "step": 21210 + }, + { + "epoch": 0.1061, + "grad_norm": 0.36416056752204895, + "learning_rate": 0.0004884661855471173, + "loss": 2.944552993774414, + "step": 21220 + }, + { + "epoch": 0.10615, + "grad_norm": 0.41668519377708435, + "learning_rate": 0.0004884542731857051, + "loss": 2.986443519592285, + "step": 21230 + }, + { + "epoch": 0.1062, + "grad_norm": 0.40528857707977295, + "learning_rate": 0.0004884423548212082, + "loss": 2.9692104339599608, + "step": 21240 + }, + { + "epoch": 0.10625, + "grad_norm": 0.4193595051765442, + "learning_rate": 0.0004884304304539265, + "loss": 2.9791385650634767, + "step": 21250 + }, + { + "epoch": 0.1063, + "grad_norm": 0.39582833647727966, + "learning_rate": 0.0004884185000841603, + "loss": 2.9342283248901366, + "step": 21260 + }, + { + "epoch": 0.10635, + "grad_norm": 0.39235034584999084, + "learning_rate": 0.0004884065637122101, + "loss": 2.992925262451172, + "step": 21270 + }, + { + "epoch": 0.1064, + "grad_norm": 0.3998987674713135, + "learning_rate": 0.0004883946213383761, + "loss": 2.945668411254883, + "step": 21280 + }, + { + "epoch": 0.10645, + "grad_norm": 0.38746389746665955, + "learning_rate": 0.0004883826729629592, + "loss": 2.953115463256836, + "step": 21290 + }, + { + "epoch": 0.1065, + "grad_norm": 0.41150856018066406, + "learning_rate": 0.0004883707185862601, + "loss": 2.9612659454345702, + "step": 21300 + }, + { + "epoch": 0.10655, + "grad_norm": 0.35538938641548157, + "learning_rate": 0.0004883587582085796, + "loss": 2.992949295043945, + "step": 21310 + }, + { + "epoch": 0.1066, + "grad_norm": 0.40850892663002014, + "learning_rate": 0.0004883467918302191, + "loss": 2.9695505142211913, + "step": 21320 + }, + { + "epoch": 0.10665, + "grad_norm": 0.40285807847976685, + "learning_rate": 0.0004883348194514798, + "loss": 2.9827117919921875, + "step": 21330 + }, + { + "epoch": 0.1067, + "grad_norm": 0.41451296210289, + "learning_rate": 0.0004883228410726628, + "loss": 2.9428874969482424, + "step": 21340 + }, + { + "epoch": 0.10675, + "grad_norm": 0.4144335687160492, + "learning_rate": 0.00048831085669407, + "loss": 2.9901412963867187, + "step": 21350 + }, + { + "epoch": 0.1068, + "grad_norm": 0.4015392065048218, + "learning_rate": 0.000488298866316003, + "loss": 2.952992820739746, + "step": 21360 + }, + { + "epoch": 0.10685, + "grad_norm": 0.3735862672328949, + "learning_rate": 0.0004882868699387635, + "loss": 2.9547908782958983, + "step": 21370 + }, + { + "epoch": 0.1069, + "grad_norm": 0.3736407458782196, + "learning_rate": 0.00048827486756265374, + "loss": 2.971299171447754, + "step": 21380 + }, + { + "epoch": 0.10695, + "grad_norm": 0.38717859983444214, + "learning_rate": 0.00048826285918797575, + "loss": 3.0295066833496094, + "step": 21390 + }, + { + "epoch": 0.107, + "grad_norm": 0.3773841857910156, + "learning_rate": 0.00048825084481503186, + "loss": 2.969883155822754, + "step": 21400 + }, + { + "epoch": 0.10705, + "grad_norm": 0.3888879418373108, + "learning_rate": 0.00048823882444412457, + "loss": 2.9425891876220702, + "step": 21410 + }, + { + "epoch": 0.1071, + "grad_norm": 0.39792105555534363, + "learning_rate": 0.0004882267980755564, + "loss": 2.949176788330078, + "step": 21420 + }, + { + "epoch": 0.10715, + "grad_norm": 0.38126370310783386, + "learning_rate": 0.0004882147657096303, + "loss": 2.974744987487793, + "step": 21430 + }, + { + "epoch": 0.1072, + "grad_norm": 0.45289894938468933, + "learning_rate": 0.0004882027273466489, + "loss": 2.967967224121094, + "step": 21440 + }, + { + "epoch": 0.10725, + "grad_norm": 0.3638432025909424, + "learning_rate": 0.0004881906829869156, + "loss": 2.938848686218262, + "step": 21450 + }, + { + "epoch": 0.1073, + "grad_norm": 0.4077974259853363, + "learning_rate": 0.00048817863263073336, + "loss": 2.9318315505981447, + "step": 21460 + }, + { + "epoch": 0.10735, + "grad_norm": 0.38824182748794556, + "learning_rate": 0.00048816657627840564, + "loss": 2.9448131561279296, + "step": 21470 + }, + { + "epoch": 0.1074, + "grad_norm": 0.36520957946777344, + "learning_rate": 0.000488154513930236, + "loss": 2.949428749084473, + "step": 21480 + }, + { + "epoch": 0.10745, + "grad_norm": 0.45751336216926575, + "learning_rate": 0.00048814244558652796, + "loss": 2.9343692779541017, + "step": 21490 + }, + { + "epoch": 0.1075, + "grad_norm": 0.39468011260032654, + "learning_rate": 0.0004881303712475856, + "loss": 2.98242244720459, + "step": 21500 + }, + { + "epoch": 0.10755, + "grad_norm": 0.4178416430950165, + "learning_rate": 0.0004881182909137126, + "loss": 3.0032428741455077, + "step": 21510 + }, + { + "epoch": 0.1076, + "grad_norm": 0.37806469202041626, + "learning_rate": 0.0004881062045852133, + "loss": 2.927627754211426, + "step": 21520 + }, + { + "epoch": 0.10765, + "grad_norm": 0.40146926045417786, + "learning_rate": 0.0004880941122623919, + "loss": 2.971565055847168, + "step": 21530 + }, + { + "epoch": 0.1077, + "grad_norm": 0.382123202085495, + "learning_rate": 0.00048808201394555275, + "loss": 2.970705032348633, + "step": 21540 + }, + { + "epoch": 0.10775, + "grad_norm": 0.3847157657146454, + "learning_rate": 0.00048806990963500045, + "loss": 2.967424011230469, + "step": 21550 + }, + { + "epoch": 0.1078, + "grad_norm": 0.38625863194465637, + "learning_rate": 0.0004880577993310399, + "loss": 2.9762384414672853, + "step": 21560 + }, + { + "epoch": 0.10785, + "grad_norm": 0.3681718409061432, + "learning_rate": 0.0004880456830339757, + "loss": 2.965622329711914, + "step": 21570 + }, + { + "epoch": 0.1079, + "grad_norm": 0.3732280731201172, + "learning_rate": 0.0004880335607441131, + "loss": 2.9437629699707033, + "step": 21580 + }, + { + "epoch": 0.10795, + "grad_norm": 0.39356061816215515, + "learning_rate": 0.0004880214324617572, + "loss": 2.909684181213379, + "step": 21590 + }, + { + "epoch": 0.108, + "grad_norm": 0.36912620067596436, + "learning_rate": 0.0004880092981872133, + "loss": 2.9447490692138674, + "step": 21600 + }, + { + "epoch": 0.10805, + "grad_norm": 0.38759535551071167, + "learning_rate": 0.0004879971579207869, + "loss": 2.904450225830078, + "step": 21610 + }, + { + "epoch": 0.1081, + "grad_norm": 0.3741895854473114, + "learning_rate": 0.0004879850116627838, + "loss": 2.92208251953125, + "step": 21620 + }, + { + "epoch": 0.10815, + "grad_norm": 0.3927638828754425, + "learning_rate": 0.0004879728594135094, + "loss": 2.9787677764892577, + "step": 21630 + }, + { + "epoch": 0.1082, + "grad_norm": 0.4296223223209381, + "learning_rate": 0.00048796070117326996, + "loss": 2.9823486328125, + "step": 21640 + }, + { + "epoch": 0.10825, + "grad_norm": 0.3848608732223511, + "learning_rate": 0.0004879485369423714, + "loss": 2.90673942565918, + "step": 21650 + }, + { + "epoch": 0.1083, + "grad_norm": 0.40072375535964966, + "learning_rate": 0.00048793636672112006, + "loss": 2.9824676513671875, + "step": 21660 + }, + { + "epoch": 0.10835, + "grad_norm": 0.3823498487472534, + "learning_rate": 0.00048792419050982226, + "loss": 2.9530012130737306, + "step": 21670 + }, + { + "epoch": 0.1084, + "grad_norm": 0.4040793180465698, + "learning_rate": 0.00048791200830878456, + "loss": 2.9128231048583983, + "step": 21680 + }, + { + "epoch": 0.10845, + "grad_norm": 0.4024946093559265, + "learning_rate": 0.00048789982011831367, + "loss": 2.96390380859375, + "step": 21690 + }, + { + "epoch": 0.1085, + "grad_norm": 0.4062943458557129, + "learning_rate": 0.0004878876259387164, + "loss": 2.937142181396484, + "step": 21700 + }, + { + "epoch": 0.10855, + "grad_norm": 0.38276001811027527, + "learning_rate": 0.00048787542577029963, + "loss": 2.9575132369995116, + "step": 21710 + }, + { + "epoch": 0.1086, + "grad_norm": 0.37770089507102966, + "learning_rate": 0.0004878632196133708, + "loss": 2.9743026733398437, + "step": 21720 + }, + { + "epoch": 0.10865, + "grad_norm": 0.3698958456516266, + "learning_rate": 0.00048785100746823683, + "loss": 2.947017860412598, + "step": 21730 + }, + { + "epoch": 0.1087, + "grad_norm": 0.41125279664993286, + "learning_rate": 0.00048783878933520537, + "loss": 2.995106887817383, + "step": 21740 + }, + { + "epoch": 0.10875, + "grad_norm": 0.3976326882839203, + "learning_rate": 0.000487826565214584, + "loss": 2.9523929595947265, + "step": 21750 + }, + { + "epoch": 0.1088, + "grad_norm": 0.38549867272377014, + "learning_rate": 0.0004878143351066805, + "loss": 2.951267623901367, + "step": 21760 + }, + { + "epoch": 0.10885, + "grad_norm": 0.3992399573326111, + "learning_rate": 0.0004878020990118026, + "loss": 2.9227811813354494, + "step": 21770 + }, + { + "epoch": 0.1089, + "grad_norm": 0.3882802128791809, + "learning_rate": 0.0004877898569302585, + "loss": 2.9476242065429688, + "step": 21780 + }, + { + "epoch": 0.10895, + "grad_norm": 0.37611743807792664, + "learning_rate": 0.0004877776088623563, + "loss": 2.9707469940185547, + "step": 21790 + }, + { + "epoch": 0.109, + "grad_norm": 0.3660087585449219, + "learning_rate": 0.0004877653548084045, + "loss": 2.936651611328125, + "step": 21800 + }, + { + "epoch": 0.10905, + "grad_norm": 0.40364310145378113, + "learning_rate": 0.00048775309476871133, + "loss": 3.000147247314453, + "step": 21810 + }, + { + "epoch": 0.1091, + "grad_norm": 0.3827480971813202, + "learning_rate": 0.0004877408287435857, + "loss": 2.939823532104492, + "step": 21820 + }, + { + "epoch": 0.10915, + "grad_norm": 0.44689881801605225, + "learning_rate": 0.0004877285567333363, + "loss": 2.978156089782715, + "step": 21830 + }, + { + "epoch": 0.1092, + "grad_norm": 0.36004093289375305, + "learning_rate": 0.00048771627873827196, + "loss": 3.001571846008301, + "step": 21840 + }, + { + "epoch": 0.10925, + "grad_norm": 0.39675381779670715, + "learning_rate": 0.000487703994758702, + "loss": 2.948797416687012, + "step": 21850 + }, + { + "epoch": 0.1093, + "grad_norm": 0.3631017208099365, + "learning_rate": 0.00048769170479493547, + "loss": 2.91392822265625, + "step": 21860 + }, + { + "epoch": 0.10935, + "grad_norm": 0.37299978733062744, + "learning_rate": 0.0004876794088472819, + "loss": 2.967456245422363, + "step": 21870 + }, + { + "epoch": 0.1094, + "grad_norm": 0.4143035113811493, + "learning_rate": 0.00048766710691605083, + "loss": 2.9533905029296874, + "step": 21880 + }, + { + "epoch": 0.10945, + "grad_norm": 0.3700999915599823, + "learning_rate": 0.0004876547990015519, + "loss": 2.950341796875, + "step": 21890 + }, + { + "epoch": 0.1095, + "grad_norm": 0.40169796347618103, + "learning_rate": 0.000487642485104095, + "loss": 2.940465545654297, + "step": 21900 + }, + { + "epoch": 0.10955, + "grad_norm": 0.36992770433425903, + "learning_rate": 0.0004876301652239901, + "loss": 3.0028812408447267, + "step": 21910 + }, + { + "epoch": 0.1096, + "grad_norm": 0.3859299123287201, + "learning_rate": 0.0004876178393615475, + "loss": 2.9647075653076174, + "step": 21920 + }, + { + "epoch": 0.10965, + "grad_norm": 0.3803994059562683, + "learning_rate": 0.0004876055075170772, + "loss": 2.9803865432739256, + "step": 21930 + }, + { + "epoch": 0.1097, + "grad_norm": 0.4155447483062744, + "learning_rate": 0.00048759316969088995, + "loss": 2.9044540405273436, + "step": 21940 + }, + { + "epoch": 0.10975, + "grad_norm": 0.3760974407196045, + "learning_rate": 0.0004875808258832962, + "loss": 2.949534225463867, + "step": 21950 + }, + { + "epoch": 0.1098, + "grad_norm": 0.3945028483867645, + "learning_rate": 0.00048756847609460674, + "loss": 2.943076324462891, + "step": 21960 + }, + { + "epoch": 0.10985, + "grad_norm": 0.39531487226486206, + "learning_rate": 0.0004875561203251325, + "loss": 2.9574337005615234, + "step": 21970 + }, + { + "epoch": 0.1099, + "grad_norm": 0.39600133895874023, + "learning_rate": 0.0004875437585751845, + "loss": 2.9843238830566405, + "step": 21980 + }, + { + "epoch": 0.10995, + "grad_norm": 0.37657928466796875, + "learning_rate": 0.00048753139084507406, + "loss": 2.995166206359863, + "step": 21990 + }, + { + "epoch": 0.11, + "grad_norm": 0.377574622631073, + "learning_rate": 0.0004875190171351124, + "loss": 2.9394145965576173, + "step": 22000 + }, + { + "epoch": 0.11005, + "grad_norm": 0.39569857716560364, + "learning_rate": 0.00048750663744561096, + "loss": 2.98294620513916, + "step": 22010 + }, + { + "epoch": 0.1101, + "grad_norm": 0.35432228446006775, + "learning_rate": 0.0004874942517768816, + "loss": 2.9893112182617188, + "step": 22020 + }, + { + "epoch": 0.11015, + "grad_norm": 0.4244389533996582, + "learning_rate": 0.000487481860129236, + "loss": 2.9959253311157226, + "step": 22030 + }, + { + "epoch": 0.1102, + "grad_norm": 0.37321096658706665, + "learning_rate": 0.00048746946250298624, + "loss": 2.9916399002075194, + "step": 22040 + }, + { + "epoch": 0.11025, + "grad_norm": 0.3791712522506714, + "learning_rate": 0.0004874570588984443, + "loss": 2.9407932281494142, + "step": 22050 + }, + { + "epoch": 0.1103, + "grad_norm": 0.3952069580554962, + "learning_rate": 0.0004874446493159225, + "loss": 2.9608436584472657, + "step": 22060 + }, + { + "epoch": 0.11035, + "grad_norm": 0.4029282331466675, + "learning_rate": 0.0004874322337557332, + "loss": 2.9283660888671874, + "step": 22070 + }, + { + "epoch": 0.1104, + "grad_norm": 0.35294750332832336, + "learning_rate": 0.000487419812218189, + "loss": 2.944174957275391, + "step": 22080 + }, + { + "epoch": 0.11045, + "grad_norm": 0.38131183385849, + "learning_rate": 0.00048740738470360266, + "loss": 2.918873405456543, + "step": 22090 + }, + { + "epoch": 0.1105, + "grad_norm": 0.38461193442344666, + "learning_rate": 0.00048739495121228693, + "loss": 3.035131645202637, + "step": 22100 + }, + { + "epoch": 0.11055, + "grad_norm": 0.3895101547241211, + "learning_rate": 0.00048738251174455495, + "loss": 2.9477563858032227, + "step": 22110 + }, + { + "epoch": 0.1106, + "grad_norm": 0.3847261667251587, + "learning_rate": 0.0004873700663007198, + "loss": 2.9745025634765625, + "step": 22120 + }, + { + "epoch": 0.11065, + "grad_norm": 0.385098934173584, + "learning_rate": 0.0004873576148810949, + "loss": 3.0253555297851564, + "step": 22130 + }, + { + "epoch": 0.1107, + "grad_norm": 0.412605881690979, + "learning_rate": 0.0004873451574859935, + "loss": 2.9262937545776366, + "step": 22140 + }, + { + "epoch": 0.11075, + "grad_norm": 0.4239142835140228, + "learning_rate": 0.0004873326941157294, + "loss": 2.923430061340332, + "step": 22150 + }, + { + "epoch": 0.1108, + "grad_norm": 0.38366594910621643, + "learning_rate": 0.0004873202247706163, + "loss": 2.9616798400878905, + "step": 22160 + }, + { + "epoch": 0.11085, + "grad_norm": 0.3610599637031555, + "learning_rate": 0.0004873077494509681, + "loss": 2.9649330139160157, + "step": 22170 + }, + { + "epoch": 0.1109, + "grad_norm": 0.3981793522834778, + "learning_rate": 0.000487295268157099, + "loss": 2.9350648880004884, + "step": 22180 + }, + { + "epoch": 0.11095, + "grad_norm": 0.40262892842292786, + "learning_rate": 0.00048728278088932305, + "loss": 2.951384353637695, + "step": 22190 + }, + { + "epoch": 0.111, + "grad_norm": 0.3669050335884094, + "learning_rate": 0.00048727028764795465, + "loss": 3.002129554748535, + "step": 22200 + }, + { + "epoch": 0.11105, + "grad_norm": 0.3864041268825531, + "learning_rate": 0.0004872577884333084, + "loss": 2.906988525390625, + "step": 22210 + }, + { + "epoch": 0.1111, + "grad_norm": 0.404325008392334, + "learning_rate": 0.00048724528324569884, + "loss": 2.988216781616211, + "step": 22220 + }, + { + "epoch": 0.11115, + "grad_norm": 0.4099927544593811, + "learning_rate": 0.0004872327720854409, + "loss": 2.971780776977539, + "step": 22230 + }, + { + "epoch": 0.1112, + "grad_norm": 0.3995363116264343, + "learning_rate": 0.0004872202549528495, + "loss": 2.9426803588867188, + "step": 22240 + }, + { + "epoch": 0.11125, + "grad_norm": 0.3557639718055725, + "learning_rate": 0.00048720773184823975, + "loss": 2.907750701904297, + "step": 22250 + }, + { + "epoch": 0.1113, + "grad_norm": 0.39334115386009216, + "learning_rate": 0.00048719520277192696, + "loss": 2.989939880371094, + "step": 22260 + }, + { + "epoch": 0.11135, + "grad_norm": 0.39065390825271606, + "learning_rate": 0.0004871826677242265, + "loss": 2.94183292388916, + "step": 22270 + }, + { + "epoch": 0.1114, + "grad_norm": 0.3758164644241333, + "learning_rate": 0.000487170126705454, + "loss": 2.9368825912475587, + "step": 22280 + }, + { + "epoch": 0.11145, + "grad_norm": 0.35368555784225464, + "learning_rate": 0.0004871575797159251, + "loss": 2.893545913696289, + "step": 22290 + }, + { + "epoch": 0.1115, + "grad_norm": 0.38709020614624023, + "learning_rate": 0.00048714502675595577, + "loss": 2.920855712890625, + "step": 22300 + }, + { + "epoch": 0.11155, + "grad_norm": 0.38967856764793396, + "learning_rate": 0.00048713246782586196, + "loss": 2.941835403442383, + "step": 22310 + }, + { + "epoch": 0.1116, + "grad_norm": 0.3903433084487915, + "learning_rate": 0.00048711990292595985, + "loss": 2.909108352661133, + "step": 22320 + }, + { + "epoch": 0.11165, + "grad_norm": 0.4048056900501251, + "learning_rate": 0.00048710733205656575, + "loss": 2.9467704772949217, + "step": 22330 + }, + { + "epoch": 0.1117, + "grad_norm": 0.4049469232559204, + "learning_rate": 0.0004870947552179962, + "loss": 3.020545768737793, + "step": 22340 + }, + { + "epoch": 0.11175, + "grad_norm": 0.3857284188270569, + "learning_rate": 0.0004870821724105677, + "loss": 2.9926658630371095, + "step": 22350 + }, + { + "epoch": 0.1118, + "grad_norm": 0.354409396648407, + "learning_rate": 0.0004870695836345972, + "loss": 2.9591073989868164, + "step": 22360 + }, + { + "epoch": 0.11185, + "grad_norm": 0.38561558723449707, + "learning_rate": 0.0004870569888904014, + "loss": 2.9509654998779298, + "step": 22370 + }, + { + "epoch": 0.1119, + "grad_norm": 0.3489324152469635, + "learning_rate": 0.00048704438817829756, + "loss": 2.9484561920166015, + "step": 22380 + }, + { + "epoch": 0.11195, + "grad_norm": 0.4270067811012268, + "learning_rate": 0.00048703178149860283, + "loss": 2.967692565917969, + "step": 22390 + }, + { + "epoch": 0.112, + "grad_norm": 0.36766839027404785, + "learning_rate": 0.00048701916885163455, + "loss": 2.985692596435547, + "step": 22400 + }, + { + "epoch": 0.11205, + "grad_norm": 0.37408071756362915, + "learning_rate": 0.00048700655023771027, + "loss": 2.963522529602051, + "step": 22410 + }, + { + "epoch": 0.1121, + "grad_norm": 0.40779533982276917, + "learning_rate": 0.00048699392565714774, + "loss": 2.935489845275879, + "step": 22420 + }, + { + "epoch": 0.11215, + "grad_norm": 0.3890751004219055, + "learning_rate": 0.0004869812951102646, + "loss": 2.9794797897338867, + "step": 22430 + }, + { + "epoch": 0.1122, + "grad_norm": 0.35025206208229065, + "learning_rate": 0.00048696865859737906, + "loss": 2.9499286651611327, + "step": 22440 + }, + { + "epoch": 0.11225, + "grad_norm": 0.34170839190483093, + "learning_rate": 0.000486956016118809, + "loss": 2.9375516891479494, + "step": 22450 + }, + { + "epoch": 0.1123, + "grad_norm": 0.4192226231098175, + "learning_rate": 0.0004869433676748729, + "loss": 3.0662506103515623, + "step": 22460 + }, + { + "epoch": 0.11235, + "grad_norm": 0.41912028193473816, + "learning_rate": 0.0004869307132658891, + "loss": 2.971514892578125, + "step": 22470 + }, + { + "epoch": 0.1124, + "grad_norm": 0.3958647549152374, + "learning_rate": 0.0004869180528921762, + "loss": 2.934498596191406, + "step": 22480 + }, + { + "epoch": 0.11245, + "grad_norm": 0.3832666575908661, + "learning_rate": 0.0004869053865540528, + "loss": 2.951586151123047, + "step": 22490 + }, + { + "epoch": 0.1125, + "grad_norm": 0.40284937620162964, + "learning_rate": 0.000486892714251838, + "loss": 2.945356559753418, + "step": 22500 + }, + { + "epoch": 0.11255, + "grad_norm": 0.39775508642196655, + "learning_rate": 0.0004868800359858506, + "loss": 2.8981231689453124, + "step": 22510 + }, + { + "epoch": 0.1126, + "grad_norm": 0.4003564715385437, + "learning_rate": 0.00048686735175640996, + "loss": 2.977340507507324, + "step": 22520 + }, + { + "epoch": 0.11265, + "grad_norm": 0.34596556425094604, + "learning_rate": 0.0004868546615638353, + "loss": 2.967356872558594, + "step": 22530 + }, + { + "epoch": 0.1127, + "grad_norm": 0.3887903690338135, + "learning_rate": 0.00048684196540844607, + "loss": 2.980998229980469, + "step": 22540 + }, + { + "epoch": 0.11275, + "grad_norm": 0.4502594769001007, + "learning_rate": 0.00048682926329056197, + "loss": 2.8992610931396485, + "step": 22550 + }, + { + "epoch": 0.1128, + "grad_norm": 0.38470083475112915, + "learning_rate": 0.0004868165552105028, + "loss": 2.9295473098754883, + "step": 22560 + }, + { + "epoch": 0.11285, + "grad_norm": 0.3724956214427948, + "learning_rate": 0.00048680384116858835, + "loss": 2.9954458236694337, + "step": 22570 + }, + { + "epoch": 0.1129, + "grad_norm": 0.36593177914619446, + "learning_rate": 0.00048679112116513883, + "loss": 2.9742679595947266, + "step": 22580 + }, + { + "epoch": 0.11295, + "grad_norm": 0.4135967493057251, + "learning_rate": 0.0004867783952004744, + "loss": 2.953957176208496, + "step": 22590 + }, + { + "epoch": 0.113, + "grad_norm": 0.41730883717536926, + "learning_rate": 0.00048676566327491547, + "loss": 2.899765396118164, + "step": 22600 + }, + { + "epoch": 0.11305, + "grad_norm": 0.44876381754875183, + "learning_rate": 0.0004867529253887826, + "loss": 2.958197021484375, + "step": 22610 + }, + { + "epoch": 0.1131, + "grad_norm": 0.39782747626304626, + "learning_rate": 0.0004867401815423963, + "loss": 2.9345603942871095, + "step": 22620 + }, + { + "epoch": 0.11315, + "grad_norm": 0.3771565854549408, + "learning_rate": 0.0004867274317360776, + "loss": 2.9840728759765627, + "step": 22630 + }, + { + "epoch": 0.1132, + "grad_norm": 0.35863417387008667, + "learning_rate": 0.0004867146759701473, + "loss": 2.9791759490966796, + "step": 22640 + }, + { + "epoch": 0.11325, + "grad_norm": 0.3602137863636017, + "learning_rate": 0.0004867019142449267, + "loss": 2.9436887741088866, + "step": 22650 + }, + { + "epoch": 0.1133, + "grad_norm": 0.3868941366672516, + "learning_rate": 0.0004866891465607369, + "loss": 2.9430599212646484, + "step": 22660 + }, + { + "epoch": 0.11335, + "grad_norm": 0.3959915041923523, + "learning_rate": 0.0004866763729178996, + "loss": 2.961963081359863, + "step": 22670 + }, + { + "epoch": 0.1134, + "grad_norm": 0.3642185926437378, + "learning_rate": 0.00048666359331673605, + "loss": 2.95705623626709, + "step": 22680 + }, + { + "epoch": 0.11345, + "grad_norm": 0.3682486116886139, + "learning_rate": 0.00048665080775756807, + "loss": 2.9639801025390624, + "step": 22690 + }, + { + "epoch": 0.1135, + "grad_norm": 0.44635507464408875, + "learning_rate": 0.00048663801624071766, + "loss": 2.934739875793457, + "step": 22700 + }, + { + "epoch": 0.11355, + "grad_norm": 0.39105626940727234, + "learning_rate": 0.0004866252187665068, + "loss": 2.955986213684082, + "step": 22710 + }, + { + "epoch": 0.1136, + "grad_norm": 0.38066306710243225, + "learning_rate": 0.0004866124153352576, + "loss": 2.9672245025634765, + "step": 22720 + }, + { + "epoch": 0.11365, + "grad_norm": 0.38635891675949097, + "learning_rate": 0.00048659960594729245, + "loss": 2.892908477783203, + "step": 22730 + }, + { + "epoch": 0.1137, + "grad_norm": 0.39761820435523987, + "learning_rate": 0.00048658679060293377, + "loss": 2.980764389038086, + "step": 22740 + }, + { + "epoch": 0.11375, + "grad_norm": 0.4016125202178955, + "learning_rate": 0.0004865739693025042, + "loss": 2.951380157470703, + "step": 22750 + }, + { + "epoch": 0.1138, + "grad_norm": 0.38061434030532837, + "learning_rate": 0.0004865611420463266, + "loss": 2.9280574798583983, + "step": 22760 + }, + { + "epoch": 0.11385, + "grad_norm": 0.3990250825881958, + "learning_rate": 0.0004865483088347238, + "loss": 2.9662712097167967, + "step": 22770 + }, + { + "epoch": 0.1139, + "grad_norm": 0.3755533695220947, + "learning_rate": 0.00048653546966801885, + "loss": 2.9493518829345704, + "step": 22780 + }, + { + "epoch": 0.11395, + "grad_norm": 0.37118011713027954, + "learning_rate": 0.0004865226245465351, + "loss": 2.9264759063720702, + "step": 22790 + }, + { + "epoch": 0.114, + "grad_norm": 0.39819690585136414, + "learning_rate": 0.0004865097734705959, + "loss": 2.916080284118652, + "step": 22800 + }, + { + "epoch": 0.11405, + "grad_norm": 0.3969298005104065, + "learning_rate": 0.00048649691644052465, + "loss": 2.9734004974365233, + "step": 22810 + }, + { + "epoch": 0.1141, + "grad_norm": 0.38079380989074707, + "learning_rate": 0.0004864840534566451, + "loss": 2.9717079162597657, + "step": 22820 + }, + { + "epoch": 0.11415, + "grad_norm": 0.3933635652065277, + "learning_rate": 0.0004864711845192811, + "loss": 2.905269432067871, + "step": 22830 + }, + { + "epoch": 0.1142, + "grad_norm": 0.3886224925518036, + "learning_rate": 0.00048645830962875666, + "loss": 2.9318042755126954, + "step": 22840 + }, + { + "epoch": 0.11425, + "grad_norm": 0.41008079051971436, + "learning_rate": 0.00048644542878539576, + "loss": 2.9427719116210938, + "step": 22850 + }, + { + "epoch": 0.1143, + "grad_norm": 0.3714187443256378, + "learning_rate": 0.00048643254198952284, + "loss": 2.9828556060791014, + "step": 22860 + }, + { + "epoch": 0.11435, + "grad_norm": 0.37692761421203613, + "learning_rate": 0.00048641964924146233, + "loss": 2.985920715332031, + "step": 22870 + }, + { + "epoch": 0.1144, + "grad_norm": 0.38081464171409607, + "learning_rate": 0.0004864067505415386, + "loss": 2.918515205383301, + "step": 22880 + }, + { + "epoch": 0.11445, + "grad_norm": 0.41948041319847107, + "learning_rate": 0.0004863938458900765, + "loss": 2.946137619018555, + "step": 22890 + }, + { + "epoch": 0.1145, + "grad_norm": 0.38375166058540344, + "learning_rate": 0.00048638093528740104, + "loss": 2.96869010925293, + "step": 22900 + }, + { + "epoch": 0.11455, + "grad_norm": 0.39907869696617126, + "learning_rate": 0.00048636801873383695, + "loss": 2.9156190872192385, + "step": 22910 + }, + { + "epoch": 0.1146, + "grad_norm": 0.39277324080467224, + "learning_rate": 0.0004863550962297097, + "loss": 2.9126720428466797, + "step": 22920 + }, + { + "epoch": 0.11465, + "grad_norm": 0.40330183506011963, + "learning_rate": 0.00048634216777534447, + "loss": 2.924030876159668, + "step": 22930 + }, + { + "epoch": 0.1147, + "grad_norm": 0.38260433077812195, + "learning_rate": 0.00048632923337106664, + "loss": 2.9561435699462892, + "step": 22940 + }, + { + "epoch": 0.11475, + "grad_norm": 0.3730204403400421, + "learning_rate": 0.000486316293017202, + "loss": 2.9046762466430662, + "step": 22950 + }, + { + "epoch": 0.1148, + "grad_norm": 0.39371350407600403, + "learning_rate": 0.0004863033467140763, + "loss": 2.9466934204101562, + "step": 22960 + }, + { + "epoch": 0.11485, + "grad_norm": 0.4101696312427521, + "learning_rate": 0.0004862903944620154, + "loss": 2.9723230361938477, + "step": 22970 + }, + { + "epoch": 0.1149, + "grad_norm": 0.3610314130783081, + "learning_rate": 0.0004862774362613454, + "loss": 2.9873069763183593, + "step": 22980 + }, + { + "epoch": 0.11495, + "grad_norm": 0.379271537065506, + "learning_rate": 0.0004862644721123925, + "loss": 2.9922197341918944, + "step": 22990 + }, + { + "epoch": 0.115, + "grad_norm": 0.4042591452598572, + "learning_rate": 0.00048625150201548305, + "loss": 2.9081132888793944, + "step": 23000 + }, + { + "epoch": 0.11505, + "grad_norm": 0.3755771517753601, + "learning_rate": 0.0004862385259709438, + "loss": 2.950638008117676, + "step": 23010 + }, + { + "epoch": 0.1151, + "grad_norm": 0.3720417022705078, + "learning_rate": 0.00048622554397910104, + "loss": 2.9970354080200194, + "step": 23020 + }, + { + "epoch": 0.11515, + "grad_norm": 0.3840938210487366, + "learning_rate": 0.0004862125560402819, + "loss": 2.919573211669922, + "step": 23030 + }, + { + "epoch": 0.1152, + "grad_norm": 0.39738839864730835, + "learning_rate": 0.0004861995621548132, + "loss": 2.9679805755615236, + "step": 23040 + }, + { + "epoch": 0.11525, + "grad_norm": 0.4355621933937073, + "learning_rate": 0.0004861865623230221, + "loss": 2.914143371582031, + "step": 23050 + }, + { + "epoch": 0.1153, + "grad_norm": 0.3629553020000458, + "learning_rate": 0.00048617355654523585, + "loss": 2.914590072631836, + "step": 23060 + }, + { + "epoch": 0.11535, + "grad_norm": 0.3999696373939514, + "learning_rate": 0.0004861605448217819, + "loss": 2.992013931274414, + "step": 23070 + }, + { + "epoch": 0.1154, + "grad_norm": 0.39955323934555054, + "learning_rate": 0.0004861475271529878, + "loss": 2.8861793518066405, + "step": 23080 + }, + { + "epoch": 0.11545, + "grad_norm": 0.4330939054489136, + "learning_rate": 0.0004861345035391813, + "loss": 2.9413761138916015, + "step": 23090 + }, + { + "epoch": 0.1155, + "grad_norm": 0.3572467565536499, + "learning_rate": 0.0004861214739806903, + "loss": 2.9979448318481445, + "step": 23100 + }, + { + "epoch": 0.11555, + "grad_norm": 0.3667079210281372, + "learning_rate": 0.0004861084384778427, + "loss": 2.9656044006347657, + "step": 23110 + }, + { + "epoch": 0.1156, + "grad_norm": 0.40228840708732605, + "learning_rate": 0.0004860953970309668, + "loss": 2.9368736267089846, + "step": 23120 + }, + { + "epoch": 0.11565, + "grad_norm": 0.4088684916496277, + "learning_rate": 0.0004860823496403908, + "loss": 2.9837600708007814, + "step": 23130 + }, + { + "epoch": 0.1157, + "grad_norm": 0.38603532314300537, + "learning_rate": 0.00048606929630644325, + "loss": 2.9375354766845705, + "step": 23140 + }, + { + "epoch": 0.11575, + "grad_norm": 0.358787477016449, + "learning_rate": 0.00048605623702945277, + "loss": 2.9577354431152343, + "step": 23150 + }, + { + "epoch": 0.1158, + "grad_norm": 0.36136284470558167, + "learning_rate": 0.00048604317180974804, + "loss": 2.9095672607421874, + "step": 23160 + }, + { + "epoch": 0.11585, + "grad_norm": 0.3829100728034973, + "learning_rate": 0.00048603010064765805, + "loss": 2.9368026733398436, + "step": 23170 + }, + { + "epoch": 0.1159, + "grad_norm": 0.4157702922821045, + "learning_rate": 0.0004860170235435119, + "loss": 2.9130746841430666, + "step": 23180 + }, + { + "epoch": 0.11595, + "grad_norm": 0.38061463832855225, + "learning_rate": 0.00048600394049763874, + "loss": 3.002023696899414, + "step": 23190 + }, + { + "epoch": 0.116, + "grad_norm": 0.3906853199005127, + "learning_rate": 0.0004859908515103679, + "loss": 2.9250940322875976, + "step": 23200 + }, + { + "epoch": 0.11605, + "grad_norm": 0.42377805709838867, + "learning_rate": 0.00048597775658202903, + "loss": 2.9638044357299806, + "step": 23210 + }, + { + "epoch": 0.1161, + "grad_norm": 0.36983227729797363, + "learning_rate": 0.0004859646557129517, + "loss": 2.968714141845703, + "step": 23220 + }, + { + "epoch": 0.11615, + "grad_norm": 0.38124361634254456, + "learning_rate": 0.0004859515489034657, + "loss": 2.9204450607299806, + "step": 23230 + }, + { + "epoch": 0.1162, + "grad_norm": 0.3651776611804962, + "learning_rate": 0.0004859384361539011, + "loss": 2.9573612213134766, + "step": 23240 + }, + { + "epoch": 0.11625, + "grad_norm": 0.42829206585884094, + "learning_rate": 0.0004859253174645879, + "loss": 2.9416990280151367, + "step": 23250 + }, + { + "epoch": 0.1163, + "grad_norm": 0.37974557280540466, + "learning_rate": 0.0004859121928358564, + "loss": 2.971574401855469, + "step": 23260 + }, + { + "epoch": 0.11635, + "grad_norm": 0.3738696575164795, + "learning_rate": 0.00048589906226803706, + "loss": 2.8797735214233398, + "step": 23270 + }, + { + "epoch": 0.1164, + "grad_norm": 0.40004852414131165, + "learning_rate": 0.00048588592576146036, + "loss": 2.910755920410156, + "step": 23280 + }, + { + "epoch": 0.11645, + "grad_norm": 0.39474329352378845, + "learning_rate": 0.0004858727833164571, + "loss": 2.906692886352539, + "step": 23290 + }, + { + "epoch": 0.1165, + "grad_norm": 0.35052114725112915, + "learning_rate": 0.00048585963493335806, + "loss": 2.909976577758789, + "step": 23300 + }, + { + "epoch": 0.11655, + "grad_norm": 0.3911208510398865, + "learning_rate": 0.00048584648061249426, + "loss": 2.901314926147461, + "step": 23310 + }, + { + "epoch": 0.1166, + "grad_norm": 0.36941349506378174, + "learning_rate": 0.0004858333203541969, + "loss": 2.90164794921875, + "step": 23320 + }, + { + "epoch": 0.11665, + "grad_norm": 0.3746105134487152, + "learning_rate": 0.0004858201541587973, + "loss": 2.922159957885742, + "step": 23330 + }, + { + "epoch": 0.1167, + "grad_norm": 0.34685373306274414, + "learning_rate": 0.00048580698202662693, + "loss": 2.985830879211426, + "step": 23340 + }, + { + "epoch": 0.11675, + "grad_norm": 0.37256237864494324, + "learning_rate": 0.00048579380395801734, + "loss": 2.9850555419921876, + "step": 23350 + }, + { + "epoch": 0.1168, + "grad_norm": 0.37193211913108826, + "learning_rate": 0.00048578061995330023, + "loss": 2.9115442276000976, + "step": 23360 + }, + { + "epoch": 0.11685, + "grad_norm": 0.38571417331695557, + "learning_rate": 0.0004857674300128077, + "loss": 2.935871696472168, + "step": 23370 + }, + { + "epoch": 0.1169, + "grad_norm": 0.40057599544525146, + "learning_rate": 0.00048575423413687165, + "loss": 2.946630859375, + "step": 23380 + }, + { + "epoch": 0.11695, + "grad_norm": 0.3644644320011139, + "learning_rate": 0.00048574103232582435, + "loss": 2.931728553771973, + "step": 23390 + }, + { + "epoch": 0.117, + "grad_norm": 0.3874429762363434, + "learning_rate": 0.00048572782457999807, + "loss": 2.9635446548461912, + "step": 23400 + }, + { + "epoch": 0.11705, + "grad_norm": 0.3839150071144104, + "learning_rate": 0.00048571461089972544, + "loss": 2.973012161254883, + "step": 23410 + }, + { + "epoch": 0.1171, + "grad_norm": 0.3986523747444153, + "learning_rate": 0.0004857013912853391, + "loss": 2.9333427429199217, + "step": 23420 + }, + { + "epoch": 0.11715, + "grad_norm": 0.3607795834541321, + "learning_rate": 0.00048568816573717167, + "loss": 2.9313261032104494, + "step": 23430 + }, + { + "epoch": 0.1172, + "grad_norm": 0.38300588726997375, + "learning_rate": 0.0004856749342555564, + "loss": 2.8781227111816405, + "step": 23440 + }, + { + "epoch": 0.11725, + "grad_norm": 0.3873383402824402, + "learning_rate": 0.0004856616968408261, + "loss": 2.932178497314453, + "step": 23450 + }, + { + "epoch": 0.1173, + "grad_norm": 0.3893912732601166, + "learning_rate": 0.0004856484534933142, + "loss": 2.9340375900268554, + "step": 23460 + }, + { + "epoch": 0.11735, + "grad_norm": 0.37243083119392395, + "learning_rate": 0.0004856352042133541, + "loss": 2.989825439453125, + "step": 23470 + }, + { + "epoch": 0.1174, + "grad_norm": 0.3627013564109802, + "learning_rate": 0.0004856219490012793, + "loss": 2.935005950927734, + "step": 23480 + }, + { + "epoch": 0.11745, + "grad_norm": 0.41394859552383423, + "learning_rate": 0.00048560868785742353, + "loss": 2.910196876525879, + "step": 23490 + }, + { + "epoch": 0.1175, + "grad_norm": 0.3688034117221832, + "learning_rate": 0.00048559542078212053, + "loss": 2.936593437194824, + "step": 23500 + }, + { + "epoch": 0.11755, + "grad_norm": 0.36628127098083496, + "learning_rate": 0.0004855821477757044, + "loss": 2.939549446105957, + "step": 23510 + }, + { + "epoch": 0.1176, + "grad_norm": 0.3988698720932007, + "learning_rate": 0.00048556886883850935, + "loss": 2.9835670471191404, + "step": 23520 + }, + { + "epoch": 0.11765, + "grad_norm": 0.3637103736400604, + "learning_rate": 0.0004855555839708696, + "loss": 2.9462404251098633, + "step": 23530 + }, + { + "epoch": 0.1177, + "grad_norm": 0.35969531536102295, + "learning_rate": 0.0004855422931731195, + "loss": 2.923902130126953, + "step": 23540 + }, + { + "epoch": 0.11775, + "grad_norm": 0.3702988922595978, + "learning_rate": 0.0004855289964455938, + "loss": 2.947884750366211, + "step": 23550 + }, + { + "epoch": 0.1178, + "grad_norm": 0.40415099263191223, + "learning_rate": 0.0004855156937886272, + "loss": 2.941353607177734, + "step": 23560 + }, + { + "epoch": 0.11785, + "grad_norm": 0.3550078272819519, + "learning_rate": 0.0004855023852025546, + "loss": 2.968246841430664, + "step": 23570 + }, + { + "epoch": 0.1179, + "grad_norm": 0.4160146117210388, + "learning_rate": 0.000485489070687711, + "loss": 2.904094696044922, + "step": 23580 + }, + { + "epoch": 0.11795, + "grad_norm": 0.40692710876464844, + "learning_rate": 0.0004854757502444316, + "loss": 2.947515678405762, + "step": 23590 + }, + { + "epoch": 0.118, + "grad_norm": 0.40972182154655457, + "learning_rate": 0.00048546242387305177, + "loss": 2.9537912368774415, + "step": 23600 + }, + { + "epoch": 0.11805, + "grad_norm": 0.4164474904537201, + "learning_rate": 0.00048544909157390703, + "loss": 2.892751121520996, + "step": 23610 + }, + { + "epoch": 0.1181, + "grad_norm": 0.3641332685947418, + "learning_rate": 0.0004854357533473329, + "loss": 3.0193897247314454, + "step": 23620 + }, + { + "epoch": 0.11815, + "grad_norm": 0.36519598960876465, + "learning_rate": 0.00048542240919366534, + "loss": 2.9338409423828127, + "step": 23630 + }, + { + "epoch": 0.1182, + "grad_norm": 0.3815874755382538, + "learning_rate": 0.0004854090591132402, + "loss": 2.935050201416016, + "step": 23640 + }, + { + "epoch": 0.11825, + "grad_norm": 0.3848799169063568, + "learning_rate": 0.0004853957031063935, + "loss": 2.959783172607422, + "step": 23650 + }, + { + "epoch": 0.1183, + "grad_norm": 0.3613007068634033, + "learning_rate": 0.0004853823411734616, + "loss": 2.9691848754882812, + "step": 23660 + }, + { + "epoch": 0.11835, + "grad_norm": 0.36339834332466125, + "learning_rate": 0.0004853689733147808, + "loss": 2.9595754623413084, + "step": 23670 + }, + { + "epoch": 0.1184, + "grad_norm": 0.36154451966285706, + "learning_rate": 0.0004853555995306877, + "loss": 2.9504983901977537, + "step": 23680 + }, + { + "epoch": 0.11845, + "grad_norm": 0.3743005096912384, + "learning_rate": 0.00048534221982151894, + "loss": 2.928909683227539, + "step": 23690 + }, + { + "epoch": 0.1185, + "grad_norm": 0.3645716607570648, + "learning_rate": 0.0004853288341876114, + "loss": 2.9102420806884766, + "step": 23700 + }, + { + "epoch": 0.11855, + "grad_norm": 0.36854860186576843, + "learning_rate": 0.0004853154426293019, + "loss": 2.9659862518310547, + "step": 23710 + }, + { + "epoch": 0.1186, + "grad_norm": 0.3864613175392151, + "learning_rate": 0.00048530204514692787, + "loss": 2.9663776397705077, + "step": 23720 + }, + { + "epoch": 0.11865, + "grad_norm": 0.40076762437820435, + "learning_rate": 0.0004852886417408264, + "loss": 2.947001647949219, + "step": 23730 + }, + { + "epoch": 0.1187, + "grad_norm": 0.38359731435775757, + "learning_rate": 0.0004852752324113349, + "loss": 2.9444751739501953, + "step": 23740 + }, + { + "epoch": 0.11875, + "grad_norm": 0.38004738092422485, + "learning_rate": 0.00048526181715879095, + "loss": 2.8802413940429688, + "step": 23750 + }, + { + "epoch": 0.1188, + "grad_norm": 0.3814530074596405, + "learning_rate": 0.00048524839598353245, + "loss": 2.9105913162231447, + "step": 23760 + }, + { + "epoch": 0.11885, + "grad_norm": 0.4320646822452545, + "learning_rate": 0.00048523496888589705, + "loss": 2.9126398086547853, + "step": 23770 + }, + { + "epoch": 0.1189, + "grad_norm": 0.40601834654808044, + "learning_rate": 0.0004852215358662229, + "loss": 2.9044706344604494, + "step": 23780 + }, + { + "epoch": 0.11895, + "grad_norm": 0.37639930844306946, + "learning_rate": 0.0004852080969248482, + "loss": 2.885761260986328, + "step": 23790 + }, + { + "epoch": 0.119, + "grad_norm": 0.3756062984466553, + "learning_rate": 0.0004851946520621112, + "loss": 2.9019138336181642, + "step": 23800 + }, + { + "epoch": 0.11905, + "grad_norm": 0.3968018591403961, + "learning_rate": 0.00048518120127835043, + "loss": 2.9190073013305664, + "step": 23810 + }, + { + "epoch": 0.1191, + "grad_norm": 0.3788958787918091, + "learning_rate": 0.00048516774457390443, + "loss": 2.917402648925781, + "step": 23820 + }, + { + "epoch": 0.11915, + "grad_norm": 0.40075474977493286, + "learning_rate": 0.000485154281949112, + "loss": 2.9984432220458985, + "step": 23830 + }, + { + "epoch": 0.1192, + "grad_norm": 0.3675633370876312, + "learning_rate": 0.00048514081340431223, + "loss": 2.9844165802001954, + "step": 23840 + }, + { + "epoch": 0.11925, + "grad_norm": 0.3741064667701721, + "learning_rate": 0.00048512733893984393, + "loss": 2.930620574951172, + "step": 23850 + }, + { + "epoch": 0.1193, + "grad_norm": 0.39405813813209534, + "learning_rate": 0.00048511385855604653, + "loss": 2.9022830963134765, + "step": 23860 + }, + { + "epoch": 0.11935, + "grad_norm": 0.38760900497436523, + "learning_rate": 0.00048510037225325933, + "loss": 2.9268646240234375, + "step": 23870 + }, + { + "epoch": 0.1194, + "grad_norm": 0.3592962622642517, + "learning_rate": 0.0004850868800318218, + "loss": 2.9306079864501955, + "step": 23880 + }, + { + "epoch": 0.11945, + "grad_norm": 0.3687079846858978, + "learning_rate": 0.00048507338189207356, + "loss": 2.940926361083984, + "step": 23890 + }, + { + "epoch": 0.1195, + "grad_norm": 0.4152320325374603, + "learning_rate": 0.00048505987783435456, + "loss": 2.8968072891235352, + "step": 23900 + }, + { + "epoch": 0.11955, + "grad_norm": 0.4256918728351593, + "learning_rate": 0.0004850463678590047, + "loss": 2.913831901550293, + "step": 23910 + }, + { + "epoch": 0.1196, + "grad_norm": 0.3589555323123932, + "learning_rate": 0.0004850328519663641, + "loss": 2.930475425720215, + "step": 23920 + }, + { + "epoch": 0.11965, + "grad_norm": 0.3686433434486389, + "learning_rate": 0.000485019330156773, + "loss": 2.964383125305176, + "step": 23930 + }, + { + "epoch": 0.1197, + "grad_norm": 0.3976680636405945, + "learning_rate": 0.0004850058024305719, + "loss": 2.976864242553711, + "step": 23940 + }, + { + "epoch": 0.11975, + "grad_norm": 0.4207930266857147, + "learning_rate": 0.0004849922687881012, + "loss": 2.9516569137573243, + "step": 23950 + }, + { + "epoch": 0.1198, + "grad_norm": 0.38804569840431213, + "learning_rate": 0.00048497872922970175, + "loss": 2.939376449584961, + "step": 23960 + }, + { + "epoch": 0.11985, + "grad_norm": 0.39082181453704834, + "learning_rate": 0.00048496518375571444, + "loss": 2.8810823440551756, + "step": 23970 + }, + { + "epoch": 0.1199, + "grad_norm": 0.3874378502368927, + "learning_rate": 0.0004849516323664801, + "loss": 2.9144094467163084, + "step": 23980 + }, + { + "epoch": 0.11995, + "grad_norm": 0.3779364824295044, + "learning_rate": 0.00048493807506234, + "loss": 2.9535638809204103, + "step": 23990 + }, + { + "epoch": 0.12, + "grad_norm": 0.3774021565914154, + "learning_rate": 0.0004849245118436355, + "loss": 2.961012268066406, + "step": 24000 + }, + { + "epoch": 0.12005, + "grad_norm": 0.3763258159160614, + "learning_rate": 0.0004849109427107079, + "loss": 2.948090744018555, + "step": 24010 + }, + { + "epoch": 0.1201, + "grad_norm": 0.3896569609642029, + "learning_rate": 0.0004848973676638989, + "loss": 2.9164657592773438, + "step": 24020 + }, + { + "epoch": 0.12015, + "grad_norm": 0.3505999445915222, + "learning_rate": 0.00048488378670355026, + "loss": 2.9219524383544924, + "step": 24030 + }, + { + "epoch": 0.1202, + "grad_norm": 0.3892377018928528, + "learning_rate": 0.0004848701998300039, + "loss": 2.928410530090332, + "step": 24040 + }, + { + "epoch": 0.12025, + "grad_norm": 0.39750561118125916, + "learning_rate": 0.0004848566070436017, + "loss": 2.9223031997680664, + "step": 24050 + }, + { + "epoch": 0.1203, + "grad_norm": 0.38982856273651123, + "learning_rate": 0.0004848430083446861, + "loss": 2.963459777832031, + "step": 24060 + }, + { + "epoch": 0.12035, + "grad_norm": 0.37163785099983215, + "learning_rate": 0.0004848294037335993, + "loss": 2.8998870849609375, + "step": 24070 + }, + { + "epoch": 0.1204, + "grad_norm": 0.38996821641921997, + "learning_rate": 0.0004848157932106838, + "loss": 2.934904098510742, + "step": 24080 + }, + { + "epoch": 0.12045, + "grad_norm": 0.37869057059288025, + "learning_rate": 0.00048480217677628236, + "loss": 2.9785871505737305, + "step": 24090 + }, + { + "epoch": 0.1205, + "grad_norm": 0.3930337727069855, + "learning_rate": 0.0004847885544307376, + "loss": 2.9211685180664064, + "step": 24100 + }, + { + "epoch": 0.12055, + "grad_norm": 0.422197163105011, + "learning_rate": 0.0004847749261743926, + "loss": 2.9669977188110352, + "step": 24110 + }, + { + "epoch": 0.1206, + "grad_norm": 0.3996433615684509, + "learning_rate": 0.0004847612920075904, + "loss": 2.9288619995117187, + "step": 24120 + }, + { + "epoch": 0.12065, + "grad_norm": 0.4038184583187103, + "learning_rate": 0.0004847476519306742, + "loss": 2.8902456283569338, + "step": 24130 + }, + { + "epoch": 0.1207, + "grad_norm": 0.37306615710258484, + "learning_rate": 0.0004847340059439875, + "loss": 2.9690519332885743, + "step": 24140 + }, + { + "epoch": 0.12075, + "grad_norm": 0.39133667945861816, + "learning_rate": 0.0004847203540478737, + "loss": 2.928641700744629, + "step": 24150 + }, + { + "epoch": 0.1208, + "grad_norm": 0.3919186294078827, + "learning_rate": 0.0004847066962426766, + "loss": 2.914582443237305, + "step": 24160 + }, + { + "epoch": 0.12085, + "grad_norm": 0.375223308801651, + "learning_rate": 0.00048469303252873997, + "loss": 2.9059513092041014, + "step": 24170 + }, + { + "epoch": 0.1209, + "grad_norm": 0.4023188650608063, + "learning_rate": 0.0004846793629064078, + "loss": 2.9342660903930664, + "step": 24180 + }, + { + "epoch": 0.12095, + "grad_norm": 0.3559700548648834, + "learning_rate": 0.00048466568737602427, + "loss": 2.9674291610717773, + "step": 24190 + }, + { + "epoch": 0.121, + "grad_norm": 0.39848896861076355, + "learning_rate": 0.00048465200593793365, + "loss": 2.9443164825439454, + "step": 24200 + }, + { + "epoch": 0.12105, + "grad_norm": 0.3556888997554779, + "learning_rate": 0.0004846383185924803, + "loss": 2.984057807922363, + "step": 24210 + }, + { + "epoch": 0.1211, + "grad_norm": 0.38507020473480225, + "learning_rate": 0.00048462462534000885, + "loss": 2.9586841583251955, + "step": 24220 + }, + { + "epoch": 0.12115, + "grad_norm": 0.3651573657989502, + "learning_rate": 0.00048461092618086406, + "loss": 2.989575958251953, + "step": 24230 + }, + { + "epoch": 0.1212, + "grad_norm": 0.40412890911102295, + "learning_rate": 0.0004845972211153907, + "loss": 2.9174898147583006, + "step": 24240 + }, + { + "epoch": 0.12125, + "grad_norm": 0.4145865738391876, + "learning_rate": 0.000484583510143934, + "loss": 2.9182018280029296, + "step": 24250 + }, + { + "epoch": 0.1213, + "grad_norm": 0.3816022574901581, + "learning_rate": 0.00048456979326683884, + "loss": 2.8722606658935548, + "step": 24260 + }, + { + "epoch": 0.12135, + "grad_norm": 0.3692860007286072, + "learning_rate": 0.0004845560704844508, + "loss": 2.9296937942504884, + "step": 24270 + }, + { + "epoch": 0.1214, + "grad_norm": 0.35266897082328796, + "learning_rate": 0.0004845423417971152, + "loss": 2.9181900024414062, + "step": 24280 + }, + { + "epoch": 0.12145, + "grad_norm": 0.3620377779006958, + "learning_rate": 0.00048452860720517776, + "loss": 2.9670841217041017, + "step": 24290 + }, + { + "epoch": 0.1215, + "grad_norm": 0.4321539103984833, + "learning_rate": 0.0004845148667089842, + "loss": 2.993732452392578, + "step": 24300 + }, + { + "epoch": 0.12155, + "grad_norm": 0.4650925397872925, + "learning_rate": 0.0004845011203088804, + "loss": 2.919995880126953, + "step": 24310 + }, + { + "epoch": 0.1216, + "grad_norm": 0.39536863565444946, + "learning_rate": 0.00048448736800521254, + "loss": 2.9334062576293944, + "step": 24320 + }, + { + "epoch": 0.12165, + "grad_norm": 0.3698311150074005, + "learning_rate": 0.00048447360979832666, + "loss": 2.96732177734375, + "step": 24330 + }, + { + "epoch": 0.1217, + "grad_norm": 0.37484464049339294, + "learning_rate": 0.00048445984568856927, + "loss": 2.9173652648925783, + "step": 24340 + }, + { + "epoch": 0.12175, + "grad_norm": 0.3840221166610718, + "learning_rate": 0.00048444607567628683, + "loss": 2.953962707519531, + "step": 24350 + }, + { + "epoch": 0.1218, + "grad_norm": 0.411245197057724, + "learning_rate": 0.00048443229976182596, + "loss": 2.936417007446289, + "step": 24360 + }, + { + "epoch": 0.12185, + "grad_norm": 0.3661956489086151, + "learning_rate": 0.00048441851794553354, + "loss": 2.909365653991699, + "step": 24370 + }, + { + "epoch": 0.1219, + "grad_norm": 0.3728189468383789, + "learning_rate": 0.00048440473022775654, + "loss": 2.915207290649414, + "step": 24380 + }, + { + "epoch": 0.12195, + "grad_norm": 0.3943656086921692, + "learning_rate": 0.00048439093660884193, + "loss": 2.9605648040771486, + "step": 24390 + }, + { + "epoch": 0.122, + "grad_norm": 0.3744719326496124, + "learning_rate": 0.0004843771370891371, + "loss": 2.915353775024414, + "step": 24400 + }, + { + "epoch": 0.12205, + "grad_norm": 0.37849751114845276, + "learning_rate": 0.0004843633316689895, + "loss": 2.9311544418334963, + "step": 24410 + }, + { + "epoch": 0.1221, + "grad_norm": 0.36280906200408936, + "learning_rate": 0.0004843495203487465, + "loss": 2.9231176376342773, + "step": 24420 + }, + { + "epoch": 0.12215, + "grad_norm": 0.3618524372577667, + "learning_rate": 0.00048433570312875587, + "loss": 2.8863298416137697, + "step": 24430 + }, + { + "epoch": 0.1222, + "grad_norm": 0.4028128385543823, + "learning_rate": 0.0004843218800093655, + "loss": 2.931708335876465, + "step": 24440 + }, + { + "epoch": 0.12225, + "grad_norm": 0.37577247619628906, + "learning_rate": 0.0004843080509909234, + "loss": 2.919857406616211, + "step": 24450 + }, + { + "epoch": 0.1223, + "grad_norm": 0.3997471034526825, + "learning_rate": 0.00048429421607377765, + "loss": 2.9929128646850587, + "step": 24460 + }, + { + "epoch": 0.12235, + "grad_norm": 0.36080238223075867, + "learning_rate": 0.0004842803752582766, + "loss": 2.9666696548461915, + "step": 24470 + }, + { + "epoch": 0.1224, + "grad_norm": 0.3827887177467346, + "learning_rate": 0.0004842665285447686, + "loss": 2.8794086456298826, + "step": 24480 + }, + { + "epoch": 0.12245, + "grad_norm": 0.38621336221694946, + "learning_rate": 0.00048425267593360245, + "loss": 2.894801139831543, + "step": 24490 + }, + { + "epoch": 0.1225, + "grad_norm": 0.39020010828971863, + "learning_rate": 0.00048423881742512667, + "loss": 2.9280973434448243, + "step": 24500 + }, + { + "epoch": 0.12255, + "grad_norm": 0.3749274015426636, + "learning_rate": 0.0004842249530196902, + "loss": 2.9027387619018556, + "step": 24510 + }, + { + "epoch": 0.1226, + "grad_norm": 0.40575850009918213, + "learning_rate": 0.0004842110827176421, + "loss": 2.901422309875488, + "step": 24520 + }, + { + "epoch": 0.12265, + "grad_norm": 0.4050185978412628, + "learning_rate": 0.0004841972065193316, + "loss": 2.9307994842529297, + "step": 24530 + }, + { + "epoch": 0.1227, + "grad_norm": 0.40645408630371094, + "learning_rate": 0.00048418332442510794, + "loss": 2.9143301010131837, + "step": 24540 + }, + { + "epoch": 0.12275, + "grad_norm": 0.3865492045879364, + "learning_rate": 0.00048416943643532067, + "loss": 2.920848274230957, + "step": 24550 + }, + { + "epoch": 0.1228, + "grad_norm": 0.3739662766456604, + "learning_rate": 0.00048415554255031945, + "loss": 2.879867935180664, + "step": 24560 + }, + { + "epoch": 0.12285, + "grad_norm": 0.35446634888648987, + "learning_rate": 0.00048414164277045394, + "loss": 2.9578237533569336, + "step": 24570 + }, + { + "epoch": 0.1229, + "grad_norm": 0.3688444495201111, + "learning_rate": 0.00048412773709607417, + "loss": 2.9337112426757814, + "step": 24580 + }, + { + "epoch": 0.12295, + "grad_norm": 0.3762110471725464, + "learning_rate": 0.00048411382552753017, + "loss": 2.907183074951172, + "step": 24590 + }, + { + "epoch": 0.123, + "grad_norm": 0.40590569376945496, + "learning_rate": 0.00048409990806517225, + "loss": 2.9530349731445313, + "step": 24600 + }, + { + "epoch": 0.12305, + "grad_norm": 0.35577550530433655, + "learning_rate": 0.0004840859847093506, + "loss": 2.943728446960449, + "step": 24610 + }, + { + "epoch": 0.1231, + "grad_norm": 0.3960273861885071, + "learning_rate": 0.00048407205546041593, + "loss": 2.911103439331055, + "step": 24620 + }, + { + "epoch": 0.12315, + "grad_norm": 0.41008448600769043, + "learning_rate": 0.00048405812031871874, + "loss": 2.9274007797241213, + "step": 24630 + }, + { + "epoch": 0.1232, + "grad_norm": 0.37311387062072754, + "learning_rate": 0.0004840441792846101, + "loss": 2.8848669052124025, + "step": 24640 + }, + { + "epoch": 0.12325, + "grad_norm": 0.4076195955276489, + "learning_rate": 0.0004840302323584407, + "loss": 2.9941146850585936, + "step": 24650 + }, + { + "epoch": 0.1233, + "grad_norm": 0.3842070996761322, + "learning_rate": 0.0004840162795405617, + "loss": 2.9229219436645506, + "step": 24660 + }, + { + "epoch": 0.12335, + "grad_norm": 0.3525528907775879, + "learning_rate": 0.00048400232083132454, + "loss": 2.9148513793945314, + "step": 24670 + }, + { + "epoch": 0.1234, + "grad_norm": 0.4143275320529938, + "learning_rate": 0.00048398835623108047, + "loss": 2.894605827331543, + "step": 24680 + }, + { + "epoch": 0.12345, + "grad_norm": 0.39702698588371277, + "learning_rate": 0.00048397438574018103, + "loss": 2.951372528076172, + "step": 24690 + }, + { + "epoch": 0.1235, + "grad_norm": 0.428440123796463, + "learning_rate": 0.0004839604093589781, + "loss": 2.9468740463256835, + "step": 24700 + }, + { + "epoch": 0.12355, + "grad_norm": 0.42993414402008057, + "learning_rate": 0.00048394642708782346, + "loss": 2.935009002685547, + "step": 24710 + }, + { + "epoch": 0.1236, + "grad_norm": 0.4044754207134247, + "learning_rate": 0.00048393243892706897, + "loss": 2.920162391662598, + "step": 24720 + }, + { + "epoch": 0.12365, + "grad_norm": 0.34793180227279663, + "learning_rate": 0.00048391844487706695, + "loss": 2.9759092330932617, + "step": 24730 + }, + { + "epoch": 0.1237, + "grad_norm": 0.3645511269569397, + "learning_rate": 0.00048390444493816954, + "loss": 2.968445587158203, + "step": 24740 + }, + { + "epoch": 0.12375, + "grad_norm": 0.3929617702960968, + "learning_rate": 0.0004838904391107294, + "loss": 2.923129081726074, + "step": 24750 + }, + { + "epoch": 0.1238, + "grad_norm": 0.39007240533828735, + "learning_rate": 0.000483876427395099, + "loss": 2.915370559692383, + "step": 24760 + }, + { + "epoch": 0.12385, + "grad_norm": 0.37806037068367004, + "learning_rate": 0.00048386240979163107, + "loss": 2.9257686614990233, + "step": 24770 + }, + { + "epoch": 0.1239, + "grad_norm": 0.4271574020385742, + "learning_rate": 0.0004838483863006786, + "loss": 2.98135986328125, + "step": 24780 + }, + { + "epoch": 0.12395, + "grad_norm": 0.40575969219207764, + "learning_rate": 0.00048383435692259455, + "loss": 2.939146041870117, + "step": 24790 + }, + { + "epoch": 0.124, + "grad_norm": 0.3795890808105469, + "learning_rate": 0.0004838203216577321, + "loss": 2.96920223236084, + "step": 24800 + }, + { + "epoch": 0.12405, + "grad_norm": 0.377956748008728, + "learning_rate": 0.0004838062805064446, + "loss": 2.9034212112426756, + "step": 24810 + }, + { + "epoch": 0.1241, + "grad_norm": 0.3846381604671478, + "learning_rate": 0.00048379223346908557, + "loss": 2.9190067291259765, + "step": 24820 + }, + { + "epoch": 0.12415, + "grad_norm": 0.3645409047603607, + "learning_rate": 0.0004837781805460086, + "loss": 2.950938415527344, + "step": 24830 + }, + { + "epoch": 0.1242, + "grad_norm": 0.38319143652915955, + "learning_rate": 0.0004837641217375676, + "loss": 2.9796985626220702, + "step": 24840 + }, + { + "epoch": 0.12425, + "grad_norm": 0.34307271242141724, + "learning_rate": 0.0004837500570441163, + "loss": 2.922085189819336, + "step": 24850 + }, + { + "epoch": 0.1243, + "grad_norm": 0.37723180651664734, + "learning_rate": 0.0004837359864660089, + "loss": 2.94166259765625, + "step": 24860 + }, + { + "epoch": 0.12435, + "grad_norm": 0.4178781807422638, + "learning_rate": 0.00048372191000359955, + "loss": 2.9443428039550783, + "step": 24870 + }, + { + "epoch": 0.1244, + "grad_norm": 0.3446047008037567, + "learning_rate": 0.00048370782765724276, + "loss": 2.907028579711914, + "step": 24880 + }, + { + "epoch": 0.12445, + "grad_norm": 0.4187827408313751, + "learning_rate": 0.00048369373942729293, + "loss": 2.9238943099975585, + "step": 24890 + }, + { + "epoch": 0.1245, + "grad_norm": 0.38719427585601807, + "learning_rate": 0.0004836796453141048, + "loss": 2.956632614135742, + "step": 24900 + }, + { + "epoch": 0.12455, + "grad_norm": 0.3935578763484955, + "learning_rate": 0.0004836655453180331, + "loss": 2.9292428970336912, + "step": 24910 + }, + { + "epoch": 0.1246, + "grad_norm": 0.3601497411727905, + "learning_rate": 0.0004836514394394329, + "loss": 2.977448272705078, + "step": 24920 + }, + { + "epoch": 0.12465, + "grad_norm": 0.40698331594467163, + "learning_rate": 0.0004836373276786593, + "loss": 2.9344858169555663, + "step": 24930 + }, + { + "epoch": 0.1247, + "grad_norm": 0.39933693408966064, + "learning_rate": 0.0004836232100360675, + "loss": 2.9379825592041016, + "step": 24940 + }, + { + "epoch": 0.12475, + "grad_norm": 0.37646812200546265, + "learning_rate": 0.00048360908651201296, + "loss": 2.9225061416625975, + "step": 24950 + }, + { + "epoch": 0.1248, + "grad_norm": 0.3653898537158966, + "learning_rate": 0.00048359495710685124, + "loss": 2.9551626205444337, + "step": 24960 + }, + { + "epoch": 0.12485, + "grad_norm": 0.36028265953063965, + "learning_rate": 0.00048358082182093796, + "loss": 2.973032760620117, + "step": 24970 + }, + { + "epoch": 0.1249, + "grad_norm": 0.4115538001060486, + "learning_rate": 0.0004835666806546292, + "loss": 2.9380069732666017, + "step": 24980 + }, + { + "epoch": 0.12495, + "grad_norm": 0.3713703453540802, + "learning_rate": 0.00048355253360828066, + "loss": 2.9226757049560548, + "step": 24990 + }, + { + "epoch": 0.125, + "grad_norm": 0.3941385746002197, + "learning_rate": 0.00048353838068224874, + "loss": 2.889577293395996, + "step": 25000 + }, + { + "epoch": 0.12505, + "grad_norm": 0.4079676866531372, + "learning_rate": 0.0004835242218768896, + "loss": 2.929517936706543, + "step": 25010 + }, + { + "epoch": 0.1251, + "grad_norm": 0.3980456590652466, + "learning_rate": 0.0004835100571925598, + "loss": 2.963794136047363, + "step": 25020 + }, + { + "epoch": 0.12515, + "grad_norm": 0.41744351387023926, + "learning_rate": 0.0004834958866296159, + "loss": 2.943686866760254, + "step": 25030 + }, + { + "epoch": 0.1252, + "grad_norm": 0.4013965427875519, + "learning_rate": 0.0004834817101884145, + "loss": 2.899832916259766, + "step": 25040 + }, + { + "epoch": 0.12525, + "grad_norm": 0.40527066588401794, + "learning_rate": 0.00048346752786931256, + "loss": 2.9321256637573243, + "step": 25050 + }, + { + "epoch": 0.1253, + "grad_norm": 0.41807302832603455, + "learning_rate": 0.0004834533396726673, + "loss": 2.970759963989258, + "step": 25060 + }, + { + "epoch": 0.12535, + "grad_norm": 0.35611963272094727, + "learning_rate": 0.0004834391455988357, + "loss": 2.968821334838867, + "step": 25070 + }, + { + "epoch": 0.1254, + "grad_norm": 0.3648330569267273, + "learning_rate": 0.00048342494564817526, + "loss": 2.8587356567382813, + "step": 25080 + }, + { + "epoch": 0.12545, + "grad_norm": 0.35936662554740906, + "learning_rate": 0.0004834107398210433, + "loss": 2.944087028503418, + "step": 25090 + }, + { + "epoch": 0.1255, + "grad_norm": 0.3573298156261444, + "learning_rate": 0.00048339652811779757, + "loss": 2.901480865478516, + "step": 25100 + }, + { + "epoch": 0.12555, + "grad_norm": 0.3999618589878082, + "learning_rate": 0.0004833823105387957, + "loss": 2.914226531982422, + "step": 25110 + }, + { + "epoch": 0.1256, + "grad_norm": 0.408074289560318, + "learning_rate": 0.0004833680870843958, + "loss": 2.9036014556884764, + "step": 25120 + }, + { + "epoch": 0.12565, + "grad_norm": 0.3958965539932251, + "learning_rate": 0.0004833538577549559, + "loss": 2.9564064025878904, + "step": 25130 + }, + { + "epoch": 0.1257, + "grad_norm": 0.3452024757862091, + "learning_rate": 0.0004833396225508341, + "loss": 2.966351318359375, + "step": 25140 + }, + { + "epoch": 0.12575, + "grad_norm": 0.35038048028945923, + "learning_rate": 0.00048332538147238894, + "loss": 2.9230175018310547, + "step": 25150 + }, + { + "epoch": 0.1258, + "grad_norm": 0.3642018735408783, + "learning_rate": 0.00048331113451997876, + "loss": 2.9541549682617188, + "step": 25160 + }, + { + "epoch": 0.12585, + "grad_norm": 0.3690115213394165, + "learning_rate": 0.00048329688169396244, + "loss": 2.9471881866455076, + "step": 25170 + }, + { + "epoch": 0.1259, + "grad_norm": 0.3646773397922516, + "learning_rate": 0.0004832826229946986, + "loss": 2.892619323730469, + "step": 25180 + }, + { + "epoch": 0.12595, + "grad_norm": 0.35427531599998474, + "learning_rate": 0.00048326835842254633, + "loss": 2.9195465087890624, + "step": 25190 + }, + { + "epoch": 0.126, + "grad_norm": 0.36220112442970276, + "learning_rate": 0.00048325408797786476, + "loss": 2.9373201370239257, + "step": 25200 + }, + { + "epoch": 0.12605, + "grad_norm": 0.3916666507720947, + "learning_rate": 0.0004832398116610129, + "loss": 2.915491485595703, + "step": 25210 + }, + { + "epoch": 0.1261, + "grad_norm": 0.38635194301605225, + "learning_rate": 0.00048322552947235056, + "loss": 2.893488121032715, + "step": 25220 + }, + { + "epoch": 0.12615, + "grad_norm": 0.34761378169059753, + "learning_rate": 0.0004832112414122369, + "loss": 2.916173553466797, + "step": 25230 + }, + { + "epoch": 0.1262, + "grad_norm": 0.41677039861679077, + "learning_rate": 0.0004831969474810319, + "loss": 2.945475387573242, + "step": 25240 + }, + { + "epoch": 0.12625, + "grad_norm": 0.38130247592926025, + "learning_rate": 0.0004831826476790953, + "loss": 2.8997859954833984, + "step": 25250 + }, + { + "epoch": 0.1263, + "grad_norm": 0.40529918670654297, + "learning_rate": 0.00048316834200678706, + "loss": 2.9605487823486327, + "step": 25260 + }, + { + "epoch": 0.12635, + "grad_norm": 0.37259137630462646, + "learning_rate": 0.0004831540304644674, + "loss": 2.9491729736328125, + "step": 25270 + }, + { + "epoch": 0.1264, + "grad_norm": 0.40515971183776855, + "learning_rate": 0.0004831397130524965, + "loss": 2.955152702331543, + "step": 25280 + }, + { + "epoch": 0.12645, + "grad_norm": 0.36482810974121094, + "learning_rate": 0.0004831253897712349, + "loss": 2.963387107849121, + "step": 25290 + }, + { + "epoch": 0.1265, + "grad_norm": 0.3842472434043884, + "learning_rate": 0.0004831110606210433, + "loss": 2.916176986694336, + "step": 25300 + }, + { + "epoch": 0.12655, + "grad_norm": 0.35783642530441284, + "learning_rate": 0.00048309672560228217, + "loss": 2.9568052291870117, + "step": 25310 + }, + { + "epoch": 0.1266, + "grad_norm": 0.38829681277275085, + "learning_rate": 0.00048308238471531263, + "loss": 2.910077667236328, + "step": 25320 + }, + { + "epoch": 0.12665, + "grad_norm": 0.38621917366981506, + "learning_rate": 0.0004830680379604955, + "loss": 2.9370986938476564, + "step": 25330 + }, + { + "epoch": 0.1267, + "grad_norm": 0.367095947265625, + "learning_rate": 0.0004830536853381922, + "loss": 2.9086864471435545, + "step": 25340 + }, + { + "epoch": 0.12675, + "grad_norm": 0.3923152983188629, + "learning_rate": 0.0004830393268487638, + "loss": 2.9065433502197267, + "step": 25350 + }, + { + "epoch": 0.1268, + "grad_norm": 0.38525381684303284, + "learning_rate": 0.000483024962492572, + "loss": 2.960127067565918, + "step": 25360 + }, + { + "epoch": 0.12685, + "grad_norm": 0.35653185844421387, + "learning_rate": 0.00048301059226997826, + "loss": 2.888186454772949, + "step": 25370 + }, + { + "epoch": 0.1269, + "grad_norm": 0.38863855600357056, + "learning_rate": 0.0004829962161813445, + "loss": 2.9761774063110353, + "step": 25380 + }, + { + "epoch": 0.12695, + "grad_norm": 0.36388564109802246, + "learning_rate": 0.00048298183422703245, + "loss": 2.933695411682129, + "step": 25390 + }, + { + "epoch": 0.127, + "grad_norm": 0.3735101819038391, + "learning_rate": 0.0004829674464074043, + "loss": 2.905988883972168, + "step": 25400 + }, + { + "epoch": 0.12705, + "grad_norm": 0.4026358127593994, + "learning_rate": 0.0004829530527228223, + "loss": 2.9233938217163087, + "step": 25410 + }, + { + "epoch": 0.1271, + "grad_norm": 0.4099673926830292, + "learning_rate": 0.0004829386531736487, + "loss": 2.914580154418945, + "step": 25420 + }, + { + "epoch": 0.12715, + "grad_norm": 0.3959093987941742, + "learning_rate": 0.0004829242477602461, + "loss": 2.9628372192382812, + "step": 25430 + }, + { + "epoch": 0.1272, + "grad_norm": 0.3556850254535675, + "learning_rate": 0.00048290983648297713, + "loss": 2.835637092590332, + "step": 25440 + }, + { + "epoch": 0.12725, + "grad_norm": 0.38089144229888916, + "learning_rate": 0.0004828954193422046, + "loss": 2.962131309509277, + "step": 25450 + }, + { + "epoch": 0.1273, + "grad_norm": 0.3645009994506836, + "learning_rate": 0.0004828809963382914, + "loss": 2.932674217224121, + "step": 25460 + }, + { + "epoch": 0.12735, + "grad_norm": 0.36985069513320923, + "learning_rate": 0.00048286656747160067, + "loss": 2.9281055450439455, + "step": 25470 + }, + { + "epoch": 0.1274, + "grad_norm": 0.38646945357322693, + "learning_rate": 0.0004828521327424957, + "loss": 2.919073295593262, + "step": 25480 + }, + { + "epoch": 0.12745, + "grad_norm": 0.3851325511932373, + "learning_rate": 0.0004828376921513398, + "loss": 2.948923873901367, + "step": 25490 + }, + { + "epoch": 0.1275, + "grad_norm": 0.4039326608181, + "learning_rate": 0.00048282324569849655, + "loss": 2.9479923248291016, + "step": 25500 + }, + { + "epoch": 0.12755, + "grad_norm": 0.35113438963890076, + "learning_rate": 0.0004828087933843297, + "loss": 2.9388395309448243, + "step": 25510 + }, + { + "epoch": 0.1276, + "grad_norm": 0.36628690361976624, + "learning_rate": 0.000482794335209203, + "loss": 2.921650695800781, + "step": 25520 + }, + { + "epoch": 0.12765, + "grad_norm": 0.35066190361976624, + "learning_rate": 0.00048277987117348043, + "loss": 2.8659873962402345, + "step": 25530 + }, + { + "epoch": 0.1277, + "grad_norm": 0.40122726559638977, + "learning_rate": 0.0004827654012775262, + "loss": 2.9750993728637694, + "step": 25540 + }, + { + "epoch": 0.12775, + "grad_norm": 0.4178300201892853, + "learning_rate": 0.0004827509255217046, + "loss": 2.905013084411621, + "step": 25550 + }, + { + "epoch": 0.1278, + "grad_norm": 0.3999781012535095, + "learning_rate": 0.00048273644390637996, + "loss": 2.9282804489135743, + "step": 25560 + }, + { + "epoch": 0.12785, + "grad_norm": 0.3853323459625244, + "learning_rate": 0.0004827219564319169, + "loss": 2.9884571075439452, + "step": 25570 + }, + { + "epoch": 0.1279, + "grad_norm": 0.3865911066532135, + "learning_rate": 0.0004827074630986801, + "loss": 2.8976982116699217, + "step": 25580 + }, + { + "epoch": 0.12795, + "grad_norm": 0.3718714714050293, + "learning_rate": 0.0004826929639070345, + "loss": 2.9841941833496093, + "step": 25590 + }, + { + "epoch": 0.128, + "grad_norm": 0.3548414409160614, + "learning_rate": 0.0004826784588573451, + "loss": 2.943600082397461, + "step": 25600 + }, + { + "epoch": 0.12805, + "grad_norm": 0.3726412057876587, + "learning_rate": 0.000482663947949977, + "loss": 2.9692853927612304, + "step": 25610 + }, + { + "epoch": 0.1281, + "grad_norm": 0.39021334052085876, + "learning_rate": 0.00048264943118529556, + "loss": 2.9247642517089845, + "step": 25620 + }, + { + "epoch": 0.12815, + "grad_norm": 0.3818647265434265, + "learning_rate": 0.0004826349085636663, + "loss": 2.8802480697631836, + "step": 25630 + }, + { + "epoch": 0.1282, + "grad_norm": 0.37887436151504517, + "learning_rate": 0.0004826203800854548, + "loss": 2.9527584075927735, + "step": 25640 + }, + { + "epoch": 0.12825, + "grad_norm": 0.3529016375541687, + "learning_rate": 0.0004826058457510267, + "loss": 2.884227180480957, + "step": 25650 + }, + { + "epoch": 0.1283, + "grad_norm": 0.3593159317970276, + "learning_rate": 0.00048259130556074813, + "loss": 2.9424612045288088, + "step": 25660 + }, + { + "epoch": 0.12835, + "grad_norm": 0.38066157698631287, + "learning_rate": 0.00048257675951498494, + "loss": 2.9112098693847654, + "step": 25670 + }, + { + "epoch": 0.1284, + "grad_norm": 0.36334919929504395, + "learning_rate": 0.0004825622076141033, + "loss": 2.9145391464233397, + "step": 25680 + }, + { + "epoch": 0.12845, + "grad_norm": 0.4002946615219116, + "learning_rate": 0.00048254764985846966, + "loss": 2.936285400390625, + "step": 25690 + }, + { + "epoch": 0.1285, + "grad_norm": 0.41038286685943604, + "learning_rate": 0.0004825330862484505, + "loss": 2.9593742370605467, + "step": 25700 + }, + { + "epoch": 0.12855, + "grad_norm": 0.39942336082458496, + "learning_rate": 0.0004825185167844125, + "loss": 2.9608592987060547, + "step": 25710 + }, + { + "epoch": 0.1286, + "grad_norm": 0.4023154079914093, + "learning_rate": 0.0004825039414667224, + "loss": 2.9312007904052733, + "step": 25720 + }, + { + "epoch": 0.12865, + "grad_norm": 0.35401004552841187, + "learning_rate": 0.000482489360295747, + "loss": 2.914395332336426, + "step": 25730 + }, + { + "epoch": 0.1287, + "grad_norm": 0.3873436152935028, + "learning_rate": 0.0004824747732718536, + "loss": 2.961992073059082, + "step": 25740 + }, + { + "epoch": 0.12875, + "grad_norm": 0.37848296761512756, + "learning_rate": 0.0004824601803954093, + "loss": 2.9403491973876954, + "step": 25750 + }, + { + "epoch": 0.1288, + "grad_norm": 0.35948893427848816, + "learning_rate": 0.00048244558166678155, + "loss": 2.9565414428710937, + "step": 25760 + }, + { + "epoch": 0.12885, + "grad_norm": 0.3903636336326599, + "learning_rate": 0.0004824309770863379, + "loss": 2.9032590866088865, + "step": 25770 + }, + { + "epoch": 0.1289, + "grad_norm": 0.3896348476409912, + "learning_rate": 0.00048241636665444584, + "loss": 2.9412906646728514, + "step": 25780 + }, + { + "epoch": 0.12895, + "grad_norm": 0.41556695103645325, + "learning_rate": 0.0004824017503714733, + "loss": 2.9468523025512696, + "step": 25790 + }, + { + "epoch": 0.129, + "grad_norm": 0.36736804246902466, + "learning_rate": 0.0004823871282377883, + "loss": 2.9044778823852537, + "step": 25800 + }, + { + "epoch": 0.12905, + "grad_norm": 0.3872169554233551, + "learning_rate": 0.0004823725002537589, + "loss": 2.9621082305908204, + "step": 25810 + }, + { + "epoch": 0.1291, + "grad_norm": 0.37101981043815613, + "learning_rate": 0.0004823578664197533, + "loss": 2.8935441970825195, + "step": 25820 + }, + { + "epoch": 0.12915, + "grad_norm": 0.4051969051361084, + "learning_rate": 0.00048234322673614005, + "loss": 2.88111572265625, + "step": 25830 + }, + { + "epoch": 0.1292, + "grad_norm": 0.3589498698711395, + "learning_rate": 0.0004823285812032876, + "loss": 2.9262786865234376, + "step": 25840 + }, + { + "epoch": 0.12925, + "grad_norm": 0.3987925946712494, + "learning_rate": 0.00048231392982156463, + "loss": 2.941716957092285, + "step": 25850 + }, + { + "epoch": 0.1293, + "grad_norm": 0.36939510703086853, + "learning_rate": 0.0004822992725913401, + "loss": 2.915989112854004, + "step": 25860 + }, + { + "epoch": 0.12935, + "grad_norm": 0.437553733587265, + "learning_rate": 0.00048228460951298287, + "loss": 2.909589958190918, + "step": 25870 + }, + { + "epoch": 0.1294, + "grad_norm": 0.3723601698875427, + "learning_rate": 0.00048226994058686215, + "loss": 2.916920471191406, + "step": 25880 + }, + { + "epoch": 0.12945, + "grad_norm": 0.36218011379241943, + "learning_rate": 0.0004822552658133472, + "loss": 2.913115310668945, + "step": 25890 + }, + { + "epoch": 0.1295, + "grad_norm": 0.3802091181278229, + "learning_rate": 0.00048224058519280756, + "loss": 2.9344160079956056, + "step": 25900 + }, + { + "epoch": 0.12955, + "grad_norm": 0.36200597882270813, + "learning_rate": 0.00048222589872561273, + "loss": 2.959825325012207, + "step": 25910 + }, + { + "epoch": 0.1296, + "grad_norm": 0.36687031388282776, + "learning_rate": 0.0004822112064121324, + "loss": 2.934452247619629, + "step": 25920 + }, + { + "epoch": 0.12965, + "grad_norm": 0.38048774003982544, + "learning_rate": 0.00048219650825273656, + "loss": 2.9636857986450194, + "step": 25930 + }, + { + "epoch": 0.1297, + "grad_norm": 0.3896191120147705, + "learning_rate": 0.00048218180424779513, + "loss": 2.906309127807617, + "step": 25940 + }, + { + "epoch": 0.12975, + "grad_norm": 0.3850070536136627, + "learning_rate": 0.00048216709439767835, + "loss": 2.895444869995117, + "step": 25950 + }, + { + "epoch": 0.1298, + "grad_norm": 0.4312337636947632, + "learning_rate": 0.0004821523787027565, + "loss": 2.842441368103027, + "step": 25960 + }, + { + "epoch": 0.12985, + "grad_norm": 0.39755547046661377, + "learning_rate": 0.0004821376571634001, + "loss": 2.9080467224121094, + "step": 25970 + }, + { + "epoch": 0.1299, + "grad_norm": 0.3715350925922394, + "learning_rate": 0.00048212292977997966, + "loss": 2.9773067474365233, + "step": 25980 + }, + { + "epoch": 0.12995, + "grad_norm": 0.3974193334579468, + "learning_rate": 0.0004821081965528661, + "loss": 2.9387767791748045, + "step": 25990 + }, + { + "epoch": 0.13, + "grad_norm": 0.39847761392593384, + "learning_rate": 0.00048209345748243015, + "loss": 2.8553083419799803, + "step": 26000 + }, + { + "epoch": 0.13005, + "grad_norm": 0.3908698558807373, + "learning_rate": 0.00048207871256904304, + "loss": 2.915854644775391, + "step": 26010 + }, + { + "epoch": 0.1301, + "grad_norm": 0.36001625657081604, + "learning_rate": 0.0004820639618130759, + "loss": 2.8996625900268556, + "step": 26020 + }, + { + "epoch": 0.13015, + "grad_norm": 0.37921369075775146, + "learning_rate": 0.00048204920521490005, + "loss": 2.8859889984130858, + "step": 26030 + }, + { + "epoch": 0.1302, + "grad_norm": 0.42857033014297485, + "learning_rate": 0.00048203444277488696, + "loss": 2.9189876556396483, + "step": 26040 + }, + { + "epoch": 0.13025, + "grad_norm": 0.39198413491249084, + "learning_rate": 0.00048201967449340837, + "loss": 2.909371185302734, + "step": 26050 + }, + { + "epoch": 0.1303, + "grad_norm": 0.34295952320098877, + "learning_rate": 0.000482004900370836, + "loss": 2.9016204833984376, + "step": 26060 + }, + { + "epoch": 0.13035, + "grad_norm": 0.3926781415939331, + "learning_rate": 0.00048199012040754186, + "loss": 2.910625457763672, + "step": 26070 + }, + { + "epoch": 0.1304, + "grad_norm": 0.38304418325424194, + "learning_rate": 0.00048197533460389797, + "loss": 2.9506900787353514, + "step": 26080 + }, + { + "epoch": 0.13045, + "grad_norm": 0.3693363070487976, + "learning_rate": 0.0004819605429602766, + "loss": 2.9768503189086912, + "step": 26090 + }, + { + "epoch": 0.1305, + "grad_norm": 0.35057973861694336, + "learning_rate": 0.00048194574547705005, + "loss": 2.929116058349609, + "step": 26100 + }, + { + "epoch": 0.13055, + "grad_norm": 0.3576221466064453, + "learning_rate": 0.00048193094215459093, + "loss": 2.952048110961914, + "step": 26110 + }, + { + "epoch": 0.1306, + "grad_norm": 0.37960636615753174, + "learning_rate": 0.0004819161329932719, + "loss": 2.930974578857422, + "step": 26120 + }, + { + "epoch": 0.13065, + "grad_norm": 0.3997630476951599, + "learning_rate": 0.00048190131799346583, + "loss": 2.9417373657226564, + "step": 26130 + }, + { + "epoch": 0.1307, + "grad_norm": 0.38749611377716064, + "learning_rate": 0.0004818864971555456, + "loss": 2.8978574752807615, + "step": 26140 + }, + { + "epoch": 0.13075, + "grad_norm": 0.36949124932289124, + "learning_rate": 0.0004818716704798842, + "loss": 2.9226810455322267, + "step": 26150 + }, + { + "epoch": 0.1308, + "grad_norm": 0.36305660009384155, + "learning_rate": 0.0004818568379668552, + "loss": 2.9190425872802734, + "step": 26160 + }, + { + "epoch": 0.13085, + "grad_norm": 0.37082019448280334, + "learning_rate": 0.00048184199961683185, + "loss": 2.943851089477539, + "step": 26170 + }, + { + "epoch": 0.1309, + "grad_norm": 0.3941890299320221, + "learning_rate": 0.0004818271554301877, + "loss": 2.900010108947754, + "step": 26180 + }, + { + "epoch": 0.13095, + "grad_norm": 0.4069499671459198, + "learning_rate": 0.00048181230540729643, + "loss": 2.9562116622924806, + "step": 26190 + }, + { + "epoch": 0.131, + "grad_norm": 0.3897918164730072, + "learning_rate": 0.0004817974495485319, + "loss": 2.934457778930664, + "step": 26200 + }, + { + "epoch": 0.13105, + "grad_norm": 0.40421611070632935, + "learning_rate": 0.0004817825878542682, + "loss": 2.92305850982666, + "step": 26210 + }, + { + "epoch": 0.1311, + "grad_norm": 0.3634985387325287, + "learning_rate": 0.00048176772032487936, + "loss": 2.989968109130859, + "step": 26220 + }, + { + "epoch": 0.13115, + "grad_norm": 0.38659584522247314, + "learning_rate": 0.0004817528469607397, + "loss": 2.9242210388183594, + "step": 26230 + }, + { + "epoch": 0.1312, + "grad_norm": 0.38555893301963806, + "learning_rate": 0.0004817379677622237, + "loss": 2.941947174072266, + "step": 26240 + }, + { + "epoch": 0.13125, + "grad_norm": 0.37181463837623596, + "learning_rate": 0.0004817230827297059, + "loss": 2.957193374633789, + "step": 26250 + }, + { + "epoch": 0.1313, + "grad_norm": 0.3887757360935211, + "learning_rate": 0.000481708191863561, + "loss": 2.967807579040527, + "step": 26260 + }, + { + "epoch": 0.13135, + "grad_norm": 0.3737315833568573, + "learning_rate": 0.00048169329516416394, + "loss": 2.9173418045043946, + "step": 26270 + }, + { + "epoch": 0.1314, + "grad_norm": 0.43761786818504333, + "learning_rate": 0.0004816783926318897, + "loss": 2.9250205993652343, + "step": 26280 + }, + { + "epoch": 0.13145, + "grad_norm": 0.38494566082954407, + "learning_rate": 0.00048166348426711356, + "loss": 2.8752952575683595, + "step": 26290 + }, + { + "epoch": 0.1315, + "grad_norm": 0.3748131990432739, + "learning_rate": 0.0004816485700702107, + "loss": 2.8636598587036133, + "step": 26300 + }, + { + "epoch": 0.13155, + "grad_norm": 0.3775962293148041, + "learning_rate": 0.0004816336500415566, + "loss": 2.917586898803711, + "step": 26310 + }, + { + "epoch": 0.1316, + "grad_norm": 0.3883666694164276, + "learning_rate": 0.000481618724181527, + "loss": 2.9238271713256836, + "step": 26320 + }, + { + "epoch": 0.13165, + "grad_norm": 0.3692477345466614, + "learning_rate": 0.0004816037924904975, + "loss": 2.9589542388916015, + "step": 26330 + }, + { + "epoch": 0.1317, + "grad_norm": 0.38963639736175537, + "learning_rate": 0.000481588854968844, + "loss": 2.843256950378418, + "step": 26340 + }, + { + "epoch": 0.13175, + "grad_norm": 0.3756103813648224, + "learning_rate": 0.00048157391161694275, + "loss": 2.8919435501098634, + "step": 26350 + }, + { + "epoch": 0.1318, + "grad_norm": 0.3586216866970062, + "learning_rate": 0.00048155896243516974, + "loss": 2.9471736907958985, + "step": 26360 + }, + { + "epoch": 0.13185, + "grad_norm": 0.3795343041419983, + "learning_rate": 0.0004815440074239015, + "loss": 2.965281677246094, + "step": 26370 + }, + { + "epoch": 0.1319, + "grad_norm": 0.3633275330066681, + "learning_rate": 0.00048152904658351424, + "loss": 2.947150993347168, + "step": 26380 + }, + { + "epoch": 0.13195, + "grad_norm": 0.34889230132102966, + "learning_rate": 0.0004815140799143849, + "loss": 2.935506820678711, + "step": 26390 + }, + { + "epoch": 0.132, + "grad_norm": 0.3896504342556, + "learning_rate": 0.0004814991074168902, + "loss": 2.9159189224243165, + "step": 26400 + }, + { + "epoch": 0.13205, + "grad_norm": 0.38388174772262573, + "learning_rate": 0.00048148412909140696, + "loss": 2.913478660583496, + "step": 26410 + }, + { + "epoch": 0.1321, + "grad_norm": 0.3620969355106354, + "learning_rate": 0.00048146914493831227, + "loss": 2.935799407958984, + "step": 26420 + }, + { + "epoch": 0.13215, + "grad_norm": 0.38049206137657166, + "learning_rate": 0.00048145415495798343, + "loss": 2.9256874084472657, + "step": 26430 + }, + { + "epoch": 0.1322, + "grad_norm": 0.33597010374069214, + "learning_rate": 0.00048143915915079786, + "loss": 2.936966896057129, + "step": 26440 + }, + { + "epoch": 0.13225, + "grad_norm": 0.3881630003452301, + "learning_rate": 0.00048142415751713285, + "loss": 2.9113115310668944, + "step": 26450 + }, + { + "epoch": 0.1323, + "grad_norm": 0.3862500786781311, + "learning_rate": 0.00048140915005736635, + "loss": 2.848940849304199, + "step": 26460 + }, + { + "epoch": 0.13235, + "grad_norm": 0.3915036916732788, + "learning_rate": 0.00048139413677187606, + "loss": 2.985984992980957, + "step": 26470 + }, + { + "epoch": 0.1324, + "grad_norm": 0.36598828434944153, + "learning_rate": 0.0004813791176610398, + "loss": 2.9263511657714845, + "step": 26480 + }, + { + "epoch": 0.13245, + "grad_norm": 0.3784525692462921, + "learning_rate": 0.0004813640927252358, + "loss": 2.9331010818481444, + "step": 26490 + }, + { + "epoch": 0.1325, + "grad_norm": 0.35784775018692017, + "learning_rate": 0.00048134906196484244, + "loss": 2.928114318847656, + "step": 26500 + }, + { + "epoch": 0.13255, + "grad_norm": 0.3611900806427002, + "learning_rate": 0.000481334025380238, + "loss": 2.936406135559082, + "step": 26510 + }, + { + "epoch": 0.1326, + "grad_norm": 0.35729193687438965, + "learning_rate": 0.00048131898297180085, + "loss": 2.9902883529663087, + "step": 26520 + }, + { + "epoch": 0.13265, + "grad_norm": 0.37757986783981323, + "learning_rate": 0.00048130393473990997, + "loss": 2.8406230926513674, + "step": 26530 + }, + { + "epoch": 0.1327, + "grad_norm": 0.359306663274765, + "learning_rate": 0.0004812888806849441, + "loss": 2.954698371887207, + "step": 26540 + }, + { + "epoch": 0.13275, + "grad_norm": 0.3711787164211273, + "learning_rate": 0.00048127382080728217, + "loss": 2.8910972595214846, + "step": 26550 + }, + { + "epoch": 0.1328, + "grad_norm": 0.36134472489356995, + "learning_rate": 0.0004812587551073033, + "loss": 2.933359909057617, + "step": 26560 + }, + { + "epoch": 0.13285, + "grad_norm": 0.39201200008392334, + "learning_rate": 0.0004812436835853868, + "loss": 2.909613037109375, + "step": 26570 + }, + { + "epoch": 0.1329, + "grad_norm": 0.3915193974971771, + "learning_rate": 0.00048122860624191224, + "loss": 2.884525680541992, + "step": 26580 + }, + { + "epoch": 0.13295, + "grad_norm": 0.39224475622177124, + "learning_rate": 0.00048121352307725895, + "loss": 2.9488689422607424, + "step": 26590 + }, + { + "epoch": 0.133, + "grad_norm": 0.37063509225845337, + "learning_rate": 0.00048119843409180684, + "loss": 2.884670639038086, + "step": 26600 + }, + { + "epoch": 0.13305, + "grad_norm": 0.3655090034008026, + "learning_rate": 0.0004811833392859356, + "loss": 2.851345443725586, + "step": 26610 + }, + { + "epoch": 0.1331, + "grad_norm": 0.36124956607818604, + "learning_rate": 0.00048116823866002547, + "loss": 2.906117630004883, + "step": 26620 + }, + { + "epoch": 0.13315, + "grad_norm": 0.40290746092796326, + "learning_rate": 0.00048115313221445635, + "loss": 2.930743408203125, + "step": 26630 + }, + { + "epoch": 0.1332, + "grad_norm": 0.39738765358924866, + "learning_rate": 0.0004811380199496087, + "loss": 2.9508214950561524, + "step": 26640 + }, + { + "epoch": 0.13325, + "grad_norm": 0.36142390966415405, + "learning_rate": 0.000481122901865863, + "loss": 2.9595382690429686, + "step": 26650 + }, + { + "epoch": 0.1333, + "grad_norm": 0.4415510892868042, + "learning_rate": 0.0004811077779635997, + "loss": 2.9213518142700194, + "step": 26660 + }, + { + "epoch": 0.13335, + "grad_norm": 0.3908076286315918, + "learning_rate": 0.00048109264824319975, + "loss": 2.9177865982055664, + "step": 26670 + }, + { + "epoch": 0.1334, + "grad_norm": 0.3782365620136261, + "learning_rate": 0.00048107751270504384, + "loss": 2.9735458374023436, + "step": 26680 + }, + { + "epoch": 0.13345, + "grad_norm": 0.3556264042854309, + "learning_rate": 0.00048106237134951313, + "loss": 2.950395202636719, + "step": 26690 + }, + { + "epoch": 0.1335, + "grad_norm": 0.3704966604709625, + "learning_rate": 0.0004810472241769888, + "loss": 2.97325439453125, + "step": 26700 + }, + { + "epoch": 0.13355, + "grad_norm": 0.37020254135131836, + "learning_rate": 0.00048103207118785205, + "loss": 2.9249797821044923, + "step": 26710 + }, + { + "epoch": 0.1336, + "grad_norm": 0.3889375329017639, + "learning_rate": 0.00048101691238248447, + "loss": 2.916792869567871, + "step": 26720 + }, + { + "epoch": 0.13365, + "grad_norm": 0.39570459723472595, + "learning_rate": 0.0004810017477612677, + "loss": 2.942928695678711, + "step": 26730 + }, + { + "epoch": 0.1337, + "grad_norm": 0.3962053954601288, + "learning_rate": 0.00048098657732458346, + "loss": 2.962717056274414, + "step": 26740 + }, + { + "epoch": 0.13375, + "grad_norm": 0.35804590582847595, + "learning_rate": 0.0004809714010728137, + "loss": 2.951587677001953, + "step": 26750 + }, + { + "epoch": 0.1338, + "grad_norm": 0.37549400329589844, + "learning_rate": 0.00048095621900634045, + "loss": 2.8899736404418945, + "step": 26760 + }, + { + "epoch": 0.13385, + "grad_norm": 0.3599599599838257, + "learning_rate": 0.00048094103112554585, + "loss": 2.8522390365600585, + "step": 26770 + }, + { + "epoch": 0.1339, + "grad_norm": 0.3833061456680298, + "learning_rate": 0.0004809258374308124, + "loss": 2.9361425399780274, + "step": 26780 + }, + { + "epoch": 0.13395, + "grad_norm": 0.36110788583755493, + "learning_rate": 0.0004809106379225226, + "loss": 2.9377838134765626, + "step": 26790 + }, + { + "epoch": 0.134, + "grad_norm": 0.3846631646156311, + "learning_rate": 0.00048089543260105893, + "loss": 2.8919200897216797, + "step": 26800 + }, + { + "epoch": 0.13405, + "grad_norm": 0.3656333088874817, + "learning_rate": 0.00048088022146680436, + "loss": 2.9440071105957033, + "step": 26810 + }, + { + "epoch": 0.1341, + "grad_norm": 0.40157777070999146, + "learning_rate": 0.0004808650045201418, + "loss": 2.868821144104004, + "step": 26820 + }, + { + "epoch": 0.13415, + "grad_norm": 0.3737427890300751, + "learning_rate": 0.00048084978176145417, + "loss": 2.890164375305176, + "step": 26830 + }, + { + "epoch": 0.1342, + "grad_norm": 0.37911325693130493, + "learning_rate": 0.0004808345531911249, + "loss": 2.864650917053223, + "step": 26840 + }, + { + "epoch": 0.13425, + "grad_norm": 0.40816012024879456, + "learning_rate": 0.0004808193188095372, + "loss": 2.966249465942383, + "step": 26850 + }, + { + "epoch": 0.1343, + "grad_norm": 0.39112454652786255, + "learning_rate": 0.0004808040786170748, + "loss": 2.897208404541016, + "step": 26860 + }, + { + "epoch": 0.13435, + "grad_norm": 0.38583019375801086, + "learning_rate": 0.00048078883261412126, + "loss": 2.925621223449707, + "step": 26870 + }, + { + "epoch": 0.1344, + "grad_norm": 0.38962322473526, + "learning_rate": 0.0004807735808010604, + "loss": 2.935319519042969, + "step": 26880 + }, + { + "epoch": 0.13445, + "grad_norm": 0.33716678619384766, + "learning_rate": 0.0004807583231782761, + "loss": 2.912581443786621, + "step": 26890 + }, + { + "epoch": 0.1345, + "grad_norm": 0.381961464881897, + "learning_rate": 0.00048074305974615263, + "loss": 2.986328887939453, + "step": 26900 + }, + { + "epoch": 0.13455, + "grad_norm": 0.36364516615867615, + "learning_rate": 0.0004807277905050742, + "loss": 2.9076911926269533, + "step": 26910 + }, + { + "epoch": 0.1346, + "grad_norm": 0.40011903643608093, + "learning_rate": 0.0004807125154554251, + "loss": 2.9621654510498048, + "step": 26920 + }, + { + "epoch": 0.13465, + "grad_norm": 0.3807370960712433, + "learning_rate": 0.00048069723459759003, + "loss": 2.8883344650268556, + "step": 26930 + }, + { + "epoch": 0.1347, + "grad_norm": 0.38479405641555786, + "learning_rate": 0.0004806819479319536, + "loss": 2.882781982421875, + "step": 26940 + }, + { + "epoch": 0.13475, + "grad_norm": 0.36968570947647095, + "learning_rate": 0.0004806666554589006, + "loss": 2.8250476837158205, + "step": 26950 + }, + { + "epoch": 0.1348, + "grad_norm": 0.3556709885597229, + "learning_rate": 0.00048065135717881627, + "loss": 2.9261764526367187, + "step": 26960 + }, + { + "epoch": 0.13485, + "grad_norm": 0.3808833062648773, + "learning_rate": 0.0004806360530920854, + "loss": 2.9083841323852537, + "step": 26970 + }, + { + "epoch": 0.1349, + "grad_norm": 0.3608994483947754, + "learning_rate": 0.0004806207431990935, + "loss": 2.9265825271606447, + "step": 26980 + }, + { + "epoch": 0.13495, + "grad_norm": 0.3888077139854431, + "learning_rate": 0.0004806054275002259, + "loss": 2.903965950012207, + "step": 26990 + }, + { + "epoch": 0.135, + "grad_norm": 0.4031972885131836, + "learning_rate": 0.0004805901059958682, + "loss": 2.8981395721435548, + "step": 27000 + }, + { + "epoch": 0.13505, + "grad_norm": 0.3814179003238678, + "learning_rate": 0.0004805747786864062, + "loss": 2.9244667053222657, + "step": 27010 + }, + { + "epoch": 0.1351, + "grad_norm": 0.38345959782600403, + "learning_rate": 0.00048055944557222563, + "loss": 2.899201774597168, + "step": 27020 + }, + { + "epoch": 0.13515, + "grad_norm": 0.3632197678089142, + "learning_rate": 0.0004805441066537125, + "loss": 2.923736000061035, + "step": 27030 + }, + { + "epoch": 0.1352, + "grad_norm": 0.41296765208244324, + "learning_rate": 0.00048052876193125315, + "loss": 2.9627838134765625, + "step": 27040 + }, + { + "epoch": 0.13525, + "grad_norm": 0.3501022458076477, + "learning_rate": 0.0004805134114052337, + "loss": 2.897196388244629, + "step": 27050 + }, + { + "epoch": 0.1353, + "grad_norm": 0.3841564953327179, + "learning_rate": 0.0004804980550760406, + "loss": 2.941621208190918, + "step": 27060 + }, + { + "epoch": 0.13535, + "grad_norm": 0.37863075733184814, + "learning_rate": 0.00048048269294406054, + "loss": 2.964263153076172, + "step": 27070 + }, + { + "epoch": 0.1354, + "grad_norm": 0.41050541400909424, + "learning_rate": 0.0004804673250096803, + "loss": 2.909626007080078, + "step": 27080 + }, + { + "epoch": 0.13545, + "grad_norm": 0.3816659450531006, + "learning_rate": 0.0004804519512732866, + "loss": 2.9671613693237306, + "step": 27090 + }, + { + "epoch": 0.1355, + "grad_norm": 0.37216612696647644, + "learning_rate": 0.0004804365717352666, + "loss": 2.902449607849121, + "step": 27100 + }, + { + "epoch": 0.13555, + "grad_norm": 0.34767961502075195, + "learning_rate": 0.0004804211863960075, + "loss": 2.9190393447875977, + "step": 27110 + }, + { + "epoch": 0.1356, + "grad_norm": 0.39962807297706604, + "learning_rate": 0.0004804057952558965, + "loss": 2.951121139526367, + "step": 27120 + }, + { + "epoch": 0.13565, + "grad_norm": 0.38647961616516113, + "learning_rate": 0.0004803903983153212, + "loss": 2.8739709854125977, + "step": 27130 + }, + { + "epoch": 0.1357, + "grad_norm": 0.39166560769081116, + "learning_rate": 0.00048037499557466914, + "loss": 2.8939870834350585, + "step": 27140 + }, + { + "epoch": 0.13575, + "grad_norm": 0.37468913197517395, + "learning_rate": 0.000480359587034328, + "loss": 2.9083087921142576, + "step": 27150 + }, + { + "epoch": 0.1358, + "grad_norm": 0.3630669414997101, + "learning_rate": 0.0004803441726946859, + "loss": 2.94659309387207, + "step": 27160 + }, + { + "epoch": 0.13585, + "grad_norm": 0.36113062500953674, + "learning_rate": 0.00048032875255613083, + "loss": 2.9079412460327148, + "step": 27170 + }, + { + "epoch": 0.1359, + "grad_norm": 0.3605669140815735, + "learning_rate": 0.00048031332661905093, + "loss": 2.921873092651367, + "step": 27180 + }, + { + "epoch": 0.13595, + "grad_norm": 0.3931385278701782, + "learning_rate": 0.0004802978948838345, + "loss": 2.948661231994629, + "step": 27190 + }, + { + "epoch": 0.136, + "grad_norm": 0.39253875613212585, + "learning_rate": 0.0004802824573508702, + "loss": 2.8938316345214843, + "step": 27200 + }, + { + "epoch": 0.13605, + "grad_norm": 0.3923114240169525, + "learning_rate": 0.0004802670140205465, + "loss": 2.870123863220215, + "step": 27210 + }, + { + "epoch": 0.1361, + "grad_norm": 0.3352682292461395, + "learning_rate": 0.0004802515648932523, + "loss": 2.9334917068481445, + "step": 27220 + }, + { + "epoch": 0.13615, + "grad_norm": 0.36503276228904724, + "learning_rate": 0.0004802361099693764, + "loss": 2.899751091003418, + "step": 27230 + }, + { + "epoch": 0.1362, + "grad_norm": 0.36352548003196716, + "learning_rate": 0.0004802206492493081, + "loss": 2.9005926132202147, + "step": 27240 + }, + { + "epoch": 0.13625, + "grad_norm": 0.3928877115249634, + "learning_rate": 0.0004802051827334364, + "loss": 2.8873647689819335, + "step": 27250 + }, + { + "epoch": 0.1363, + "grad_norm": 0.3784511685371399, + "learning_rate": 0.0004801897104221508, + "loss": 2.9047340393066405, + "step": 27260 + }, + { + "epoch": 0.13635, + "grad_norm": 0.3582150638103485, + "learning_rate": 0.00048017423231584076, + "loss": 2.8637262344360352, + "step": 27270 + }, + { + "epoch": 0.1364, + "grad_norm": 0.38398104906082153, + "learning_rate": 0.00048015874841489604, + "loss": 2.875265121459961, + "step": 27280 + }, + { + "epoch": 0.13645, + "grad_norm": 0.4232466518878937, + "learning_rate": 0.00048014325871970634, + "loss": 2.893733596801758, + "step": 27290 + }, + { + "epoch": 0.1365, + "grad_norm": 0.36737996339797974, + "learning_rate": 0.00048012776323066157, + "loss": 2.9684755325317385, + "step": 27300 + }, + { + "epoch": 0.13655, + "grad_norm": 0.39892077445983887, + "learning_rate": 0.00048011226194815196, + "loss": 2.932980537414551, + "step": 27310 + }, + { + "epoch": 0.1366, + "grad_norm": 0.40332475304603577, + "learning_rate": 0.00048009675487256767, + "loss": 2.887357711791992, + "step": 27320 + }, + { + "epoch": 0.13665, + "grad_norm": 0.3910316526889801, + "learning_rate": 0.0004800812420042991, + "loss": 2.855747413635254, + "step": 27330 + }, + { + "epoch": 0.1367, + "grad_norm": 0.3637521266937256, + "learning_rate": 0.00048006572334373675, + "loss": 2.8606643676757812, + "step": 27340 + }, + { + "epoch": 0.13675, + "grad_norm": 0.36200204491615295, + "learning_rate": 0.00048005019889127144, + "loss": 2.918485641479492, + "step": 27350 + }, + { + "epoch": 0.1368, + "grad_norm": 0.3634633719921112, + "learning_rate": 0.00048003466864729395, + "loss": 2.878544235229492, + "step": 27360 + }, + { + "epoch": 0.13685, + "grad_norm": 0.36086541414260864, + "learning_rate": 0.0004800191326121951, + "loss": 2.9007205963134766, + "step": 27370 + }, + { + "epoch": 0.1369, + "grad_norm": 0.37968501448631287, + "learning_rate": 0.0004800035907863662, + "loss": 2.9409984588623046, + "step": 27380 + }, + { + "epoch": 0.13695, + "grad_norm": 0.36860981583595276, + "learning_rate": 0.0004799880431701984, + "loss": 2.9158956527709963, + "step": 27390 + }, + { + "epoch": 0.137, + "grad_norm": 0.3800581097602844, + "learning_rate": 0.00047997248976408314, + "loss": 2.954301452636719, + "step": 27400 + }, + { + "epoch": 0.13705, + "grad_norm": 0.3856695890426636, + "learning_rate": 0.00047995693056841207, + "loss": 2.934916687011719, + "step": 27410 + }, + { + "epoch": 0.1371, + "grad_norm": 0.3519247770309448, + "learning_rate": 0.0004799413655835767, + "loss": 2.938766098022461, + "step": 27420 + }, + { + "epoch": 0.13715, + "grad_norm": 0.38515251874923706, + "learning_rate": 0.000479925794809969, + "loss": 2.929876136779785, + "step": 27430 + }, + { + "epoch": 0.1372, + "grad_norm": 0.3451438844203949, + "learning_rate": 0.00047991021824798105, + "loss": 2.929775047302246, + "step": 27440 + }, + { + "epoch": 0.13725, + "grad_norm": 0.3635682463645935, + "learning_rate": 0.0004798946358980048, + "loss": 2.917681121826172, + "step": 27450 + }, + { + "epoch": 0.1373, + "grad_norm": 0.38713160157203674, + "learning_rate": 0.00047987904776043266, + "loss": 2.9165323257446287, + "step": 27460 + }, + { + "epoch": 0.13735, + "grad_norm": 0.4038919508457184, + "learning_rate": 0.00047986345383565707, + "loss": 2.8854997634887694, + "step": 27470 + }, + { + "epoch": 0.1374, + "grad_norm": 0.37676146626472473, + "learning_rate": 0.0004798478541240705, + "loss": 2.913122367858887, + "step": 27480 + }, + { + "epoch": 0.13745, + "grad_norm": 0.3866631090641022, + "learning_rate": 0.0004798322486260658, + "loss": 2.8846403121948243, + "step": 27490 + }, + { + "epoch": 0.1375, + "grad_norm": 0.3643092215061188, + "learning_rate": 0.0004798166373420357, + "loss": 2.9238061904907227, + "step": 27500 + }, + { + "epoch": 0.13755, + "grad_norm": 0.3653920590877533, + "learning_rate": 0.0004798010202723733, + "loss": 2.894057273864746, + "step": 27510 + }, + { + "epoch": 0.1376, + "grad_norm": 0.3509802520275116, + "learning_rate": 0.0004797853974174718, + "loss": 2.8797185897827147, + "step": 27520 + }, + { + "epoch": 0.13765, + "grad_norm": 0.3761005103588104, + "learning_rate": 0.0004797697687777244, + "loss": 2.889048957824707, + "step": 27530 + }, + { + "epoch": 0.1377, + "grad_norm": 0.375191867351532, + "learning_rate": 0.00047975413435352466, + "loss": 2.9374523162841797, + "step": 27540 + }, + { + "epoch": 0.13775, + "grad_norm": 0.3544359505176544, + "learning_rate": 0.0004797384941452661, + "loss": 2.9429084777832033, + "step": 27550 + }, + { + "epoch": 0.1378, + "grad_norm": 0.4020138084888458, + "learning_rate": 0.0004797228481533425, + "loss": 2.8802087783813475, + "step": 27560 + }, + { + "epoch": 0.13785, + "grad_norm": 0.395637184381485, + "learning_rate": 0.0004797071963781477, + "loss": 2.9279945373535154, + "step": 27570 + }, + { + "epoch": 0.1379, + "grad_norm": 0.37564709782600403, + "learning_rate": 0.00047969153882007577, + "loss": 2.965615081787109, + "step": 27580 + }, + { + "epoch": 0.13795, + "grad_norm": 0.36955785751342773, + "learning_rate": 0.0004796758754795209, + "loss": 2.8859363555908204, + "step": 27590 + }, + { + "epoch": 0.138, + "grad_norm": 0.3916771113872528, + "learning_rate": 0.00047966020635687747, + "loss": 2.870256042480469, + "step": 27600 + }, + { + "epoch": 0.13805, + "grad_norm": 0.376367449760437, + "learning_rate": 0.00047964453145253974, + "loss": 2.910803031921387, + "step": 27610 + }, + { + "epoch": 0.1381, + "grad_norm": 0.3748619258403778, + "learning_rate": 0.0004796288507669025, + "loss": 2.9082990646362306, + "step": 27620 + }, + { + "epoch": 0.13815, + "grad_norm": 0.38071534037590027, + "learning_rate": 0.0004796131643003606, + "loss": 2.8994136810302735, + "step": 27630 + }, + { + "epoch": 0.1382, + "grad_norm": 0.35427722334861755, + "learning_rate": 0.0004795974720533087, + "loss": 2.9013858795166017, + "step": 27640 + }, + { + "epoch": 0.13825, + "grad_norm": 0.3372753858566284, + "learning_rate": 0.000479581774026142, + "loss": 2.9143035888671873, + "step": 27650 + }, + { + "epoch": 0.1383, + "grad_norm": 0.35398122668266296, + "learning_rate": 0.0004795660702192557, + "loss": 2.8503511428833006, + "step": 27660 + }, + { + "epoch": 0.13835, + "grad_norm": 0.384951114654541, + "learning_rate": 0.0004795503606330451, + "loss": 2.9298933029174803, + "step": 27670 + }, + { + "epoch": 0.1384, + "grad_norm": 0.3823627531528473, + "learning_rate": 0.00047953464526790567, + "loss": 2.938493347167969, + "step": 27680 + }, + { + "epoch": 0.13845, + "grad_norm": 0.37075287103652954, + "learning_rate": 0.00047951892412423315, + "loss": 2.9221673965454102, + "step": 27690 + }, + { + "epoch": 0.1385, + "grad_norm": 0.39440250396728516, + "learning_rate": 0.00047950319720242317, + "loss": 2.8516525268554687, + "step": 27700 + }, + { + "epoch": 0.13855, + "grad_norm": 0.37477606534957886, + "learning_rate": 0.0004794874645028718, + "loss": 2.9392206192016603, + "step": 27710 + }, + { + "epoch": 0.1386, + "grad_norm": 0.38380172848701477, + "learning_rate": 0.000479471726025975, + "loss": 2.891842079162598, + "step": 27720 + }, + { + "epoch": 0.13865, + "grad_norm": 0.3857271075248718, + "learning_rate": 0.0004794559817721291, + "loss": 2.8986095428466796, + "step": 27730 + }, + { + "epoch": 0.1387, + "grad_norm": 0.34787070751190186, + "learning_rate": 0.00047944023174173033, + "loss": 2.9794708251953126, + "step": 27740 + }, + { + "epoch": 0.13875, + "grad_norm": 0.36182665824890137, + "learning_rate": 0.0004794244759351752, + "loss": 2.8806995391845702, + "step": 27750 + }, + { + "epoch": 0.1388, + "grad_norm": 0.40278950333595276, + "learning_rate": 0.0004794087143528605, + "loss": 2.8640010833740233, + "step": 27760 + }, + { + "epoch": 0.13885, + "grad_norm": 0.37960338592529297, + "learning_rate": 0.000479392946995183, + "loss": 2.9286077499389647, + "step": 27770 + }, + { + "epoch": 0.1389, + "grad_norm": 0.4207148551940918, + "learning_rate": 0.00047937717386253956, + "loss": 2.92098445892334, + "step": 27780 + }, + { + "epoch": 0.13895, + "grad_norm": 0.4049313962459564, + "learning_rate": 0.00047936139495532727, + "loss": 2.9199651718139648, + "step": 27790 + }, + { + "epoch": 0.139, + "grad_norm": 0.37705549597740173, + "learning_rate": 0.00047934561027394336, + "loss": 2.961600494384766, + "step": 27800 + }, + { + "epoch": 0.13905, + "grad_norm": 0.33895227313041687, + "learning_rate": 0.0004793298198187853, + "loss": 2.9056705474853515, + "step": 27810 + }, + { + "epoch": 0.1391, + "grad_norm": 0.38305342197418213, + "learning_rate": 0.0004793140235902506, + "loss": 2.850033187866211, + "step": 27820 + }, + { + "epoch": 0.13915, + "grad_norm": 0.35673263669013977, + "learning_rate": 0.00047929822158873674, + "loss": 2.8767980575561523, + "step": 27830 + }, + { + "epoch": 0.1392, + "grad_norm": 0.41097748279571533, + "learning_rate": 0.00047928241381464177, + "loss": 2.9189348220825195, + "step": 27840 + }, + { + "epoch": 0.13925, + "grad_norm": 0.3667374849319458, + "learning_rate": 0.0004792666002683636, + "loss": 2.9175273895263674, + "step": 27850 + }, + { + "epoch": 0.1393, + "grad_norm": 0.3636946678161621, + "learning_rate": 0.00047925078095030027, + "loss": 2.8579912185668945, + "step": 27860 + }, + { + "epoch": 0.13935, + "grad_norm": 0.37459203600883484, + "learning_rate": 0.0004792349558608501, + "loss": 2.9488590240478514, + "step": 27870 + }, + { + "epoch": 0.1394, + "grad_norm": 0.41264691948890686, + "learning_rate": 0.0004792191250004113, + "loss": 2.886056137084961, + "step": 27880 + }, + { + "epoch": 0.13945, + "grad_norm": 0.36674368381500244, + "learning_rate": 0.0004792032883693828, + "loss": 2.8778432846069335, + "step": 27890 + }, + { + "epoch": 0.1395, + "grad_norm": 0.37744247913360596, + "learning_rate": 0.0004791874459681629, + "loss": 2.9236322402954102, + "step": 27900 + }, + { + "epoch": 0.13955, + "grad_norm": 0.4022049307823181, + "learning_rate": 0.0004791715977971505, + "loss": 2.9397985458374025, + "step": 27910 + }, + { + "epoch": 0.1396, + "grad_norm": 0.3553810119628906, + "learning_rate": 0.00047915574385674474, + "loss": 2.906529998779297, + "step": 27920 + }, + { + "epoch": 0.13965, + "grad_norm": 0.35604238510131836, + "learning_rate": 0.0004791398841473447, + "loss": 2.9514408111572266, + "step": 27930 + }, + { + "epoch": 0.1397, + "grad_norm": 0.3470006585121155, + "learning_rate": 0.00047912401866934964, + "loss": 2.9034631729125975, + "step": 27940 + }, + { + "epoch": 0.13975, + "grad_norm": 0.3833562433719635, + "learning_rate": 0.0004791081474231588, + "loss": 2.908356475830078, + "step": 27950 + }, + { + "epoch": 0.1398, + "grad_norm": 0.4115186631679535, + "learning_rate": 0.000479092270409172, + "loss": 2.970598030090332, + "step": 27960 + }, + { + "epoch": 0.13985, + "grad_norm": 0.36268287897109985, + "learning_rate": 0.00047907638762778875, + "loss": 2.8665117263793944, + "step": 27970 + }, + { + "epoch": 0.1399, + "grad_norm": 0.3772282302379608, + "learning_rate": 0.000479060499079409, + "loss": 2.9301742553710937, + "step": 27980 + }, + { + "epoch": 0.13995, + "grad_norm": 0.34590861201286316, + "learning_rate": 0.00047904460476443274, + "loss": 2.9258182525634764, + "step": 27990 + }, + { + "epoch": 0.14, + "grad_norm": 0.3507023751735687, + "learning_rate": 0.0004790287046832601, + "loss": 2.960640525817871, + "step": 28000 + }, + { + "epoch": 0.14005, + "grad_norm": 0.3814890682697296, + "learning_rate": 0.0004790127988362913, + "loss": 2.895982360839844, + "step": 28010 + }, + { + "epoch": 0.1401, + "grad_norm": 0.4019654393196106, + "learning_rate": 0.0004789968872239268, + "loss": 2.9003412246704103, + "step": 28020 + }, + { + "epoch": 0.14015, + "grad_norm": 0.3939569592475891, + "learning_rate": 0.00047898096984656726, + "loss": 2.8939207077026365, + "step": 28030 + }, + { + "epoch": 0.1402, + "grad_norm": 0.3734310269355774, + "learning_rate": 0.00047896504670461325, + "loss": 2.9268896102905275, + "step": 28040 + }, + { + "epoch": 0.14025, + "grad_norm": 0.412850022315979, + "learning_rate": 0.00047894911779846586, + "loss": 2.8941131591796876, + "step": 28050 + }, + { + "epoch": 0.1403, + "grad_norm": 0.38735997676849365, + "learning_rate": 0.00047893318312852587, + "loss": 2.8885847091674806, + "step": 28060 + }, + { + "epoch": 0.14035, + "grad_norm": 0.34762802720069885, + "learning_rate": 0.0004789172426951945, + "loss": 2.9166244506835937, + "step": 28070 + }, + { + "epoch": 0.1404, + "grad_norm": 0.3703233599662781, + "learning_rate": 0.0004789012964988731, + "loss": 2.8605876922607423, + "step": 28080 + }, + { + "epoch": 0.14045, + "grad_norm": 0.4055093228816986, + "learning_rate": 0.0004788853445399631, + "loss": 2.9120510101318358, + "step": 28090 + }, + { + "epoch": 0.1405, + "grad_norm": 0.3800540268421173, + "learning_rate": 0.00047886938681886604, + "loss": 2.8828361511230467, + "step": 28100 + }, + { + "epoch": 0.14055, + "grad_norm": 0.3875916302204132, + "learning_rate": 0.0004788534233359837, + "loss": 2.923063850402832, + "step": 28110 + }, + { + "epoch": 0.1406, + "grad_norm": 0.3547535836696625, + "learning_rate": 0.000478837454091718, + "loss": 2.8476932525634764, + "step": 28120 + }, + { + "epoch": 0.14065, + "grad_norm": 0.3585605323314667, + "learning_rate": 0.00047882147908647095, + "loss": 2.935593605041504, + "step": 28130 + }, + { + "epoch": 0.1407, + "grad_norm": 0.42855849862098694, + "learning_rate": 0.00047880549832064463, + "loss": 2.9138036727905274, + "step": 28140 + }, + { + "epoch": 0.14075, + "grad_norm": 0.3713895082473755, + "learning_rate": 0.00047878951179464147, + "loss": 2.8871082305908202, + "step": 28150 + }, + { + "epoch": 0.1408, + "grad_norm": 0.38041380047798157, + "learning_rate": 0.0004787735195088638, + "loss": 2.9619693756103516, + "step": 28160 + }, + { + "epoch": 0.14085, + "grad_norm": 0.4369162321090698, + "learning_rate": 0.0004787575214637144, + "loss": 2.9465850830078124, + "step": 28170 + }, + { + "epoch": 0.1409, + "grad_norm": 0.3765942454338074, + "learning_rate": 0.00047874151765959583, + "loss": 2.9200321197509767, + "step": 28180 + }, + { + "epoch": 0.14095, + "grad_norm": 0.38500985503196716, + "learning_rate": 0.00047872550809691116, + "loss": 2.8902732849121096, + "step": 28190 + }, + { + "epoch": 0.141, + "grad_norm": 0.3627932667732239, + "learning_rate": 0.00047870949277606335, + "loss": 2.8970298767089844, + "step": 28200 + }, + { + "epoch": 0.14105, + "grad_norm": 0.34555110335350037, + "learning_rate": 0.00047869347169745554, + "loss": 2.9213134765625, + "step": 28210 + }, + { + "epoch": 0.1411, + "grad_norm": 0.3589212894439697, + "learning_rate": 0.0004786774448614911, + "loss": 2.9149730682373045, + "step": 28220 + }, + { + "epoch": 0.14115, + "grad_norm": 0.3503093719482422, + "learning_rate": 0.0004786614122685735, + "loss": 2.9171504974365234, + "step": 28230 + }, + { + "epoch": 0.1412, + "grad_norm": 0.3842655122280121, + "learning_rate": 0.00047864537391910643, + "loss": 2.872216796875, + "step": 28240 + }, + { + "epoch": 0.14125, + "grad_norm": 0.3859809935092926, + "learning_rate": 0.00047862932981349363, + "loss": 2.9624256134033202, + "step": 28250 + }, + { + "epoch": 0.1413, + "grad_norm": 0.35915055871009827, + "learning_rate": 0.0004786132799521389, + "loss": 2.945933723449707, + "step": 28260 + }, + { + "epoch": 0.14135, + "grad_norm": 0.4037717282772064, + "learning_rate": 0.00047859722433544647, + "loss": 2.892999839782715, + "step": 28270 + }, + { + "epoch": 0.1414, + "grad_norm": 0.3936765193939209, + "learning_rate": 0.0004785811629638204, + "loss": 2.8916391372680663, + "step": 28280 + }, + { + "epoch": 0.14145, + "grad_norm": 0.38370636105537415, + "learning_rate": 0.00047856509583766503, + "loss": 2.926483726501465, + "step": 28290 + }, + { + "epoch": 0.1415, + "grad_norm": 0.3520573675632477, + "learning_rate": 0.0004785490229573849, + "loss": 2.907196044921875, + "step": 28300 + }, + { + "epoch": 0.14155, + "grad_norm": 0.3537794053554535, + "learning_rate": 0.00047853294432338467, + "loss": 2.899775505065918, + "step": 28310 + }, + { + "epoch": 0.1416, + "grad_norm": 0.359220415353775, + "learning_rate": 0.00047851685993606914, + "loss": 2.8679130554199217, + "step": 28320 + }, + { + "epoch": 0.14165, + "grad_norm": 0.3733152747154236, + "learning_rate": 0.0004785007697958431, + "loss": 2.9016483306884764, + "step": 28330 + }, + { + "epoch": 0.1417, + "grad_norm": 0.38241711258888245, + "learning_rate": 0.0004784846739031118, + "loss": 2.9309270858764647, + "step": 28340 + }, + { + "epoch": 0.14175, + "grad_norm": 0.38607266545295715, + "learning_rate": 0.00047846857225828033, + "loss": 2.8577812194824217, + "step": 28350 + }, + { + "epoch": 0.1418, + "grad_norm": 0.3628918528556824, + "learning_rate": 0.0004784524648617541, + "loss": 2.944838523864746, + "step": 28360 + }, + { + "epoch": 0.14185, + "grad_norm": 0.389614075422287, + "learning_rate": 0.0004784363517139385, + "loss": 2.8902000427246093, + "step": 28370 + }, + { + "epoch": 0.1419, + "grad_norm": 0.39060795307159424, + "learning_rate": 0.0004784202328152394, + "loss": 2.956539535522461, + "step": 28380 + }, + { + "epoch": 0.14195, + "grad_norm": 0.3793882727622986, + "learning_rate": 0.00047840410816606235, + "loss": 2.892274284362793, + "step": 28390 + }, + { + "epoch": 0.142, + "grad_norm": 0.3853428065776825, + "learning_rate": 0.0004783879777668134, + "loss": 2.9314775466918945, + "step": 28400 + }, + { + "epoch": 0.14205, + "grad_norm": 0.3828856348991394, + "learning_rate": 0.0004783718416178987, + "loss": 2.8898284912109373, + "step": 28410 + }, + { + "epoch": 0.1421, + "grad_norm": 0.3770284950733185, + "learning_rate": 0.00047835569971972446, + "loss": 2.93453426361084, + "step": 28420 + }, + { + "epoch": 0.14215, + "grad_norm": 0.37070176005363464, + "learning_rate": 0.0004783395520726969, + "loss": 2.9153110504150392, + "step": 28430 + }, + { + "epoch": 0.1422, + "grad_norm": 0.3742634654045105, + "learning_rate": 0.00047832339867722275, + "loss": 2.896531105041504, + "step": 28440 + }, + { + "epoch": 0.14225, + "grad_norm": 0.35622638463974, + "learning_rate": 0.0004783072395337085, + "loss": 2.8250394821166993, + "step": 28450 + }, + { + "epoch": 0.1423, + "grad_norm": 0.397617369890213, + "learning_rate": 0.000478291074642561, + "loss": 2.924175834655762, + "step": 28460 + }, + { + "epoch": 0.14235, + "grad_norm": 0.38167867064476013, + "learning_rate": 0.0004782749040041872, + "loss": 2.9366079330444337, + "step": 28470 + }, + { + "epoch": 0.1424, + "grad_norm": 0.35878345370292664, + "learning_rate": 0.0004782587276189943, + "loss": 2.875445747375488, + "step": 28480 + }, + { + "epoch": 0.14245, + "grad_norm": 0.3729279935359955, + "learning_rate": 0.00047824254548738945, + "loss": 2.9435699462890623, + "step": 28490 + }, + { + "epoch": 0.1425, + "grad_norm": 0.40313008427619934, + "learning_rate": 0.00047822635760977995, + "loss": 2.9147869110107423, + "step": 28500 + }, + { + "epoch": 0.14255, + "grad_norm": 0.38750892877578735, + "learning_rate": 0.0004782101639865735, + "loss": 2.8963863372802736, + "step": 28510 + }, + { + "epoch": 0.1426, + "grad_norm": 0.3510226309299469, + "learning_rate": 0.0004781939646181777, + "loss": 2.894538688659668, + "step": 28520 + }, + { + "epoch": 0.14265, + "grad_norm": 0.36190885305404663, + "learning_rate": 0.00047817775950500033, + "loss": 2.8941673278808593, + "step": 28530 + }, + { + "epoch": 0.1427, + "grad_norm": 0.37290632724761963, + "learning_rate": 0.0004781615486474494, + "loss": 2.9503007888793946, + "step": 28540 + }, + { + "epoch": 0.14275, + "grad_norm": 0.38454946875572205, + "learning_rate": 0.00047814533204593304, + "loss": 2.8669071197509766, + "step": 28550 + }, + { + "epoch": 0.1428, + "grad_norm": 0.3473680019378662, + "learning_rate": 0.00047812910970085944, + "loss": 2.883683776855469, + "step": 28560 + }, + { + "epoch": 0.14285, + "grad_norm": 0.3738240897655487, + "learning_rate": 0.000478112881612637, + "loss": 2.963277816772461, + "step": 28570 + }, + { + "epoch": 0.1429, + "grad_norm": 0.38191094994544983, + "learning_rate": 0.00047809664778167434, + "loss": 2.901020050048828, + "step": 28580 + }, + { + "epoch": 0.14295, + "grad_norm": 0.345314621925354, + "learning_rate": 0.00047808040820838005, + "loss": 2.8774967193603516, + "step": 28590 + }, + { + "epoch": 0.143, + "grad_norm": 0.4549427628517151, + "learning_rate": 0.00047806416289316304, + "loss": 2.8715984344482424, + "step": 28600 + }, + { + "epoch": 0.14305, + "grad_norm": 0.339175820350647, + "learning_rate": 0.00047804791183643226, + "loss": 2.8615018844604494, + "step": 28610 + }, + { + "epoch": 0.1431, + "grad_norm": 0.3508572280406952, + "learning_rate": 0.0004780316550385968, + "loss": 2.9224531173706056, + "step": 28620 + }, + { + "epoch": 0.14315, + "grad_norm": 0.334559828042984, + "learning_rate": 0.00047801539250006596, + "loss": 2.9344966888427733, + "step": 28630 + }, + { + "epoch": 0.1432, + "grad_norm": 0.35820361971855164, + "learning_rate": 0.0004779991242212491, + "loss": 2.9127260208129884, + "step": 28640 + }, + { + "epoch": 0.14325, + "grad_norm": 0.3570467233657837, + "learning_rate": 0.0004779828502025559, + "loss": 2.9488704681396483, + "step": 28650 + }, + { + "epoch": 0.1433, + "grad_norm": 0.3916756510734558, + "learning_rate": 0.0004779665704443959, + "loss": 2.884560775756836, + "step": 28660 + }, + { + "epoch": 0.14335, + "grad_norm": 0.3621651530265808, + "learning_rate": 0.00047795028494717907, + "loss": 2.9039432525634767, + "step": 28670 + }, + { + "epoch": 0.1434, + "grad_norm": 0.34772980213165283, + "learning_rate": 0.0004779339937113153, + "loss": 2.9441806793212892, + "step": 28680 + }, + { + "epoch": 0.14345, + "grad_norm": 0.3634345829486847, + "learning_rate": 0.00047791769673721476, + "loss": 2.8770830154418947, + "step": 28690 + }, + { + "epoch": 0.1435, + "grad_norm": 0.38859453797340393, + "learning_rate": 0.00047790139402528764, + "loss": 2.9097885131835937, + "step": 28700 + }, + { + "epoch": 0.14355, + "grad_norm": 0.37200936675071716, + "learning_rate": 0.00047788508557594455, + "loss": 2.881402587890625, + "step": 28710 + }, + { + "epoch": 0.1436, + "grad_norm": 0.36005493998527527, + "learning_rate": 0.00047786877138959597, + "loss": 2.909657287597656, + "step": 28720 + }, + { + "epoch": 0.14365, + "grad_norm": 0.3624618947505951, + "learning_rate": 0.00047785245146665255, + "loss": 2.927195739746094, + "step": 28730 + }, + { + "epoch": 0.1437, + "grad_norm": 0.3452727496623993, + "learning_rate": 0.00047783612580752524, + "loss": 2.8455488204956056, + "step": 28740 + }, + { + "epoch": 0.14375, + "grad_norm": 0.3992403447628021, + "learning_rate": 0.0004778197944126249, + "loss": 2.9470775604248045, + "step": 28750 + }, + { + "epoch": 0.1438, + "grad_norm": 0.36592555046081543, + "learning_rate": 0.00047780345728236286, + "loss": 2.9450551986694338, + "step": 28760 + }, + { + "epoch": 0.14385, + "grad_norm": 0.3709334135055542, + "learning_rate": 0.00047778711441715015, + "loss": 2.9022184371948243, + "step": 28770 + }, + { + "epoch": 0.1439, + "grad_norm": 0.3627029359340668, + "learning_rate": 0.0004777707658173985, + "loss": 2.974868583679199, + "step": 28780 + }, + { + "epoch": 0.14395, + "grad_norm": 0.3828754425048828, + "learning_rate": 0.0004777544114835193, + "loss": 2.9165199279785154, + "step": 28790 + }, + { + "epoch": 0.144, + "grad_norm": 0.38393259048461914, + "learning_rate": 0.0004777380514159243, + "loss": 2.957480239868164, + "step": 28800 + }, + { + "epoch": 0.14405, + "grad_norm": 0.3679179847240448, + "learning_rate": 0.0004777216856150254, + "loss": 2.949268913269043, + "step": 28810 + }, + { + "epoch": 0.1441, + "grad_norm": 0.35596275329589844, + "learning_rate": 0.0004777053140812346, + "loss": 2.883640670776367, + "step": 28820 + }, + { + "epoch": 0.14415, + "grad_norm": 0.3869820237159729, + "learning_rate": 0.00047768893681496397, + "loss": 2.8560998916625975, + "step": 28830 + }, + { + "epoch": 0.1442, + "grad_norm": 0.3866129219532013, + "learning_rate": 0.00047767255381662586, + "loss": 2.861763572692871, + "step": 28840 + }, + { + "epoch": 0.14425, + "grad_norm": 0.3609616160392761, + "learning_rate": 0.00047765616508663286, + "loss": 2.8605186462402346, + "step": 28850 + }, + { + "epoch": 0.1443, + "grad_norm": 0.34508419036865234, + "learning_rate": 0.00047763977062539734, + "loss": 2.885706329345703, + "step": 28860 + }, + { + "epoch": 0.14435, + "grad_norm": 0.37175124883651733, + "learning_rate": 0.00047762337043333217, + "loss": 2.9228105545043945, + "step": 28870 + }, + { + "epoch": 0.1444, + "grad_norm": 0.35626253485679626, + "learning_rate": 0.0004776069645108501, + "loss": 2.8686004638671876, + "step": 28880 + }, + { + "epoch": 0.14445, + "grad_norm": 0.39847826957702637, + "learning_rate": 0.0004775905528583643, + "loss": 2.8723663330078124, + "step": 28890 + }, + { + "epoch": 0.1445, + "grad_norm": 0.35412368178367615, + "learning_rate": 0.00047757413547628777, + "loss": 2.909596824645996, + "step": 28900 + }, + { + "epoch": 0.14455, + "grad_norm": 0.3861222267150879, + "learning_rate": 0.00047755771236503386, + "loss": 2.906645393371582, + "step": 28910 + }, + { + "epoch": 0.1446, + "grad_norm": 0.3515689969062805, + "learning_rate": 0.00047754128352501613, + "loss": 2.8520153045654295, + "step": 28920 + }, + { + "epoch": 0.14465, + "grad_norm": 0.3643410801887512, + "learning_rate": 0.00047752484895664815, + "loss": 2.8693187713623045, + "step": 28930 + }, + { + "epoch": 0.1447, + "grad_norm": 0.3578280806541443, + "learning_rate": 0.0004775084086603436, + "loss": 2.8981468200683596, + "step": 28940 + }, + { + "epoch": 0.14475, + "grad_norm": 0.38203635811805725, + "learning_rate": 0.0004774919626365163, + "loss": 2.8767608642578124, + "step": 28950 + }, + { + "epoch": 0.1448, + "grad_norm": 0.37871336936950684, + "learning_rate": 0.0004774755108855805, + "loss": 2.890627861022949, + "step": 28960 + }, + { + "epoch": 0.14485, + "grad_norm": 0.389232873916626, + "learning_rate": 0.00047745905340795016, + "loss": 2.945455551147461, + "step": 28970 + }, + { + "epoch": 0.1449, + "grad_norm": 0.3349314332008362, + "learning_rate": 0.0004774425902040396, + "loss": 2.9399553298950196, + "step": 28980 + }, + { + "epoch": 0.14495, + "grad_norm": 0.3706963360309601, + "learning_rate": 0.00047742612127426345, + "loss": 2.939855766296387, + "step": 28990 + }, + { + "epoch": 0.145, + "grad_norm": 0.37393248081207275, + "learning_rate": 0.0004774096466190362, + "loss": 2.884865570068359, + "step": 29000 + }, + { + "epoch": 0.14505, + "grad_norm": 0.3524548411369324, + "learning_rate": 0.0004773931662387726, + "loss": 2.871634292602539, + "step": 29010 + }, + { + "epoch": 0.1451, + "grad_norm": 0.3315848112106323, + "learning_rate": 0.0004773766801338876, + "loss": 2.907744216918945, + "step": 29020 + }, + { + "epoch": 0.14515, + "grad_norm": 0.36816394329071045, + "learning_rate": 0.0004773601883047961, + "loss": 2.8999792098999024, + "step": 29030 + }, + { + "epoch": 0.1452, + "grad_norm": 0.39312678575515747, + "learning_rate": 0.00047734369075191343, + "loss": 2.964163970947266, + "step": 29040 + }, + { + "epoch": 0.14525, + "grad_norm": 0.37550392746925354, + "learning_rate": 0.00047732718747565484, + "loss": 2.8917259216308593, + "step": 29050 + }, + { + "epoch": 0.1453, + "grad_norm": 0.345310777425766, + "learning_rate": 0.00047731067847643584, + "loss": 2.9624452590942383, + "step": 29060 + }, + { + "epoch": 0.14535, + "grad_norm": 0.36814481019973755, + "learning_rate": 0.00047729416375467205, + "loss": 2.948072814941406, + "step": 29070 + }, + { + "epoch": 0.1454, + "grad_norm": 0.35620585083961487, + "learning_rate": 0.00047727764331077923, + "loss": 2.9209644317626955, + "step": 29080 + }, + { + "epoch": 0.14545, + "grad_norm": 0.3617779612541199, + "learning_rate": 0.0004772611171451732, + "loss": 2.844098663330078, + "step": 29090 + }, + { + "epoch": 0.1455, + "grad_norm": 0.36899736523628235, + "learning_rate": 0.00047724458525827006, + "loss": 2.8984216690063476, + "step": 29100 + }, + { + "epoch": 0.14555, + "grad_norm": 0.3630231022834778, + "learning_rate": 0.0004772280476504861, + "loss": 2.922854995727539, + "step": 29110 + }, + { + "epoch": 0.1456, + "grad_norm": 0.3784343898296356, + "learning_rate": 0.00047721150432223746, + "loss": 2.8919448852539062, + "step": 29120 + }, + { + "epoch": 0.14565, + "grad_norm": 0.3500450849533081, + "learning_rate": 0.00047719495527394075, + "loss": 2.9026113510131837, + "step": 29130 + }, + { + "epoch": 0.1457, + "grad_norm": 0.38839590549468994, + "learning_rate": 0.00047717840050601255, + "loss": 2.8462114334106445, + "step": 29140 + }, + { + "epoch": 0.14575, + "grad_norm": 0.4071058928966522, + "learning_rate": 0.00047716184001886966, + "loss": 2.9446269989013674, + "step": 29150 + }, + { + "epoch": 0.1458, + "grad_norm": 0.3367745280265808, + "learning_rate": 0.00047714527381292893, + "loss": 2.9217788696289064, + "step": 29160 + }, + { + "epoch": 0.14585, + "grad_norm": 0.3765797019004822, + "learning_rate": 0.00047712870188860746, + "loss": 2.8713428497314455, + "step": 29170 + }, + { + "epoch": 0.1459, + "grad_norm": 0.3582843840122223, + "learning_rate": 0.00047711212424632245, + "loss": 2.9649877548217773, + "step": 29180 + }, + { + "epoch": 0.14595, + "grad_norm": 0.3562053442001343, + "learning_rate": 0.0004770955408864913, + "loss": 2.8860931396484375, + "step": 29190 + }, + { + "epoch": 0.146, + "grad_norm": 0.38674482703208923, + "learning_rate": 0.00047707895180953134, + "loss": 2.9034385681152344, + "step": 29200 + }, + { + "epoch": 0.14605, + "grad_norm": 0.3663107752799988, + "learning_rate": 0.0004770623570158603, + "loss": 2.910167694091797, + "step": 29210 + }, + { + "epoch": 0.1461, + "grad_norm": 0.3509627878665924, + "learning_rate": 0.0004770457565058959, + "loss": 2.899915885925293, + "step": 29220 + }, + { + "epoch": 0.14615, + "grad_norm": 0.37992408871650696, + "learning_rate": 0.0004770291502800561, + "loss": 2.907041549682617, + "step": 29230 + }, + { + "epoch": 0.1462, + "grad_norm": 0.3400565981864929, + "learning_rate": 0.000477012538338759, + "loss": 2.8923982620239257, + "step": 29240 + }, + { + "epoch": 0.14625, + "grad_norm": 0.37101536989212036, + "learning_rate": 0.0004769959206824228, + "loss": 2.9407220840454102, + "step": 29250 + }, + { + "epoch": 0.1463, + "grad_norm": 0.33395859599113464, + "learning_rate": 0.00047697929731146575, + "loss": 2.8333335876464845, + "step": 29260 + }, + { + "epoch": 0.14635, + "grad_norm": 0.3616665005683899, + "learning_rate": 0.0004769626682263064, + "loss": 2.921405792236328, + "step": 29270 + }, + { + "epoch": 0.1464, + "grad_norm": 0.3645349144935608, + "learning_rate": 0.0004769460334273634, + "loss": 2.879375457763672, + "step": 29280 + }, + { + "epoch": 0.14645, + "grad_norm": 0.3500014841556549, + "learning_rate": 0.0004769293929150556, + "loss": 2.8670604705810545, + "step": 29290 + }, + { + "epoch": 0.1465, + "grad_norm": 0.37222471833229065, + "learning_rate": 0.0004769127466898018, + "loss": 2.883197021484375, + "step": 29300 + }, + { + "epoch": 0.14655, + "grad_norm": 0.3387391269207001, + "learning_rate": 0.0004768960947520211, + "loss": 2.9152830123901365, + "step": 29310 + }, + { + "epoch": 0.1466, + "grad_norm": 0.35786792635917664, + "learning_rate": 0.0004768794371021328, + "loss": 2.8944700241088865, + "step": 29320 + }, + { + "epoch": 0.14665, + "grad_norm": 0.3692586421966553, + "learning_rate": 0.0004768627737405561, + "loss": 2.9275461196899415, + "step": 29330 + }, + { + "epoch": 0.1467, + "grad_norm": 0.418256938457489, + "learning_rate": 0.0004768461046677106, + "loss": 2.8937402725219727, + "step": 29340 + }, + { + "epoch": 0.14675, + "grad_norm": 0.3597736358642578, + "learning_rate": 0.00047682942988401593, + "loss": 2.8967288970947265, + "step": 29350 + }, + { + "epoch": 0.1468, + "grad_norm": 0.3811469078063965, + "learning_rate": 0.00047681274938989195, + "loss": 2.851347541809082, + "step": 29360 + }, + { + "epoch": 0.14685, + "grad_norm": 0.3590770959854126, + "learning_rate": 0.0004767960631857584, + "loss": 2.8815134048461912, + "step": 29370 + }, + { + "epoch": 0.1469, + "grad_norm": 0.381028413772583, + "learning_rate": 0.00047677937127203556, + "loss": 2.900274085998535, + "step": 29380 + }, + { + "epoch": 0.14695, + "grad_norm": 0.3456946313381195, + "learning_rate": 0.0004767626736491436, + "loss": 2.913640022277832, + "step": 29390 + }, + { + "epoch": 0.147, + "grad_norm": 0.3890123963356018, + "learning_rate": 0.00047674597031750277, + "loss": 2.931261444091797, + "step": 29400 + }, + { + "epoch": 0.14705, + "grad_norm": 0.36680173873901367, + "learning_rate": 0.0004767292612775337, + "loss": 2.910574722290039, + "step": 29410 + }, + { + "epoch": 0.1471, + "grad_norm": 0.35043564438819885, + "learning_rate": 0.0004767125465296569, + "loss": 2.909130096435547, + "step": 29420 + }, + { + "epoch": 0.14715, + "grad_norm": 0.3664659559726715, + "learning_rate": 0.00047669582607429335, + "loss": 2.9469024658203127, + "step": 29430 + }, + { + "epoch": 0.1472, + "grad_norm": 0.40098077058792114, + "learning_rate": 0.00047667909991186387, + "loss": 2.8570074081420898, + "step": 29440 + }, + { + "epoch": 0.14725, + "grad_norm": 0.3464270532131195, + "learning_rate": 0.0004766623680427896, + "loss": 2.8442163467407227, + "step": 29450 + }, + { + "epoch": 0.1473, + "grad_norm": 0.36884447932243347, + "learning_rate": 0.0004766456304674917, + "loss": 2.8498056411743162, + "step": 29460 + }, + { + "epoch": 0.14735, + "grad_norm": 0.43602126836776733, + "learning_rate": 0.0004766288871863915, + "loss": 2.9260597229003906, + "step": 29470 + }, + { + "epoch": 0.1474, + "grad_norm": 0.3817833960056305, + "learning_rate": 0.00047661213819991067, + "loss": 2.891306686401367, + "step": 29480 + }, + { + "epoch": 0.14745, + "grad_norm": 0.36539995670318604, + "learning_rate": 0.00047659538350847076, + "loss": 2.885457992553711, + "step": 29490 + }, + { + "epoch": 0.1475, + "grad_norm": 0.35087233781814575, + "learning_rate": 0.0004765786231124936, + "loss": 2.9280881881713867, + "step": 29500 + }, + { + "epoch": 0.14755, + "grad_norm": 0.38009124994277954, + "learning_rate": 0.000476561857012401, + "loss": 2.925203323364258, + "step": 29510 + }, + { + "epoch": 0.1476, + "grad_norm": 0.3548792600631714, + "learning_rate": 0.00047654508520861527, + "loss": 2.8425235748291016, + "step": 29520 + }, + { + "epoch": 0.14765, + "grad_norm": 0.37618696689605713, + "learning_rate": 0.0004765283077015585, + "loss": 2.833555221557617, + "step": 29530 + }, + { + "epoch": 0.1477, + "grad_norm": 0.36406198143959045, + "learning_rate": 0.0004765115244916531, + "loss": 2.8681772232055662, + "step": 29540 + }, + { + "epoch": 0.14775, + "grad_norm": 0.4121563136577606, + "learning_rate": 0.0004764947355793216, + "loss": 2.9155826568603516, + "step": 29550 + }, + { + "epoch": 0.1478, + "grad_norm": 0.3396623134613037, + "learning_rate": 0.00047647794096498665, + "loss": 2.8822839736938475, + "step": 29560 + }, + { + "epoch": 0.14785, + "grad_norm": 0.3770158886909485, + "learning_rate": 0.00047646114064907097, + "loss": 2.932499313354492, + "step": 29570 + }, + { + "epoch": 0.1479, + "grad_norm": 0.38873788714408875, + "learning_rate": 0.0004764443346319977, + "loss": 2.9193805694580077, + "step": 29580 + }, + { + "epoch": 0.14795, + "grad_norm": 0.4218592345714569, + "learning_rate": 0.0004764275229141898, + "loss": 2.9087852478027343, + "step": 29590 + }, + { + "epoch": 0.148, + "grad_norm": 0.36959636211395264, + "learning_rate": 0.00047641070549607045, + "loss": 2.9190597534179688, + "step": 29600 + }, + { + "epoch": 0.14805, + "grad_norm": 0.33926311135292053, + "learning_rate": 0.0004763938823780631, + "loss": 2.8736289978027343, + "step": 29610 + }, + { + "epoch": 0.1481, + "grad_norm": 0.3691504895687103, + "learning_rate": 0.00047637705356059134, + "loss": 2.907720375061035, + "step": 29620 + }, + { + "epoch": 0.14815, + "grad_norm": 0.3584154546260834, + "learning_rate": 0.0004763602190440788, + "loss": 2.877670478820801, + "step": 29630 + }, + { + "epoch": 0.1482, + "grad_norm": 0.3797438442707062, + "learning_rate": 0.00047634337882894917, + "loss": 2.917009735107422, + "step": 29640 + }, + { + "epoch": 0.14825, + "grad_norm": 0.35309895873069763, + "learning_rate": 0.00047632653291562656, + "loss": 2.872242736816406, + "step": 29650 + }, + { + "epoch": 0.1483, + "grad_norm": 0.41351985931396484, + "learning_rate": 0.000476309681304535, + "loss": 2.8577404022216797, + "step": 29660 + }, + { + "epoch": 0.14835, + "grad_norm": 0.34278273582458496, + "learning_rate": 0.0004762928239960987, + "loss": 2.94249267578125, + "step": 29670 + }, + { + "epoch": 0.1484, + "grad_norm": 0.39708468317985535, + "learning_rate": 0.000476275960990742, + "loss": 2.898551177978516, + "step": 29680 + }, + { + "epoch": 0.14845, + "grad_norm": 0.3719024956226349, + "learning_rate": 0.00047625909228888957, + "loss": 2.9215736389160156, + "step": 29690 + }, + { + "epoch": 0.1485, + "grad_norm": 0.3566451370716095, + "learning_rate": 0.000476242217890966, + "loss": 2.901099967956543, + "step": 29700 + }, + { + "epoch": 0.14855, + "grad_norm": 0.3589199483394623, + "learning_rate": 0.00047622533779739605, + "loss": 2.8539051055908202, + "step": 29710 + }, + { + "epoch": 0.1486, + "grad_norm": 0.35329771041870117, + "learning_rate": 0.00047620845200860485, + "loss": 2.9607110977172852, + "step": 29720 + }, + { + "epoch": 0.14865, + "grad_norm": 0.3623623847961426, + "learning_rate": 0.0004761915605250173, + "loss": 2.887783432006836, + "step": 29730 + }, + { + "epoch": 0.1487, + "grad_norm": 0.38750821352005005, + "learning_rate": 0.0004761746633470587, + "loss": 2.865289497375488, + "step": 29740 + }, + { + "epoch": 0.14875, + "grad_norm": 0.3789031505584717, + "learning_rate": 0.00047615776047515453, + "loss": 2.9391979217529296, + "step": 29750 + }, + { + "epoch": 0.1488, + "grad_norm": 0.3555009961128235, + "learning_rate": 0.0004761408519097302, + "loss": 2.870024871826172, + "step": 29760 + }, + { + "epoch": 0.14885, + "grad_norm": 0.3499075174331665, + "learning_rate": 0.0004761239376512114, + "loss": 2.9096090316772463, + "step": 29770 + }, + { + "epoch": 0.1489, + "grad_norm": 0.3935737609863281, + "learning_rate": 0.0004761070177000241, + "loss": 2.8453304290771486, + "step": 29780 + }, + { + "epoch": 0.14895, + "grad_norm": 0.3484538495540619, + "learning_rate": 0.00047609009205659404, + "loss": 2.8719814300537108, + "step": 29790 + }, + { + "epoch": 0.149, + "grad_norm": 0.4013405740261078, + "learning_rate": 0.00047607316072134745, + "loss": 2.8414846420288087, + "step": 29800 + }, + { + "epoch": 0.14905, + "grad_norm": 0.36196014285087585, + "learning_rate": 0.0004760562236947106, + "loss": 2.880062294006348, + "step": 29810 + }, + { + "epoch": 0.1491, + "grad_norm": 0.3675706088542938, + "learning_rate": 0.0004760392809771098, + "loss": 2.889408302307129, + "step": 29820 + }, + { + "epoch": 0.14915, + "grad_norm": 0.33114901185035706, + "learning_rate": 0.00047602233256897154, + "loss": 2.906082534790039, + "step": 29830 + }, + { + "epoch": 0.1492, + "grad_norm": 0.3725619912147522, + "learning_rate": 0.0004760053784707227, + "loss": 2.919314956665039, + "step": 29840 + }, + { + "epoch": 0.14925, + "grad_norm": 0.37527456879615784, + "learning_rate": 0.0004759884186827899, + "loss": 2.8563976287841797, + "step": 29850 + }, + { + "epoch": 0.1493, + "grad_norm": 0.3648639917373657, + "learning_rate": 0.0004759714532056001, + "loss": 2.9212469100952148, + "step": 29860 + }, + { + "epoch": 0.14935, + "grad_norm": 0.363153338432312, + "learning_rate": 0.0004759544820395805, + "loss": 2.912179183959961, + "step": 29870 + }, + { + "epoch": 0.1494, + "grad_norm": 0.37588489055633545, + "learning_rate": 0.00047593750518515836, + "loss": 2.8742488861083983, + "step": 29880 + }, + { + "epoch": 0.14945, + "grad_norm": 0.3655509054660797, + "learning_rate": 0.0004759205226427611, + "loss": 2.904946708679199, + "step": 29890 + }, + { + "epoch": 0.1495, + "grad_norm": 0.3361775279045105, + "learning_rate": 0.0004759035344128161, + "loss": 2.8864023208618166, + "step": 29900 + }, + { + "epoch": 0.14955, + "grad_norm": 0.3520027995109558, + "learning_rate": 0.00047588654049575115, + "loss": 2.9023277282714846, + "step": 29910 + }, + { + "epoch": 0.1496, + "grad_norm": 0.38592228293418884, + "learning_rate": 0.0004758695408919941, + "loss": 2.920467758178711, + "step": 29920 + }, + { + "epoch": 0.14965, + "grad_norm": 0.39483392238616943, + "learning_rate": 0.0004758525356019728, + "loss": 2.8335033416748048, + "step": 29930 + }, + { + "epoch": 0.1497, + "grad_norm": 0.3991139829158783, + "learning_rate": 0.00047583552462611547, + "loss": 2.8875974655151366, + "step": 29940 + }, + { + "epoch": 0.14975, + "grad_norm": 0.4135494530200958, + "learning_rate": 0.0004758185079648503, + "loss": 2.9113264083862305, + "step": 29950 + }, + { + "epoch": 0.1498, + "grad_norm": 0.40162476897239685, + "learning_rate": 0.0004758014856186057, + "loss": 2.862501335144043, + "step": 29960 + }, + { + "epoch": 0.14985, + "grad_norm": 0.3809100389480591, + "learning_rate": 0.0004757844575878102, + "loss": 2.9240486145019533, + "step": 29970 + }, + { + "epoch": 0.1499, + "grad_norm": 0.3744357228279114, + "learning_rate": 0.0004757674238728925, + "loss": 2.8305925369262694, + "step": 29980 + }, + { + "epoch": 0.14995, + "grad_norm": 0.3689877390861511, + "learning_rate": 0.00047575038447428145, + "loss": 2.9360061645507813, + "step": 29990 + }, + { + "epoch": 0.15, + "grad_norm": 0.3963874876499176, + "learning_rate": 0.00047573333939240594, + "loss": 2.9167367935180666, + "step": 30000 + }, + { + "epoch": 0.15005, + "grad_norm": 0.3609394431114197, + "learning_rate": 0.00047571628862769504, + "loss": 2.865688705444336, + "step": 30010 + }, + { + "epoch": 0.1501, + "grad_norm": 0.38758817315101624, + "learning_rate": 0.0004756992321805782, + "loss": 2.8898683547973634, + "step": 30020 + }, + { + "epoch": 0.15015, + "grad_norm": 0.34626561403274536, + "learning_rate": 0.0004756821700514846, + "loss": 2.889671516418457, + "step": 30030 + }, + { + "epoch": 0.1502, + "grad_norm": 0.36643269658088684, + "learning_rate": 0.00047566510224084405, + "loss": 2.9190742492675783, + "step": 30040 + }, + { + "epoch": 0.15025, + "grad_norm": 0.36682450771331787, + "learning_rate": 0.00047564802874908586, + "loss": 2.934504508972168, + "step": 30050 + }, + { + "epoch": 0.1503, + "grad_norm": 0.3794460892677307, + "learning_rate": 0.00047563094957664014, + "loss": 2.8781543731689454, + "step": 30060 + }, + { + "epoch": 0.15035, + "grad_norm": 0.37122905254364014, + "learning_rate": 0.0004756138647239368, + "loss": 2.8783937454223634, + "step": 30070 + }, + { + "epoch": 0.1504, + "grad_norm": 0.36332499980926514, + "learning_rate": 0.0004755967741914059, + "loss": 2.952390670776367, + "step": 30080 + }, + { + "epoch": 0.15045, + "grad_norm": 0.35668322443962097, + "learning_rate": 0.0004755796779794777, + "loss": 2.906431770324707, + "step": 30090 + }, + { + "epoch": 0.1505, + "grad_norm": 0.351093590259552, + "learning_rate": 0.00047556257608858267, + "loss": 2.930133819580078, + "step": 30100 + }, + { + "epoch": 0.15055, + "grad_norm": 0.3620196282863617, + "learning_rate": 0.0004755454685191513, + "loss": 2.90014762878418, + "step": 30110 + }, + { + "epoch": 0.1506, + "grad_norm": 0.4002264738082886, + "learning_rate": 0.0004755283552716141, + "loss": 2.8750398635864256, + "step": 30120 + }, + { + "epoch": 0.15065, + "grad_norm": 0.3720206618309021, + "learning_rate": 0.00047551123634640227, + "loss": 2.9028980255126955, + "step": 30130 + }, + { + "epoch": 0.1507, + "grad_norm": 0.42239245772361755, + "learning_rate": 0.0004754941117439464, + "loss": 2.9187292098999023, + "step": 30140 + }, + { + "epoch": 0.15075, + "grad_norm": 0.3781891167163849, + "learning_rate": 0.0004754769814646779, + "loss": 2.841886520385742, + "step": 30150 + }, + { + "epoch": 0.1508, + "grad_norm": 0.340076744556427, + "learning_rate": 0.0004754598455090279, + "loss": 2.879165267944336, + "step": 30160 + }, + { + "epoch": 0.15085, + "grad_norm": 0.3441670536994934, + "learning_rate": 0.0004754427038774277, + "loss": 2.880447006225586, + "step": 30170 + }, + { + "epoch": 0.1509, + "grad_norm": 0.3878057301044464, + "learning_rate": 0.00047542555657030896, + "loss": 2.9196062088012695, + "step": 30180 + }, + { + "epoch": 0.15095, + "grad_norm": 0.36107301712036133, + "learning_rate": 0.0004754084035881034, + "loss": 2.888232421875, + "step": 30190 + }, + { + "epoch": 0.151, + "grad_norm": 0.3785131573677063, + "learning_rate": 0.00047539124493124274, + "loss": 2.8832269668579102, + "step": 30200 + }, + { + "epoch": 0.15105, + "grad_norm": 0.35633593797683716, + "learning_rate": 0.00047537408060015896, + "loss": 2.90970516204834, + "step": 30210 + }, + { + "epoch": 0.1511, + "grad_norm": 0.3369220495223999, + "learning_rate": 0.0004753569105952843, + "loss": 2.8797065734863283, + "step": 30220 + }, + { + "epoch": 0.15115, + "grad_norm": 0.36534926295280457, + "learning_rate": 0.00047533973491705085, + "loss": 2.8915838241577148, + "step": 30230 + }, + { + "epoch": 0.1512, + "grad_norm": 0.3542662262916565, + "learning_rate": 0.0004753225535658912, + "loss": 2.922660064697266, + "step": 30240 + }, + { + "epoch": 0.15125, + "grad_norm": 0.32586246728897095, + "learning_rate": 0.0004753053665422376, + "loss": 2.8987607955932617, + "step": 30250 + }, + { + "epoch": 0.1513, + "grad_norm": 0.36731430888175964, + "learning_rate": 0.000475288173846523, + "loss": 2.905399513244629, + "step": 30260 + }, + { + "epoch": 0.15135, + "grad_norm": 0.3690316379070282, + "learning_rate": 0.0004752709754791801, + "loss": 2.8902608871459963, + "step": 30270 + }, + { + "epoch": 0.1514, + "grad_norm": 0.363203763961792, + "learning_rate": 0.00047525377144064186, + "loss": 2.9342105865478514, + "step": 30280 + }, + { + "epoch": 0.15145, + "grad_norm": 0.35208481550216675, + "learning_rate": 0.0004752365617313415, + "loss": 2.9168636322021486, + "step": 30290 + }, + { + "epoch": 0.1515, + "grad_norm": 0.3723727762699127, + "learning_rate": 0.00047521934635171215, + "loss": 2.9070035934448244, + "step": 30300 + }, + { + "epoch": 0.15155, + "grad_norm": 0.368895947933197, + "learning_rate": 0.00047520212530218734, + "loss": 2.898011016845703, + "step": 30310 + }, + { + "epoch": 0.1516, + "grad_norm": 0.35908111929893494, + "learning_rate": 0.00047518489858320044, + "loss": 2.888205146789551, + "step": 30320 + }, + { + "epoch": 0.15165, + "grad_norm": 0.3648492395877838, + "learning_rate": 0.00047516766619518525, + "loss": 2.917765426635742, + "step": 30330 + }, + { + "epoch": 0.1517, + "grad_norm": 0.35729771852493286, + "learning_rate": 0.00047515042813857557, + "loss": 2.898690032958984, + "step": 30340 + }, + { + "epoch": 0.15175, + "grad_norm": 0.3792121410369873, + "learning_rate": 0.0004751331844138054, + "loss": 2.8873102188110353, + "step": 30350 + }, + { + "epoch": 0.1518, + "grad_norm": 0.364906907081604, + "learning_rate": 0.00047511593502130877, + "loss": 2.886184310913086, + "step": 30360 + }, + { + "epoch": 0.15185, + "grad_norm": 0.3568670153617859, + "learning_rate": 0.00047509867996152, + "loss": 2.9018426895141602, + "step": 30370 + }, + { + "epoch": 0.1519, + "grad_norm": 0.3405749797821045, + "learning_rate": 0.00047508141923487346, + "loss": 2.9288841247558595, + "step": 30380 + }, + { + "epoch": 0.15195, + "grad_norm": 0.3623667061328888, + "learning_rate": 0.0004750641528418037, + "loss": 2.8880529403686523, + "step": 30390 + }, + { + "epoch": 0.152, + "grad_norm": 0.4123450219631195, + "learning_rate": 0.0004750468807827454, + "loss": 2.9075721740722655, + "step": 30400 + }, + { + "epoch": 0.15205, + "grad_norm": 0.3600562810897827, + "learning_rate": 0.0004750296030581334, + "loss": 2.8747095108032226, + "step": 30410 + }, + { + "epoch": 0.1521, + "grad_norm": 0.35464876890182495, + "learning_rate": 0.0004750123196684026, + "loss": 2.904144287109375, + "step": 30420 + }, + { + "epoch": 0.15215, + "grad_norm": 0.3604324758052826, + "learning_rate": 0.0004749950306139882, + "loss": 2.9585922241210936, + "step": 30430 + }, + { + "epoch": 0.1522, + "grad_norm": 0.3446435034275055, + "learning_rate": 0.0004749777358953254, + "loss": 2.936117935180664, + "step": 30440 + }, + { + "epoch": 0.15225, + "grad_norm": 0.3576695919036865, + "learning_rate": 0.0004749604355128496, + "loss": 2.905727767944336, + "step": 30450 + }, + { + "epoch": 0.1523, + "grad_norm": 0.33843761682510376, + "learning_rate": 0.00047494312946699634, + "loss": 2.9355045318603517, + "step": 30460 + }, + { + "epoch": 0.15235, + "grad_norm": 0.400594562292099, + "learning_rate": 0.0004749258177582013, + "loss": 2.878104019165039, + "step": 30470 + }, + { + "epoch": 0.1524, + "grad_norm": 0.3471226692199707, + "learning_rate": 0.0004749085003869003, + "loss": 2.86026611328125, + "step": 30480 + }, + { + "epoch": 0.15245, + "grad_norm": 0.3362470865249634, + "learning_rate": 0.0004748911773535293, + "loss": 2.869831657409668, + "step": 30490 + }, + { + "epoch": 0.1525, + "grad_norm": 0.3619175851345062, + "learning_rate": 0.0004748738486585243, + "loss": 2.8788190841674806, + "step": 30500 + }, + { + "epoch": 0.15255, + "grad_norm": 0.3611898720264435, + "learning_rate": 0.0004748565143023218, + "loss": 2.903660774230957, + "step": 30510 + }, + { + "epoch": 0.1526, + "grad_norm": 0.36823156476020813, + "learning_rate": 0.0004748391742853581, + "loss": 2.8693172454833986, + "step": 30520 + }, + { + "epoch": 0.15265, + "grad_norm": 0.3518320322036743, + "learning_rate": 0.00047482182860806953, + "loss": 2.9208398818969727, + "step": 30530 + }, + { + "epoch": 0.1527, + "grad_norm": 0.35802802443504333, + "learning_rate": 0.00047480447727089306, + "loss": 2.877161407470703, + "step": 30540 + }, + { + "epoch": 0.15275, + "grad_norm": 0.36364641785621643, + "learning_rate": 0.00047478712027426527, + "loss": 2.8575532913208006, + "step": 30550 + }, + { + "epoch": 0.1528, + "grad_norm": 0.37395039200782776, + "learning_rate": 0.0004747697576186233, + "loss": 2.8329185485839843, + "step": 30560 + }, + { + "epoch": 0.15285, + "grad_norm": 0.3735373616218567, + "learning_rate": 0.00047475238930440425, + "loss": 2.915997314453125, + "step": 30570 + }, + { + "epoch": 0.1529, + "grad_norm": 0.35031965374946594, + "learning_rate": 0.0004747350153320452, + "loss": 2.909429359436035, + "step": 30580 + }, + { + "epoch": 0.15295, + "grad_norm": 0.36210551857948303, + "learning_rate": 0.00047471763570198366, + "loss": 2.9106107711791993, + "step": 30590 + }, + { + "epoch": 0.153, + "grad_norm": 0.40287062525749207, + "learning_rate": 0.00047470025041465716, + "loss": 2.8394344329833983, + "step": 30600 + }, + { + "epoch": 0.15305, + "grad_norm": 0.3607267439365387, + "learning_rate": 0.0004746828594705033, + "loss": 2.8842226028442384, + "step": 30610 + }, + { + "epoch": 0.1531, + "grad_norm": 0.3575437664985657, + "learning_rate": 0.00047466546286996006, + "loss": 2.8881805419921873, + "step": 30620 + }, + { + "epoch": 0.15315, + "grad_norm": 0.3483883738517761, + "learning_rate": 0.00047464806061346524, + "loss": 2.8751804351806642, + "step": 30630 + }, + { + "epoch": 0.1532, + "grad_norm": 0.3588925004005432, + "learning_rate": 0.00047463065270145693, + "loss": 2.8891017913818358, + "step": 30640 + }, + { + "epoch": 0.15325, + "grad_norm": 0.39071205258369446, + "learning_rate": 0.00047461323913437347, + "loss": 2.9169490814208983, + "step": 30650 + }, + { + "epoch": 0.1533, + "grad_norm": 0.366230845451355, + "learning_rate": 0.00047459581991265325, + "loss": 2.880338668823242, + "step": 30660 + }, + { + "epoch": 0.15335, + "grad_norm": 0.34918802976608276, + "learning_rate": 0.0004745783950367348, + "loss": 2.877543067932129, + "step": 30670 + }, + { + "epoch": 0.1534, + "grad_norm": 0.36267057061195374, + "learning_rate": 0.00047456096450705664, + "loss": 2.894327163696289, + "step": 30680 + }, + { + "epoch": 0.15345, + "grad_norm": 0.3831796944141388, + "learning_rate": 0.0004745435283240578, + "loss": 2.8744438171386717, + "step": 30690 + }, + { + "epoch": 0.1535, + "grad_norm": 0.3425635099411011, + "learning_rate": 0.00047452608648817715, + "loss": 2.8959991455078127, + "step": 30700 + }, + { + "epoch": 0.15355, + "grad_norm": 0.3917529582977295, + "learning_rate": 0.00047450863899985374, + "loss": 2.890713119506836, + "step": 30710 + }, + { + "epoch": 0.1536, + "grad_norm": 0.3882993757724762, + "learning_rate": 0.0004744911858595268, + "loss": 2.862107849121094, + "step": 30720 + }, + { + "epoch": 0.15365, + "grad_norm": 0.3743784427642822, + "learning_rate": 0.0004744737270676358, + "loss": 2.8668455123901366, + "step": 30730 + }, + { + "epoch": 0.1537, + "grad_norm": 0.3564930558204651, + "learning_rate": 0.0004744562626246202, + "loss": 2.8689002990722656, + "step": 30740 + }, + { + "epoch": 0.15375, + "grad_norm": 0.343288779258728, + "learning_rate": 0.00047443879253091973, + "loss": 2.832417106628418, + "step": 30750 + }, + { + "epoch": 0.1538, + "grad_norm": 0.3537216782569885, + "learning_rate": 0.0004744213167869741, + "loss": 2.89721794128418, + "step": 30760 + }, + { + "epoch": 0.15385, + "grad_norm": 0.35423052310943604, + "learning_rate": 0.00047440383539322327, + "loss": 2.9099979400634766, + "step": 30770 + }, + { + "epoch": 0.1539, + "grad_norm": 0.3606443405151367, + "learning_rate": 0.00047438634835010753, + "loss": 2.938638687133789, + "step": 30780 + }, + { + "epoch": 0.15395, + "grad_norm": 0.3988078236579895, + "learning_rate": 0.0004743688556580669, + "loss": 2.9082914352416993, + "step": 30790 + }, + { + "epoch": 0.154, + "grad_norm": 0.4116831421852112, + "learning_rate": 0.0004743513573175418, + "loss": 2.863520050048828, + "step": 30800 + }, + { + "epoch": 0.15405, + "grad_norm": 0.36728259921073914, + "learning_rate": 0.0004743338533289728, + "loss": 2.9199520111083985, + "step": 30810 + }, + { + "epoch": 0.1541, + "grad_norm": 0.36922192573547363, + "learning_rate": 0.00047431634369280053, + "loss": 2.8509410858154296, + "step": 30820 + }, + { + "epoch": 0.15415, + "grad_norm": 0.3636712431907654, + "learning_rate": 0.0004742988284094658, + "loss": 2.9045875549316404, + "step": 30830 + }, + { + "epoch": 0.1542, + "grad_norm": 0.340957909822464, + "learning_rate": 0.00047428130747940967, + "loss": 2.872606086730957, + "step": 30840 + }, + { + "epoch": 0.15425, + "grad_norm": 0.3838902413845062, + "learning_rate": 0.000474263780903073, + "loss": 2.909902572631836, + "step": 30850 + }, + { + "epoch": 0.1543, + "grad_norm": 0.3636667728424072, + "learning_rate": 0.00047424624868089716, + "loss": 2.933151435852051, + "step": 30860 + }, + { + "epoch": 0.15435, + "grad_norm": 0.3604026734828949, + "learning_rate": 0.00047422871081332355, + "loss": 2.896988296508789, + "step": 30870 + }, + { + "epoch": 0.1544, + "grad_norm": 0.368845671415329, + "learning_rate": 0.0004742111673007936, + "loss": 2.897374153137207, + "step": 30880 + }, + { + "epoch": 0.15445, + "grad_norm": 0.338442862033844, + "learning_rate": 0.00047419361814374906, + "loss": 2.912685775756836, + "step": 30890 + }, + { + "epoch": 0.1545, + "grad_norm": 0.36140304803848267, + "learning_rate": 0.00047417606334263163, + "loss": 2.926785469055176, + "step": 30900 + }, + { + "epoch": 0.15455, + "grad_norm": 0.3833834230899811, + "learning_rate": 0.00047415850289788335, + "loss": 2.8650079727172852, + "step": 30910 + }, + { + "epoch": 0.1546, + "grad_norm": 0.33828869462013245, + "learning_rate": 0.00047414093680994625, + "loss": 2.891884994506836, + "step": 30920 + }, + { + "epoch": 0.15465, + "grad_norm": 0.3222410976886749, + "learning_rate": 0.00047412336507926257, + "loss": 2.9115741729736326, + "step": 30930 + }, + { + "epoch": 0.1547, + "grad_norm": 0.3685593903064728, + "learning_rate": 0.00047410578770627467, + "loss": 2.8643112182617188, + "step": 30940 + }, + { + "epoch": 0.15475, + "grad_norm": 0.3666355609893799, + "learning_rate": 0.000474088204691425, + "loss": 2.8886877059936524, + "step": 30950 + }, + { + "epoch": 0.1548, + "grad_norm": 0.3624544143676758, + "learning_rate": 0.0004740706160351563, + "loss": 2.807780075073242, + "step": 30960 + }, + { + "epoch": 0.15485, + "grad_norm": 0.37207698822021484, + "learning_rate": 0.0004740530217379114, + "loss": 2.8983434677124023, + "step": 30970 + }, + { + "epoch": 0.1549, + "grad_norm": 0.37499454617500305, + "learning_rate": 0.00047403542180013316, + "loss": 2.9291049957275392, + "step": 30980 + }, + { + "epoch": 0.15495, + "grad_norm": 0.3653201758861542, + "learning_rate": 0.0004740178162222646, + "loss": 2.8689592361450194, + "step": 30990 + }, + { + "epoch": 0.155, + "grad_norm": 0.37462544441223145, + "learning_rate": 0.00047400020500474905, + "loss": 2.890121078491211, + "step": 31000 + }, + { + "epoch": 0.15505, + "grad_norm": 0.34331178665161133, + "learning_rate": 0.00047398258814802986, + "loss": 2.955076217651367, + "step": 31010 + }, + { + "epoch": 0.1551, + "grad_norm": 0.3836670219898224, + "learning_rate": 0.0004739649656525505, + "loss": 2.8622655868530273, + "step": 31020 + }, + { + "epoch": 0.15515, + "grad_norm": 0.3962303400039673, + "learning_rate": 0.00047394733751875465, + "loss": 2.8900304794311524, + "step": 31030 + }, + { + "epoch": 0.1552, + "grad_norm": 0.3614294230937958, + "learning_rate": 0.000473929703747086, + "loss": 2.8437221527099608, + "step": 31040 + }, + { + "epoch": 0.15525, + "grad_norm": 0.3867725729942322, + "learning_rate": 0.0004739120643379886, + "loss": 2.8904518127441405, + "step": 31050 + }, + { + "epoch": 0.1553, + "grad_norm": 0.37520521879196167, + "learning_rate": 0.00047389441929190647, + "loss": 2.8661176681518556, + "step": 31060 + }, + { + "epoch": 0.15535, + "grad_norm": 0.4207322895526886, + "learning_rate": 0.0004738767686092839, + "loss": 2.9211606979370117, + "step": 31070 + }, + { + "epoch": 0.1554, + "grad_norm": 0.3688398599624634, + "learning_rate": 0.00047385911229056506, + "loss": 2.8480112075805666, + "step": 31080 + }, + { + "epoch": 0.15545, + "grad_norm": 0.3607187569141388, + "learning_rate": 0.0004738414503361946, + "loss": 2.866745948791504, + "step": 31090 + }, + { + "epoch": 0.1555, + "grad_norm": 0.3905791640281677, + "learning_rate": 0.00047382378274661715, + "loss": 2.9219425201416014, + "step": 31100 + }, + { + "epoch": 0.15555, + "grad_norm": 0.35826393961906433, + "learning_rate": 0.00047380610952227743, + "loss": 2.869588851928711, + "step": 31110 + }, + { + "epoch": 0.1556, + "grad_norm": 0.37525349855422974, + "learning_rate": 0.00047378843066362043, + "loss": 2.839267349243164, + "step": 31120 + }, + { + "epoch": 0.15565, + "grad_norm": 0.36594054102897644, + "learning_rate": 0.00047377074617109115, + "loss": 2.899544906616211, + "step": 31130 + }, + { + "epoch": 0.1557, + "grad_norm": 0.38381436467170715, + "learning_rate": 0.0004737530560451349, + "loss": 2.9115219116210938, + "step": 31140 + }, + { + "epoch": 0.15575, + "grad_norm": 0.3869333863258362, + "learning_rate": 0.0004737353602861969, + "loss": 2.8924514770507814, + "step": 31150 + }, + { + "epoch": 0.1558, + "grad_norm": 0.37311649322509766, + "learning_rate": 0.0004737176588947227, + "loss": 2.9178794860839843, + "step": 31160 + }, + { + "epoch": 0.15585, + "grad_norm": 0.367973268032074, + "learning_rate": 0.000473699951871158, + "loss": 2.857040596008301, + "step": 31170 + }, + { + "epoch": 0.1559, + "grad_norm": 0.3712767958641052, + "learning_rate": 0.00047368223921594844, + "loss": 2.8119102478027345, + "step": 31180 + }, + { + "epoch": 0.15595, + "grad_norm": 0.3610530197620392, + "learning_rate": 0.00047366452092954006, + "loss": 2.890353775024414, + "step": 31190 + }, + { + "epoch": 0.156, + "grad_norm": 0.3566649854183197, + "learning_rate": 0.00047364679701237877, + "loss": 2.9414155960083006, + "step": 31200 + }, + { + "epoch": 0.15605, + "grad_norm": 0.3495974540710449, + "learning_rate": 0.0004736290674649109, + "loss": 2.9020551681518554, + "step": 31210 + }, + { + "epoch": 0.1561, + "grad_norm": 0.37794560194015503, + "learning_rate": 0.0004736113322875828, + "loss": 2.865007019042969, + "step": 31220 + }, + { + "epoch": 0.15615, + "grad_norm": 0.36229407787323, + "learning_rate": 0.0004735935914808408, + "loss": 2.8862565994262694, + "step": 31230 + }, + { + "epoch": 0.1562, + "grad_norm": 0.3538976013660431, + "learning_rate": 0.00047357584504513175, + "loss": 2.921207809448242, + "step": 31240 + }, + { + "epoch": 0.15625, + "grad_norm": 0.40084919333457947, + "learning_rate": 0.0004735580929809022, + "loss": 2.8577404022216797, + "step": 31250 + }, + { + "epoch": 0.1563, + "grad_norm": 0.34350067377090454, + "learning_rate": 0.0004735403352885992, + "loss": 2.9081005096435546, + "step": 31260 + }, + { + "epoch": 0.15635, + "grad_norm": 0.39393097162246704, + "learning_rate": 0.0004735225719686698, + "loss": 2.8959136962890626, + "step": 31270 + }, + { + "epoch": 0.1564, + "grad_norm": 0.39995962381362915, + "learning_rate": 0.000473504803021561, + "loss": 2.9127714157104494, + "step": 31280 + }, + { + "epoch": 0.15645, + "grad_norm": 0.34104689955711365, + "learning_rate": 0.0004734870284477204, + "loss": 2.9031986236572265, + "step": 31290 + }, + { + "epoch": 0.1565, + "grad_norm": 0.36090540885925293, + "learning_rate": 0.0004734692482475953, + "loss": 2.882879066467285, + "step": 31300 + }, + { + "epoch": 0.15655, + "grad_norm": 0.35500720143318176, + "learning_rate": 0.0004734514624216334, + "loss": 2.936303901672363, + "step": 31310 + }, + { + "epoch": 0.1566, + "grad_norm": 0.33433911204338074, + "learning_rate": 0.0004734336709702824, + "loss": 2.9010940551757813, + "step": 31320 + }, + { + "epoch": 0.15665, + "grad_norm": 0.32997530698776245, + "learning_rate": 0.0004734158738939902, + "loss": 2.9100549697875975, + "step": 31330 + }, + { + "epoch": 0.1567, + "grad_norm": 0.35647398233413696, + "learning_rate": 0.00047339807119320496, + "loss": 2.9071046829223635, + "step": 31340 + }, + { + "epoch": 0.15675, + "grad_norm": 0.3991349935531616, + "learning_rate": 0.00047338026286837466, + "loss": 2.9042856216430666, + "step": 31350 + }, + { + "epoch": 0.1568, + "grad_norm": 0.35112717747688293, + "learning_rate": 0.00047336244891994783, + "loss": 2.842451286315918, + "step": 31360 + }, + { + "epoch": 0.15685, + "grad_norm": 0.360321044921875, + "learning_rate": 0.00047334462934837273, + "loss": 2.830997848510742, + "step": 31370 + }, + { + "epoch": 0.1569, + "grad_norm": 0.3370412290096283, + "learning_rate": 0.0004733268041540982, + "loss": 2.8914695739746095, + "step": 31380 + }, + { + "epoch": 0.15695, + "grad_norm": 0.3740256428718567, + "learning_rate": 0.00047330897333757284, + "loss": 2.8692216873168945, + "step": 31390 + }, + { + "epoch": 0.157, + "grad_norm": 0.3463596999645233, + "learning_rate": 0.0004732911368992455, + "loss": 2.8551042556762694, + "step": 31400 + }, + { + "epoch": 0.15705, + "grad_norm": 0.3530009984970093, + "learning_rate": 0.0004732732948395653, + "loss": 2.895857048034668, + "step": 31410 + }, + { + "epoch": 0.1571, + "grad_norm": 0.3657478988170624, + "learning_rate": 0.0004732554471589815, + "loss": 2.833382225036621, + "step": 31420 + }, + { + "epoch": 0.15715, + "grad_norm": 0.37972885370254517, + "learning_rate": 0.0004732375938579432, + "loss": 2.8496328353881837, + "step": 31430 + }, + { + "epoch": 0.1572, + "grad_norm": 0.3795199394226074, + "learning_rate": 0.00047321973493689995, + "loss": 2.899027633666992, + "step": 31440 + }, + { + "epoch": 0.15725, + "grad_norm": 0.3787808120250702, + "learning_rate": 0.00047320187039630145, + "loss": 2.8930789947509767, + "step": 31450 + }, + { + "epoch": 0.1573, + "grad_norm": 0.36156147718429565, + "learning_rate": 0.0004731840002365973, + "loss": 2.9376916885375977, + "step": 31460 + }, + { + "epoch": 0.15735, + "grad_norm": 0.35532811284065247, + "learning_rate": 0.00047316612445823746, + "loss": 2.8618843078613283, + "step": 31470 + }, + { + "epoch": 0.1574, + "grad_norm": 0.3570187985897064, + "learning_rate": 0.0004731482430616719, + "loss": 2.87581729888916, + "step": 31480 + }, + { + "epoch": 0.15745, + "grad_norm": 0.3432493805885315, + "learning_rate": 0.00047313035604735083, + "loss": 2.8892677307128904, + "step": 31490 + }, + { + "epoch": 0.1575, + "grad_norm": 0.36092421412467957, + "learning_rate": 0.00047311246341572456, + "loss": 2.8949581146240235, + "step": 31500 + }, + { + "epoch": 0.15755, + "grad_norm": 0.3730868399143219, + "learning_rate": 0.0004730945651672435, + "loss": 2.8840301513671873, + "step": 31510 + }, + { + "epoch": 0.1576, + "grad_norm": 0.3700283169746399, + "learning_rate": 0.0004730766613023582, + "loss": 2.8811742782592775, + "step": 31520 + }, + { + "epoch": 0.15765, + "grad_norm": 0.3629869818687439, + "learning_rate": 0.00047305875182151953, + "loss": 2.889894485473633, + "step": 31530 + }, + { + "epoch": 0.1577, + "grad_norm": 0.36586254835128784, + "learning_rate": 0.0004730408367251782, + "loss": 2.84090518951416, + "step": 31540 + }, + { + "epoch": 0.15775, + "grad_norm": 0.3496834933757782, + "learning_rate": 0.00047302291601378536, + "loss": 2.8973834991455076, + "step": 31550 + }, + { + "epoch": 0.1578, + "grad_norm": 0.34187740087509155, + "learning_rate": 0.000473004989687792, + "loss": 2.952738952636719, + "step": 31560 + }, + { + "epoch": 0.15785, + "grad_norm": 0.3635886311531067, + "learning_rate": 0.0004729870577476496, + "loss": 2.939321517944336, + "step": 31570 + }, + { + "epoch": 0.1579, + "grad_norm": 0.37050265073776245, + "learning_rate": 0.00047296912019380955, + "loss": 2.906153106689453, + "step": 31580 + }, + { + "epoch": 0.15795, + "grad_norm": 0.34684354066848755, + "learning_rate": 0.0004729511770267233, + "loss": 2.903788185119629, + "step": 31590 + }, + { + "epoch": 0.158, + "grad_norm": 0.35707247257232666, + "learning_rate": 0.0004729332282468427, + "loss": 2.929742431640625, + "step": 31600 + }, + { + "epoch": 0.15805, + "grad_norm": 0.3882448673248291, + "learning_rate": 0.0004729152738546195, + "loss": 2.936435508728027, + "step": 31610 + }, + { + "epoch": 0.1581, + "grad_norm": 0.38275983929634094, + "learning_rate": 0.00047289731385050587, + "loss": 2.9554168701171877, + "step": 31620 + }, + { + "epoch": 0.15815, + "grad_norm": 0.350411981344223, + "learning_rate": 0.00047287934823495384, + "loss": 2.8486440658569334, + "step": 31630 + }, + { + "epoch": 0.1582, + "grad_norm": 0.35367387533187866, + "learning_rate": 0.00047286137700841563, + "loss": 2.8513731002807616, + "step": 31640 + }, + { + "epoch": 0.15825, + "grad_norm": 0.38809311389923096, + "learning_rate": 0.0004728434001713438, + "loss": 2.882701110839844, + "step": 31650 + }, + { + "epoch": 0.1583, + "grad_norm": 0.40513256192207336, + "learning_rate": 0.00047282541772419086, + "loss": 2.865403747558594, + "step": 31660 + }, + { + "epoch": 0.15835, + "grad_norm": 0.3608868718147278, + "learning_rate": 0.0004728074296674095, + "loss": 2.8787166595458986, + "step": 31670 + }, + { + "epoch": 0.1584, + "grad_norm": 0.3921208381652832, + "learning_rate": 0.00047278943600145265, + "loss": 2.9077648162841796, + "step": 31680 + }, + { + "epoch": 0.15845, + "grad_norm": 0.371097207069397, + "learning_rate": 0.00047277143672677316, + "loss": 2.8666349411010743, + "step": 31690 + }, + { + "epoch": 0.1585, + "grad_norm": 0.3803045153617859, + "learning_rate": 0.00047275343184382423, + "loss": 2.864592170715332, + "step": 31700 + }, + { + "epoch": 0.15855, + "grad_norm": 0.38606661558151245, + "learning_rate": 0.00047273542135305917, + "loss": 2.8642562866210937, + "step": 31710 + }, + { + "epoch": 0.1586, + "grad_norm": 0.36100220680236816, + "learning_rate": 0.00047271740525493133, + "loss": 2.890386199951172, + "step": 31720 + }, + { + "epoch": 0.15865, + "grad_norm": 0.3634166419506073, + "learning_rate": 0.0004726993835498944, + "loss": 2.8980201721191405, + "step": 31730 + }, + { + "epoch": 0.1587, + "grad_norm": 0.3467034101486206, + "learning_rate": 0.0004726813562384019, + "loss": 2.899850082397461, + "step": 31740 + }, + { + "epoch": 0.15875, + "grad_norm": 0.34678101539611816, + "learning_rate": 0.00047266332332090776, + "loss": 2.8750612258911135, + "step": 31750 + }, + { + "epoch": 0.1588, + "grad_norm": 0.38339483737945557, + "learning_rate": 0.00047264528479786595, + "loss": 2.8755010604858398, + "step": 31760 + }, + { + "epoch": 0.15885, + "grad_norm": 0.3655926585197449, + "learning_rate": 0.0004726272406697306, + "loss": 2.903726577758789, + "step": 31770 + }, + { + "epoch": 0.1589, + "grad_norm": 0.3509470224380493, + "learning_rate": 0.0004726091909369559, + "loss": 2.87332820892334, + "step": 31780 + }, + { + "epoch": 0.15895, + "grad_norm": 0.3346630930900574, + "learning_rate": 0.00047259113559999636, + "loss": 2.8762796401977537, + "step": 31790 + }, + { + "epoch": 0.159, + "grad_norm": 0.3337944746017456, + "learning_rate": 0.0004725730746593064, + "loss": 2.8313404083251954, + "step": 31800 + }, + { + "epoch": 0.15905, + "grad_norm": 0.37404730916023254, + "learning_rate": 0.00047255500811534086, + "loss": 2.8354619979858398, + "step": 31810 + }, + { + "epoch": 0.1591, + "grad_norm": 0.36992835998535156, + "learning_rate": 0.0004725369359685544, + "loss": 2.8762832641601563, + "step": 31820 + }, + { + "epoch": 0.15915, + "grad_norm": 0.3898414969444275, + "learning_rate": 0.0004725188582194021, + "loss": 2.9062658309936524, + "step": 31830 + }, + { + "epoch": 0.1592, + "grad_norm": 0.3643850088119507, + "learning_rate": 0.0004725007748683391, + "loss": 2.829255485534668, + "step": 31840 + }, + { + "epoch": 0.15925, + "grad_norm": 0.3828677833080292, + "learning_rate": 0.00047248268591582043, + "loss": 2.9022125244140624, + "step": 31850 + }, + { + "epoch": 0.1593, + "grad_norm": 0.3576291501522064, + "learning_rate": 0.00047246459136230176, + "loss": 2.8975042343139648, + "step": 31860 + }, + { + "epoch": 0.15935, + "grad_norm": 0.33482152223587036, + "learning_rate": 0.0004724464912082385, + "loss": 2.9022214889526365, + "step": 31870 + }, + { + "epoch": 0.1594, + "grad_norm": 0.33900001645088196, + "learning_rate": 0.0004724283854540862, + "loss": 2.905984878540039, + "step": 31880 + }, + { + "epoch": 0.15945, + "grad_norm": 0.36863431334495544, + "learning_rate": 0.0004724102741003008, + "loss": 2.8543912887573244, + "step": 31890 + }, + { + "epoch": 0.1595, + "grad_norm": 0.3388134241104126, + "learning_rate": 0.0004723921571473383, + "loss": 2.8226253509521486, + "step": 31900 + }, + { + "epoch": 0.15955, + "grad_norm": 0.38003894686698914, + "learning_rate": 0.00047237403459565464, + "loss": 2.86840934753418, + "step": 31910 + }, + { + "epoch": 0.1596, + "grad_norm": 0.34327104687690735, + "learning_rate": 0.00047235590644570624, + "loss": 2.913662338256836, + "step": 31920 + }, + { + "epoch": 0.15965, + "grad_norm": 0.3168874680995941, + "learning_rate": 0.00047233777269794933, + "loss": 2.9518285751342774, + "step": 31930 + }, + { + "epoch": 0.1597, + "grad_norm": 0.37042123079299927, + "learning_rate": 0.0004723196333528405, + "loss": 2.8598606109619142, + "step": 31940 + }, + { + "epoch": 0.15975, + "grad_norm": 0.35344940423965454, + "learning_rate": 0.00047230148841083635, + "loss": 2.839743804931641, + "step": 31950 + }, + { + "epoch": 0.1598, + "grad_norm": 0.34573569893836975, + "learning_rate": 0.00047228333787239374, + "loss": 2.9304178237915037, + "step": 31960 + }, + { + "epoch": 0.15985, + "grad_norm": 0.38310015201568604, + "learning_rate": 0.0004722651817379696, + "loss": 2.886104393005371, + "step": 31970 + }, + { + "epoch": 0.1599, + "grad_norm": 0.357001394033432, + "learning_rate": 0.000472247020008021, + "loss": 2.8651065826416016, + "step": 31980 + }, + { + "epoch": 0.15995, + "grad_norm": 0.36843836307525635, + "learning_rate": 0.00047222885268300514, + "loss": 2.9189485549926757, + "step": 31990 + }, + { + "epoch": 0.16, + "grad_norm": 0.3369367718696594, + "learning_rate": 0.0004722106797633794, + "loss": 2.916453742980957, + "step": 32000 + }, + { + "epoch": 0.16005, + "grad_norm": 0.3567599952220917, + "learning_rate": 0.0004721925012496012, + "loss": 2.8698457717895507, + "step": 32010 + }, + { + "epoch": 0.1601, + "grad_norm": 0.3809739947319031, + "learning_rate": 0.00047217431714212843, + "loss": 2.848765182495117, + "step": 32020 + }, + { + "epoch": 0.16015, + "grad_norm": 0.3409351706504822, + "learning_rate": 0.0004721561274414186, + "loss": 2.9157583236694338, + "step": 32030 + }, + { + "epoch": 0.1602, + "grad_norm": 0.3885153532028198, + "learning_rate": 0.0004721379321479298, + "loss": 2.9186332702636717, + "step": 32040 + }, + { + "epoch": 0.16025, + "grad_norm": 0.37899234890937805, + "learning_rate": 0.00047211973126212, + "loss": 2.8514358520507814, + "step": 32050 + }, + { + "epoch": 0.1603, + "grad_norm": 0.3449757695198059, + "learning_rate": 0.00047210152478444745, + "loss": 2.8593544006347655, + "step": 32060 + }, + { + "epoch": 0.16035, + "grad_norm": 0.3424462080001831, + "learning_rate": 0.0004720833127153705, + "loss": 2.8383684158325195, + "step": 32070 + }, + { + "epoch": 0.1604, + "grad_norm": 0.35383179783821106, + "learning_rate": 0.0004720650950553477, + "loss": 2.8746910095214844, + "step": 32080 + }, + { + "epoch": 0.16045, + "grad_norm": 0.3496774435043335, + "learning_rate": 0.0004720468718048375, + "loss": 2.873851203918457, + "step": 32090 + }, + { + "epoch": 0.1605, + "grad_norm": 0.37769970297813416, + "learning_rate": 0.0004720286429642988, + "loss": 2.851235771179199, + "step": 32100 + }, + { + "epoch": 0.16055, + "grad_norm": 0.3383680582046509, + "learning_rate": 0.00047201040853419066, + "loss": 2.874506378173828, + "step": 32110 + }, + { + "epoch": 0.1606, + "grad_norm": 0.503634512424469, + "learning_rate": 0.0004719921685149718, + "loss": 2.8594615936279295, + "step": 32120 + }, + { + "epoch": 0.16065, + "grad_norm": 0.3877790868282318, + "learning_rate": 0.00047197392290710164, + "loss": 2.9157888412475588, + "step": 32130 + }, + { + "epoch": 0.1607, + "grad_norm": 0.3477957844734192, + "learning_rate": 0.0004719556717110394, + "loss": 2.8510643005371095, + "step": 32140 + }, + { + "epoch": 0.16075, + "grad_norm": 0.3550337851047516, + "learning_rate": 0.00047193741492724465, + "loss": 2.8072513580322265, + "step": 32150 + }, + { + "epoch": 0.1608, + "grad_norm": 0.36055779457092285, + "learning_rate": 0.000471919152556177, + "loss": 2.922962951660156, + "step": 32160 + }, + { + "epoch": 0.16085, + "grad_norm": 0.35376110672950745, + "learning_rate": 0.0004719008845982962, + "loss": 2.8426198959350586, + "step": 32170 + }, + { + "epoch": 0.1609, + "grad_norm": 0.35543879866600037, + "learning_rate": 0.00047188261105406197, + "loss": 2.8519344329833984, + "step": 32180 + }, + { + "epoch": 0.16095, + "grad_norm": 0.3397212326526642, + "learning_rate": 0.0004718643319239346, + "loss": 2.9258956909179688, + "step": 32190 + }, + { + "epoch": 0.161, + "grad_norm": 0.416713148355484, + "learning_rate": 0.0004718460472083741, + "loss": 2.8919198989868162, + "step": 32200 + }, + { + "epoch": 0.16105, + "grad_norm": 0.3356587290763855, + "learning_rate": 0.00047182775690784084, + "loss": 2.813313293457031, + "step": 32210 + }, + { + "epoch": 0.1611, + "grad_norm": 0.3637433350086212, + "learning_rate": 0.0004718094610227953, + "loss": 2.871041107177734, + "step": 32220 + }, + { + "epoch": 0.16115, + "grad_norm": 0.37584784626960754, + "learning_rate": 0.0004717911595536981, + "loss": 2.9155193328857423, + "step": 32230 + }, + { + "epoch": 0.1612, + "grad_norm": 0.375392347574234, + "learning_rate": 0.0004717728525010099, + "loss": 2.880098342895508, + "step": 32240 + }, + { + "epoch": 0.16125, + "grad_norm": 0.3428201675415039, + "learning_rate": 0.0004717545398651917, + "loss": 2.8859931945800783, + "step": 32250 + }, + { + "epoch": 0.1613, + "grad_norm": 0.36967119574546814, + "learning_rate": 0.00047173622164670435, + "loss": 2.839415168762207, + "step": 32260 + }, + { + "epoch": 0.16135, + "grad_norm": 0.3310029208660126, + "learning_rate": 0.0004717178978460091, + "loss": 2.847917366027832, + "step": 32270 + }, + { + "epoch": 0.1614, + "grad_norm": 0.35709744691848755, + "learning_rate": 0.0004716995684635673, + "loss": 2.907064437866211, + "step": 32280 + }, + { + "epoch": 0.16145, + "grad_norm": 0.4291013479232788, + "learning_rate": 0.00047168123349984034, + "loss": 2.875534248352051, + "step": 32290 + }, + { + "epoch": 0.1615, + "grad_norm": 0.35880807042121887, + "learning_rate": 0.00047166289295528976, + "loss": 2.9248483657836912, + "step": 32300 + }, + { + "epoch": 0.16155, + "grad_norm": 0.3257029950618744, + "learning_rate": 0.0004716445468303774, + "loss": 2.861863899230957, + "step": 32310 + }, + { + "epoch": 0.1616, + "grad_norm": 0.3446526825428009, + "learning_rate": 0.00047162619512556496, + "loss": 2.8774417877197265, + "step": 32320 + }, + { + "epoch": 0.16165, + "grad_norm": 0.363967627286911, + "learning_rate": 0.00047160783784131464, + "loss": 2.863408088684082, + "step": 32330 + }, + { + "epoch": 0.1617, + "grad_norm": 0.342174768447876, + "learning_rate": 0.0004715894749780884, + "loss": 2.874293899536133, + "step": 32340 + }, + { + "epoch": 0.16175, + "grad_norm": 0.33518534898757935, + "learning_rate": 0.00047157110653634865, + "loss": 2.855508232116699, + "step": 32350 + }, + { + "epoch": 0.1618, + "grad_norm": 0.353695273399353, + "learning_rate": 0.0004715527325165578, + "loss": 2.8914819717407227, + "step": 32360 + }, + { + "epoch": 0.16185, + "grad_norm": 0.36778244376182556, + "learning_rate": 0.0004715343529191783, + "loss": 2.9064022064208985, + "step": 32370 + }, + { + "epoch": 0.1619, + "grad_norm": 0.35460057854652405, + "learning_rate": 0.000471515967744673, + "loss": 2.855073356628418, + "step": 32380 + }, + { + "epoch": 0.16195, + "grad_norm": 0.402747243642807, + "learning_rate": 0.0004714975769935047, + "loss": 2.85408992767334, + "step": 32390 + }, + { + "epoch": 0.162, + "grad_norm": 0.40484192967414856, + "learning_rate": 0.0004714791806661364, + "loss": 2.8859481811523438, + "step": 32400 + }, + { + "epoch": 0.16205, + "grad_norm": 0.39356890320777893, + "learning_rate": 0.00047146077876303116, + "loss": 2.882750129699707, + "step": 32410 + }, + { + "epoch": 0.1621, + "grad_norm": 0.3983581066131592, + "learning_rate": 0.0004714423712846524, + "loss": 2.9301761627197265, + "step": 32420 + }, + { + "epoch": 0.16215, + "grad_norm": 0.34799903631210327, + "learning_rate": 0.0004714239582314633, + "loss": 2.8918691635131837, + "step": 32430 + }, + { + "epoch": 0.1622, + "grad_norm": 0.36694779992103577, + "learning_rate": 0.00047140553960392756, + "loss": 2.8191831588745115, + "step": 32440 + }, + { + "epoch": 0.16225, + "grad_norm": 0.36104363203048706, + "learning_rate": 0.00047138711540250887, + "loss": 2.9138660430908203, + "step": 32450 + }, + { + "epoch": 0.1623, + "grad_norm": 0.35287001729011536, + "learning_rate": 0.000471368685627671, + "loss": 2.9119085311889648, + "step": 32460 + }, + { + "epoch": 0.16235, + "grad_norm": 0.3770354390144348, + "learning_rate": 0.000471350250279878, + "loss": 2.9364341735839843, + "step": 32470 + }, + { + "epoch": 0.1624, + "grad_norm": 0.37521445751190186, + "learning_rate": 0.00047133180935959385, + "loss": 2.902358818054199, + "step": 32480 + }, + { + "epoch": 0.16245, + "grad_norm": 0.352951318025589, + "learning_rate": 0.0004713133628672829, + "loss": 2.906671714782715, + "step": 32490 + }, + { + "epoch": 0.1625, + "grad_norm": 0.33887606859207153, + "learning_rate": 0.0004712949108034095, + "loss": 2.8562498092651367, + "step": 32500 + }, + { + "epoch": 0.16255, + "grad_norm": 0.36549752950668335, + "learning_rate": 0.00047127645316843823, + "loss": 2.8137569427490234, + "step": 32510 + }, + { + "epoch": 0.1626, + "grad_norm": 0.3344549536705017, + "learning_rate": 0.00047125798996283376, + "loss": 2.818418502807617, + "step": 32520 + }, + { + "epoch": 0.16265, + "grad_norm": 0.3593219518661499, + "learning_rate": 0.0004712395211870609, + "loss": 2.8950761795043944, + "step": 32530 + }, + { + "epoch": 0.1627, + "grad_norm": 0.3704797327518463, + "learning_rate": 0.0004712210468415845, + "loss": 2.893121337890625, + "step": 32540 + }, + { + "epoch": 0.16275, + "grad_norm": 0.369098424911499, + "learning_rate": 0.0004712025669268697, + "loss": 2.9032020568847656, + "step": 32550 + }, + { + "epoch": 0.1628, + "grad_norm": 0.3562659025192261, + "learning_rate": 0.0004711840814433818, + "loss": 2.8190032958984377, + "step": 32560 + }, + { + "epoch": 0.16285, + "grad_norm": 0.3247053325176239, + "learning_rate": 0.0004711655903915861, + "loss": 2.8485010147094725, + "step": 32570 + }, + { + "epoch": 0.1629, + "grad_norm": 0.37667402625083923, + "learning_rate": 0.0004711470937719482, + "loss": 2.8672035217285154, + "step": 32580 + }, + { + "epoch": 0.16295, + "grad_norm": 0.3936837613582611, + "learning_rate": 0.0004711285915849337, + "loss": 2.8360570907592773, + "step": 32590 + }, + { + "epoch": 0.163, + "grad_norm": 0.3571334481239319, + "learning_rate": 0.0004711100838310083, + "loss": 2.858846664428711, + "step": 32600 + }, + { + "epoch": 0.16305, + "grad_norm": 0.35498639941215515, + "learning_rate": 0.00047109157051063806, + "loss": 2.900436592102051, + "step": 32610 + }, + { + "epoch": 0.1631, + "grad_norm": 0.3644285500049591, + "learning_rate": 0.00047107305162428896, + "loss": 2.8809476852416993, + "step": 32620 + }, + { + "epoch": 0.16315, + "grad_norm": 0.41239485144615173, + "learning_rate": 0.00047105452717242734, + "loss": 2.8632080078125, + "step": 32630 + }, + { + "epoch": 0.1632, + "grad_norm": 0.3659605383872986, + "learning_rate": 0.00047103599715551947, + "loss": 2.8888774871826173, + "step": 32640 + }, + { + "epoch": 0.16325, + "grad_norm": 0.35326099395751953, + "learning_rate": 0.0004710174615740318, + "loss": 2.848659896850586, + "step": 32650 + }, + { + "epoch": 0.1633, + "grad_norm": 0.3664577901363373, + "learning_rate": 0.00047099892042843104, + "loss": 2.855691146850586, + "step": 32660 + }, + { + "epoch": 0.16335, + "grad_norm": 0.3761183023452759, + "learning_rate": 0.0004709803737191839, + "loss": 2.8321823120117187, + "step": 32670 + }, + { + "epoch": 0.1634, + "grad_norm": 0.3295232057571411, + "learning_rate": 0.00047096182144675734, + "loss": 2.8779317855834963, + "step": 32680 + }, + { + "epoch": 0.16345, + "grad_norm": 0.34463006258010864, + "learning_rate": 0.0004709432636116184, + "loss": 2.893660545349121, + "step": 32690 + }, + { + "epoch": 0.1635, + "grad_norm": 0.37636980414390564, + "learning_rate": 0.00047092470021423437, + "loss": 2.9151174545288088, + "step": 32700 + }, + { + "epoch": 0.16355, + "grad_norm": 0.3604608476161957, + "learning_rate": 0.0004709061312550724, + "loss": 2.8962095260620115, + "step": 32710 + }, + { + "epoch": 0.1636, + "grad_norm": 0.3310059905052185, + "learning_rate": 0.00047088755673460006, + "loss": 2.894745445251465, + "step": 32720 + }, + { + "epoch": 0.16365, + "grad_norm": 0.3770290017127991, + "learning_rate": 0.000470868976653285, + "loss": 2.850640869140625, + "step": 32730 + }, + { + "epoch": 0.1637, + "grad_norm": 0.3529275953769684, + "learning_rate": 0.0004708503910115949, + "loss": 2.8641090393066406, + "step": 32740 + }, + { + "epoch": 0.16375, + "grad_norm": 0.3578837215900421, + "learning_rate": 0.00047083179980999764, + "loss": 2.872178840637207, + "step": 32750 + }, + { + "epoch": 0.1638, + "grad_norm": 0.3427543640136719, + "learning_rate": 0.0004708132030489613, + "loss": 2.8855960845947264, + "step": 32760 + }, + { + "epoch": 0.16385, + "grad_norm": 0.342026025056839, + "learning_rate": 0.00047079460072895407, + "loss": 2.8728870391845702, + "step": 32770 + }, + { + "epoch": 0.1639, + "grad_norm": 0.3861854076385498, + "learning_rate": 0.00047077599285044427, + "loss": 2.880876922607422, + "step": 32780 + }, + { + "epoch": 0.16395, + "grad_norm": 0.3473995327949524, + "learning_rate": 0.0004707573794139003, + "loss": 2.8500051498413086, + "step": 32790 + }, + { + "epoch": 0.164, + "grad_norm": 0.35751083493232727, + "learning_rate": 0.0004707387604197907, + "loss": 2.8547426223754884, + "step": 32800 + }, + { + "epoch": 0.16405, + "grad_norm": 0.36307087540626526, + "learning_rate": 0.00047072013586858444, + "loss": 2.935780715942383, + "step": 32810 + }, + { + "epoch": 0.1641, + "grad_norm": 0.3570255935192108, + "learning_rate": 0.0004707015057607501, + "loss": 2.8623016357421873, + "step": 32820 + }, + { + "epoch": 0.16415, + "grad_norm": 0.3572100102901459, + "learning_rate": 0.0004706828700967569, + "loss": 2.869291877746582, + "step": 32830 + }, + { + "epoch": 0.1642, + "grad_norm": 0.3634357452392578, + "learning_rate": 0.00047066422887707396, + "loss": 2.830093574523926, + "step": 32840 + }, + { + "epoch": 0.16425, + "grad_norm": 0.3396090567111969, + "learning_rate": 0.00047064558210217045, + "loss": 2.936289405822754, + "step": 32850 + }, + { + "epoch": 0.1643, + "grad_norm": 0.39903202652931213, + "learning_rate": 0.00047062692977251584, + "loss": 2.830102729797363, + "step": 32860 + }, + { + "epoch": 0.16435, + "grad_norm": 0.36128783226013184, + "learning_rate": 0.0004706082718885799, + "loss": 2.8398326873779296, + "step": 32870 + }, + { + "epoch": 0.1644, + "grad_norm": 0.3894776701927185, + "learning_rate": 0.0004705896084508321, + "loss": 2.8896457672119142, + "step": 32880 + }, + { + "epoch": 0.16445, + "grad_norm": 0.3601222634315491, + "learning_rate": 0.0004705709394597424, + "loss": 2.90982608795166, + "step": 32890 + }, + { + "epoch": 0.1645, + "grad_norm": 0.37544262409210205, + "learning_rate": 0.0004705522649157807, + "loss": 2.912486267089844, + "step": 32900 + }, + { + "epoch": 0.16455, + "grad_norm": 0.38186976313591003, + "learning_rate": 0.0004705335848194173, + "loss": 2.9352079391479493, + "step": 32910 + }, + { + "epoch": 0.1646, + "grad_norm": 0.3414611518383026, + "learning_rate": 0.0004705148991711223, + "loss": 2.8560249328613283, + "step": 32920 + }, + { + "epoch": 0.16465, + "grad_norm": 0.3492908477783203, + "learning_rate": 0.0004704962079713663, + "loss": 2.8668390274047852, + "step": 32930 + }, + { + "epoch": 0.1647, + "grad_norm": 0.3393482267856598, + "learning_rate": 0.00047047751122061967, + "loss": 2.908764457702637, + "step": 32940 + }, + { + "epoch": 0.16475, + "grad_norm": 0.3731714189052582, + "learning_rate": 0.0004704588089193531, + "loss": 2.837943458557129, + "step": 32950 + }, + { + "epoch": 0.1648, + "grad_norm": 0.3538128733634949, + "learning_rate": 0.0004704401010680376, + "loss": 2.8945058822631835, + "step": 32960 + }, + { + "epoch": 0.16485, + "grad_norm": 0.37358948588371277, + "learning_rate": 0.00047042138766714394, + "loss": 2.8777099609375, + "step": 32970 + }, + { + "epoch": 0.1649, + "grad_norm": 0.3700099289417267, + "learning_rate": 0.00047040266871714337, + "loss": 2.938925933837891, + "step": 32980 + }, + { + "epoch": 0.16495, + "grad_norm": 0.3334321677684784, + "learning_rate": 0.00047038394421850704, + "loss": 2.871377372741699, + "step": 32990 + }, + { + "epoch": 0.165, + "grad_norm": 0.3805723488330841, + "learning_rate": 0.00047036521417170645, + "loss": 2.8820621490478517, + "step": 33000 + }, + { + "epoch": 0.16505, + "grad_norm": 0.36282020807266235, + "learning_rate": 0.00047034647857721303, + "loss": 2.874891662597656, + "step": 33010 + }, + { + "epoch": 0.1651, + "grad_norm": 0.36534643173217773, + "learning_rate": 0.0004703277374354985, + "loss": 2.8311723709106444, + "step": 33020 + }, + { + "epoch": 0.16515, + "grad_norm": 0.38072484731674194, + "learning_rate": 0.00047030899074703466, + "loss": 2.861786460876465, + "step": 33030 + }, + { + "epoch": 0.1652, + "grad_norm": 0.38000062108039856, + "learning_rate": 0.00047029023851229346, + "loss": 2.8578121185302736, + "step": 33040 + }, + { + "epoch": 0.16525, + "grad_norm": 0.34835490584373474, + "learning_rate": 0.00047027148073174697, + "loss": 2.89776554107666, + "step": 33050 + }, + { + "epoch": 0.1653, + "grad_norm": 0.36063292622566223, + "learning_rate": 0.0004702527174058674, + "loss": 2.8012805938720704, + "step": 33060 + }, + { + "epoch": 0.16535, + "grad_norm": 0.37643373012542725, + "learning_rate": 0.00047023394853512715, + "loss": 2.907540130615234, + "step": 33070 + }, + { + "epoch": 0.1654, + "grad_norm": 0.3674621880054474, + "learning_rate": 0.0004702151741199987, + "loss": 2.8923870086669923, + "step": 33080 + }, + { + "epoch": 0.16545, + "grad_norm": 0.40026894211769104, + "learning_rate": 0.0004701963941609547, + "loss": 2.8028621673583984, + "step": 33090 + }, + { + "epoch": 0.1655, + "grad_norm": 0.35676321387290955, + "learning_rate": 0.000470177608658468, + "loss": 2.904555320739746, + "step": 33100 + }, + { + "epoch": 0.16555, + "grad_norm": 0.3698141276836395, + "learning_rate": 0.00047015881761301145, + "loss": 2.8711830139160157, + "step": 33110 + }, + { + "epoch": 0.1656, + "grad_norm": 0.3624369204044342, + "learning_rate": 0.0004701400210250581, + "loss": 2.828721046447754, + "step": 33120 + }, + { + "epoch": 0.16565, + "grad_norm": 0.3857244551181793, + "learning_rate": 0.0004701212188950813, + "loss": 2.9377681732177736, + "step": 33130 + }, + { + "epoch": 0.1657, + "grad_norm": 0.3718077838420868, + "learning_rate": 0.0004701024112235542, + "loss": 2.832421875, + "step": 33140 + }, + { + "epoch": 0.16575, + "grad_norm": 0.3542257845401764, + "learning_rate": 0.0004700835980109504, + "loss": 2.9144174575805666, + "step": 33150 + }, + { + "epoch": 0.1658, + "grad_norm": 0.35441118478775024, + "learning_rate": 0.0004700647792577434, + "loss": 2.876732063293457, + "step": 33160 + }, + { + "epoch": 0.16585, + "grad_norm": 0.37356311082839966, + "learning_rate": 0.0004700459549644072, + "loss": 2.8361087799072267, + "step": 33170 + }, + { + "epoch": 0.1659, + "grad_norm": 0.3507006764411926, + "learning_rate": 0.00047002712513141546, + "loss": 2.9088619232177733, + "step": 33180 + }, + { + "epoch": 0.16595, + "grad_norm": 0.33504629135131836, + "learning_rate": 0.0004700082897592423, + "loss": 2.8895763397216796, + "step": 33190 + }, + { + "epoch": 0.166, + "grad_norm": 0.3312457501888275, + "learning_rate": 0.000469989448848362, + "loss": 2.9022619247436525, + "step": 33200 + }, + { + "epoch": 0.16605, + "grad_norm": 0.37319114804267883, + "learning_rate": 0.0004699706023992487, + "loss": 2.8590274810791017, + "step": 33210 + }, + { + "epoch": 0.1661, + "grad_norm": 0.3540584146976471, + "learning_rate": 0.000469951750412377, + "loss": 2.888365936279297, + "step": 33220 + }, + { + "epoch": 0.16615, + "grad_norm": 0.36240264773368835, + "learning_rate": 0.0004699328928882215, + "loss": 2.835518646240234, + "step": 33230 + }, + { + "epoch": 0.1662, + "grad_norm": 0.3266156315803528, + "learning_rate": 0.00046991402982725686, + "loss": 2.88997745513916, + "step": 33240 + }, + { + "epoch": 0.16625, + "grad_norm": 0.4354550838470459, + "learning_rate": 0.00046989516122995795, + "loss": 2.951532554626465, + "step": 33250 + }, + { + "epoch": 0.1663, + "grad_norm": 0.36658114194869995, + "learning_rate": 0.0004698762870967999, + "loss": 2.9111865997314452, + "step": 33260 + }, + { + "epoch": 0.16635, + "grad_norm": 0.37289246916770935, + "learning_rate": 0.00046985740742825765, + "loss": 2.848533058166504, + "step": 33270 + }, + { + "epoch": 0.1664, + "grad_norm": 0.3373710811138153, + "learning_rate": 0.00046983852222480673, + "loss": 2.893321990966797, + "step": 33280 + }, + { + "epoch": 0.16645, + "grad_norm": 0.38067278265953064, + "learning_rate": 0.00046981963148692255, + "loss": 2.8911582946777346, + "step": 33290 + }, + { + "epoch": 0.1665, + "grad_norm": 0.36226537823677063, + "learning_rate": 0.0004698007352150805, + "loss": 2.8694286346435547, + "step": 33300 + }, + { + "epoch": 0.16655, + "grad_norm": 0.3868003487586975, + "learning_rate": 0.00046978183340975647, + "loss": 2.9222307205200195, + "step": 33310 + }, + { + "epoch": 0.1666, + "grad_norm": 0.38155412673950195, + "learning_rate": 0.00046976292607142623, + "loss": 2.8700397491455076, + "step": 33320 + }, + { + "epoch": 0.16665, + "grad_norm": 0.3567661643028259, + "learning_rate": 0.0004697440132005658, + "loss": 2.9206048965454103, + "step": 33330 + }, + { + "epoch": 0.1667, + "grad_norm": 0.3888932764530182, + "learning_rate": 0.0004697250947976513, + "loss": 2.8662315368652345, + "step": 33340 + }, + { + "epoch": 0.16675, + "grad_norm": 0.36754608154296875, + "learning_rate": 0.00046970617086315904, + "loss": 2.902561569213867, + "step": 33350 + }, + { + "epoch": 0.1668, + "grad_norm": 0.3562617301940918, + "learning_rate": 0.00046968724139756534, + "loss": 2.8346187591552736, + "step": 33360 + }, + { + "epoch": 0.16685, + "grad_norm": 0.3782777786254883, + "learning_rate": 0.00046966830640134686, + "loss": 2.877008247375488, + "step": 33370 + }, + { + "epoch": 0.1669, + "grad_norm": 0.33692583441734314, + "learning_rate": 0.0004696493658749802, + "loss": 2.9002532958984375, + "step": 33380 + }, + { + "epoch": 0.16695, + "grad_norm": 0.34636110067367554, + "learning_rate": 0.00046963041981894227, + "loss": 2.8916473388671875, + "step": 33390 + }, + { + "epoch": 0.167, + "grad_norm": 0.3643942177295685, + "learning_rate": 0.00046961146823370993, + "loss": 2.898698425292969, + "step": 33400 + }, + { + "epoch": 0.16705, + "grad_norm": 0.3607020080089569, + "learning_rate": 0.00046959251111976035, + "loss": 2.8341144561767577, + "step": 33410 + }, + { + "epoch": 0.1671, + "grad_norm": 0.35277801752090454, + "learning_rate": 0.00046957354847757085, + "loss": 2.8426851272583007, + "step": 33420 + }, + { + "epoch": 0.16715, + "grad_norm": 0.36254891753196716, + "learning_rate": 0.0004695545803076186, + "loss": 2.8444828033447265, + "step": 33430 + }, + { + "epoch": 0.1672, + "grad_norm": 0.3763853907585144, + "learning_rate": 0.0004695356066103814, + "loss": 2.928886604309082, + "step": 33440 + }, + { + "epoch": 0.16725, + "grad_norm": 0.3876681327819824, + "learning_rate": 0.0004695166273863367, + "loss": 2.8564325332641602, + "step": 33450 + }, + { + "epoch": 0.1673, + "grad_norm": 0.37522169947624207, + "learning_rate": 0.0004694976426359623, + "loss": 2.882516860961914, + "step": 33460 + }, + { + "epoch": 0.16735, + "grad_norm": 0.36327406764030457, + "learning_rate": 0.0004694786523597363, + "loss": 2.907136154174805, + "step": 33470 + }, + { + "epoch": 0.1674, + "grad_norm": 0.3729734420776367, + "learning_rate": 0.0004694596565581367, + "loss": 2.9024208068847654, + "step": 33480 + }, + { + "epoch": 0.16745, + "grad_norm": 0.3617364466190338, + "learning_rate": 0.0004694406552316417, + "loss": 2.872275543212891, + "step": 33490 + }, + { + "epoch": 0.1675, + "grad_norm": 0.3821735084056854, + "learning_rate": 0.00046942164838072967, + "loss": 2.909000778198242, + "step": 33500 + }, + { + "epoch": 0.16755, + "grad_norm": 0.3727242648601532, + "learning_rate": 0.0004694026360058791, + "loss": 2.8583885192871095, + "step": 33510 + }, + { + "epoch": 0.1676, + "grad_norm": 0.3401758372783661, + "learning_rate": 0.0004693836181075686, + "loss": 2.879804992675781, + "step": 33520 + }, + { + "epoch": 0.16765, + "grad_norm": 0.356801301240921, + "learning_rate": 0.000469364594686277, + "loss": 2.83564453125, + "step": 33530 + }, + { + "epoch": 0.1677, + "grad_norm": 0.3781443238258362, + "learning_rate": 0.0004693455657424832, + "loss": 2.920326042175293, + "step": 33540 + }, + { + "epoch": 0.16775, + "grad_norm": 0.3917240500450134, + "learning_rate": 0.00046932653127666627, + "loss": 2.868600082397461, + "step": 33550 + }, + { + "epoch": 0.1678, + "grad_norm": 0.3746403455734253, + "learning_rate": 0.0004693074912893054, + "loss": 2.882185363769531, + "step": 33560 + }, + { + "epoch": 0.16785, + "grad_norm": 0.3831437826156616, + "learning_rate": 0.00046928844578087987, + "loss": 2.8795846939086913, + "step": 33570 + }, + { + "epoch": 0.1679, + "grad_norm": 0.34182673692703247, + "learning_rate": 0.00046926939475186913, + "loss": 2.878773880004883, + "step": 33580 + }, + { + "epoch": 0.16795, + "grad_norm": 0.3958249092102051, + "learning_rate": 0.0004692503382027529, + "loss": 2.943048095703125, + "step": 33590 + }, + { + "epoch": 0.168, + "grad_norm": 0.3188331127166748, + "learning_rate": 0.0004692312761340109, + "loss": 2.8311105728149415, + "step": 33600 + }, + { + "epoch": 0.16805, + "grad_norm": 0.37639880180358887, + "learning_rate": 0.00046921220854612296, + "loss": 2.8699758529663084, + "step": 33610 + }, + { + "epoch": 0.1681, + "grad_norm": 0.37278783321380615, + "learning_rate": 0.0004691931354395691, + "loss": 2.8952518463134767, + "step": 33620 + }, + { + "epoch": 0.16815, + "grad_norm": 0.3637174665927887, + "learning_rate": 0.0004691740568148296, + "loss": 2.886565017700195, + "step": 33630 + }, + { + "epoch": 0.1682, + "grad_norm": 0.3796730637550354, + "learning_rate": 0.0004691549726723846, + "loss": 2.8769222259521485, + "step": 33640 + }, + { + "epoch": 0.16825, + "grad_norm": 0.3637452721595764, + "learning_rate": 0.0004691358830127147, + "loss": 2.829718017578125, + "step": 33650 + }, + { + "epoch": 0.1683, + "grad_norm": 0.37838438153266907, + "learning_rate": 0.0004691167878363003, + "loss": 2.93759822845459, + "step": 33660 + }, + { + "epoch": 0.16835, + "grad_norm": 0.3508177101612091, + "learning_rate": 0.00046909768714362236, + "loss": 2.8778865814208983, + "step": 33670 + }, + { + "epoch": 0.1684, + "grad_norm": 0.3811466097831726, + "learning_rate": 0.0004690785809351615, + "loss": 2.867452621459961, + "step": 33680 + }, + { + "epoch": 0.16845, + "grad_norm": 0.35132062435150146, + "learning_rate": 0.0004690594692113989, + "loss": 2.855402183532715, + "step": 33690 + }, + { + "epoch": 0.1685, + "grad_norm": 0.38478538393974304, + "learning_rate": 0.0004690403519728156, + "loss": 2.8482643127441407, + "step": 33700 + }, + { + "epoch": 0.16855, + "grad_norm": 0.3599494397640228, + "learning_rate": 0.0004690212292198929, + "loss": 2.8570133209228517, + "step": 33710 + }, + { + "epoch": 0.1686, + "grad_norm": 0.3879190981388092, + "learning_rate": 0.00046900210095311224, + "loss": 2.8936763763427735, + "step": 33720 + }, + { + "epoch": 0.16865, + "grad_norm": 0.39148688316345215, + "learning_rate": 0.0004689829671729551, + "loss": 2.86712760925293, + "step": 33730 + }, + { + "epoch": 0.1687, + "grad_norm": 0.3168189227581024, + "learning_rate": 0.00046896382787990333, + "loss": 2.878936767578125, + "step": 33740 + }, + { + "epoch": 0.16875, + "grad_norm": 0.32902792096138, + "learning_rate": 0.0004689446830744386, + "loss": 2.8966629028320314, + "step": 33750 + }, + { + "epoch": 0.1688, + "grad_norm": 0.37276217341423035, + "learning_rate": 0.0004689255327570429, + "loss": 2.848961067199707, + "step": 33760 + }, + { + "epoch": 0.16885, + "grad_norm": 0.33403703570365906, + "learning_rate": 0.0004689063769281984, + "loss": 2.854051208496094, + "step": 33770 + }, + { + "epoch": 0.1689, + "grad_norm": 0.36744368076324463, + "learning_rate": 0.00046888721558838734, + "loss": 2.862761116027832, + "step": 33780 + }, + { + "epoch": 0.16895, + "grad_norm": 0.3389984369277954, + "learning_rate": 0.00046886804873809206, + "loss": 2.8726749420166016, + "step": 33790 + }, + { + "epoch": 0.169, + "grad_norm": 0.3417995572090149, + "learning_rate": 0.00046884887637779516, + "loss": 2.8283790588378905, + "step": 33800 + }, + { + "epoch": 0.16905, + "grad_norm": 0.36391663551330566, + "learning_rate": 0.0004688296985079792, + "loss": 2.843199348449707, + "step": 33810 + }, + { + "epoch": 0.1691, + "grad_norm": 0.36998072266578674, + "learning_rate": 0.0004688105151291271, + "loss": 2.8884239196777344, + "step": 33820 + }, + { + "epoch": 0.16915, + "grad_norm": 0.34842783212661743, + "learning_rate": 0.0004687913262417217, + "loss": 2.872066307067871, + "step": 33830 + }, + { + "epoch": 0.1692, + "grad_norm": 0.3843635320663452, + "learning_rate": 0.0004687721318462461, + "loss": 2.9328792572021483, + "step": 33840 + }, + { + "epoch": 0.16925, + "grad_norm": 0.33155885338783264, + "learning_rate": 0.00046875293194318355, + "loss": 2.904124069213867, + "step": 33850 + }, + { + "epoch": 0.1693, + "grad_norm": 0.3453630805015564, + "learning_rate": 0.0004687337265330175, + "loss": 2.84224796295166, + "step": 33860 + }, + { + "epoch": 0.16935, + "grad_norm": 0.4139905571937561, + "learning_rate": 0.00046871451561623126, + "loss": 2.849345397949219, + "step": 33870 + }, + { + "epoch": 0.1694, + "grad_norm": 0.338795930147171, + "learning_rate": 0.00046869529919330857, + "loss": 2.8335371017456055, + "step": 33880 + }, + { + "epoch": 0.16945, + "grad_norm": 0.41172078251838684, + "learning_rate": 0.00046867607726473314, + "loss": 2.8598407745361327, + "step": 33890 + }, + { + "epoch": 0.1695, + "grad_norm": 0.37777140736579895, + "learning_rate": 0.000468656849830989, + "loss": 2.8658281326293946, + "step": 33900 + }, + { + "epoch": 0.16955, + "grad_norm": 0.3869510591030121, + "learning_rate": 0.00046863761689256014, + "loss": 2.833609390258789, + "step": 33910 + }, + { + "epoch": 0.1696, + "grad_norm": 0.33651968836784363, + "learning_rate": 0.00046861837844993065, + "loss": 2.832840919494629, + "step": 33920 + }, + { + "epoch": 0.16965, + "grad_norm": 0.37961432337760925, + "learning_rate": 0.000468599134503585, + "loss": 2.8538700103759767, + "step": 33930 + }, + { + "epoch": 0.1697, + "grad_norm": 0.33605465292930603, + "learning_rate": 0.0004685798850540076, + "loss": 2.862836456298828, + "step": 33940 + }, + { + "epoch": 0.16975, + "grad_norm": 0.377586305141449, + "learning_rate": 0.00046856063010168304, + "loss": 2.883519744873047, + "step": 33950 + }, + { + "epoch": 0.1698, + "grad_norm": 0.371600866317749, + "learning_rate": 0.000468541369647096, + "loss": 2.886541748046875, + "step": 33960 + }, + { + "epoch": 0.16985, + "grad_norm": 0.3612309992313385, + "learning_rate": 0.00046852210369073154, + "loss": 2.898946189880371, + "step": 33970 + }, + { + "epoch": 0.1699, + "grad_norm": 0.3971792161464691, + "learning_rate": 0.0004685028322330746, + "loss": 2.8439313888549806, + "step": 33980 + }, + { + "epoch": 0.16995, + "grad_norm": 0.3260321617126465, + "learning_rate": 0.00046848355527461026, + "loss": 2.822545051574707, + "step": 33990 + }, + { + "epoch": 0.17, + "grad_norm": 0.35389766097068787, + "learning_rate": 0.00046846427281582395, + "loss": 2.8328792572021486, + "step": 34000 + }, + { + "epoch": 0.17005, + "grad_norm": 0.37607404589653015, + "learning_rate": 0.00046844498485720094, + "loss": 2.845869445800781, + "step": 34010 + }, + { + "epoch": 0.1701, + "grad_norm": 0.34417924284935, + "learning_rate": 0.0004684256913992269, + "loss": 2.853196716308594, + "step": 34020 + }, + { + "epoch": 0.17015, + "grad_norm": 0.3816951513290405, + "learning_rate": 0.0004684063924423876, + "loss": 2.872451591491699, + "step": 34030 + }, + { + "epoch": 0.1702, + "grad_norm": 0.3865385353565216, + "learning_rate": 0.0004683870879871688, + "loss": 2.900624084472656, + "step": 34040 + }, + { + "epoch": 0.17025, + "grad_norm": 0.3619687557220459, + "learning_rate": 0.0004683677780340565, + "loss": 2.9058500289916993, + "step": 34050 + }, + { + "epoch": 0.1703, + "grad_norm": 0.3795410096645355, + "learning_rate": 0.00046834846258353685, + "loss": 2.8569103240966798, + "step": 34060 + }, + { + "epoch": 0.17035, + "grad_norm": 0.3346679210662842, + "learning_rate": 0.00046832914163609616, + "loss": 2.883707046508789, + "step": 34070 + }, + { + "epoch": 0.1704, + "grad_norm": 0.3611302971839905, + "learning_rate": 0.0004683098151922207, + "loss": 2.843779754638672, + "step": 34080 + }, + { + "epoch": 0.17045, + "grad_norm": 0.37046417593955994, + "learning_rate": 0.0004682904832523972, + "loss": 2.8155961990356446, + "step": 34090 + }, + { + "epoch": 0.1705, + "grad_norm": 0.33567890524864197, + "learning_rate": 0.00046827114581711216, + "loss": 2.9117507934570312, + "step": 34100 + }, + { + "epoch": 0.17055, + "grad_norm": 0.36695483326911926, + "learning_rate": 0.00046825180288685253, + "loss": 2.8531417846679688, + "step": 34110 + }, + { + "epoch": 0.1706, + "grad_norm": 0.36299067735671997, + "learning_rate": 0.0004682324544621053, + "loss": 2.8395322799682616, + "step": 34120 + }, + { + "epoch": 0.17065, + "grad_norm": 0.3714747726917267, + "learning_rate": 0.0004682131005433573, + "loss": 2.845713233947754, + "step": 34130 + }, + { + "epoch": 0.1707, + "grad_norm": 0.3382774889469147, + "learning_rate": 0.0004681937411310961, + "loss": 2.836009407043457, + "step": 34140 + }, + { + "epoch": 0.17075, + "grad_norm": 0.3621121942996979, + "learning_rate": 0.00046817437622580883, + "loss": 2.8635955810546876, + "step": 34150 + }, + { + "epoch": 0.1708, + "grad_norm": 0.33840325474739075, + "learning_rate": 0.0004681550058279832, + "loss": 2.8545530319213865, + "step": 34160 + }, + { + "epoch": 0.17085, + "grad_norm": 0.31764858961105347, + "learning_rate": 0.00046813562993810667, + "loss": 2.903290939331055, + "step": 34170 + }, + { + "epoch": 0.1709, + "grad_norm": 0.40023884177207947, + "learning_rate": 0.00046811624855666714, + "loss": 2.9005596160888674, + "step": 34180 + }, + { + "epoch": 0.17095, + "grad_norm": 0.3499501347541809, + "learning_rate": 0.00046809686168415256, + "loss": 2.872231674194336, + "step": 34190 + }, + { + "epoch": 0.171, + "grad_norm": 0.3538975119590759, + "learning_rate": 0.0004680774693210509, + "loss": 2.8208494186401367, + "step": 34200 + }, + { + "epoch": 0.17105, + "grad_norm": 0.38244694471359253, + "learning_rate": 0.0004680580714678504, + "loss": 2.923151969909668, + "step": 34210 + }, + { + "epoch": 0.1711, + "grad_norm": 0.34338560700416565, + "learning_rate": 0.0004680386681250394, + "loss": 2.889633560180664, + "step": 34220 + }, + { + "epoch": 0.17115, + "grad_norm": 0.3586474359035492, + "learning_rate": 0.00046801925929310643, + "loss": 2.876970672607422, + "step": 34230 + }, + { + "epoch": 0.1712, + "grad_norm": 0.36648884415626526, + "learning_rate": 0.00046799984497254, + "loss": 2.8535228729248048, + "step": 34240 + }, + { + "epoch": 0.17125, + "grad_norm": 0.37201350927352905, + "learning_rate": 0.00046798042516382895, + "loss": 2.883792686462402, + "step": 34250 + }, + { + "epoch": 0.1713, + "grad_norm": 0.371008038520813, + "learning_rate": 0.00046796099986746217, + "loss": 2.892206573486328, + "step": 34260 + }, + { + "epoch": 0.17135, + "grad_norm": 0.37394121289253235, + "learning_rate": 0.0004679415690839287, + "loss": 2.890748977661133, + "step": 34270 + }, + { + "epoch": 0.1714, + "grad_norm": 0.3547057807445526, + "learning_rate": 0.0004679221328137176, + "loss": 2.8719377517700195, + "step": 34280 + }, + { + "epoch": 0.17145, + "grad_norm": 0.338276207447052, + "learning_rate": 0.0004679026910573183, + "loss": 2.878486442565918, + "step": 34290 + }, + { + "epoch": 0.1715, + "grad_norm": 0.3592245280742645, + "learning_rate": 0.00046788324381522023, + "loss": 2.9007612228393556, + "step": 34300 + }, + { + "epoch": 0.17155, + "grad_norm": 0.3648377060890198, + "learning_rate": 0.000467863791087913, + "loss": 2.8731388092041015, + "step": 34310 + }, + { + "epoch": 0.1716, + "grad_norm": 0.3528133034706116, + "learning_rate": 0.0004678443328758862, + "loss": 2.8956121444702148, + "step": 34320 + }, + { + "epoch": 0.17165, + "grad_norm": 0.353880375623703, + "learning_rate": 0.0004678248691796298, + "loss": 2.9032608032226563, + "step": 34330 + }, + { + "epoch": 0.1717, + "grad_norm": 0.36613017320632935, + "learning_rate": 0.0004678053999996338, + "loss": 2.863454246520996, + "step": 34340 + }, + { + "epoch": 0.17175, + "grad_norm": 0.3429817259311676, + "learning_rate": 0.0004677859253363883, + "loss": 2.88584041595459, + "step": 34350 + }, + { + "epoch": 0.1718, + "grad_norm": 0.42278507351875305, + "learning_rate": 0.0004677664451903836, + "loss": 2.8743207931518553, + "step": 34360 + }, + { + "epoch": 0.17185, + "grad_norm": 0.3411770761013031, + "learning_rate": 0.00046774695956211, + "loss": 2.87374267578125, + "step": 34370 + }, + { + "epoch": 0.1719, + "grad_norm": 0.3541390001773834, + "learning_rate": 0.00046772746845205826, + "loss": 2.864146041870117, + "step": 34380 + }, + { + "epoch": 0.17195, + "grad_norm": 0.3673609793186188, + "learning_rate": 0.0004677079718607189, + "loss": 2.8613677978515626, + "step": 34390 + }, + { + "epoch": 0.172, + "grad_norm": 0.3588378131389618, + "learning_rate": 0.00046768846978858283, + "loss": 2.847056579589844, + "step": 34400 + }, + { + "epoch": 0.17205, + "grad_norm": 0.36910420656204224, + "learning_rate": 0.000467668962236141, + "loss": 2.8368804931640623, + "step": 34410 + }, + { + "epoch": 0.1721, + "grad_norm": 0.35983800888061523, + "learning_rate": 0.0004676494492038845, + "loss": 2.854792022705078, + "step": 34420 + }, + { + "epoch": 0.17215, + "grad_norm": 0.3963858485221863, + "learning_rate": 0.00046762993069230453, + "loss": 2.9127643585205076, + "step": 34430 + }, + { + "epoch": 0.1722, + "grad_norm": 0.34875982999801636, + "learning_rate": 0.0004676104067018925, + "loss": 2.819740867614746, + "step": 34440 + }, + { + "epoch": 0.17225, + "grad_norm": 0.35709765553474426, + "learning_rate": 0.00046759087723313996, + "loss": 2.9343698501586912, + "step": 34450 + }, + { + "epoch": 0.1723, + "grad_norm": 0.34992364048957825, + "learning_rate": 0.0004675713422865386, + "loss": 2.900133514404297, + "step": 34460 + }, + { + "epoch": 0.17235, + "grad_norm": 0.36395159363746643, + "learning_rate": 0.00046755180186258005, + "loss": 2.856146049499512, + "step": 34470 + }, + { + "epoch": 0.1724, + "grad_norm": 0.33210885524749756, + "learning_rate": 0.0004675322559617564, + "loss": 2.9196529388427734, + "step": 34480 + }, + { + "epoch": 0.17245, + "grad_norm": 0.3462127447128296, + "learning_rate": 0.00046751270458455965, + "loss": 2.848832702636719, + "step": 34490 + }, + { + "epoch": 0.1725, + "grad_norm": 0.35729101300239563, + "learning_rate": 0.000467493147731482, + "loss": 2.8615455627441406, + "step": 34500 + }, + { + "epoch": 0.17255, + "grad_norm": 0.3606840968132019, + "learning_rate": 0.00046747358540301587, + "loss": 2.905168342590332, + "step": 34510 + }, + { + "epoch": 0.1726, + "grad_norm": 0.3544435501098633, + "learning_rate": 0.00046745401759965363, + "loss": 2.838613510131836, + "step": 34520 + }, + { + "epoch": 0.17265, + "grad_norm": 0.39623233675956726, + "learning_rate": 0.00046743444432188795, + "loss": 2.8501255035400392, + "step": 34530 + }, + { + "epoch": 0.1727, + "grad_norm": 0.33549192547798157, + "learning_rate": 0.00046741486557021157, + "loss": 2.8820068359375, + "step": 34540 + }, + { + "epoch": 0.17275, + "grad_norm": 0.3312508165836334, + "learning_rate": 0.00046739528134511746, + "loss": 2.8147933959960936, + "step": 34550 + }, + { + "epoch": 0.1728, + "grad_norm": 0.34172406792640686, + "learning_rate": 0.00046737569164709855, + "loss": 2.858746337890625, + "step": 34560 + }, + { + "epoch": 0.17285, + "grad_norm": 0.3524666428565979, + "learning_rate": 0.000467356096476648, + "loss": 2.8494194030761717, + "step": 34570 + }, + { + "epoch": 0.1729, + "grad_norm": 0.35976743698120117, + "learning_rate": 0.0004673364958342593, + "loss": 2.9036231994628907, + "step": 34580 + }, + { + "epoch": 0.17295, + "grad_norm": 0.3561556935310364, + "learning_rate": 0.0004673168897204257, + "loss": 2.890438270568848, + "step": 34590 + }, + { + "epoch": 0.173, + "grad_norm": 0.35742563009262085, + "learning_rate": 0.0004672972781356408, + "loss": 2.813694381713867, + "step": 34600 + }, + { + "epoch": 0.17305, + "grad_norm": 0.33499598503112793, + "learning_rate": 0.0004672776610803985, + "loss": 2.8364234924316407, + "step": 34610 + }, + { + "epoch": 0.1731, + "grad_norm": 0.3706107437610626, + "learning_rate": 0.0004672580385551924, + "loss": 2.9130474090576173, + "step": 34620 + }, + { + "epoch": 0.17315, + "grad_norm": 0.3738437294960022, + "learning_rate": 0.00046723841056051676, + "loss": 2.8362077713012694, + "step": 34630 + }, + { + "epoch": 0.1732, + "grad_norm": 0.40391597151756287, + "learning_rate": 0.0004672187770968656, + "loss": 2.8446367263793944, + "step": 34640 + }, + { + "epoch": 0.17325, + "grad_norm": 0.34431561827659607, + "learning_rate": 0.00046719913816473303, + "loss": 2.844122886657715, + "step": 34650 + }, + { + "epoch": 0.1733, + "grad_norm": 0.370770126581192, + "learning_rate": 0.00046717949376461364, + "loss": 2.8823064804077148, + "step": 34660 + }, + { + "epoch": 0.17335, + "grad_norm": 0.3858480453491211, + "learning_rate": 0.00046715984389700206, + "loss": 2.8683708190917967, + "step": 34670 + }, + { + "epoch": 0.1734, + "grad_norm": 0.3516501188278198, + "learning_rate": 0.0004671401885623927, + "loss": 2.84317569732666, + "step": 34680 + }, + { + "epoch": 0.17345, + "grad_norm": 0.3600732684135437, + "learning_rate": 0.00046712052776128065, + "loss": 2.8251312255859373, + "step": 34690 + }, + { + "epoch": 0.1735, + "grad_norm": 0.3374788761138916, + "learning_rate": 0.00046710086149416075, + "loss": 2.8710962295532227, + "step": 34700 + }, + { + "epoch": 0.17355, + "grad_norm": 0.3461740016937256, + "learning_rate": 0.00046708118976152816, + "loss": 2.861971855163574, + "step": 34710 + }, + { + "epoch": 0.1736, + "grad_norm": 0.3448116183280945, + "learning_rate": 0.00046706151256387796, + "loss": 2.8484947204589846, + "step": 34720 + }, + { + "epoch": 0.17365, + "grad_norm": 0.36544492840766907, + "learning_rate": 0.00046704182990170575, + "loss": 2.8920949935913085, + "step": 34730 + }, + { + "epoch": 0.1737, + "grad_norm": 0.37691596150398254, + "learning_rate": 0.00046702214177550685, + "loss": 2.882939910888672, + "step": 34740 + }, + { + "epoch": 0.17375, + "grad_norm": 0.34896883368492126, + "learning_rate": 0.00046700244818577696, + "loss": 2.9707351684570313, + "step": 34750 + }, + { + "epoch": 0.1738, + "grad_norm": 0.3697916269302368, + "learning_rate": 0.000466982749133012, + "loss": 2.9020463943481447, + "step": 34760 + }, + { + "epoch": 0.17385, + "grad_norm": 0.36851105093955994, + "learning_rate": 0.0004669630446177077, + "loss": 2.8620330810546877, + "step": 34770 + }, + { + "epoch": 0.1739, + "grad_norm": 0.3666597604751587, + "learning_rate": 0.00046694333464036017, + "loss": 2.8391490936279298, + "step": 34780 + }, + { + "epoch": 0.17395, + "grad_norm": 0.3592590093612671, + "learning_rate": 0.00046692361920146566, + "loss": 2.8886390686035157, + "step": 34790 + }, + { + "epoch": 0.174, + "grad_norm": 0.38388922810554504, + "learning_rate": 0.0004669038983015205, + "loss": 2.8384780883789062, + "step": 34800 + }, + { + "epoch": 0.17405, + "grad_norm": 0.3452869951725006, + "learning_rate": 0.0004668841719410212, + "loss": 2.86505126953125, + "step": 34810 + }, + { + "epoch": 0.1741, + "grad_norm": 0.3658875823020935, + "learning_rate": 0.0004668644401204642, + "loss": 2.894541358947754, + "step": 34820 + }, + { + "epoch": 0.17415, + "grad_norm": 0.38083064556121826, + "learning_rate": 0.0004668447028403464, + "loss": 2.8646717071533203, + "step": 34830 + }, + { + "epoch": 0.1742, + "grad_norm": 0.3504948019981384, + "learning_rate": 0.0004668249601011647, + "loss": 2.8015756607055664, + "step": 34840 + }, + { + "epoch": 0.17425, + "grad_norm": 0.37611445784568787, + "learning_rate": 0.0004668052119034161, + "loss": 2.8791122436523438, + "step": 34850 + }, + { + "epoch": 0.1743, + "grad_norm": 0.35822606086730957, + "learning_rate": 0.00046678545824759774, + "loss": 2.869371223449707, + "step": 34860 + }, + { + "epoch": 0.17435, + "grad_norm": 0.3835788667201996, + "learning_rate": 0.00046676569913420686, + "loss": 2.839106559753418, + "step": 34870 + }, + { + "epoch": 0.1744, + "grad_norm": 0.36951011419296265, + "learning_rate": 0.0004667459345637409, + "loss": 2.8939563751220705, + "step": 34880 + }, + { + "epoch": 0.17445, + "grad_norm": 0.32826709747314453, + "learning_rate": 0.0004667261645366976, + "loss": 2.8904912948608397, + "step": 34890 + }, + { + "epoch": 0.1745, + "grad_norm": 0.3858826458454132, + "learning_rate": 0.00046670638905357457, + "loss": 2.8847461700439454, + "step": 34900 + }, + { + "epoch": 0.17455, + "grad_norm": 0.3379826843738556, + "learning_rate": 0.00046668660811486956, + "loss": 2.822005271911621, + "step": 34910 + }, + { + "epoch": 0.1746, + "grad_norm": 0.3487701117992401, + "learning_rate": 0.00046666682172108063, + "loss": 2.8747215270996094, + "step": 34920 + }, + { + "epoch": 0.17465, + "grad_norm": 0.3467901051044464, + "learning_rate": 0.000466647029872706, + "loss": 2.8585838317871093, + "step": 34930 + }, + { + "epoch": 0.1747, + "grad_norm": 0.3632333278656006, + "learning_rate": 0.00046662723257024383, + "loss": 2.8963274002075194, + "step": 34940 + }, + { + "epoch": 0.17475, + "grad_norm": 0.33132418990135193, + "learning_rate": 0.00046660742981419244, + "loss": 2.89648551940918, + "step": 34950 + }, + { + "epoch": 0.1748, + "grad_norm": 0.3353152573108673, + "learning_rate": 0.00046658762160505055, + "loss": 2.8519514083862303, + "step": 34960 + }, + { + "epoch": 0.17485, + "grad_norm": 0.34757012128829956, + "learning_rate": 0.00046656780794331667, + "loss": 2.887396240234375, + "step": 34970 + }, + { + "epoch": 0.1749, + "grad_norm": 0.355753093957901, + "learning_rate": 0.0004665479888294897, + "loss": 2.9051277160644533, + "step": 34980 + }, + { + "epoch": 0.17495, + "grad_norm": 0.36207833886146545, + "learning_rate": 0.0004665281642640685, + "loss": 2.9007122039794924, + "step": 34990 + }, + { + "epoch": 0.175, + "grad_norm": 0.34922757744789124, + "learning_rate": 0.00046650833424755235, + "loss": 2.934766387939453, + "step": 35000 + }, + { + "epoch": 0.17505, + "grad_norm": 0.4021916091442108, + "learning_rate": 0.0004664884987804402, + "loss": 2.8134811401367186, + "step": 35010 + }, + { + "epoch": 0.1751, + "grad_norm": 0.3671434819698334, + "learning_rate": 0.00046646865786323155, + "loss": 2.8804515838623046, + "step": 35020 + }, + { + "epoch": 0.17515, + "grad_norm": 0.34408038854599, + "learning_rate": 0.00046644881149642594, + "loss": 2.879179000854492, + "step": 35030 + }, + { + "epoch": 0.1752, + "grad_norm": 0.3787565529346466, + "learning_rate": 0.0004664289596805229, + "loss": 2.8582128524780273, + "step": 35040 + }, + { + "epoch": 0.17525, + "grad_norm": 0.33475834131240845, + "learning_rate": 0.0004664091024160223, + "loss": 2.863444137573242, + "step": 35050 + }, + { + "epoch": 0.1753, + "grad_norm": 0.3313619792461395, + "learning_rate": 0.00046638923970342394, + "loss": 2.9053012847900392, + "step": 35060 + }, + { + "epoch": 0.17535, + "grad_norm": 0.3110974133014679, + "learning_rate": 0.00046636937154322794, + "loss": 2.8885272979736327, + "step": 35070 + }, + { + "epoch": 0.1754, + "grad_norm": 0.3329101800918579, + "learning_rate": 0.0004663494979359345, + "loss": 2.861586570739746, + "step": 35080 + }, + { + "epoch": 0.17545, + "grad_norm": 0.37014371156692505, + "learning_rate": 0.00046632961888204384, + "loss": 2.9507984161376952, + "step": 35090 + }, + { + "epoch": 0.1755, + "grad_norm": 0.33941519260406494, + "learning_rate": 0.0004663097343820565, + "loss": 2.8664608001708984, + "step": 35100 + }, + { + "epoch": 0.17555, + "grad_norm": 0.38293683528900146, + "learning_rate": 0.000466289844436473, + "loss": 2.8678220748901366, + "step": 35110 + }, + { + "epoch": 0.1756, + "grad_norm": 0.3598805367946625, + "learning_rate": 0.0004662699490457941, + "loss": 2.931573486328125, + "step": 35120 + }, + { + "epoch": 0.17565, + "grad_norm": 0.38759756088256836, + "learning_rate": 0.00046625004821052073, + "loss": 2.893564987182617, + "step": 35130 + }, + { + "epoch": 0.1757, + "grad_norm": 0.37073275446891785, + "learning_rate": 0.00046623014193115383, + "loss": 2.851879692077637, + "step": 35140 + }, + { + "epoch": 0.17575, + "grad_norm": 0.33898478746414185, + "learning_rate": 0.0004662102302081945, + "loss": 2.812484550476074, + "step": 35150 + }, + { + "epoch": 0.1758, + "grad_norm": 0.34878766536712646, + "learning_rate": 0.00046619031304214413, + "loss": 2.8562156677246096, + "step": 35160 + }, + { + "epoch": 0.17585, + "grad_norm": 0.35143354535102844, + "learning_rate": 0.00046617039043350407, + "loss": 2.8594131469726562, + "step": 35170 + }, + { + "epoch": 0.1759, + "grad_norm": 0.3541794419288635, + "learning_rate": 0.0004661504623827759, + "loss": 2.8002655029296877, + "step": 35180 + }, + { + "epoch": 0.17595, + "grad_norm": 0.3908710479736328, + "learning_rate": 0.00046613052889046125, + "loss": 2.872665023803711, + "step": 35190 + }, + { + "epoch": 0.176, + "grad_norm": 0.34175392985343933, + "learning_rate": 0.000466110589957062, + "loss": 2.848458480834961, + "step": 35200 + }, + { + "epoch": 0.17605, + "grad_norm": 0.38249629735946655, + "learning_rate": 0.00046609064558308, + "loss": 2.8337379455566407, + "step": 35210 + }, + { + "epoch": 0.1761, + "grad_norm": 0.32345378398895264, + "learning_rate": 0.0004660706957690175, + "loss": 2.8369436264038086, + "step": 35220 + }, + { + "epoch": 0.17615, + "grad_norm": 0.31643155217170715, + "learning_rate": 0.0004660507405153767, + "loss": 2.8253009796142576, + "step": 35230 + }, + { + "epoch": 0.1762, + "grad_norm": 0.3983455002307892, + "learning_rate": 0.00046603077982265993, + "loss": 2.7923154830932617, + "step": 35240 + }, + { + "epoch": 0.17625, + "grad_norm": 0.3185589611530304, + "learning_rate": 0.00046601081369136975, + "loss": 2.8713184356689454, + "step": 35250 + }, + { + "epoch": 0.1763, + "grad_norm": 0.3646845817565918, + "learning_rate": 0.00046599084212200874, + "loss": 2.838506889343262, + "step": 35260 + }, + { + "epoch": 0.17635, + "grad_norm": 0.3523193597793579, + "learning_rate": 0.0004659708651150798, + "loss": 2.8510063171386717, + "step": 35270 + }, + { + "epoch": 0.1764, + "grad_norm": 0.39954373240470886, + "learning_rate": 0.00046595088267108566, + "loss": 2.880158615112305, + "step": 35280 + }, + { + "epoch": 0.17645, + "grad_norm": 0.35042038559913635, + "learning_rate": 0.0004659308947905296, + "loss": 2.8365001678466797, + "step": 35290 + }, + { + "epoch": 0.1765, + "grad_norm": 0.3627223074436188, + "learning_rate": 0.0004659109014739146, + "loss": 2.863566589355469, + "step": 35300 + }, + { + "epoch": 0.17655, + "grad_norm": 0.36123403906822205, + "learning_rate": 0.00046589090272174414, + "loss": 2.8764875411987303, + "step": 35310 + }, + { + "epoch": 0.1766, + "grad_norm": 0.36399152874946594, + "learning_rate": 0.0004658708985345217, + "loss": 2.900775909423828, + "step": 35320 + }, + { + "epoch": 0.17665, + "grad_norm": 0.3608197569847107, + "learning_rate": 0.0004658508889127508, + "loss": 2.85025634765625, + "step": 35330 + }, + { + "epoch": 0.1767, + "grad_norm": 0.36838585138320923, + "learning_rate": 0.0004658308738569352, + "loss": 2.7944477081298826, + "step": 35340 + }, + { + "epoch": 0.17675, + "grad_norm": 0.35068079829216003, + "learning_rate": 0.0004658108533675788, + "loss": 2.865572929382324, + "step": 35350 + }, + { + "epoch": 0.1768, + "grad_norm": 0.35138610005378723, + "learning_rate": 0.00046579082744518566, + "loss": 2.8456029891967773, + "step": 35360 + }, + { + "epoch": 0.17685, + "grad_norm": 0.3520863652229309, + "learning_rate": 0.00046577079609025975, + "loss": 2.8279895782470703, + "step": 35370 + }, + { + "epoch": 0.1769, + "grad_norm": 0.39019787311553955, + "learning_rate": 0.0004657507593033056, + "loss": 2.8281497955322266, + "step": 35380 + }, + { + "epoch": 0.17695, + "grad_norm": 0.3302905559539795, + "learning_rate": 0.0004657307170848275, + "loss": 2.8690523147583007, + "step": 35390 + }, + { + "epoch": 0.177, + "grad_norm": 0.35701635479927063, + "learning_rate": 0.00046571066943533, + "loss": 2.851780891418457, + "step": 35400 + }, + { + "epoch": 0.17705, + "grad_norm": 0.3772318363189697, + "learning_rate": 0.00046569061635531785, + "loss": 2.877532386779785, + "step": 35410 + }, + { + "epoch": 0.1771, + "grad_norm": 0.3491792678833008, + "learning_rate": 0.00046567055784529595, + "loss": 2.9308048248291017, + "step": 35420 + }, + { + "epoch": 0.17715, + "grad_norm": 0.38268738985061646, + "learning_rate": 0.00046565049390576906, + "loss": 2.867306709289551, + "step": 35430 + }, + { + "epoch": 0.1772, + "grad_norm": 0.37545937299728394, + "learning_rate": 0.00046563042453724257, + "loss": 2.855211639404297, + "step": 35440 + }, + { + "epoch": 0.17725, + "grad_norm": 0.3278750777244568, + "learning_rate": 0.0004656103497402215, + "loss": 2.881388854980469, + "step": 35450 + }, + { + "epoch": 0.1773, + "grad_norm": 0.3889175057411194, + "learning_rate": 0.0004655902695152113, + "loss": 2.887151527404785, + "step": 35460 + }, + { + "epoch": 0.17735, + "grad_norm": 0.3417208790779114, + "learning_rate": 0.0004655701838627176, + "loss": 2.8701759338378907, + "step": 35470 + }, + { + "epoch": 0.1774, + "grad_norm": 0.349649041891098, + "learning_rate": 0.0004655500927832459, + "loss": 2.8424554824829102, + "step": 35480 + }, + { + "epoch": 0.17745, + "grad_norm": 0.3615107834339142, + "learning_rate": 0.0004655299962773021, + "loss": 2.869821548461914, + "step": 35490 + }, + { + "epoch": 0.1775, + "grad_norm": 0.35524314641952515, + "learning_rate": 0.00046550989434539203, + "loss": 2.8914169311523437, + "step": 35500 + }, + { + "epoch": 0.17755, + "grad_norm": 0.3633536398410797, + "learning_rate": 0.0004654897869880218, + "loss": 2.900125503540039, + "step": 35510 + }, + { + "epoch": 0.1776, + "grad_norm": 0.3637497127056122, + "learning_rate": 0.00046546967420569766, + "loss": 2.8948394775390627, + "step": 35520 + }, + { + "epoch": 0.17765, + "grad_norm": 0.3570825457572937, + "learning_rate": 0.00046544955599892593, + "loss": 2.8411333084106447, + "step": 35530 + }, + { + "epoch": 0.1777, + "grad_norm": 0.3751680552959442, + "learning_rate": 0.000465429432368213, + "loss": 2.8819528579711915, + "step": 35540 + }, + { + "epoch": 0.17775, + "grad_norm": 0.33762651681900024, + "learning_rate": 0.0004654093033140656, + "loss": 2.831355094909668, + "step": 35550 + }, + { + "epoch": 0.1778, + "grad_norm": 0.3599300980567932, + "learning_rate": 0.0004653891688369904, + "loss": 2.8920719146728517, + "step": 35560 + }, + { + "epoch": 0.17785, + "grad_norm": 0.348459392786026, + "learning_rate": 0.0004653690289374943, + "loss": 2.8399332046508787, + "step": 35570 + }, + { + "epoch": 0.1779, + "grad_norm": 0.37389859557151794, + "learning_rate": 0.0004653488836160844, + "loss": 2.860540580749512, + "step": 35580 + }, + { + "epoch": 0.17795, + "grad_norm": 0.34501925110816956, + "learning_rate": 0.0004653287328732677, + "loss": 2.869792938232422, + "step": 35590 + }, + { + "epoch": 0.178, + "grad_norm": 0.3527596592903137, + "learning_rate": 0.0004653085767095517, + "loss": 2.8832815170288084, + "step": 35600 + }, + { + "epoch": 0.17805, + "grad_norm": 0.3882737457752228, + "learning_rate": 0.00046528841512544363, + "loss": 2.8318912506103517, + "step": 35610 + }, + { + "epoch": 0.1781, + "grad_norm": 0.36807918548583984, + "learning_rate": 0.0004652682481214512, + "loss": 2.8364093780517576, + "step": 35620 + }, + { + "epoch": 0.17815, + "grad_norm": 0.3764789402484894, + "learning_rate": 0.00046524807569808205, + "loss": 2.842586708068848, + "step": 35630 + }, + { + "epoch": 0.1782, + "grad_norm": 0.38613957166671753, + "learning_rate": 0.000465227897855844, + "loss": 2.848115921020508, + "step": 35640 + }, + { + "epoch": 0.17825, + "grad_norm": 0.3625898063182831, + "learning_rate": 0.00046520771459524515, + "loss": 2.8690313339233398, + "step": 35650 + }, + { + "epoch": 0.1783, + "grad_norm": 0.3480524718761444, + "learning_rate": 0.0004651875259167934, + "loss": 2.8576276779174803, + "step": 35660 + }, + { + "epoch": 0.17835, + "grad_norm": 0.347021222114563, + "learning_rate": 0.0004651673318209972, + "loss": 2.8734289169311524, + "step": 35670 + }, + { + "epoch": 0.1784, + "grad_norm": 0.3919488489627838, + "learning_rate": 0.0004651471323083648, + "loss": 2.9094018936157227, + "step": 35680 + }, + { + "epoch": 0.17845, + "grad_norm": 0.32834646105766296, + "learning_rate": 0.0004651269273794049, + "loss": 2.8311023712158203, + "step": 35690 + }, + { + "epoch": 0.1785, + "grad_norm": 0.37902265787124634, + "learning_rate": 0.0004651067170346259, + "loss": 2.913722038269043, + "step": 35700 + }, + { + "epoch": 0.17855, + "grad_norm": 0.38403117656707764, + "learning_rate": 0.0004650865012745368, + "loss": 2.8982769012451173, + "step": 35710 + }, + { + "epoch": 0.1786, + "grad_norm": 0.3807055950164795, + "learning_rate": 0.0004650662800996465, + "loss": 2.890576934814453, + "step": 35720 + }, + { + "epoch": 0.17865, + "grad_norm": 0.40200045704841614, + "learning_rate": 0.000465046053510464, + "loss": 2.8167566299438476, + "step": 35730 + }, + { + "epoch": 0.1787, + "grad_norm": 0.32211408019065857, + "learning_rate": 0.00046502582150749857, + "loss": 2.8551435470581055, + "step": 35740 + }, + { + "epoch": 0.17875, + "grad_norm": 0.3171333074569702, + "learning_rate": 0.00046500558409125946, + "loss": 2.8191596984863283, + "step": 35750 + }, + { + "epoch": 0.1788, + "grad_norm": 0.3641412556171417, + "learning_rate": 0.00046498534126225625, + "loss": 2.8964649200439454, + "step": 35760 + }, + { + "epoch": 0.17885, + "grad_norm": 0.3970263600349426, + "learning_rate": 0.00046496509302099853, + "loss": 2.8556095123291017, + "step": 35770 + }, + { + "epoch": 0.1789, + "grad_norm": 0.34453317523002625, + "learning_rate": 0.00046494483936799597, + "loss": 2.867622947692871, + "step": 35780 + }, + { + "epoch": 0.17895, + "grad_norm": 0.3477690517902374, + "learning_rate": 0.0004649245803037586, + "loss": 2.934531021118164, + "step": 35790 + }, + { + "epoch": 0.179, + "grad_norm": 0.36035820841789246, + "learning_rate": 0.00046490431582879625, + "loss": 2.8853374481201173, + "step": 35800 + }, + { + "epoch": 0.17905, + "grad_norm": 0.35474956035614014, + "learning_rate": 0.0004648840459436193, + "loss": 2.901642608642578, + "step": 35810 + }, + { + "epoch": 0.1791, + "grad_norm": 0.37676921486854553, + "learning_rate": 0.00046486377064873784, + "loss": 2.921804428100586, + "step": 35820 + }, + { + "epoch": 0.17915, + "grad_norm": 0.3409601151943207, + "learning_rate": 0.0004648434899446624, + "loss": 2.8977920532226564, + "step": 35830 + }, + { + "epoch": 0.1792, + "grad_norm": 0.32157081365585327, + "learning_rate": 0.00046482320383190356, + "loss": 2.807149887084961, + "step": 35840 + }, + { + "epoch": 0.17925, + "grad_norm": 0.34273427724838257, + "learning_rate": 0.000464802912310972, + "loss": 2.8654949188232424, + "step": 35850 + }, + { + "epoch": 0.1793, + "grad_norm": 0.3589465320110321, + "learning_rate": 0.00046478261538237853, + "loss": 2.8062957763671874, + "step": 35860 + }, + { + "epoch": 0.17935, + "grad_norm": 0.3637753129005432, + "learning_rate": 0.0004647623130466342, + "loss": 2.784908103942871, + "step": 35870 + }, + { + "epoch": 0.1794, + "grad_norm": 0.35176897048950195, + "learning_rate": 0.00046474200530425, + "loss": 2.8379873275756835, + "step": 35880 + }, + { + "epoch": 0.17945, + "grad_norm": 0.3973831236362457, + "learning_rate": 0.0004647216921557373, + "loss": 2.8786251068115236, + "step": 35890 + }, + { + "epoch": 0.1795, + "grad_norm": 0.3494340777397156, + "learning_rate": 0.0004647013736016075, + "loss": 2.8385623931884765, + "step": 35900 + }, + { + "epoch": 0.17955, + "grad_norm": 0.3377736806869507, + "learning_rate": 0.00046468104964237196, + "loss": 2.8239620208740233, + "step": 35910 + }, + { + "epoch": 0.1796, + "grad_norm": 0.3748203217983246, + "learning_rate": 0.00046466072027854247, + "loss": 2.8489185333251954, + "step": 35920 + }, + { + "epoch": 0.17965, + "grad_norm": 0.31387490034103394, + "learning_rate": 0.00046464038551063074, + "loss": 2.8824533462524413, + "step": 35930 + }, + { + "epoch": 0.1797, + "grad_norm": 0.4171691834926605, + "learning_rate": 0.00046462004533914884, + "loss": 2.8493566513061523, + "step": 35940 + }, + { + "epoch": 0.17975, + "grad_norm": 0.3845033049583435, + "learning_rate": 0.00046459969976460856, + "loss": 2.8568082809448243, + "step": 35950 + }, + { + "epoch": 0.1798, + "grad_norm": 0.33021748065948486, + "learning_rate": 0.00046457934878752237, + "loss": 2.9047866821289063, + "step": 35960 + }, + { + "epoch": 0.17985, + "grad_norm": 0.3450514078140259, + "learning_rate": 0.00046455899240840253, + "loss": 2.8371788024902345, + "step": 35970 + }, + { + "epoch": 0.1799, + "grad_norm": 0.3480536639690399, + "learning_rate": 0.00046453863062776145, + "loss": 2.844780158996582, + "step": 35980 + }, + { + "epoch": 0.17995, + "grad_norm": 0.34492969512939453, + "learning_rate": 0.0004645182634461118, + "loss": 2.7975223541259764, + "step": 35990 + }, + { + "epoch": 0.18, + "grad_norm": 0.38983550667762756, + "learning_rate": 0.00046449789086396633, + "loss": 2.807935905456543, + "step": 36000 + }, + { + "epoch": 0.18005, + "grad_norm": 0.33175718784332275, + "learning_rate": 0.0004644775128818379, + "loss": 2.8025800704956056, + "step": 36010 + }, + { + "epoch": 0.1801, + "grad_norm": 0.3447435796260834, + "learning_rate": 0.00046445712950023946, + "loss": 2.847788429260254, + "step": 36020 + }, + { + "epoch": 0.18015, + "grad_norm": 0.34510424733161926, + "learning_rate": 0.00046443674071968425, + "loss": 2.8546785354614257, + "step": 36030 + }, + { + "epoch": 0.1802, + "grad_norm": 0.34879693388938904, + "learning_rate": 0.00046441634654068546, + "loss": 2.8331300735473635, + "step": 36040 + }, + { + "epoch": 0.18025, + "grad_norm": 0.3666203022003174, + "learning_rate": 0.00046439594696375664, + "loss": 2.896413040161133, + "step": 36050 + }, + { + "epoch": 0.1803, + "grad_norm": 0.33348315954208374, + "learning_rate": 0.0004643755419894113, + "loss": 2.880052947998047, + "step": 36060 + }, + { + "epoch": 0.18035, + "grad_norm": 0.3324209451675415, + "learning_rate": 0.0004643551316181631, + "loss": 2.889217185974121, + "step": 36070 + }, + { + "epoch": 0.1804, + "grad_norm": 0.3625341057777405, + "learning_rate": 0.0004643347158505259, + "loss": 2.8914003372192383, + "step": 36080 + }, + { + "epoch": 0.18045, + "grad_norm": 0.3670724332332611, + "learning_rate": 0.00046431429468701363, + "loss": 2.822999954223633, + "step": 36090 + }, + { + "epoch": 0.1805, + "grad_norm": 0.3305692672729492, + "learning_rate": 0.00046429386812814045, + "loss": 2.81317195892334, + "step": 36100 + }, + { + "epoch": 0.18055, + "grad_norm": 0.3394930064678192, + "learning_rate": 0.0004642734361744206, + "loss": 2.8530467987060546, + "step": 36110 + }, + { + "epoch": 0.1806, + "grad_norm": 0.3558415472507477, + "learning_rate": 0.0004642529988263684, + "loss": 2.870347595214844, + "step": 36120 + }, + { + "epoch": 0.18065, + "grad_norm": 0.36770880222320557, + "learning_rate": 0.00046423255608449835, + "loss": 2.8298433303833006, + "step": 36130 + }, + { + "epoch": 0.1807, + "grad_norm": 0.3326603174209595, + "learning_rate": 0.0004642121079493252, + "loss": 2.887615203857422, + "step": 36140 + }, + { + "epoch": 0.18075, + "grad_norm": 0.3507646322250366, + "learning_rate": 0.0004641916544213636, + "loss": 2.868510055541992, + "step": 36150 + }, + { + "epoch": 0.1808, + "grad_norm": 0.3802492022514343, + "learning_rate": 0.0004641711955011285, + "loss": 2.8584836959838866, + "step": 36160 + }, + { + "epoch": 0.18085, + "grad_norm": 0.3531288504600525, + "learning_rate": 0.0004641507311891351, + "loss": 2.884345054626465, + "step": 36170 + }, + { + "epoch": 0.1809, + "grad_norm": 0.3455961346626282, + "learning_rate": 0.0004641302614858984, + "loss": 2.871829605102539, + "step": 36180 + }, + { + "epoch": 0.18095, + "grad_norm": 0.35027557611465454, + "learning_rate": 0.0004641097863919338, + "loss": 2.851584625244141, + "step": 36190 + }, + { + "epoch": 0.181, + "grad_norm": 0.3834189176559448, + "learning_rate": 0.0004640893059077568, + "loss": 2.8519027709960936, + "step": 36200 + }, + { + "epoch": 0.18105, + "grad_norm": 0.3536381721496582, + "learning_rate": 0.00046406882003388296, + "loss": 2.848900032043457, + "step": 36210 + }, + { + "epoch": 0.1811, + "grad_norm": 0.38055136799812317, + "learning_rate": 0.00046404832877082794, + "loss": 2.800638771057129, + "step": 36220 + }, + { + "epoch": 0.18115, + "grad_norm": 0.31970450282096863, + "learning_rate": 0.0004640278321191077, + "loss": 2.8249521255493164, + "step": 36230 + }, + { + "epoch": 0.1812, + "grad_norm": 0.3612990379333496, + "learning_rate": 0.00046400733007923825, + "loss": 2.844767379760742, + "step": 36240 + }, + { + "epoch": 0.18125, + "grad_norm": 0.38712841272354126, + "learning_rate": 0.00046398682265173563, + "loss": 2.818136787414551, + "step": 36250 + }, + { + "epoch": 0.1813, + "grad_norm": 0.34881284832954407, + "learning_rate": 0.00046396630983711626, + "loss": 2.7910247802734376, + "step": 36260 + }, + { + "epoch": 0.18135, + "grad_norm": 0.34865301847457886, + "learning_rate": 0.0004639457916358964, + "loss": 2.8384963989257814, + "step": 36270 + }, + { + "epoch": 0.1814, + "grad_norm": 0.3340891897678375, + "learning_rate": 0.0004639252680485927, + "loss": 2.8981151580810547, + "step": 36280 + }, + { + "epoch": 0.18145, + "grad_norm": 0.33383268117904663, + "learning_rate": 0.0004639047390757218, + "loss": 2.8131135940551757, + "step": 36290 + }, + { + "epoch": 0.1815, + "grad_norm": 0.39135947823524475, + "learning_rate": 0.0004638842047178005, + "loss": 2.8313983917236327, + "step": 36300 + }, + { + "epoch": 0.18155, + "grad_norm": 0.343888521194458, + "learning_rate": 0.0004638636649753458, + "loss": 2.847258377075195, + "step": 36310 + }, + { + "epoch": 0.1816, + "grad_norm": 0.35883083939552307, + "learning_rate": 0.0004638431198488747, + "loss": 2.9056236267089846, + "step": 36320 + }, + { + "epoch": 0.18165, + "grad_norm": 0.34024789929389954, + "learning_rate": 0.0004638225693389046, + "loss": 2.8785587310791017, + "step": 36330 + }, + { + "epoch": 0.1817, + "grad_norm": 0.3684893250465393, + "learning_rate": 0.00046380201344595277, + "loss": 2.8767810821533204, + "step": 36340 + }, + { + "epoch": 0.18175, + "grad_norm": 0.3334956765174866, + "learning_rate": 0.0004637814521705365, + "loss": 2.8768171310424804, + "step": 36350 + }, + { + "epoch": 0.1818, + "grad_norm": 0.31359609961509705, + "learning_rate": 0.00046376088551317377, + "loss": 2.8678722381591797, + "step": 36360 + }, + { + "epoch": 0.18185, + "grad_norm": 0.37054598331451416, + "learning_rate": 0.00046374031347438206, + "loss": 2.8244205474853517, + "step": 36370 + }, + { + "epoch": 0.1819, + "grad_norm": 0.3252423107624054, + "learning_rate": 0.0004637197360546794, + "loss": 2.793994331359863, + "step": 36380 + }, + { + "epoch": 0.18195, + "grad_norm": 0.3679979145526886, + "learning_rate": 0.0004636991532545839, + "loss": 2.832419204711914, + "step": 36390 + }, + { + "epoch": 0.182, + "grad_norm": 0.33480438590049744, + "learning_rate": 0.0004636785650746136, + "loss": 2.874131965637207, + "step": 36400 + }, + { + "epoch": 0.18205, + "grad_norm": 0.35229620337486267, + "learning_rate": 0.0004636579715152868, + "loss": 2.828114318847656, + "step": 36410 + }, + { + "epoch": 0.1821, + "grad_norm": 0.3417867422103882, + "learning_rate": 0.0004636373725771221, + "loss": 2.9068672180175783, + "step": 36420 + }, + { + "epoch": 0.18215, + "grad_norm": 0.3490261137485504, + "learning_rate": 0.00046361676826063786, + "loss": 2.8746316909790037, + "step": 36430 + }, + { + "epoch": 0.1822, + "grad_norm": 0.3703453540802002, + "learning_rate": 0.00046359615856635296, + "loss": 2.888620376586914, + "step": 36440 + }, + { + "epoch": 0.18225, + "grad_norm": 0.34699079394340515, + "learning_rate": 0.00046357554349478626, + "loss": 2.857135009765625, + "step": 36450 + }, + { + "epoch": 0.1823, + "grad_norm": 0.3558257818222046, + "learning_rate": 0.00046355492304645657, + "loss": 2.8932432174682616, + "step": 36460 + }, + { + "epoch": 0.18235, + "grad_norm": 0.37459996342658997, + "learning_rate": 0.0004635342972218831, + "loss": 2.8989028930664062, + "step": 36470 + }, + { + "epoch": 0.1824, + "grad_norm": 0.3761593997478485, + "learning_rate": 0.00046351366602158525, + "loss": 2.8772161483764647, + "step": 36480 + }, + { + "epoch": 0.18245, + "grad_norm": 0.34658437967300415, + "learning_rate": 0.00046349302944608217, + "loss": 2.893310546875, + "step": 36490 + }, + { + "epoch": 0.1825, + "grad_norm": 0.3653164505958557, + "learning_rate": 0.0004634723874958936, + "loss": 2.820954513549805, + "step": 36500 + }, + { + "epoch": 0.18255, + "grad_norm": 0.36889123916625977, + "learning_rate": 0.000463451740171539, + "loss": 2.840134048461914, + "step": 36510 + }, + { + "epoch": 0.1826, + "grad_norm": 0.3588324189186096, + "learning_rate": 0.00046343108747353833, + "loss": 2.85176944732666, + "step": 36520 + }, + { + "epoch": 0.18265, + "grad_norm": 0.3425031304359436, + "learning_rate": 0.0004634104294024115, + "loss": 2.8051307678222654, + "step": 36530 + }, + { + "epoch": 0.1827, + "grad_norm": 0.38319602608680725, + "learning_rate": 0.00046338976595867846, + "loss": 2.832447624206543, + "step": 36540 + }, + { + "epoch": 0.18275, + "grad_norm": 0.3392985165119171, + "learning_rate": 0.0004633690971428595, + "loss": 2.9050064086914062, + "step": 36550 + }, + { + "epoch": 0.1828, + "grad_norm": 0.37056243419647217, + "learning_rate": 0.000463348422955475, + "loss": 2.8950420379638673, + "step": 36560 + }, + { + "epoch": 0.18285, + "grad_norm": 0.3484647572040558, + "learning_rate": 0.0004633277433970453, + "loss": 2.825170707702637, + "step": 36570 + }, + { + "epoch": 0.1829, + "grad_norm": 0.3421938419342041, + "learning_rate": 0.0004633070584680912, + "loss": 2.846047210693359, + "step": 36580 + }, + { + "epoch": 0.18295, + "grad_norm": 0.37613070011138916, + "learning_rate": 0.0004632863681691332, + "loss": 2.883365249633789, + "step": 36590 + }, + { + "epoch": 0.183, + "grad_norm": 0.37446966767311096, + "learning_rate": 0.00046326567250069237, + "loss": 2.911395263671875, + "step": 36600 + }, + { + "epoch": 0.18305, + "grad_norm": 0.34381353855133057, + "learning_rate": 0.0004632449714632896, + "loss": 2.8615846633911133, + "step": 36610 + }, + { + "epoch": 0.1831, + "grad_norm": 0.3685382902622223, + "learning_rate": 0.0004632242650574462, + "loss": 2.8354381561279296, + "step": 36620 + }, + { + "epoch": 0.18315, + "grad_norm": 0.3799106180667877, + "learning_rate": 0.0004632035532836832, + "loss": 2.8379165649414064, + "step": 36630 + }, + { + "epoch": 0.1832, + "grad_norm": 0.3461941182613373, + "learning_rate": 0.0004631828361425223, + "loss": 2.8859931945800783, + "step": 36640 + }, + { + "epoch": 0.18325, + "grad_norm": 0.35535120964050293, + "learning_rate": 0.0004631621136344849, + "loss": 2.874314880371094, + "step": 36650 + }, + { + "epoch": 0.1833, + "grad_norm": 0.3536674976348877, + "learning_rate": 0.00046314138576009274, + "loss": 2.847574806213379, + "step": 36660 + }, + { + "epoch": 0.18335, + "grad_norm": 0.3485620319843292, + "learning_rate": 0.0004631206525198676, + "loss": 2.9084043502807617, + "step": 36670 + }, + { + "epoch": 0.1834, + "grad_norm": 0.3164786994457245, + "learning_rate": 0.00046309991391433144, + "loss": 2.8794071197509767, + "step": 36680 + }, + { + "epoch": 0.18345, + "grad_norm": 0.3284541070461273, + "learning_rate": 0.0004630791699440063, + "loss": 2.847591209411621, + "step": 36690 + }, + { + "epoch": 0.1835, + "grad_norm": 0.3244602382183075, + "learning_rate": 0.0004630584206094146, + "loss": 2.8522838592529296, + "step": 36700 + }, + { + "epoch": 0.18355, + "grad_norm": 0.35176074504852295, + "learning_rate": 0.0004630376659110785, + "loss": 2.869820404052734, + "step": 36710 + }, + { + "epoch": 0.1836, + "grad_norm": 0.33184537291526794, + "learning_rate": 0.0004630169058495207, + "loss": 2.8564720153808594, + "step": 36720 + }, + { + "epoch": 0.18365, + "grad_norm": 0.3476483225822449, + "learning_rate": 0.0004629961404252636, + "loss": 2.79791316986084, + "step": 36730 + }, + { + "epoch": 0.1837, + "grad_norm": 0.34958648681640625, + "learning_rate": 0.00046297536963883014, + "loss": 2.8315494537353514, + "step": 36740 + }, + { + "epoch": 0.18375, + "grad_norm": 0.3624802827835083, + "learning_rate": 0.00046295459349074316, + "loss": 2.88323860168457, + "step": 36750 + }, + { + "epoch": 0.1838, + "grad_norm": 0.33011963963508606, + "learning_rate": 0.00046293381198152575, + "loss": 2.82605094909668, + "step": 36760 + }, + { + "epoch": 0.18385, + "grad_norm": 0.34822916984558105, + "learning_rate": 0.000462913025111701, + "loss": 2.8600561141967775, + "step": 36770 + }, + { + "epoch": 0.1839, + "grad_norm": 0.3423764407634735, + "learning_rate": 0.0004628922328817922, + "loss": 2.871921920776367, + "step": 36780 + }, + { + "epoch": 0.18395, + "grad_norm": 0.35145485401153564, + "learning_rate": 0.000462871435292323, + "loss": 2.845098114013672, + "step": 36790 + }, + { + "epoch": 0.184, + "grad_norm": 0.3529909551143646, + "learning_rate": 0.00046285063234381674, + "loss": 2.871572494506836, + "step": 36800 + }, + { + "epoch": 0.18405, + "grad_norm": 0.36794114112854004, + "learning_rate": 0.0004628298240367973, + "loss": 2.7973812103271483, + "step": 36810 + }, + { + "epoch": 0.1841, + "grad_norm": 0.37657827138900757, + "learning_rate": 0.00046280901037178834, + "loss": 2.796398735046387, + "step": 36820 + }, + { + "epoch": 0.18415, + "grad_norm": 0.36082029342651367, + "learning_rate": 0.0004627881913493141, + "loss": 2.8863958358764648, + "step": 36830 + }, + { + "epoch": 0.1842, + "grad_norm": 0.34642452001571655, + "learning_rate": 0.00046276736696989853, + "loss": 2.8795480728149414, + "step": 36840 + }, + { + "epoch": 0.18425, + "grad_norm": 0.3501076102256775, + "learning_rate": 0.0004627465372340659, + "loss": 2.8693620681762697, + "step": 36850 + }, + { + "epoch": 0.1843, + "grad_norm": 0.35853666067123413, + "learning_rate": 0.0004627257021423406, + "loss": 2.8461374282836913, + "step": 36860 + }, + { + "epoch": 0.18435, + "grad_norm": 0.3631095588207245, + "learning_rate": 0.0004627048616952472, + "loss": 2.8561784744262697, + "step": 36870 + }, + { + "epoch": 0.1844, + "grad_norm": 0.34762081503868103, + "learning_rate": 0.0004626840158933103, + "loss": 2.846373748779297, + "step": 36880 + }, + { + "epoch": 0.18445, + "grad_norm": 0.35966992378234863, + "learning_rate": 0.00046266316473705477, + "loss": 2.8849609375, + "step": 36890 + }, + { + "epoch": 0.1845, + "grad_norm": 0.3593134880065918, + "learning_rate": 0.0004626423082270055, + "loss": 2.860037422180176, + "step": 36900 + }, + { + "epoch": 0.18455, + "grad_norm": 0.3402668237686157, + "learning_rate": 0.00046262144636368747, + "loss": 2.8629297256469726, + "step": 36910 + }, + { + "epoch": 0.1846, + "grad_norm": 0.34717071056365967, + "learning_rate": 0.000462600579147626, + "loss": 2.8830820083618165, + "step": 36920 + }, + { + "epoch": 0.18465, + "grad_norm": 0.4019257426261902, + "learning_rate": 0.0004625797065793463, + "loss": 2.797211456298828, + "step": 36930 + }, + { + "epoch": 0.1847, + "grad_norm": 0.3470756709575653, + "learning_rate": 0.000462558828659374, + "loss": 2.8714372634887697, + "step": 36940 + }, + { + "epoch": 0.18475, + "grad_norm": 0.3458661139011383, + "learning_rate": 0.00046253794538823455, + "loss": 2.8775251388549803, + "step": 36950 + }, + { + "epoch": 0.1848, + "grad_norm": 0.33619940280914307, + "learning_rate": 0.00046251705676645375, + "loss": 2.862895965576172, + "step": 36960 + }, + { + "epoch": 0.18485, + "grad_norm": 0.3603318929672241, + "learning_rate": 0.00046249616279455747, + "loss": 2.811944770812988, + "step": 36970 + }, + { + "epoch": 0.1849, + "grad_norm": 0.3373560309410095, + "learning_rate": 0.00046247526347307176, + "loss": 2.827017402648926, + "step": 36980 + }, + { + "epoch": 0.18495, + "grad_norm": 0.33021917939186096, + "learning_rate": 0.00046245435880252263, + "loss": 2.8290748596191406, + "step": 36990 + }, + { + "epoch": 0.185, + "grad_norm": 0.32315197587013245, + "learning_rate": 0.0004624334487834365, + "loss": 2.8292842864990235, + "step": 37000 + }, + { + "epoch": 0.18505, + "grad_norm": 0.3364740014076233, + "learning_rate": 0.00046241253341633973, + "loss": 2.820978355407715, + "step": 37010 + }, + { + "epoch": 0.1851, + "grad_norm": 0.35852980613708496, + "learning_rate": 0.0004623916127017588, + "loss": 2.8845054626464846, + "step": 37020 + }, + { + "epoch": 0.18515, + "grad_norm": 0.36213722825050354, + "learning_rate": 0.0004623706866402204, + "loss": 2.859622764587402, + "step": 37030 + }, + { + "epoch": 0.1852, + "grad_norm": 0.33978790044784546, + "learning_rate": 0.00046234975523225144, + "loss": 2.8347291946411133, + "step": 37040 + }, + { + "epoch": 0.18525, + "grad_norm": 0.35536402463912964, + "learning_rate": 0.00046232881847837885, + "loss": 2.847237968444824, + "step": 37050 + }, + { + "epoch": 0.1853, + "grad_norm": 0.3265399932861328, + "learning_rate": 0.00046230787637912966, + "loss": 2.879253387451172, + "step": 37060 + }, + { + "epoch": 0.18535, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.000462286928935031, + "loss": 2.867985153198242, + "step": 37070 + }, + { + "epoch": 0.1854, + "grad_norm": 0.365257203578949, + "learning_rate": 0.00046226597614661044, + "loss": 2.8773429870605467, + "step": 37080 + }, + { + "epoch": 0.18545, + "grad_norm": 0.3661104142665863, + "learning_rate": 0.0004622450180143953, + "loss": 2.880403518676758, + "step": 37090 + }, + { + "epoch": 0.1855, + "grad_norm": 0.37116286158561707, + "learning_rate": 0.0004622240545389133, + "loss": 2.897838592529297, + "step": 37100 + }, + { + "epoch": 0.18555, + "grad_norm": 0.37751924991607666, + "learning_rate": 0.00046220308572069205, + "loss": 2.8450477600097654, + "step": 37110 + }, + { + "epoch": 0.1856, + "grad_norm": 0.3228621482849121, + "learning_rate": 0.00046218211156025957, + "loss": 2.877583885192871, + "step": 37120 + }, + { + "epoch": 0.18565, + "grad_norm": 0.35921239852905273, + "learning_rate": 0.0004621611320581439, + "loss": 2.7827264785766603, + "step": 37130 + }, + { + "epoch": 0.1857, + "grad_norm": 0.3424564301967621, + "learning_rate": 0.0004621401472148731, + "loss": 2.822139549255371, + "step": 37140 + }, + { + "epoch": 0.18575, + "grad_norm": 0.3812296688556671, + "learning_rate": 0.0004621191570309755, + "loss": 2.857160186767578, + "step": 37150 + }, + { + "epoch": 0.1858, + "grad_norm": 0.3319356143474579, + "learning_rate": 0.0004620981615069796, + "loss": 2.8672510147094727, + "step": 37160 + }, + { + "epoch": 0.18585, + "grad_norm": 0.3176444172859192, + "learning_rate": 0.0004620771606434138, + "loss": 2.856711578369141, + "step": 37170 + }, + { + "epoch": 0.1859, + "grad_norm": 0.374905526638031, + "learning_rate": 0.00046205615444080695, + "loss": 2.8480766296386717, + "step": 37180 + }, + { + "epoch": 0.18595, + "grad_norm": 0.3612973988056183, + "learning_rate": 0.00046203514289968787, + "loss": 2.846395492553711, + "step": 37190 + }, + { + "epoch": 0.186, + "grad_norm": 0.35341230034828186, + "learning_rate": 0.00046201412602058547, + "loss": 2.829328155517578, + "step": 37200 + }, + { + "epoch": 0.18605, + "grad_norm": 0.3618084192276001, + "learning_rate": 0.0004619931038040288, + "loss": 2.8628875732421877, + "step": 37210 + }, + { + "epoch": 0.1861, + "grad_norm": 0.4116677939891815, + "learning_rate": 0.00046197207625054727, + "loss": 2.87137508392334, + "step": 37220 + }, + { + "epoch": 0.18615, + "grad_norm": 0.3681493401527405, + "learning_rate": 0.0004619510433606701, + "loss": 2.8892406463623046, + "step": 37230 + }, + { + "epoch": 0.1862, + "grad_norm": 0.364080548286438, + "learning_rate": 0.00046193000513492676, + "loss": 2.8048984527587892, + "step": 37240 + }, + { + "epoch": 0.18625, + "grad_norm": 0.31775474548339844, + "learning_rate": 0.000461908961573847, + "loss": 2.8435457229614256, + "step": 37250 + }, + { + "epoch": 0.1863, + "grad_norm": 0.3324628472328186, + "learning_rate": 0.00046188791267796057, + "loss": 2.871057891845703, + "step": 37260 + }, + { + "epoch": 0.18635, + "grad_norm": 0.35119345784187317, + "learning_rate": 0.0004618668584477973, + "loss": 2.8053102493286133, + "step": 37270 + }, + { + "epoch": 0.1864, + "grad_norm": 0.3298032879829407, + "learning_rate": 0.00046184579888388733, + "loss": 2.832433319091797, + "step": 37280 + }, + { + "epoch": 0.18645, + "grad_norm": 0.34449660778045654, + "learning_rate": 0.00046182473398676075, + "loss": 2.8388233184814453, + "step": 37290 + }, + { + "epoch": 0.1865, + "grad_norm": 0.35273727774620056, + "learning_rate": 0.00046180366375694795, + "loss": 2.8273502349853517, + "step": 37300 + }, + { + "epoch": 0.18655, + "grad_norm": 0.3471861779689789, + "learning_rate": 0.0004617825881949793, + "loss": 2.880775260925293, + "step": 37310 + }, + { + "epoch": 0.1866, + "grad_norm": 0.3394084870815277, + "learning_rate": 0.00046176150730138546, + "loss": 2.815764236450195, + "step": 37320 + }, + { + "epoch": 0.18665, + "grad_norm": 0.3958697021007538, + "learning_rate": 0.000461740421076697, + "loss": 2.9039751052856446, + "step": 37330 + }, + { + "epoch": 0.1867, + "grad_norm": 0.36606529355049133, + "learning_rate": 0.0004617193295214449, + "loss": 2.8724010467529295, + "step": 37340 + }, + { + "epoch": 0.18675, + "grad_norm": 0.3431398868560791, + "learning_rate": 0.00046169823263616014, + "loss": 2.8742923736572266, + "step": 37350 + }, + { + "epoch": 0.1868, + "grad_norm": 0.3188226521015167, + "learning_rate": 0.0004616771304213737, + "loss": 2.871502685546875, + "step": 37360 + }, + { + "epoch": 0.18685, + "grad_norm": 0.39920973777770996, + "learning_rate": 0.00046165602287761697, + "loss": 2.8511171340942383, + "step": 37370 + }, + { + "epoch": 0.1869, + "grad_norm": 0.3966017961502075, + "learning_rate": 0.0004616349100054212, + "loss": 2.7867046356201173, + "step": 37380 + }, + { + "epoch": 0.18695, + "grad_norm": 0.3574140965938568, + "learning_rate": 0.00046161379180531813, + "loss": 2.846399116516113, + "step": 37390 + }, + { + "epoch": 0.187, + "grad_norm": 0.36331191658973694, + "learning_rate": 0.0004615926682778392, + "loss": 2.843675422668457, + "step": 37400 + }, + { + "epoch": 0.18705, + "grad_norm": 0.3437931537628174, + "learning_rate": 0.00046157153942351625, + "loss": 2.8671531677246094, + "step": 37410 + }, + { + "epoch": 0.1871, + "grad_norm": 0.3228244483470917, + "learning_rate": 0.00046155040524288116, + "loss": 2.8308238983154297, + "step": 37420 + }, + { + "epoch": 0.18715, + "grad_norm": 0.3543795049190521, + "learning_rate": 0.00046152926573646616, + "loss": 2.8404109954833983, + "step": 37430 + }, + { + "epoch": 0.1872, + "grad_norm": 0.3737463355064392, + "learning_rate": 0.0004615081209048032, + "loss": 2.8313404083251954, + "step": 37440 + }, + { + "epoch": 0.18725, + "grad_norm": 0.33747681975364685, + "learning_rate": 0.0004614869707484247, + "loss": 2.8697883605957033, + "step": 37450 + }, + { + "epoch": 0.1873, + "grad_norm": 0.3369773030281067, + "learning_rate": 0.0004614658152678632, + "loss": 2.8753593444824217, + "step": 37460 + }, + { + "epoch": 0.18735, + "grad_norm": 0.3504437208175659, + "learning_rate": 0.0004614446544636511, + "loss": 2.8544477462768554, + "step": 37470 + }, + { + "epoch": 0.1874, + "grad_norm": 0.37139081954956055, + "learning_rate": 0.00046142348833632144, + "loss": 2.8790138244628904, + "step": 37480 + }, + { + "epoch": 0.18745, + "grad_norm": 0.33805355429649353, + "learning_rate": 0.0004614023168864067, + "loss": 2.8991439819335936, + "step": 37490 + }, + { + "epoch": 0.1875, + "grad_norm": 0.3609998822212219, + "learning_rate": 0.0004613811401144401, + "loss": 2.83953800201416, + "step": 37500 + }, + { + "epoch": 0.18755, + "grad_norm": 0.3347666263580322, + "learning_rate": 0.00046135995802095467, + "loss": 2.8567861557006835, + "step": 37510 + }, + { + "epoch": 0.1876, + "grad_norm": 0.37966570258140564, + "learning_rate": 0.0004613387706064838, + "loss": 2.8499401092529295, + "step": 37520 + }, + { + "epoch": 0.18765, + "grad_norm": 0.365071564912796, + "learning_rate": 0.0004613175778715607, + "loss": 2.818370819091797, + "step": 37530 + }, + { + "epoch": 0.1877, + "grad_norm": 0.35801324248313904, + "learning_rate": 0.000461296379816719, + "loss": 2.8867420196533202, + "step": 37540 + }, + { + "epoch": 0.18775, + "grad_norm": 0.36637330055236816, + "learning_rate": 0.0004612751764424924, + "loss": 2.840646171569824, + "step": 37550 + }, + { + "epoch": 0.1878, + "grad_norm": 0.32699498534202576, + "learning_rate": 0.00046125396774941463, + "loss": 2.827714538574219, + "step": 37560 + }, + { + "epoch": 0.18785, + "grad_norm": 0.3425405025482178, + "learning_rate": 0.0004612327537380196, + "loss": 2.8778671264648437, + "step": 37570 + }, + { + "epoch": 0.1879, + "grad_norm": 0.35749581456184387, + "learning_rate": 0.00046121153440884145, + "loss": 2.862747573852539, + "step": 37580 + }, + { + "epoch": 0.18795, + "grad_norm": 0.35886573791503906, + "learning_rate": 0.0004611903097624143, + "loss": 2.8013584136962892, + "step": 37590 + }, + { + "epoch": 0.188, + "grad_norm": 0.37807944416999817, + "learning_rate": 0.0004611690797992725, + "loss": 2.8542142868041993, + "step": 37600 + }, + { + "epoch": 0.18805, + "grad_norm": 0.36229053139686584, + "learning_rate": 0.0004611478445199505, + "loss": 2.832365608215332, + "step": 37610 + }, + { + "epoch": 0.1881, + "grad_norm": 0.3454887866973877, + "learning_rate": 0.000461126603924983, + "loss": 2.8769569396972656, + "step": 37620 + }, + { + "epoch": 0.18815, + "grad_norm": 0.35969334840774536, + "learning_rate": 0.0004611053580149047, + "loss": 2.816790962219238, + "step": 37630 + }, + { + "epoch": 0.1882, + "grad_norm": 0.3576090931892395, + "learning_rate": 0.0004610841067902503, + "loss": 2.858123207092285, + "step": 37640 + }, + { + "epoch": 0.18825, + "grad_norm": 0.336041122674942, + "learning_rate": 0.00046106285025155496, + "loss": 2.808339309692383, + "step": 37650 + }, + { + "epoch": 0.1883, + "grad_norm": 0.3611387014389038, + "learning_rate": 0.00046104158839935385, + "loss": 2.779623794555664, + "step": 37660 + }, + { + "epoch": 0.18835, + "grad_norm": 0.39656999707221985, + "learning_rate": 0.000461020321234182, + "loss": 2.8161678314208984, + "step": 37670 + }, + { + "epoch": 0.1884, + "grad_norm": 0.43750014901161194, + "learning_rate": 0.0004609990487565752, + "loss": 2.8527753829956053, + "step": 37680 + }, + { + "epoch": 0.18845, + "grad_norm": 0.32209277153015137, + "learning_rate": 0.0004609777709670686, + "loss": 2.8655832290649412, + "step": 37690 + }, + { + "epoch": 0.1885, + "grad_norm": 0.3637388050556183, + "learning_rate": 0.000460956487866198, + "loss": 2.848017120361328, + "step": 37700 + }, + { + "epoch": 0.18855, + "grad_norm": 0.319944828748703, + "learning_rate": 0.00046093519945449935, + "loss": 2.8560646057128904, + "step": 37710 + }, + { + "epoch": 0.1886, + "grad_norm": 0.32945242524147034, + "learning_rate": 0.00046091390573250833, + "loss": 2.8623287200927736, + "step": 37720 + }, + { + "epoch": 0.18865, + "grad_norm": 0.3265266716480255, + "learning_rate": 0.00046089260670076126, + "loss": 2.8875860214233398, + "step": 37730 + }, + { + "epoch": 0.1887, + "grad_norm": 0.3351488411426544, + "learning_rate": 0.0004608713023597941, + "loss": 2.8556421279907225, + "step": 37740 + }, + { + "epoch": 0.18875, + "grad_norm": 0.38284286856651306, + "learning_rate": 0.0004608499927101435, + "loss": 2.7894636154174806, + "step": 37750 + }, + { + "epoch": 0.1888, + "grad_norm": 0.37427741289138794, + "learning_rate": 0.0004608286777523456, + "loss": 2.8718639373779298, + "step": 37760 + }, + { + "epoch": 0.18885, + "grad_norm": 0.3387341797351837, + "learning_rate": 0.0004608073574869371, + "loss": 2.8340036392211916, + "step": 37770 + }, + { + "epoch": 0.1889, + "grad_norm": 0.32482555508613586, + "learning_rate": 0.00046078603191445494, + "loss": 2.7812625885009767, + "step": 37780 + }, + { + "epoch": 0.18895, + "grad_norm": 0.3704550564289093, + "learning_rate": 0.00046076470103543577, + "loss": 2.8390607833862305, + "step": 37790 + }, + { + "epoch": 0.189, + "grad_norm": 0.36005789041519165, + "learning_rate": 0.00046074336485041665, + "loss": 2.853083038330078, + "step": 37800 + }, + { + "epoch": 0.18905, + "grad_norm": 0.37339505553245544, + "learning_rate": 0.00046072202335993475, + "loss": 2.8767139434814455, + "step": 37810 + }, + { + "epoch": 0.1891, + "grad_norm": 0.3452443480491638, + "learning_rate": 0.00046070067656452725, + "loss": 2.8884077072143555, + "step": 37820 + }, + { + "epoch": 0.18915, + "grad_norm": 0.3598582148551941, + "learning_rate": 0.0004606793244647317, + "loss": 2.8493379592895507, + "step": 37830 + }, + { + "epoch": 0.1892, + "grad_norm": 0.3690253496170044, + "learning_rate": 0.0004606579670610855, + "loss": 2.8779928207397463, + "step": 37840 + }, + { + "epoch": 0.18925, + "grad_norm": 0.3595035970211029, + "learning_rate": 0.0004606366043541264, + "loss": 2.831578826904297, + "step": 37850 + }, + { + "epoch": 0.1893, + "grad_norm": 0.36415407061576843, + "learning_rate": 0.0004606152363443923, + "loss": 2.809413528442383, + "step": 37860 + }, + { + "epoch": 0.18935, + "grad_norm": 0.36779236793518066, + "learning_rate": 0.0004605938630324209, + "loss": 2.833542823791504, + "step": 37870 + }, + { + "epoch": 0.1894, + "grad_norm": 0.3312298059463501, + "learning_rate": 0.0004605724844187504, + "loss": 2.836895751953125, + "step": 37880 + }, + { + "epoch": 0.18945, + "grad_norm": 0.33025795221328735, + "learning_rate": 0.0004605511005039191, + "loss": 2.8958616256713867, + "step": 37890 + }, + { + "epoch": 0.1895, + "grad_norm": 0.37289512157440186, + "learning_rate": 0.0004605297112884652, + "loss": 2.8610076904296875, + "step": 37900 + }, + { + "epoch": 0.18955, + "grad_norm": 0.3587748110294342, + "learning_rate": 0.00046050831677292727, + "loss": 2.794106674194336, + "step": 37910 + }, + { + "epoch": 0.1896, + "grad_norm": 0.4209056496620178, + "learning_rate": 0.0004604869169578438, + "loss": 2.8488037109375, + "step": 37920 + }, + { + "epoch": 0.18965, + "grad_norm": 0.3558873236179352, + "learning_rate": 0.0004604655118437537, + "loss": 2.8595279693603515, + "step": 37930 + }, + { + "epoch": 0.1897, + "grad_norm": 0.3468804955482483, + "learning_rate": 0.00046044410143119565, + "loss": 2.8646345138549805, + "step": 37940 + }, + { + "epoch": 0.18975, + "grad_norm": 0.35978490114212036, + "learning_rate": 0.0004604226857207088, + "loss": 2.888427734375, + "step": 37950 + }, + { + "epoch": 0.1898, + "grad_norm": 0.3788745701313019, + "learning_rate": 0.00046040126471283226, + "loss": 2.891963577270508, + "step": 37960 + }, + { + "epoch": 0.18985, + "grad_norm": 0.4267536997795105, + "learning_rate": 0.0004603798384081053, + "loss": 2.8618886947631834, + "step": 37970 + }, + { + "epoch": 0.1899, + "grad_norm": 0.3454354703426361, + "learning_rate": 0.0004603584068070672, + "loss": 2.837969207763672, + "step": 37980 + }, + { + "epoch": 0.18995, + "grad_norm": 0.3585532307624817, + "learning_rate": 0.00046033696991025773, + "loss": 2.825738525390625, + "step": 37990 + }, + { + "epoch": 0.19, + "grad_norm": 0.40294477343559265, + "learning_rate": 0.0004603155277182164, + "loss": 2.8570476531982423, + "step": 38000 + }, + { + "epoch": 0.19005, + "grad_norm": 0.3341706097126007, + "learning_rate": 0.0004602940802314831, + "loss": 2.7951019287109373, + "step": 38010 + }, + { + "epoch": 0.1901, + "grad_norm": 0.3412943184375763, + "learning_rate": 0.00046027262745059773, + "loss": 2.834110641479492, + "step": 38020 + }, + { + "epoch": 0.19015, + "grad_norm": 0.3628341853618622, + "learning_rate": 0.00046025116937610034, + "loss": 2.804364013671875, + "step": 38030 + }, + { + "epoch": 0.1902, + "grad_norm": 0.35416552424430847, + "learning_rate": 0.00046022970600853114, + "loss": 2.787668991088867, + "step": 38040 + }, + { + "epoch": 0.19025, + "grad_norm": 0.3174043595790863, + "learning_rate": 0.00046020823734843055, + "loss": 2.863397216796875, + "step": 38050 + }, + { + "epoch": 0.1903, + "grad_norm": 0.37627291679382324, + "learning_rate": 0.0004601867633963389, + "loss": 2.8774139404296877, + "step": 38060 + }, + { + "epoch": 0.19035, + "grad_norm": 0.367354154586792, + "learning_rate": 0.000460165284152797, + "loss": 2.8596588134765626, + "step": 38070 + }, + { + "epoch": 0.1904, + "grad_norm": 0.3626890480518341, + "learning_rate": 0.0004601437996183455, + "loss": 2.832733726501465, + "step": 38080 + }, + { + "epoch": 0.19045, + "grad_norm": 0.37853801250457764, + "learning_rate": 0.00046012230979352507, + "loss": 2.912740707397461, + "step": 38090 + }, + { + "epoch": 0.1905, + "grad_norm": 0.35075700283050537, + "learning_rate": 0.00046010081467887706, + "loss": 2.8478103637695313, + "step": 38100 + }, + { + "epoch": 0.19055, + "grad_norm": 0.3573131263256073, + "learning_rate": 0.0004600793142749424, + "loss": 2.8458835601806642, + "step": 38110 + }, + { + "epoch": 0.1906, + "grad_norm": 0.3745926022529602, + "learning_rate": 0.00046005780858226244, + "loss": 2.8453460693359376, + "step": 38120 + }, + { + "epoch": 0.19065, + "grad_norm": 0.32159051299095154, + "learning_rate": 0.0004600362976013784, + "loss": 2.855428123474121, + "step": 38130 + }, + { + "epoch": 0.1907, + "grad_norm": 0.34157443046569824, + "learning_rate": 0.00046001478133283216, + "loss": 2.834274101257324, + "step": 38140 + }, + { + "epoch": 0.19075, + "grad_norm": 0.33654728531837463, + "learning_rate": 0.0004599932597771651, + "loss": 2.8451305389404298, + "step": 38150 + }, + { + "epoch": 0.1908, + "grad_norm": 0.3536951541900635, + "learning_rate": 0.00045997173293491913, + "loss": 2.8279273986816404, + "step": 38160 + }, + { + "epoch": 0.19085, + "grad_norm": 0.3400050699710846, + "learning_rate": 0.00045995020080663617, + "loss": 2.827983856201172, + "step": 38170 + }, + { + "epoch": 0.1909, + "grad_norm": 0.333223432302475, + "learning_rate": 0.00045992866339285846, + "loss": 2.8651168823242186, + "step": 38180 + }, + { + "epoch": 0.19095, + "grad_norm": 0.33160164952278137, + "learning_rate": 0.0004599071206941279, + "loss": 2.8532333374023438, + "step": 38190 + }, + { + "epoch": 0.191, + "grad_norm": 0.33447787165641785, + "learning_rate": 0.00045988557271098697, + "loss": 2.882658004760742, + "step": 38200 + }, + { + "epoch": 0.19105, + "grad_norm": 0.36412709951400757, + "learning_rate": 0.00045986401944397826, + "loss": 2.9287797927856447, + "step": 38210 + }, + { + "epoch": 0.1911, + "grad_norm": 0.34448087215423584, + "learning_rate": 0.0004598424608936441, + "loss": 2.8893985748291016, + "step": 38220 + }, + { + "epoch": 0.19115, + "grad_norm": 0.3369698226451874, + "learning_rate": 0.00045982089706052753, + "loss": 2.815079689025879, + "step": 38230 + }, + { + "epoch": 0.1912, + "grad_norm": 0.32998716831207275, + "learning_rate": 0.0004597993279451713, + "loss": 2.8869733810424805, + "step": 38240 + }, + { + "epoch": 0.19125, + "grad_norm": 0.329700767993927, + "learning_rate": 0.0004597777535481183, + "loss": 2.8805767059326173, + "step": 38250 + }, + { + "epoch": 0.1913, + "grad_norm": 0.36369627714157104, + "learning_rate": 0.00045975617386991176, + "loss": 2.883289909362793, + "step": 38260 + }, + { + "epoch": 0.19135, + "grad_norm": 0.34488481283187866, + "learning_rate": 0.00045973458891109497, + "loss": 2.7970703125, + "step": 38270 + }, + { + "epoch": 0.1914, + "grad_norm": 0.4246232509613037, + "learning_rate": 0.00045971299867221126, + "loss": 2.8030527114868162, + "step": 38280 + }, + { + "epoch": 0.19145, + "grad_norm": 0.3563256561756134, + "learning_rate": 0.0004596914031538043, + "loss": 2.8267162322998045, + "step": 38290 + }, + { + "epoch": 0.1915, + "grad_norm": 0.3530254662036896, + "learning_rate": 0.00045966980235641764, + "loss": 2.840110969543457, + "step": 38300 + }, + { + "epoch": 0.19155, + "grad_norm": 0.33481553196907043, + "learning_rate": 0.0004596481962805951, + "loss": 2.882917022705078, + "step": 38310 + }, + { + "epoch": 0.1916, + "grad_norm": 0.34993091225624084, + "learning_rate": 0.00045962658492688057, + "loss": 2.8520612716674805, + "step": 38320 + }, + { + "epoch": 0.19165, + "grad_norm": 0.3357253074645996, + "learning_rate": 0.0004596049682958182, + "loss": 2.8425624847412108, + "step": 38330 + }, + { + "epoch": 0.1917, + "grad_norm": 0.3610876500606537, + "learning_rate": 0.0004595833463879522, + "loss": 2.8690690994262695, + "step": 38340 + }, + { + "epoch": 0.19175, + "grad_norm": 0.3694901168346405, + "learning_rate": 0.0004595617192038268, + "loss": 2.8441017150878904, + "step": 38350 + }, + { + "epoch": 0.1918, + "grad_norm": 0.34114137291908264, + "learning_rate": 0.0004595400867439865, + "loss": 2.8442115783691406, + "step": 38360 + }, + { + "epoch": 0.19185, + "grad_norm": 0.3514702618122101, + "learning_rate": 0.000459518449008976, + "loss": 2.850543212890625, + "step": 38370 + }, + { + "epoch": 0.1919, + "grad_norm": 0.3464371860027313, + "learning_rate": 0.00045949680599933985, + "loss": 2.8519941329956056, + "step": 38380 + }, + { + "epoch": 0.19195, + "grad_norm": 0.3701815605163574, + "learning_rate": 0.0004594751577156231, + "loss": 2.8449775695800783, + "step": 38390 + }, + { + "epoch": 0.192, + "grad_norm": 0.34201398491859436, + "learning_rate": 0.00045945350415837056, + "loss": 2.878372001647949, + "step": 38400 + }, + { + "epoch": 0.19205, + "grad_norm": 0.3409600555896759, + "learning_rate": 0.00045943184532812754, + "loss": 2.85513916015625, + "step": 38410 + }, + { + "epoch": 0.1921, + "grad_norm": 0.3369607925415039, + "learning_rate": 0.00045941018122543916, + "loss": 2.821485710144043, + "step": 38420 + }, + { + "epoch": 0.19215, + "grad_norm": 0.3481343388557434, + "learning_rate": 0.00045938851185085085, + "loss": 2.918476104736328, + "step": 38430 + }, + { + "epoch": 0.1922, + "grad_norm": 0.3536899983882904, + "learning_rate": 0.00045936683720490816, + "loss": 2.8211542129516602, + "step": 38440 + }, + { + "epoch": 0.19225, + "grad_norm": 0.34131187200546265, + "learning_rate": 0.0004593451572881568, + "loss": 2.8398202896118163, + "step": 38450 + }, + { + "epoch": 0.1923, + "grad_norm": 0.3676207363605499, + "learning_rate": 0.00045932347210114245, + "loss": 2.741354751586914, + "step": 38460 + }, + { + "epoch": 0.19235, + "grad_norm": 0.34500712156295776, + "learning_rate": 0.00045930178164441116, + "loss": 2.866720771789551, + "step": 38470 + }, + { + "epoch": 0.1924, + "grad_norm": 0.3774857521057129, + "learning_rate": 0.0004592800859185088, + "loss": 2.8952407836914062, + "step": 38480 + }, + { + "epoch": 0.19245, + "grad_norm": 0.37018299102783203, + "learning_rate": 0.0004592583849239817, + "loss": 2.883538246154785, + "step": 38490 + }, + { + "epoch": 0.1925, + "grad_norm": 0.3657027781009674, + "learning_rate": 0.00045923667866137625, + "loss": 2.8643268585205077, + "step": 38500 + }, + { + "epoch": 0.19255, + "grad_norm": 0.33769533038139343, + "learning_rate": 0.0004592149671312388, + "loss": 2.838809776306152, + "step": 38510 + }, + { + "epoch": 0.1926, + "grad_norm": 0.33940163254737854, + "learning_rate": 0.00045919325033411584, + "loss": 2.952041816711426, + "step": 38520 + }, + { + "epoch": 0.19265, + "grad_norm": 0.3420262634754181, + "learning_rate": 0.0004591715282705543, + "loss": 2.850127601623535, + "step": 38530 + }, + { + "epoch": 0.1927, + "grad_norm": 0.3636834919452667, + "learning_rate": 0.00045914980094110094, + "loss": 2.854897308349609, + "step": 38540 + }, + { + "epoch": 0.19275, + "grad_norm": 0.33680376410484314, + "learning_rate": 0.0004591280683463027, + "loss": 2.8316320419311523, + "step": 38550 + }, + { + "epoch": 0.1928, + "grad_norm": 0.3783840239048004, + "learning_rate": 0.0004591063304867067, + "loss": 2.830095100402832, + "step": 38560 + }, + { + "epoch": 0.19285, + "grad_norm": 0.3357202708721161, + "learning_rate": 0.0004590845873628603, + "loss": 2.812094879150391, + "step": 38570 + }, + { + "epoch": 0.1929, + "grad_norm": 0.350166380405426, + "learning_rate": 0.00045906283897531076, + "loss": 2.781584548950195, + "step": 38580 + }, + { + "epoch": 0.19295, + "grad_norm": 0.3456714451313019, + "learning_rate": 0.00045904108532460566, + "loss": 2.8181161880493164, + "step": 38590 + }, + { + "epoch": 0.193, + "grad_norm": 0.3291044533252716, + "learning_rate": 0.00045901932641129265, + "loss": 2.83661003112793, + "step": 38600 + }, + { + "epoch": 0.19305, + "grad_norm": 0.34832826256752014, + "learning_rate": 0.0004589975622359195, + "loss": 2.8612260818481445, + "step": 38610 + }, + { + "epoch": 0.1931, + "grad_norm": 0.34657683968544006, + "learning_rate": 0.0004589757927990341, + "loss": 2.891806793212891, + "step": 38620 + }, + { + "epoch": 0.19315, + "grad_norm": 0.3629623055458069, + "learning_rate": 0.0004589540181011845, + "loss": 2.8784557342529298, + "step": 38630 + }, + { + "epoch": 0.1932, + "grad_norm": 0.3503754138946533, + "learning_rate": 0.00045893223814291885, + "loss": 2.862186622619629, + "step": 38640 + }, + { + "epoch": 0.19325, + "grad_norm": 0.37800246477127075, + "learning_rate": 0.00045891045292478555, + "loss": 2.846821975708008, + "step": 38650 + }, + { + "epoch": 0.1933, + "grad_norm": 0.3522031605243683, + "learning_rate": 0.000458888662447333, + "loss": 2.8231008529663084, + "step": 38660 + }, + { + "epoch": 0.19335, + "grad_norm": 0.3686130940914154, + "learning_rate": 0.0004588668667111098, + "loss": 2.8451869964599608, + "step": 38670 + }, + { + "epoch": 0.1934, + "grad_norm": 0.34186285734176636, + "learning_rate": 0.0004588450657166646, + "loss": 2.858975410461426, + "step": 38680 + }, + { + "epoch": 0.19345, + "grad_norm": 0.4232251048088074, + "learning_rate": 0.0004588232594645462, + "loss": 2.808083343505859, + "step": 38690 + }, + { + "epoch": 0.1935, + "grad_norm": 0.37703996896743774, + "learning_rate": 0.0004588014479553037, + "loss": 2.854382133483887, + "step": 38700 + }, + { + "epoch": 0.19355, + "grad_norm": 0.3553779125213623, + "learning_rate": 0.0004587796311894862, + "loss": 2.8382102966308596, + "step": 38710 + }, + { + "epoch": 0.1936, + "grad_norm": 0.35454949736595154, + "learning_rate": 0.00045875780916764273, + "loss": 2.8588531494140623, + "step": 38720 + }, + { + "epoch": 0.19365, + "grad_norm": 0.3525417745113373, + "learning_rate": 0.00045873598189032295, + "loss": 2.838178825378418, + "step": 38730 + }, + { + "epoch": 0.1937, + "grad_norm": 0.35396963357925415, + "learning_rate": 0.00045871414935807613, + "loss": 2.853485870361328, + "step": 38740 + }, + { + "epoch": 0.19375, + "grad_norm": 0.35924020409584045, + "learning_rate": 0.0004586923115714521, + "loss": 2.8357402801513674, + "step": 38750 + }, + { + "epoch": 0.1938, + "grad_norm": 0.3696933388710022, + "learning_rate": 0.00045867046853100034, + "loss": 2.858484649658203, + "step": 38760 + }, + { + "epoch": 0.19385, + "grad_norm": 0.34034740924835205, + "learning_rate": 0.00045864862023727113, + "loss": 2.804395294189453, + "step": 38770 + }, + { + "epoch": 0.1939, + "grad_norm": 0.34251317381858826, + "learning_rate": 0.00045862676669081414, + "loss": 2.851711654663086, + "step": 38780 + }, + { + "epoch": 0.19395, + "grad_norm": 0.40035560727119446, + "learning_rate": 0.00045860490789217977, + "loss": 2.8730403900146486, + "step": 38790 + }, + { + "epoch": 0.194, + "grad_norm": 0.3566685914993286, + "learning_rate": 0.0004585830438419183, + "loss": 2.831101417541504, + "step": 38800 + }, + { + "epoch": 0.19405, + "grad_norm": 0.3483644127845764, + "learning_rate": 0.00045856117454058, + "loss": 2.790033721923828, + "step": 38810 + }, + { + "epoch": 0.1941, + "grad_norm": 0.3304741084575653, + "learning_rate": 0.00045853929998871556, + "loss": 2.8500410079956056, + "step": 38820 + }, + { + "epoch": 0.19415, + "grad_norm": 0.3935031592845917, + "learning_rate": 0.0004585174201868756, + "loss": 2.8249490737915037, + "step": 38830 + }, + { + "epoch": 0.1942, + "grad_norm": 0.3258267939090729, + "learning_rate": 0.000458495535135611, + "loss": 2.854292869567871, + "step": 38840 + }, + { + "epoch": 0.19425, + "grad_norm": 0.33895567059516907, + "learning_rate": 0.0004584736448354727, + "loss": 2.845300483703613, + "step": 38850 + }, + { + "epoch": 0.1943, + "grad_norm": 0.3436082601547241, + "learning_rate": 0.00045845174928701185, + "loss": 2.889860916137695, + "step": 38860 + }, + { + "epoch": 0.19435, + "grad_norm": 0.3256368339061737, + "learning_rate": 0.00045842984849077953, + "loss": 2.844677734375, + "step": 38870 + }, + { + "epoch": 0.1944, + "grad_norm": 0.3599049150943756, + "learning_rate": 0.0004584079424473272, + "loss": 2.880068778991699, + "step": 38880 + }, + { + "epoch": 0.19445, + "grad_norm": 0.34493786096572876, + "learning_rate": 0.00045838603115720624, + "loss": 2.8522096633911134, + "step": 38890 + }, + { + "epoch": 0.1945, + "grad_norm": 0.3316764533519745, + "learning_rate": 0.0004583641146209684, + "loss": 2.8560630798339846, + "step": 38900 + }, + { + "epoch": 0.19455, + "grad_norm": 0.3852003216743469, + "learning_rate": 0.00045834219283916536, + "loss": 2.8893171310424806, + "step": 38910 + }, + { + "epoch": 0.1946, + "grad_norm": 0.36831212043762207, + "learning_rate": 0.000458320265812349, + "loss": 2.8502267837524413, + "step": 38920 + }, + { + "epoch": 0.19465, + "grad_norm": 0.3324825167655945, + "learning_rate": 0.00045829833354107133, + "loss": 2.863662910461426, + "step": 38930 + }, + { + "epoch": 0.1947, + "grad_norm": 0.39460262656211853, + "learning_rate": 0.00045827639602588453, + "loss": 2.8603933334350584, + "step": 38940 + }, + { + "epoch": 0.19475, + "grad_norm": 0.3666238486766815, + "learning_rate": 0.0004582544532673408, + "loss": 2.8947994232177736, + "step": 38950 + }, + { + "epoch": 0.1948, + "grad_norm": 0.33583125472068787, + "learning_rate": 0.0004582325052659927, + "loss": 2.8384014129638673, + "step": 38960 + }, + { + "epoch": 0.19485, + "grad_norm": 0.3105217516422272, + "learning_rate": 0.0004582105520223926, + "loss": 2.872724723815918, + "step": 38970 + }, + { + "epoch": 0.1949, + "grad_norm": 0.3729989528656006, + "learning_rate": 0.0004581885935370932, + "loss": 2.9107334136962892, + "step": 38980 + }, + { + "epoch": 0.19495, + "grad_norm": 0.3678586483001709, + "learning_rate": 0.00045816662981064746, + "loss": 2.8424680709838865, + "step": 38990 + }, + { + "epoch": 0.195, + "grad_norm": 0.3351372480392456, + "learning_rate": 0.0004581446608436082, + "loss": 2.849211311340332, + "step": 39000 + }, + { + "epoch": 0.19505, + "grad_norm": 0.42423710227012634, + "learning_rate": 0.00045812268663652833, + "loss": 2.852702331542969, + "step": 39010 + }, + { + "epoch": 0.1951, + "grad_norm": 0.3574000895023346, + "learning_rate": 0.00045810070718996134, + "loss": 2.882835578918457, + "step": 39020 + }, + { + "epoch": 0.19515, + "grad_norm": 0.35171669721603394, + "learning_rate": 0.0004580787225044605, + "loss": 2.8438789367675783, + "step": 39030 + }, + { + "epoch": 0.1952, + "grad_norm": 0.3500445485115051, + "learning_rate": 0.00045805673258057913, + "loss": 2.9095438003540037, + "step": 39040 + }, + { + "epoch": 0.19525, + "grad_norm": 0.3552171289920807, + "learning_rate": 0.00045803473741887093, + "loss": 2.858315277099609, + "step": 39050 + }, + { + "epoch": 0.1953, + "grad_norm": 0.3683996796607971, + "learning_rate": 0.00045801273701988955, + "loss": 2.8973480224609376, + "step": 39060 + }, + { + "epoch": 0.19535, + "grad_norm": 0.3396645784378052, + "learning_rate": 0.00045799073138418897, + "loss": 2.84689998626709, + "step": 39070 + }, + { + "epoch": 0.1954, + "grad_norm": 0.3647054135799408, + "learning_rate": 0.0004579687205123231, + "loss": 2.857808494567871, + "step": 39080 + }, + { + "epoch": 0.19545, + "grad_norm": 0.3315117061138153, + "learning_rate": 0.000457946704404846, + "loss": 2.9028148651123047, + "step": 39090 + }, + { + "epoch": 0.1955, + "grad_norm": 0.3701636493206024, + "learning_rate": 0.00045792468306231213, + "loss": 2.8366247177124024, + "step": 39100 + }, + { + "epoch": 0.19555, + "grad_norm": 0.34960880875587463, + "learning_rate": 0.0004579026564852757, + "loss": 2.8609455108642576, + "step": 39110 + }, + { + "epoch": 0.1956, + "grad_norm": 0.3487541675567627, + "learning_rate": 0.00045788062467429127, + "loss": 2.8663658142089843, + "step": 39120 + }, + { + "epoch": 0.19565, + "grad_norm": 0.33619365096092224, + "learning_rate": 0.0004578585876299135, + "loss": 2.8857000350952147, + "step": 39130 + }, + { + "epoch": 0.1957, + "grad_norm": 0.3431452512741089, + "learning_rate": 0.0004578365453526972, + "loss": 2.8381542205810546, + "step": 39140 + }, + { + "epoch": 0.19575, + "grad_norm": 0.33372044563293457, + "learning_rate": 0.0004578144978431973, + "loss": 2.8262157440185547, + "step": 39150 + }, + { + "epoch": 0.1958, + "grad_norm": 0.3988279104232788, + "learning_rate": 0.00045779244510196873, + "loss": 2.8723276138305662, + "step": 39160 + }, + { + "epoch": 0.19585, + "grad_norm": 0.38754984736442566, + "learning_rate": 0.00045777038712956677, + "loss": 2.7890108108520506, + "step": 39170 + }, + { + "epoch": 0.1959, + "grad_norm": 0.3226994276046753, + "learning_rate": 0.0004577483239265467, + "loss": 2.818580436706543, + "step": 39180 + }, + { + "epoch": 0.19595, + "grad_norm": 0.3296877145767212, + "learning_rate": 0.000457726255493464, + "loss": 2.8629695892333986, + "step": 39190 + }, + { + "epoch": 0.196, + "grad_norm": 0.33507195115089417, + "learning_rate": 0.00045770418183087416, + "loss": 2.82034912109375, + "step": 39200 + }, + { + "epoch": 0.19605, + "grad_norm": 0.36471185088157654, + "learning_rate": 0.0004576821029393329, + "loss": 2.9188634872436525, + "step": 39210 + }, + { + "epoch": 0.1961, + "grad_norm": 0.349333792924881, + "learning_rate": 0.0004576600188193961, + "loss": 2.876717185974121, + "step": 39220 + }, + { + "epoch": 0.19615, + "grad_norm": 0.3808686435222626, + "learning_rate": 0.0004576379294716198, + "loss": 2.836391258239746, + "step": 39230 + }, + { + "epoch": 0.1962, + "grad_norm": 0.323930948972702, + "learning_rate": 0.00045761583489656, + "loss": 2.8641019821166993, + "step": 39240 + }, + { + "epoch": 0.19625, + "grad_norm": 0.3555019199848175, + "learning_rate": 0.0004575937350947729, + "loss": 2.819235420227051, + "step": 39250 + }, + { + "epoch": 0.1963, + "grad_norm": 0.3720868229866028, + "learning_rate": 0.00045757163006681484, + "loss": 2.8383529663085936, + "step": 39260 + }, + { + "epoch": 0.19635, + "grad_norm": 0.35251063108444214, + "learning_rate": 0.00045754951981324244, + "loss": 2.8512781143188475, + "step": 39270 + }, + { + "epoch": 0.1964, + "grad_norm": 0.36221328377723694, + "learning_rate": 0.00045752740433461234, + "loss": 2.874264717102051, + "step": 39280 + }, + { + "epoch": 0.19645, + "grad_norm": 0.3605222702026367, + "learning_rate": 0.0004575052836314811, + "loss": 2.875098991394043, + "step": 39290 + }, + { + "epoch": 0.1965, + "grad_norm": 0.35363277792930603, + "learning_rate": 0.0004574831577044058, + "loss": 2.867971420288086, + "step": 39300 + }, + { + "epoch": 0.19655, + "grad_norm": 0.41844215989112854, + "learning_rate": 0.0004574610265539434, + "loss": 2.90916748046875, + "step": 39310 + }, + { + "epoch": 0.1966, + "grad_norm": 0.3215714693069458, + "learning_rate": 0.000457438890180651, + "loss": 2.846781921386719, + "step": 39320 + }, + { + "epoch": 0.19665, + "grad_norm": 0.35437190532684326, + "learning_rate": 0.000457416748585086, + "loss": 2.9383663177490233, + "step": 39330 + }, + { + "epoch": 0.1967, + "grad_norm": 0.3425946831703186, + "learning_rate": 0.0004573946017678057, + "loss": 2.801084518432617, + "step": 39340 + }, + { + "epoch": 0.19675, + "grad_norm": 0.3857385218143463, + "learning_rate": 0.0004573724497293676, + "loss": 2.8429969787597655, + "step": 39350 + }, + { + "epoch": 0.1968, + "grad_norm": 0.35941895842552185, + "learning_rate": 0.00045735029247032956, + "loss": 2.874639701843262, + "step": 39360 + }, + { + "epoch": 0.19685, + "grad_norm": 0.3792363107204437, + "learning_rate": 0.0004573281299912492, + "loss": 2.841995048522949, + "step": 39370 + }, + { + "epoch": 0.1969, + "grad_norm": 0.32867005467414856, + "learning_rate": 0.0004573059622926847, + "loss": 2.863136863708496, + "step": 39380 + }, + { + "epoch": 0.19695, + "grad_norm": 0.3962661325931549, + "learning_rate": 0.0004572837893751939, + "loss": 2.8419471740722657, + "step": 39390 + }, + { + "epoch": 0.197, + "grad_norm": 0.36858272552490234, + "learning_rate": 0.00045726161123933506, + "loss": 2.8578752517700194, + "step": 39400 + }, + { + "epoch": 0.19705, + "grad_norm": 0.33612653613090515, + "learning_rate": 0.0004572394278856666, + "loss": 2.8837751388549804, + "step": 39410 + }, + { + "epoch": 0.1971, + "grad_norm": 0.36950773000717163, + "learning_rate": 0.0004572172393147469, + "loss": 2.8237941741943358, + "step": 39420 + }, + { + "epoch": 0.19715, + "grad_norm": 0.3367069661617279, + "learning_rate": 0.00045719504552713456, + "loss": 2.862497901916504, + "step": 39430 + }, + { + "epoch": 0.1972, + "grad_norm": 0.3536355197429657, + "learning_rate": 0.0004571728465233884, + "loss": 2.8582345962524416, + "step": 39440 + }, + { + "epoch": 0.19725, + "grad_norm": 0.3510274291038513, + "learning_rate": 0.0004571506423040671, + "loss": 2.7846349716186523, + "step": 39450 + }, + { + "epoch": 0.1973, + "grad_norm": 0.3563089966773987, + "learning_rate": 0.00045712843286972987, + "loss": 2.8359657287597657, + "step": 39460 + }, + { + "epoch": 0.19735, + "grad_norm": 0.3584759831428528, + "learning_rate": 0.00045710621822093566, + "loss": 2.8279129028320313, + "step": 39470 + }, + { + "epoch": 0.1974, + "grad_norm": 0.44362369179725647, + "learning_rate": 0.0004570839983582439, + "loss": 2.862477111816406, + "step": 39480 + }, + { + "epoch": 0.19745, + "grad_norm": 0.32596081495285034, + "learning_rate": 0.00045706177328221374, + "loss": 2.835947799682617, + "step": 39490 + }, + { + "epoch": 0.1975, + "grad_norm": 0.35267937183380127, + "learning_rate": 0.00045703954299340486, + "loss": 2.8312398910522463, + "step": 39500 + }, + { + "epoch": 0.19755, + "grad_norm": 0.32649290561676025, + "learning_rate": 0.00045701730749237693, + "loss": 2.848710632324219, + "step": 39510 + }, + { + "epoch": 0.1976, + "grad_norm": 0.3374091684818268, + "learning_rate": 0.0004569950667796896, + "loss": 2.8475570678710938, + "step": 39520 + }, + { + "epoch": 0.19765, + "grad_norm": 0.33321937918663025, + "learning_rate": 0.0004569728208559029, + "loss": 2.898405647277832, + "step": 39530 + }, + { + "epoch": 0.1977, + "grad_norm": 0.32683122158050537, + "learning_rate": 0.0004569505697215768, + "loss": 2.8380022048950195, + "step": 39540 + }, + { + "epoch": 0.19775, + "grad_norm": 0.35552650690078735, + "learning_rate": 0.0004569283133772714, + "loss": 2.8094318389892576, + "step": 39550 + }, + { + "epoch": 0.1978, + "grad_norm": 0.3578189015388489, + "learning_rate": 0.00045690605182354717, + "loss": 2.837823486328125, + "step": 39560 + }, + { + "epoch": 0.19785, + "grad_norm": 0.4195692837238312, + "learning_rate": 0.00045688378506096453, + "loss": 2.86431770324707, + "step": 39570 + }, + { + "epoch": 0.1979, + "grad_norm": 0.33111006021499634, + "learning_rate": 0.00045686151309008385, + "loss": 2.8141153335571287, + "step": 39580 + }, + { + "epoch": 0.19795, + "grad_norm": 0.34890058636665344, + "learning_rate": 0.00045683923591146604, + "loss": 2.8161170959472654, + "step": 39590 + }, + { + "epoch": 0.198, + "grad_norm": 0.35623541474342346, + "learning_rate": 0.0004568169535256718, + "loss": 2.838735008239746, + "step": 39600 + }, + { + "epoch": 0.19805, + "grad_norm": 0.3725087642669678, + "learning_rate": 0.00045679466593326226, + "loss": 2.8788993835449217, + "step": 39610 + }, + { + "epoch": 0.1981, + "grad_norm": 0.3526809811592102, + "learning_rate": 0.00045677237313479827, + "loss": 2.8542585372924805, + "step": 39620 + }, + { + "epoch": 0.19815, + "grad_norm": 0.37528282403945923, + "learning_rate": 0.0004567500751308412, + "loss": 2.8737464904785157, + "step": 39630 + }, + { + "epoch": 0.1982, + "grad_norm": 0.3747260272502899, + "learning_rate": 0.00045672777192195234, + "loss": 2.8918657302856445, + "step": 39640 + }, + { + "epoch": 0.19825, + "grad_norm": 0.3215908110141754, + "learning_rate": 0.00045670546350869324, + "loss": 2.9225383758544923, + "step": 39650 + }, + { + "epoch": 0.1983, + "grad_norm": 0.3382894694805145, + "learning_rate": 0.00045668314989162547, + "loss": 2.8546207427978514, + "step": 39660 + }, + { + "epoch": 0.19835, + "grad_norm": 0.33842551708221436, + "learning_rate": 0.00045666083107131074, + "loss": 2.834523010253906, + "step": 39670 + }, + { + "epoch": 0.1984, + "grad_norm": 0.3457135558128357, + "learning_rate": 0.00045663850704831104, + "loss": 2.881685256958008, + "step": 39680 + }, + { + "epoch": 0.19845, + "grad_norm": 0.3375546634197235, + "learning_rate": 0.0004566161778231883, + "loss": 2.831229400634766, + "step": 39690 + }, + { + "epoch": 0.1985, + "grad_norm": 0.35175007581710815, + "learning_rate": 0.00045659384339650464, + "loss": 2.8850419998168944, + "step": 39700 + }, + { + "epoch": 0.19855, + "grad_norm": 0.34342870116233826, + "learning_rate": 0.0004565715037688223, + "loss": 2.8600317001342774, + "step": 39710 + }, + { + "epoch": 0.1986, + "grad_norm": 0.36717531085014343, + "learning_rate": 0.0004565491589407038, + "loss": 2.7863731384277344, + "step": 39720 + }, + { + "epoch": 0.19865, + "grad_norm": 0.3464389443397522, + "learning_rate": 0.00045652680891271157, + "loss": 2.8335647583007812, + "step": 39730 + }, + { + "epoch": 0.1987, + "grad_norm": 0.35148659348487854, + "learning_rate": 0.0004565044536854084, + "loss": 2.834770393371582, + "step": 39740 + }, + { + "epoch": 0.19875, + "grad_norm": 0.3520524203777313, + "learning_rate": 0.00045648209325935686, + "loss": 2.8661705017089845, + "step": 39750 + }, + { + "epoch": 0.1988, + "grad_norm": 0.3771170675754547, + "learning_rate": 0.00045645972763512006, + "loss": 2.8798477172851564, + "step": 39760 + }, + { + "epoch": 0.19885, + "grad_norm": 0.3514363765716553, + "learning_rate": 0.000456437356813261, + "loss": 2.8632133483886717, + "step": 39770 + }, + { + "epoch": 0.1989, + "grad_norm": 0.36735978722572327, + "learning_rate": 0.00045641498079434286, + "loss": 2.8685365676879884, + "step": 39780 + }, + { + "epoch": 0.19895, + "grad_norm": 0.35614365339279175, + "learning_rate": 0.00045639259957892895, + "loss": 2.835688018798828, + "step": 39790 + }, + { + "epoch": 0.199, + "grad_norm": 0.3753197491168976, + "learning_rate": 0.0004563702131675827, + "loss": 2.8423154830932615, + "step": 39800 + }, + { + "epoch": 0.19905, + "grad_norm": 0.35190606117248535, + "learning_rate": 0.00045634782156086774, + "loss": 2.823918342590332, + "step": 39810 + }, + { + "epoch": 0.1991, + "grad_norm": 0.38449564576148987, + "learning_rate": 0.00045632542475934775, + "loss": 2.8517112731933594, + "step": 39820 + }, + { + "epoch": 0.19915, + "grad_norm": 0.3508298397064209, + "learning_rate": 0.0004563030227635866, + "loss": 2.870138740539551, + "step": 39830 + }, + { + "epoch": 0.1992, + "grad_norm": 0.3363278806209564, + "learning_rate": 0.0004562806155741481, + "loss": 2.7965213775634767, + "step": 39840 + }, + { + "epoch": 0.19925, + "grad_norm": 0.3440210223197937, + "learning_rate": 0.0004562582031915966, + "loss": 2.8760587692260744, + "step": 39850 + }, + { + "epoch": 0.1993, + "grad_norm": 0.37253737449645996, + "learning_rate": 0.00045623578561649627, + "loss": 2.8878597259521483, + "step": 39860 + }, + { + "epoch": 0.19935, + "grad_norm": 0.3351092040538788, + "learning_rate": 0.0004562133628494113, + "loss": 2.9124021530151367, + "step": 39870 + }, + { + "epoch": 0.1994, + "grad_norm": 0.3633771240711212, + "learning_rate": 0.00045619093489090637, + "loss": 2.873111915588379, + "step": 39880 + }, + { + "epoch": 0.19945, + "grad_norm": 0.33578550815582275, + "learning_rate": 0.000456168501741546, + "loss": 2.894091796875, + "step": 39890 + }, + { + "epoch": 0.1995, + "grad_norm": 0.3617364764213562, + "learning_rate": 0.0004561460634018949, + "loss": 2.800567626953125, + "step": 39900 + }, + { + "epoch": 0.19955, + "grad_norm": 0.3407825827598572, + "learning_rate": 0.0004561236198725182, + "loss": 2.8573013305664063, + "step": 39910 + }, + { + "epoch": 0.1996, + "grad_norm": 0.3473286032676697, + "learning_rate": 0.0004561011711539806, + "loss": 2.8316999435424806, + "step": 39920 + }, + { + "epoch": 0.19965, + "grad_norm": 0.3538297712802887, + "learning_rate": 0.0004560787172468475, + "loss": 2.817875099182129, + "step": 39930 + }, + { + "epoch": 0.1997, + "grad_norm": 0.33591732382774353, + "learning_rate": 0.00045605625815168393, + "loss": 2.8439769744873047, + "step": 39940 + }, + { + "epoch": 0.19975, + "grad_norm": 0.36127132177352905, + "learning_rate": 0.00045603379386905557, + "loss": 2.8701641082763674, + "step": 39950 + }, + { + "epoch": 0.1998, + "grad_norm": 0.35122135281562805, + "learning_rate": 0.00045601132439952773, + "loss": 2.833110237121582, + "step": 39960 + }, + { + "epoch": 0.19985, + "grad_norm": 0.3338288366794586, + "learning_rate": 0.00045598884974366627, + "loss": 2.8530305862426757, + "step": 39970 + }, + { + "epoch": 0.1999, + "grad_norm": 0.3802099823951721, + "learning_rate": 0.0004559663699020369, + "loss": 2.8599103927612304, + "step": 39980 + }, + { + "epoch": 0.19995, + "grad_norm": 0.36109259724617004, + "learning_rate": 0.00045594388487520544, + "loss": 2.8811458587646483, + "step": 39990 + }, + { + "epoch": 0.2, + "grad_norm": 0.3678721785545349, + "learning_rate": 0.00045592139466373813, + "loss": 2.891261100769043, + "step": 40000 + }, + { + "epoch": 0.20005, + "grad_norm": 0.3348548412322998, + "learning_rate": 0.0004558988992682011, + "loss": 2.81421012878418, + "step": 40010 + }, + { + "epoch": 0.2001, + "grad_norm": 0.3872324824333191, + "learning_rate": 0.00045587639868916054, + "loss": 2.821419906616211, + "step": 40020 + }, + { + "epoch": 0.20015, + "grad_norm": 0.37680885195732117, + "learning_rate": 0.00045585389292718314, + "loss": 2.8540515899658203, + "step": 40030 + }, + { + "epoch": 0.2002, + "grad_norm": 0.33739838004112244, + "learning_rate": 0.0004558313819828353, + "loss": 2.887360191345215, + "step": 40040 + }, + { + "epoch": 0.20025, + "grad_norm": 0.34190964698791504, + "learning_rate": 0.0004558088658566838, + "loss": 2.863615608215332, + "step": 40050 + }, + { + "epoch": 0.2003, + "grad_norm": 0.34489670395851135, + "learning_rate": 0.0004557863445492955, + "loss": 2.875055503845215, + "step": 40060 + }, + { + "epoch": 0.20035, + "grad_norm": 0.36569082736968994, + "learning_rate": 0.00045576381806123733, + "loss": 2.8745538711547853, + "step": 40070 + }, + { + "epoch": 0.2004, + "grad_norm": 0.36706265807151794, + "learning_rate": 0.0004557412863930764, + "loss": 2.9439334869384766, + "step": 40080 + }, + { + "epoch": 0.20045, + "grad_norm": 0.34528815746307373, + "learning_rate": 0.00045571874954538006, + "loss": 2.8773242950439455, + "step": 40090 + }, + { + "epoch": 0.2005, + "grad_norm": 0.3554857671260834, + "learning_rate": 0.0004556962075187155, + "loss": 2.880145454406738, + "step": 40100 + }, + { + "epoch": 0.20055, + "grad_norm": 0.36604785919189453, + "learning_rate": 0.0004556736603136503, + "loss": 2.883439636230469, + "step": 40110 + }, + { + "epoch": 0.2006, + "grad_norm": 0.3452981412410736, + "learning_rate": 0.000455651107930752, + "loss": 2.8840980529785156, + "step": 40120 + }, + { + "epoch": 0.20065, + "grad_norm": 0.3378477394580841, + "learning_rate": 0.00045562855037058854, + "loss": 2.8423595428466797, + "step": 40130 + }, + { + "epoch": 0.2007, + "grad_norm": 0.3531666696071625, + "learning_rate": 0.00045560598763372766, + "loss": 2.8211553573608397, + "step": 40140 + }, + { + "epoch": 0.20075, + "grad_norm": 0.3386712074279785, + "learning_rate": 0.0004555834197207374, + "loss": 2.861226272583008, + "step": 40150 + }, + { + "epoch": 0.2008, + "grad_norm": 0.3816453516483307, + "learning_rate": 0.00045556084663218594, + "loss": 2.876871871948242, + "step": 40160 + }, + { + "epoch": 0.20085, + "grad_norm": 0.37466585636138916, + "learning_rate": 0.00045553826836864156, + "loss": 2.834865188598633, + "step": 40170 + }, + { + "epoch": 0.2009, + "grad_norm": 0.3513108789920807, + "learning_rate": 0.00045551568493067265, + "loss": 2.8832525253295898, + "step": 40180 + }, + { + "epoch": 0.20095, + "grad_norm": 0.3550271987915039, + "learning_rate": 0.0004554930963188477, + "loss": 2.8172813415527345, + "step": 40190 + }, + { + "epoch": 0.201, + "grad_norm": 0.4473459720611572, + "learning_rate": 0.00045547050253373546, + "loss": 2.8374664306640627, + "step": 40200 + }, + { + "epoch": 0.20105, + "grad_norm": 0.33891764283180237, + "learning_rate": 0.0004554479035759046, + "loss": 2.8482484817504883, + "step": 40210 + }, + { + "epoch": 0.2011, + "grad_norm": 0.34160110354423523, + "learning_rate": 0.00045542529944592424, + "loss": 2.8114015579223635, + "step": 40220 + }, + { + "epoch": 0.20115, + "grad_norm": 0.383076936006546, + "learning_rate": 0.0004554026901443633, + "loss": 2.844976043701172, + "step": 40230 + }, + { + "epoch": 0.2012, + "grad_norm": 0.3503985106945038, + "learning_rate": 0.00045538007567179107, + "loss": 2.819466209411621, + "step": 40240 + }, + { + "epoch": 0.20125, + "grad_norm": 0.34893783926963806, + "learning_rate": 0.00045535745602877674, + "loss": 2.800282096862793, + "step": 40250 + }, + { + "epoch": 0.2013, + "grad_norm": 0.3619614839553833, + "learning_rate": 0.0004553348312158898, + "loss": 2.857919120788574, + "step": 40260 + }, + { + "epoch": 0.20135, + "grad_norm": 0.36281144618988037, + "learning_rate": 0.00045531220123369993, + "loss": 2.8395370483398437, + "step": 40270 + }, + { + "epoch": 0.2014, + "grad_norm": 0.33258289098739624, + "learning_rate": 0.00045528956608277673, + "loss": 2.857342338562012, + "step": 40280 + }, + { + "epoch": 0.20145, + "grad_norm": 0.3682108521461487, + "learning_rate": 0.0004552669257636901, + "loss": 2.903404426574707, + "step": 40290 + }, + { + "epoch": 0.2015, + "grad_norm": 0.33979642391204834, + "learning_rate": 0.00045524428027700993, + "loss": 2.878773498535156, + "step": 40300 + }, + { + "epoch": 0.20155, + "grad_norm": 0.3451957404613495, + "learning_rate": 0.0004552216296233064, + "loss": 2.8637392044067385, + "step": 40310 + }, + { + "epoch": 0.2016, + "grad_norm": 0.3369506299495697, + "learning_rate": 0.0004551989738031498, + "loss": 2.825057792663574, + "step": 40320 + }, + { + "epoch": 0.20165, + "grad_norm": 0.35282352566719055, + "learning_rate": 0.00045517631281711033, + "loss": 2.792440414428711, + "step": 40330 + }, + { + "epoch": 0.2017, + "grad_norm": 0.39605939388275146, + "learning_rate": 0.0004551536466657585, + "loss": 2.859220504760742, + "step": 40340 + }, + { + "epoch": 0.20175, + "grad_norm": 0.3205952048301697, + "learning_rate": 0.00045513097534966506, + "loss": 2.7885978698730467, + "step": 40350 + }, + { + "epoch": 0.2018, + "grad_norm": 0.382217139005661, + "learning_rate": 0.0004551082988694007, + "loss": 2.8784793853759765, + "step": 40360 + }, + { + "epoch": 0.20185, + "grad_norm": 0.3340085744857788, + "learning_rate": 0.0004550856172255363, + "loss": 2.8410598754882814, + "step": 40370 + }, + { + "epoch": 0.2019, + "grad_norm": 0.3354496955871582, + "learning_rate": 0.0004550629304186428, + "loss": 2.826278495788574, + "step": 40380 + }, + { + "epoch": 0.20195, + "grad_norm": 0.3574986755847931, + "learning_rate": 0.0004550402384492914, + "loss": 2.910732078552246, + "step": 40390 + }, + { + "epoch": 0.202, + "grad_norm": 0.35273677110671997, + "learning_rate": 0.00045501754131805337, + "loss": 2.8353130340576174, + "step": 40400 + }, + { + "epoch": 0.20205, + "grad_norm": 0.3566650152206421, + "learning_rate": 0.0004549948390255001, + "loss": 2.8053489685058595, + "step": 40410 + }, + { + "epoch": 0.2021, + "grad_norm": 0.3118414282798767, + "learning_rate": 0.00045497213157220316, + "loss": 2.8505184173583986, + "step": 40420 + }, + { + "epoch": 0.20215, + "grad_norm": 0.3451545834541321, + "learning_rate": 0.00045494941895873414, + "loss": 2.818693923950195, + "step": 40430 + }, + { + "epoch": 0.2022, + "grad_norm": 0.36248406767845154, + "learning_rate": 0.00045492670118566495, + "loss": 2.824573516845703, + "step": 40440 + }, + { + "epoch": 0.20225, + "grad_norm": 0.34796273708343506, + "learning_rate": 0.00045490397825356734, + "loss": 2.8093374252319334, + "step": 40450 + }, + { + "epoch": 0.2023, + "grad_norm": 0.37134483456611633, + "learning_rate": 0.00045488125016301343, + "loss": 2.8565067291259765, + "step": 40460 + }, + { + "epoch": 0.20235, + "grad_norm": 0.35652443766593933, + "learning_rate": 0.0004548585169145755, + "loss": 2.8177934646606446, + "step": 40470 + }, + { + "epoch": 0.2024, + "grad_norm": 0.3502378761768341, + "learning_rate": 0.0004548357785088257, + "loss": 2.888774871826172, + "step": 40480 + }, + { + "epoch": 0.20245, + "grad_norm": 0.3312751352787018, + "learning_rate": 0.00045481303494633655, + "loss": 2.806667709350586, + "step": 40490 + }, + { + "epoch": 0.2025, + "grad_norm": 0.31163275241851807, + "learning_rate": 0.0004547902862276807, + "loss": 2.889775848388672, + "step": 40500 + }, + { + "epoch": 0.20255, + "grad_norm": 0.3890340030193329, + "learning_rate": 0.0004547675323534307, + "loss": 2.8312387466430664, + "step": 40510 + }, + { + "epoch": 0.2026, + "grad_norm": 0.3407190442085266, + "learning_rate": 0.0004547447733241594, + "loss": 2.853321838378906, + "step": 40520 + }, + { + "epoch": 0.20265, + "grad_norm": 0.3587588965892792, + "learning_rate": 0.0004547220091404399, + "loss": 2.803757667541504, + "step": 40530 + }, + { + "epoch": 0.2027, + "grad_norm": 0.3390841484069824, + "learning_rate": 0.0004546992398028451, + "loss": 2.8246671676635744, + "step": 40540 + }, + { + "epoch": 0.20275, + "grad_norm": 0.3679295480251312, + "learning_rate": 0.00045467646531194836, + "loss": 2.7807973861694335, + "step": 40550 + }, + { + "epoch": 0.2028, + "grad_norm": 0.3388402462005615, + "learning_rate": 0.0004546536856683229, + "loss": 2.85223388671875, + "step": 40560 + }, + { + "epoch": 0.20285, + "grad_norm": 0.37668660283088684, + "learning_rate": 0.00045463090087254234, + "loss": 2.8230716705322267, + "step": 40570 + }, + { + "epoch": 0.2029, + "grad_norm": 0.3370462656021118, + "learning_rate": 0.00045460811092518024, + "loss": 2.8684741973876955, + "step": 40580 + }, + { + "epoch": 0.20295, + "grad_norm": 0.34979668259620667, + "learning_rate": 0.00045458531582681026, + "loss": 2.8491704940795897, + "step": 40590 + }, + { + "epoch": 0.203, + "grad_norm": 0.35629093647003174, + "learning_rate": 0.00045456251557800633, + "loss": 2.8454620361328127, + "step": 40600 + }, + { + "epoch": 0.20305, + "grad_norm": 0.3536173403263092, + "learning_rate": 0.0004545397101793425, + "loss": 2.90665340423584, + "step": 40610 + }, + { + "epoch": 0.2031, + "grad_norm": 0.3221631944179535, + "learning_rate": 0.00045451689963139273, + "loss": 2.898701286315918, + "step": 40620 + }, + { + "epoch": 0.20315, + "grad_norm": 0.3674144148826599, + "learning_rate": 0.00045449408393473146, + "loss": 2.834543228149414, + "step": 40630 + }, + { + "epoch": 0.2032, + "grad_norm": 0.35243383049964905, + "learning_rate": 0.00045447126308993294, + "loss": 2.7905006408691406, + "step": 40640 + }, + { + "epoch": 0.20325, + "grad_norm": 0.3343481421470642, + "learning_rate": 0.00045444843709757177, + "loss": 2.8364086151123047, + "step": 40650 + }, + { + "epoch": 0.2033, + "grad_norm": 0.35260137915611267, + "learning_rate": 0.0004544256059582225, + "loss": 2.884627342224121, + "step": 40660 + }, + { + "epoch": 0.20335, + "grad_norm": 0.3880082368850708, + "learning_rate": 0.00045440276967246006, + "loss": 2.8519752502441404, + "step": 40670 + }, + { + "epoch": 0.2034, + "grad_norm": 0.35593995451927185, + "learning_rate": 0.00045437992824085916, + "loss": 2.8607646942138674, + "step": 40680 + }, + { + "epoch": 0.20345, + "grad_norm": 0.36025354266166687, + "learning_rate": 0.000454357081663995, + "loss": 2.7958303451538087, + "step": 40690 + }, + { + "epoch": 0.2035, + "grad_norm": 0.3296757638454437, + "learning_rate": 0.00045433422994244267, + "loss": 2.9053340911865235, + "step": 40700 + }, + { + "epoch": 0.20355, + "grad_norm": 0.3525979220867157, + "learning_rate": 0.00045431137307677753, + "loss": 2.846640777587891, + "step": 40710 + }, + { + "epoch": 0.2036, + "grad_norm": 0.3573268949985504, + "learning_rate": 0.0004542885110675748, + "loss": 2.868569564819336, + "step": 40720 + }, + { + "epoch": 0.20365, + "grad_norm": 0.3780444860458374, + "learning_rate": 0.0004542656439154103, + "loss": 2.8571739196777344, + "step": 40730 + }, + { + "epoch": 0.2037, + "grad_norm": 0.3551792800426483, + "learning_rate": 0.0004542427716208595, + "loss": 2.8277463912963867, + "step": 40740 + }, + { + "epoch": 0.20375, + "grad_norm": 0.3427170515060425, + "learning_rate": 0.00045421989418449827, + "loss": 2.8092893600463866, + "step": 40750 + }, + { + "epoch": 0.2038, + "grad_norm": 0.32568472623825073, + "learning_rate": 0.0004541970116069026, + "loss": 2.803614616394043, + "step": 40760 + }, + { + "epoch": 0.20385, + "grad_norm": 0.3434043228626251, + "learning_rate": 0.00045417412388864853, + "loss": 2.8085235595703124, + "step": 40770 + }, + { + "epoch": 0.2039, + "grad_norm": 0.37255674600601196, + "learning_rate": 0.00045415123103031225, + "loss": 2.864155960083008, + "step": 40780 + }, + { + "epoch": 0.20395, + "grad_norm": 0.3372601568698883, + "learning_rate": 0.0004541283330324701, + "loss": 2.8198274612426757, + "step": 40790 + }, + { + "epoch": 0.204, + "grad_norm": 0.34980708360671997, + "learning_rate": 0.0004541054298956985, + "loss": 2.8807573318481445, + "step": 40800 + }, + { + "epoch": 0.20405, + "grad_norm": 0.3282723128795624, + "learning_rate": 0.0004540825216205741, + "loss": 2.822195053100586, + "step": 40810 + }, + { + "epoch": 0.2041, + "grad_norm": 0.36256736516952515, + "learning_rate": 0.00045405960820767355, + "loss": 2.845446014404297, + "step": 40820 + }, + { + "epoch": 0.20415, + "grad_norm": 0.33430609107017517, + "learning_rate": 0.00045403668965757383, + "loss": 2.8502920150756834, + "step": 40830 + }, + { + "epoch": 0.2042, + "grad_norm": 0.3825342059135437, + "learning_rate": 0.0004540137659708517, + "loss": 2.8688453674316405, + "step": 40840 + }, + { + "epoch": 0.20425, + "grad_norm": 0.33080023527145386, + "learning_rate": 0.0004539908371480844, + "loss": 2.864778518676758, + "step": 40850 + }, + { + "epoch": 0.2043, + "grad_norm": 0.31677520275115967, + "learning_rate": 0.00045396790318984914, + "loss": 2.7965259552001953, + "step": 40860 + }, + { + "epoch": 0.20435, + "grad_norm": 0.3633684515953064, + "learning_rate": 0.0004539449640967233, + "loss": 2.84210262298584, + "step": 40870 + }, + { + "epoch": 0.2044, + "grad_norm": 0.3199993968009949, + "learning_rate": 0.00045392201986928433, + "loss": 2.8003889083862306, + "step": 40880 + }, + { + "epoch": 0.20445, + "grad_norm": 0.3857916593551636, + "learning_rate": 0.00045389907050810983, + "loss": 2.853664207458496, + "step": 40890 + }, + { + "epoch": 0.2045, + "grad_norm": 0.39124536514282227, + "learning_rate": 0.0004538761160137777, + "loss": 2.819746971130371, + "step": 40900 + }, + { + "epoch": 0.20455, + "grad_norm": 0.30358511209487915, + "learning_rate": 0.00045385315638686564, + "loss": 2.7927267074584963, + "step": 40910 + }, + { + "epoch": 0.2046, + "grad_norm": 0.3887353241443634, + "learning_rate": 0.0004538301916279518, + "loss": 2.8514001846313475, + "step": 40920 + }, + { + "epoch": 0.20465, + "grad_norm": 0.3569748103618622, + "learning_rate": 0.00045380722173761414, + "loss": 2.8477169036865235, + "step": 40930 + }, + { + "epoch": 0.2047, + "grad_norm": 0.398926317691803, + "learning_rate": 0.00045378424671643113, + "loss": 2.828413391113281, + "step": 40940 + }, + { + "epoch": 0.20475, + "grad_norm": 0.37312084436416626, + "learning_rate": 0.00045376126656498103, + "loss": 2.842312049865723, + "step": 40950 + }, + { + "epoch": 0.2048, + "grad_norm": 0.355481892824173, + "learning_rate": 0.0004537382812838424, + "loss": 2.8399139404296876, + "step": 40960 + }, + { + "epoch": 0.20485, + "grad_norm": 0.35498690605163574, + "learning_rate": 0.00045371529087359386, + "loss": 2.864211082458496, + "step": 40970 + }, + { + "epoch": 0.2049, + "grad_norm": 0.3352961242198944, + "learning_rate": 0.00045369229533481434, + "loss": 2.877581787109375, + "step": 40980 + }, + { + "epoch": 0.20495, + "grad_norm": 0.3678177297115326, + "learning_rate": 0.00045366929466808255, + "loss": 2.842845153808594, + "step": 40990 + }, + { + "epoch": 0.205, + "grad_norm": 0.3465934097766876, + "learning_rate": 0.0004536462888739776, + "loss": 2.837766647338867, + "step": 41000 + }, + { + "epoch": 0.20505, + "grad_norm": 0.34135666489601135, + "learning_rate": 0.0004536232779530788, + "loss": 2.807243537902832, + "step": 41010 + }, + { + "epoch": 0.2051, + "grad_norm": 0.3520164489746094, + "learning_rate": 0.00045360026190596525, + "loss": 2.849888801574707, + "step": 41020 + }, + { + "epoch": 0.20515, + "grad_norm": 0.3455839455127716, + "learning_rate": 0.00045357724073321655, + "loss": 2.8271581649780275, + "step": 41030 + }, + { + "epoch": 0.2052, + "grad_norm": 0.34239429235458374, + "learning_rate": 0.00045355421443541214, + "loss": 2.8585556030273436, + "step": 41040 + }, + { + "epoch": 0.20525, + "grad_norm": 0.33649998903274536, + "learning_rate": 0.00045353118301313174, + "loss": 2.8147945404052734, + "step": 41050 + }, + { + "epoch": 0.2053, + "grad_norm": 0.33625367283821106, + "learning_rate": 0.00045350814646695514, + "loss": 2.8416887283325196, + "step": 41060 + }, + { + "epoch": 0.20535, + "grad_norm": 0.3470055162906647, + "learning_rate": 0.0004534851047974623, + "loss": 2.903325653076172, + "step": 41070 + }, + { + "epoch": 0.2054, + "grad_norm": 0.3179188370704651, + "learning_rate": 0.00045346205800523334, + "loss": 2.8436233520507814, + "step": 41080 + }, + { + "epoch": 0.20545, + "grad_norm": 0.35633155703544617, + "learning_rate": 0.00045343900609084845, + "loss": 2.7989282608032227, + "step": 41090 + }, + { + "epoch": 0.2055, + "grad_norm": 0.3716106712818146, + "learning_rate": 0.00045341594905488794, + "loss": 2.8281326293945312, + "step": 41100 + }, + { + "epoch": 0.20555, + "grad_norm": 0.33949658274650574, + "learning_rate": 0.0004533928868979322, + "loss": 2.853445816040039, + "step": 41110 + }, + { + "epoch": 0.2056, + "grad_norm": 0.35412195324897766, + "learning_rate": 0.00045336981962056203, + "loss": 2.8602216720581053, + "step": 41120 + }, + { + "epoch": 0.20565, + "grad_norm": 0.36113741993904114, + "learning_rate": 0.0004533467472233579, + "loss": 2.8498620986938477, + "step": 41130 + }, + { + "epoch": 0.2057, + "grad_norm": 0.34338679909706116, + "learning_rate": 0.00045332366970690077, + "loss": 2.8239526748657227, + "step": 41140 + }, + { + "epoch": 0.20575, + "grad_norm": 0.32929229736328125, + "learning_rate": 0.0004533005870717716, + "loss": 2.8394283294677733, + "step": 41150 + }, + { + "epoch": 0.2058, + "grad_norm": 0.3687082827091217, + "learning_rate": 0.00045327749931855154, + "loss": 2.8422216415405273, + "step": 41160 + }, + { + "epoch": 0.20585, + "grad_norm": 0.36561551690101624, + "learning_rate": 0.0004532544064478218, + "loss": 2.8304601669311524, + "step": 41170 + }, + { + "epoch": 0.2059, + "grad_norm": 0.3613913357257843, + "learning_rate": 0.00045323130846016373, + "loss": 2.833769416809082, + "step": 41180 + }, + { + "epoch": 0.20595, + "grad_norm": 0.3381495475769043, + "learning_rate": 0.0004532082053561588, + "loss": 2.853791427612305, + "step": 41190 + }, + { + "epoch": 0.206, + "grad_norm": 0.36370542645454407, + "learning_rate": 0.0004531850971363887, + "loss": 2.8266250610351564, + "step": 41200 + }, + { + "epoch": 0.20605, + "grad_norm": 0.34818506240844727, + "learning_rate": 0.00045316198380143517, + "loss": 2.852571487426758, + "step": 41210 + }, + { + "epoch": 0.2061, + "grad_norm": 0.33712369203567505, + "learning_rate": 0.00045313886535188, + "loss": 2.8573347091674806, + "step": 41220 + }, + { + "epoch": 0.20615, + "grad_norm": 0.3625553846359253, + "learning_rate": 0.00045311574178830523, + "loss": 2.8528072357177736, + "step": 41230 + }, + { + "epoch": 0.2062, + "grad_norm": 0.38526299595832825, + "learning_rate": 0.00045309261311129306, + "loss": 2.8439332962036135, + "step": 41240 + }, + { + "epoch": 0.20625, + "grad_norm": 0.34860971570014954, + "learning_rate": 0.0004530694793214257, + "loss": 2.832173156738281, + "step": 41250 + }, + { + "epoch": 0.2063, + "grad_norm": 0.3723740875720978, + "learning_rate": 0.00045304634041928556, + "loss": 2.836155319213867, + "step": 41260 + }, + { + "epoch": 0.20635, + "grad_norm": 0.3565866947174072, + "learning_rate": 0.00045302319640545515, + "loss": 2.8322164535522463, + "step": 41270 + }, + { + "epoch": 0.2064, + "grad_norm": 0.3335269093513489, + "learning_rate": 0.00045300004728051713, + "loss": 2.8487545013427735, + "step": 41280 + }, + { + "epoch": 0.20645, + "grad_norm": 0.31474122405052185, + "learning_rate": 0.0004529768930450543, + "loss": 2.8477706909179688, + "step": 41290 + }, + { + "epoch": 0.2065, + "grad_norm": 0.31198692321777344, + "learning_rate": 0.0004529537336996495, + "loss": 2.8437826156616213, + "step": 41300 + }, + { + "epoch": 0.20655, + "grad_norm": 0.3619883060455322, + "learning_rate": 0.00045293056924488586, + "loss": 2.8832704544067385, + "step": 41310 + }, + { + "epoch": 0.2066, + "grad_norm": 0.34992310404777527, + "learning_rate": 0.0004529073996813464, + "loss": 2.8700374603271483, + "step": 41320 + }, + { + "epoch": 0.20665, + "grad_norm": 0.3341822326183319, + "learning_rate": 0.0004528842250096146, + "loss": 2.842588996887207, + "step": 41330 + }, + { + "epoch": 0.2067, + "grad_norm": 0.38242053985595703, + "learning_rate": 0.0004528610452302737, + "loss": 2.8019155502319335, + "step": 41340 + }, + { + "epoch": 0.20675, + "grad_norm": 0.33897867798805237, + "learning_rate": 0.00045283786034390743, + "loss": 2.81378288269043, + "step": 41350 + }, + { + "epoch": 0.2068, + "grad_norm": 0.34328413009643555, + "learning_rate": 0.00045281467035109934, + "loss": 2.8516914367675783, + "step": 41360 + }, + { + "epoch": 0.20685, + "grad_norm": 0.3280213475227356, + "learning_rate": 0.00045279147525243335, + "loss": 2.8110803604125976, + "step": 41370 + }, + { + "epoch": 0.2069, + "grad_norm": 0.3192994296550751, + "learning_rate": 0.0004527682750484933, + "loss": 2.8338722229003905, + "step": 41380 + }, + { + "epoch": 0.20695, + "grad_norm": 0.3407706022262573, + "learning_rate": 0.00045274506973986325, + "loss": 2.8080190658569335, + "step": 41390 + }, + { + "epoch": 0.207, + "grad_norm": 0.3715844750404358, + "learning_rate": 0.0004527218593271274, + "loss": 2.8938247680664064, + "step": 41400 + }, + { + "epoch": 0.20705, + "grad_norm": 0.3604896664619446, + "learning_rate": 0.0004526986438108701, + "loss": 2.9121994018554687, + "step": 41410 + }, + { + "epoch": 0.2071, + "grad_norm": 0.34150412678718567, + "learning_rate": 0.00045267542319167586, + "loss": 2.8086544036865235, + "step": 41420 + }, + { + "epoch": 0.20715, + "grad_norm": 0.3856627643108368, + "learning_rate": 0.00045265219747012917, + "loss": 2.910566520690918, + "step": 41430 + }, + { + "epoch": 0.2072, + "grad_norm": 0.36143597960472107, + "learning_rate": 0.00045262896664681473, + "loss": 2.85259952545166, + "step": 41440 + }, + { + "epoch": 0.20725, + "grad_norm": 0.3168695271015167, + "learning_rate": 0.00045260573072231737, + "loss": 2.7875812530517576, + "step": 41450 + }, + { + "epoch": 0.2073, + "grad_norm": 0.3424910008907318, + "learning_rate": 0.00045258248969722216, + "loss": 2.8482852935791017, + "step": 41460 + }, + { + "epoch": 0.20735, + "grad_norm": 0.3443971872329712, + "learning_rate": 0.00045255924357211414, + "loss": 2.825539779663086, + "step": 41470 + }, + { + "epoch": 0.2074, + "grad_norm": 0.34344354271888733, + "learning_rate": 0.00045253599234757846, + "loss": 2.8576223373413088, + "step": 41480 + }, + { + "epoch": 0.20745, + "grad_norm": 0.3912670910358429, + "learning_rate": 0.00045251273602420065, + "loss": 2.8321325302124025, + "step": 41490 + }, + { + "epoch": 0.2075, + "grad_norm": 0.32235369086265564, + "learning_rate": 0.0004524894746025659, + "loss": 2.840361404418945, + "step": 41500 + }, + { + "epoch": 0.20755, + "grad_norm": 0.36832287907600403, + "learning_rate": 0.00045246620808326004, + "loss": 2.908772277832031, + "step": 41510 + }, + { + "epoch": 0.2076, + "grad_norm": 0.37938034534454346, + "learning_rate": 0.00045244293646686873, + "loss": 2.8363555908203124, + "step": 41520 + }, + { + "epoch": 0.20765, + "grad_norm": 0.34772789478302, + "learning_rate": 0.0004524196597539779, + "loss": 2.8479122161865233, + "step": 41530 + }, + { + "epoch": 0.2077, + "grad_norm": 0.33842554688453674, + "learning_rate": 0.0004523963779451735, + "loss": 2.8268699645996094, + "step": 41540 + }, + { + "epoch": 0.20775, + "grad_norm": 0.3050522208213806, + "learning_rate": 0.00045237309104104157, + "loss": 2.854047966003418, + "step": 41550 + }, + { + "epoch": 0.2078, + "grad_norm": 0.34437745809555054, + "learning_rate": 0.0004523497990421684, + "loss": 2.8772342681884764, + "step": 41560 + }, + { + "epoch": 0.20785, + "grad_norm": 0.3389931619167328, + "learning_rate": 0.0004523265019491405, + "loss": 2.832992744445801, + "step": 41570 + }, + { + "epoch": 0.2079, + "grad_norm": 0.35487428307533264, + "learning_rate": 0.0004523031997625442, + "loss": 2.8369882583618162, + "step": 41580 + }, + { + "epoch": 0.20795, + "grad_norm": 0.33915743231773376, + "learning_rate": 0.0004522798924829662, + "loss": 2.842449951171875, + "step": 41590 + }, + { + "epoch": 0.208, + "grad_norm": 0.3339273929595947, + "learning_rate": 0.00045225658011099326, + "loss": 2.7994312286376952, + "step": 41600 + }, + { + "epoch": 0.20805, + "grad_norm": 0.39016467332839966, + "learning_rate": 0.0004522332626472123, + "loss": 2.846811294555664, + "step": 41610 + }, + { + "epoch": 0.2081, + "grad_norm": 0.33738985657691956, + "learning_rate": 0.00045220994009221026, + "loss": 2.828877639770508, + "step": 41620 + }, + { + "epoch": 0.20815, + "grad_norm": 0.34962642192840576, + "learning_rate": 0.00045218661244657434, + "loss": 2.8052999496459963, + "step": 41630 + }, + { + "epoch": 0.2082, + "grad_norm": 0.36697855591773987, + "learning_rate": 0.0004521632797108918, + "loss": 2.9099363327026366, + "step": 41640 + }, + { + "epoch": 0.20825, + "grad_norm": 0.37439730763435364, + "learning_rate": 0.00045213994188575006, + "loss": 2.817440414428711, + "step": 41650 + }, + { + "epoch": 0.2083, + "grad_norm": 0.3432646691799164, + "learning_rate": 0.00045211659897173665, + "loss": 2.8045766830444334, + "step": 41660 + }, + { + "epoch": 0.20835, + "grad_norm": 0.35347166657447815, + "learning_rate": 0.0004520932509694392, + "loss": 2.875701141357422, + "step": 41670 + }, + { + "epoch": 0.2084, + "grad_norm": 0.33030709624290466, + "learning_rate": 0.00045206989787944553, + "loss": 2.7649953842163084, + "step": 41680 + }, + { + "epoch": 0.20845, + "grad_norm": 0.3392440974712372, + "learning_rate": 0.0004520465397023434, + "loss": 2.8671966552734376, + "step": 41690 + }, + { + "epoch": 0.2085, + "grad_norm": 0.3735390305519104, + "learning_rate": 0.00045202317643872113, + "loss": 2.867182731628418, + "step": 41700 + }, + { + "epoch": 0.20855, + "grad_norm": 0.3479546308517456, + "learning_rate": 0.0004519998080891667, + "loss": 2.849447250366211, + "step": 41710 + }, + { + "epoch": 0.2086, + "grad_norm": 0.3437989056110382, + "learning_rate": 0.0004519764346542684, + "loss": 2.8745731353759765, + "step": 41720 + }, + { + "epoch": 0.20865, + "grad_norm": 0.3509753942489624, + "learning_rate": 0.0004519530561346148, + "loss": 2.8888431549072267, + "step": 41730 + }, + { + "epoch": 0.2087, + "grad_norm": 0.340424120426178, + "learning_rate": 0.0004519296725307943, + "loss": 2.9005207061767577, + "step": 41740 + }, + { + "epoch": 0.20875, + "grad_norm": 0.3637494146823883, + "learning_rate": 0.00045190628384339564, + "loss": 2.8585332870483398, + "step": 41750 + }, + { + "epoch": 0.2088, + "grad_norm": 0.37488606572151184, + "learning_rate": 0.00045188289007300767, + "loss": 2.8519487380981445, + "step": 41760 + }, + { + "epoch": 0.20885, + "grad_norm": 0.3521476089954376, + "learning_rate": 0.0004518594912202193, + "loss": 2.857629585266113, + "step": 41770 + }, + { + "epoch": 0.2089, + "grad_norm": 0.35440343618392944, + "learning_rate": 0.00045183608728561954, + "loss": 2.8298599243164064, + "step": 41780 + }, + { + "epoch": 0.20895, + "grad_norm": 0.38414937257766724, + "learning_rate": 0.00045181267826979764, + "loss": 2.868828773498535, + "step": 41790 + }, + { + "epoch": 0.209, + "grad_norm": 0.3464762270450592, + "learning_rate": 0.00045178926417334294, + "loss": 2.811982536315918, + "step": 41800 + }, + { + "epoch": 0.20905, + "grad_norm": 0.3400702476501465, + "learning_rate": 0.0004517658449968448, + "loss": 2.881213569641113, + "step": 41810 + }, + { + "epoch": 0.2091, + "grad_norm": 0.3136358857154846, + "learning_rate": 0.00045174242074089296, + "loss": 2.8658096313476564, + "step": 41820 + }, + { + "epoch": 0.20915, + "grad_norm": 0.3291848301887512, + "learning_rate": 0.00045171899140607696, + "loss": 2.811798858642578, + "step": 41830 + }, + { + "epoch": 0.2092, + "grad_norm": 0.35601624846458435, + "learning_rate": 0.0004516955569929867, + "loss": 2.8496631622314452, + "step": 41840 + }, + { + "epoch": 0.20925, + "grad_norm": 0.3608093559741974, + "learning_rate": 0.00045167211750221214, + "loss": 2.813472557067871, + "step": 41850 + }, + { + "epoch": 0.2093, + "grad_norm": 0.3850418031215668, + "learning_rate": 0.0004516486729343434, + "loss": 2.8757747650146483, + "step": 41860 + }, + { + "epoch": 0.20935, + "grad_norm": 0.36655041575431824, + "learning_rate": 0.0004516252232899707, + "loss": 2.858357620239258, + "step": 41870 + }, + { + "epoch": 0.2094, + "grad_norm": 0.3076498806476593, + "learning_rate": 0.0004516017685696843, + "loss": 2.814997100830078, + "step": 41880 + }, + { + "epoch": 0.20945, + "grad_norm": 0.3509843647480011, + "learning_rate": 0.0004515783087740747, + "loss": 2.8086593627929686, + "step": 41890 + }, + { + "epoch": 0.2095, + "grad_norm": 0.35701784491539, + "learning_rate": 0.00045155484390373255, + "loss": 2.792435073852539, + "step": 41900 + }, + { + "epoch": 0.20955, + "grad_norm": 0.36247992515563965, + "learning_rate": 0.00045153137395924847, + "loss": 2.8566240310668944, + "step": 41910 + }, + { + "epoch": 0.2096, + "grad_norm": 0.36954084038734436, + "learning_rate": 0.0004515078989412135, + "loss": 2.8543529510498047, + "step": 41920 + }, + { + "epoch": 0.20965, + "grad_norm": 0.32860302925109863, + "learning_rate": 0.0004514844188502185, + "loss": 2.8276298522949217, + "step": 41930 + }, + { + "epoch": 0.2097, + "grad_norm": 0.331917405128479, + "learning_rate": 0.0004514609336868546, + "loss": 2.8819772720336916, + "step": 41940 + }, + { + "epoch": 0.20975, + "grad_norm": 0.3579454720020294, + "learning_rate": 0.000451437443451713, + "loss": 2.854771041870117, + "step": 41950 + }, + { + "epoch": 0.2098, + "grad_norm": 0.31874364614486694, + "learning_rate": 0.0004514139481453852, + "loss": 2.822149467468262, + "step": 41960 + }, + { + "epoch": 0.20985, + "grad_norm": 0.36804530024528503, + "learning_rate": 0.00045139044776846247, + "loss": 2.8514270782470703, + "step": 41970 + }, + { + "epoch": 0.2099, + "grad_norm": 0.36228224635124207, + "learning_rate": 0.0004513669423215367, + "loss": 2.870855140686035, + "step": 41980 + }, + { + "epoch": 0.20995, + "grad_norm": 0.35565391182899475, + "learning_rate": 0.00045134343180519944, + "loss": 2.8289188385009765, + "step": 41990 + }, + { + "epoch": 0.21, + "grad_norm": 0.364735871553421, + "learning_rate": 0.00045131991622004264, + "loss": 2.854927635192871, + "step": 42000 + }, + { + "epoch": 0.21005, + "grad_norm": 0.365027517080307, + "learning_rate": 0.00045129639556665827, + "loss": 2.8284017562866213, + "step": 42010 + }, + { + "epoch": 0.2101, + "grad_norm": 0.3374779224395752, + "learning_rate": 0.0004512728698456385, + "loss": 2.829415512084961, + "step": 42020 + }, + { + "epoch": 0.21015, + "grad_norm": 0.3628893494606018, + "learning_rate": 0.0004512493390575756, + "loss": 2.810393714904785, + "step": 42030 + }, + { + "epoch": 0.2102, + "grad_norm": 0.3268432915210724, + "learning_rate": 0.00045122580320306187, + "loss": 2.835218620300293, + "step": 42040 + }, + { + "epoch": 0.21025, + "grad_norm": 0.35572364926338196, + "learning_rate": 0.00045120226228268997, + "loss": 2.832113838195801, + "step": 42050 + }, + { + "epoch": 0.2103, + "grad_norm": 0.3205704092979431, + "learning_rate": 0.00045117871629705243, + "loss": 2.778306007385254, + "step": 42060 + }, + { + "epoch": 0.21035, + "grad_norm": 0.3559894859790802, + "learning_rate": 0.00045115516524674203, + "loss": 2.8292203903198243, + "step": 42070 + }, + { + "epoch": 0.2104, + "grad_norm": 0.3700430691242218, + "learning_rate": 0.0004511316091323517, + "loss": 2.8322422027587892, + "step": 42080 + }, + { + "epoch": 0.21045, + "grad_norm": 0.3734930753707886, + "learning_rate": 0.00045110804795447446, + "loss": 2.83651180267334, + "step": 42090 + }, + { + "epoch": 0.2105, + "grad_norm": 0.3368440270423889, + "learning_rate": 0.0004510844817137034, + "loss": 2.840473175048828, + "step": 42100 + }, + { + "epoch": 0.21055, + "grad_norm": 0.3874538540840149, + "learning_rate": 0.00045106091041063195, + "loss": 2.873932647705078, + "step": 42110 + }, + { + "epoch": 0.2106, + "grad_norm": 0.3493232727050781, + "learning_rate": 0.0004510373340458534, + "loss": 2.8666349411010743, + "step": 42120 + }, + { + "epoch": 0.21065, + "grad_norm": 0.3540133237838745, + "learning_rate": 0.00045101375261996125, + "loss": 2.8731210708618162, + "step": 42130 + }, + { + "epoch": 0.2107, + "grad_norm": 0.3531893789768219, + "learning_rate": 0.00045099016613354926, + "loss": 2.8176427841186524, + "step": 42140 + }, + { + "epoch": 0.21075, + "grad_norm": 0.35506972670555115, + "learning_rate": 0.0004509665745872112, + "loss": 2.800887680053711, + "step": 42150 + }, + { + "epoch": 0.2108, + "grad_norm": 0.3643660247325897, + "learning_rate": 0.000450942977981541, + "loss": 2.836701011657715, + "step": 42160 + }, + { + "epoch": 0.21085, + "grad_norm": 0.350869357585907, + "learning_rate": 0.0004509193763171326, + "loss": 2.8478925704956053, + "step": 42170 + }, + { + "epoch": 0.2109, + "grad_norm": 0.342495322227478, + "learning_rate": 0.0004508957695945802, + "loss": 2.837643623352051, + "step": 42180 + }, + { + "epoch": 0.21095, + "grad_norm": 0.3809254765510559, + "learning_rate": 0.00045087215781447834, + "loss": 2.8844385147094727, + "step": 42190 + }, + { + "epoch": 0.211, + "grad_norm": 0.3352200984954834, + "learning_rate": 0.00045084854097742107, + "loss": 2.7910205841064455, + "step": 42200 + }, + { + "epoch": 0.21105, + "grad_norm": 0.37117165327072144, + "learning_rate": 0.00045082491908400314, + "loss": 2.8219728469848633, + "step": 42210 + }, + { + "epoch": 0.2111, + "grad_norm": 0.35443800687789917, + "learning_rate": 0.00045080129213481923, + "loss": 2.8258934020996094, + "step": 42220 + }, + { + "epoch": 0.21115, + "grad_norm": 0.32840245962142944, + "learning_rate": 0.00045077766013046415, + "loss": 2.813466262817383, + "step": 42230 + }, + { + "epoch": 0.2112, + "grad_norm": 0.3417123854160309, + "learning_rate": 0.00045075402307153284, + "loss": 2.8119586944580077, + "step": 42240 + }, + { + "epoch": 0.21125, + "grad_norm": 0.3401896357536316, + "learning_rate": 0.00045073038095862027, + "loss": 2.8389556884765623, + "step": 42250 + }, + { + "epoch": 0.2113, + "grad_norm": 0.3404863178730011, + "learning_rate": 0.0004507067337923218, + "loss": 2.7970302581787108, + "step": 42260 + }, + { + "epoch": 0.21135, + "grad_norm": 0.33325308561325073, + "learning_rate": 0.0004506830815732326, + "loss": 2.875151443481445, + "step": 42270 + }, + { + "epoch": 0.2114, + "grad_norm": 0.3672698736190796, + "learning_rate": 0.0004506594243019482, + "loss": 2.81700439453125, + "step": 42280 + }, + { + "epoch": 0.21145, + "grad_norm": 0.33069974184036255, + "learning_rate": 0.00045063576197906406, + "loss": 2.8336280822753905, + "step": 42290 + }, + { + "epoch": 0.2115, + "grad_norm": 0.3476468324661255, + "learning_rate": 0.000450612094605176, + "loss": 2.8929460525512694, + "step": 42300 + }, + { + "epoch": 0.21155, + "grad_norm": 0.37120285630226135, + "learning_rate": 0.0004505884221808798, + "loss": 2.9186212539672853, + "step": 42310 + }, + { + "epoch": 0.2116, + "grad_norm": 0.35366886854171753, + "learning_rate": 0.0004505647447067713, + "loss": 2.8423883438110353, + "step": 42320 + }, + { + "epoch": 0.21165, + "grad_norm": 0.3530419170856476, + "learning_rate": 0.0004505410621834468, + "loss": 2.8139081954956056, + "step": 42330 + }, + { + "epoch": 0.2117, + "grad_norm": 0.33595970273017883, + "learning_rate": 0.00045051737461150246, + "loss": 2.8118629455566406, + "step": 42340 + }, + { + "epoch": 0.21175, + "grad_norm": 0.3646652102470398, + "learning_rate": 0.0004504936819915345, + "loss": 2.8261924743652345, + "step": 42350 + }, + { + "epoch": 0.2118, + "grad_norm": 0.3471113443374634, + "learning_rate": 0.0004504699843241394, + "loss": 2.8245162963867188, + "step": 42360 + }, + { + "epoch": 0.21185, + "grad_norm": 0.3656593859195709, + "learning_rate": 0.00045044628160991374, + "loss": 2.8377910614013673, + "step": 42370 + }, + { + "epoch": 0.2119, + "grad_norm": 0.333457350730896, + "learning_rate": 0.0004504225738494544, + "loss": 2.819989013671875, + "step": 42380 + }, + { + "epoch": 0.21195, + "grad_norm": 0.32372331619262695, + "learning_rate": 0.00045039886104335805, + "loss": 2.857356643676758, + "step": 42390 + }, + { + "epoch": 0.212, + "grad_norm": 0.34708452224731445, + "learning_rate": 0.0004503751431922217, + "loss": 2.8636825561523436, + "step": 42400 + }, + { + "epoch": 0.21205, + "grad_norm": 0.35025152564048767, + "learning_rate": 0.0004503514202966424, + "loss": 2.856828308105469, + "step": 42410 + }, + { + "epoch": 0.2121, + "grad_norm": 0.3429851830005646, + "learning_rate": 0.00045032769235721755, + "loss": 2.884417915344238, + "step": 42420 + }, + { + "epoch": 0.21215, + "grad_norm": 0.3775425851345062, + "learning_rate": 0.0004503039593745443, + "loss": 2.8507579803466796, + "step": 42430 + }, + { + "epoch": 0.2122, + "grad_norm": 0.32880520820617676, + "learning_rate": 0.00045028022134922026, + "loss": 2.8496259689331054, + "step": 42440 + }, + { + "epoch": 0.21225, + "grad_norm": 0.3445592224597931, + "learning_rate": 0.00045025647828184296, + "loss": 2.848603057861328, + "step": 42450 + }, + { + "epoch": 0.2123, + "grad_norm": 0.39191845059394836, + "learning_rate": 0.00045023273017301017, + "loss": 2.8451635360717775, + "step": 42460 + }, + { + "epoch": 0.21235, + "grad_norm": 0.3625827729701996, + "learning_rate": 0.00045020897702331973, + "loss": 2.852548027038574, + "step": 42470 + }, + { + "epoch": 0.2124, + "grad_norm": 0.38367971777915955, + "learning_rate": 0.0004501852188333696, + "loss": 2.794620132446289, + "step": 42480 + }, + { + "epoch": 0.21245, + "grad_norm": 0.3467719554901123, + "learning_rate": 0.000450161455603758, + "loss": 2.855988883972168, + "step": 42490 + }, + { + "epoch": 0.2125, + "grad_norm": 0.30904421210289, + "learning_rate": 0.0004501376873350831, + "loss": 2.8741615295410154, + "step": 42500 + }, + { + "epoch": 0.21255, + "grad_norm": 0.3380681872367859, + "learning_rate": 0.00045011391402794315, + "loss": 2.857814598083496, + "step": 42510 + }, + { + "epoch": 0.2126, + "grad_norm": 0.33677709102630615, + "learning_rate": 0.0004500901356829368, + "loss": 2.8434661865234374, + "step": 42520 + }, + { + "epoch": 0.21265, + "grad_norm": 0.3557634949684143, + "learning_rate": 0.0004500663523006626, + "loss": 2.8844619750976563, + "step": 42530 + }, + { + "epoch": 0.2127, + "grad_norm": 0.35165882110595703, + "learning_rate": 0.00045004256388171937, + "loss": 2.8214439392089843, + "step": 42540 + }, + { + "epoch": 0.21275, + "grad_norm": 0.3804748058319092, + "learning_rate": 0.00045001877042670594, + "loss": 2.908976173400879, + "step": 42550 + }, + { + "epoch": 0.2128, + "grad_norm": 0.3612578809261322, + "learning_rate": 0.0004499949719362213, + "loss": 2.8370826721191404, + "step": 42560 + }, + { + "epoch": 0.21285, + "grad_norm": 0.3435121178627014, + "learning_rate": 0.0004499711684108645, + "loss": 2.8477928161621096, + "step": 42570 + }, + { + "epoch": 0.2129, + "grad_norm": 0.3445492684841156, + "learning_rate": 0.00044994735985123493, + "loss": 2.84923095703125, + "step": 42580 + }, + { + "epoch": 0.21295, + "grad_norm": 0.33270716667175293, + "learning_rate": 0.00044992354625793187, + "loss": 2.896504783630371, + "step": 42590 + }, + { + "epoch": 0.213, + "grad_norm": 0.37119555473327637, + "learning_rate": 0.00044989972763155497, + "loss": 2.883552360534668, + "step": 42600 + }, + { + "epoch": 0.21305, + "grad_norm": 0.35604193806648254, + "learning_rate": 0.0004498759039727037, + "loss": 2.8398685455322266, + "step": 42610 + }, + { + "epoch": 0.2131, + "grad_norm": 0.31351521611213684, + "learning_rate": 0.0004498520752819778, + "loss": 2.8493629455566407, + "step": 42620 + }, + { + "epoch": 0.21315, + "grad_norm": 0.3772552013397217, + "learning_rate": 0.00044982824155997736, + "loss": 2.842407989501953, + "step": 42630 + }, + { + "epoch": 0.2132, + "grad_norm": 0.3304370939731598, + "learning_rate": 0.0004498044028073022, + "loss": 2.827780914306641, + "step": 42640 + }, + { + "epoch": 0.21325, + "grad_norm": 0.33956822752952576, + "learning_rate": 0.0004497805590245526, + "loss": 2.8106733322143556, + "step": 42650 + }, + { + "epoch": 0.2133, + "grad_norm": 0.3407596945762634, + "learning_rate": 0.0004497567102123287, + "loss": 2.808293914794922, + "step": 42660 + }, + { + "epoch": 0.21335, + "grad_norm": 0.3578781485557556, + "learning_rate": 0.00044973285637123103, + "loss": 2.8108348846435547, + "step": 42670 + }, + { + "epoch": 0.2134, + "grad_norm": 0.39356449246406555, + "learning_rate": 0.00044970899750185997, + "loss": 2.87674560546875, + "step": 42680 + }, + { + "epoch": 0.21345, + "grad_norm": 0.34496402740478516, + "learning_rate": 0.00044968513360481624, + "loss": 2.821781349182129, + "step": 42690 + }, + { + "epoch": 0.2135, + "grad_norm": 0.34533974528312683, + "learning_rate": 0.0004496612646807006, + "loss": 2.837067222595215, + "step": 42700 + }, + { + "epoch": 0.21355, + "grad_norm": 0.33312657475471497, + "learning_rate": 0.000449637390730114, + "loss": 2.8230056762695312, + "step": 42710 + }, + { + "epoch": 0.2136, + "grad_norm": 0.39875003695487976, + "learning_rate": 0.0004496135117536574, + "loss": 2.8029075622558595, + "step": 42720 + }, + { + "epoch": 0.21365, + "grad_norm": 0.4084235429763794, + "learning_rate": 0.00044958962775193195, + "loss": 2.828691864013672, + "step": 42730 + }, + { + "epoch": 0.2137, + "grad_norm": 0.3621795177459717, + "learning_rate": 0.000449565738725539, + "loss": 2.7849897384643554, + "step": 42740 + }, + { + "epoch": 0.21375, + "grad_norm": 0.3504452407360077, + "learning_rate": 0.00044954184467507985, + "loss": 2.865599822998047, + "step": 42750 + }, + { + "epoch": 0.2138, + "grad_norm": 0.3457004129886627, + "learning_rate": 0.0004495179456011561, + "loss": 2.8329389572143553, + "step": 42760 + }, + { + "epoch": 0.21385, + "grad_norm": 0.33272773027420044, + "learning_rate": 0.00044949404150436943, + "loss": 2.8847312927246094, + "step": 42770 + }, + { + "epoch": 0.2139, + "grad_norm": 0.3669862449169159, + "learning_rate": 0.00044947013238532166, + "loss": 2.8834632873535155, + "step": 42780 + }, + { + "epoch": 0.21395, + "grad_norm": 0.33879587054252625, + "learning_rate": 0.00044944621824461454, + "loss": 2.8021171569824217, + "step": 42790 + }, + { + "epoch": 0.214, + "grad_norm": 0.3496924638748169, + "learning_rate": 0.0004494222990828503, + "loss": 2.875373649597168, + "step": 42800 + }, + { + "epoch": 0.21405, + "grad_norm": 0.39702072739601135, + "learning_rate": 0.00044939837490063093, + "loss": 2.789503288269043, + "step": 42810 + }, + { + "epoch": 0.2141, + "grad_norm": 0.38151654601097107, + "learning_rate": 0.0004493744456985588, + "loss": 2.8108686447143554, + "step": 42820 + }, + { + "epoch": 0.21415, + "grad_norm": 0.32415667176246643, + "learning_rate": 0.0004493505114772364, + "loss": 2.82027587890625, + "step": 42830 + }, + { + "epoch": 0.2142, + "grad_norm": 0.3471766710281372, + "learning_rate": 0.00044932657223726614, + "loss": 2.808719444274902, + "step": 42840 + }, + { + "epoch": 0.21425, + "grad_norm": 0.33629995584487915, + "learning_rate": 0.0004493026279792508, + "loss": 2.8422508239746094, + "step": 42850 + }, + { + "epoch": 0.2143, + "grad_norm": 0.3613933026790619, + "learning_rate": 0.00044927867870379315, + "loss": 2.819402885437012, + "step": 42860 + }, + { + "epoch": 0.21435, + "grad_norm": 0.3391580581665039, + "learning_rate": 0.00044925472441149607, + "loss": 2.8605535507202147, + "step": 42870 + }, + { + "epoch": 0.2144, + "grad_norm": 0.32444292306900024, + "learning_rate": 0.00044923076510296255, + "loss": 2.8535388946533202, + "step": 42880 + }, + { + "epoch": 0.21445, + "grad_norm": 0.3505062758922577, + "learning_rate": 0.00044920680077879593, + "loss": 2.826309394836426, + "step": 42890 + }, + { + "epoch": 0.2145, + "grad_norm": 0.33672457933425903, + "learning_rate": 0.00044918283143959947, + "loss": 2.8201578140258787, + "step": 42900 + }, + { + "epoch": 0.21455, + "grad_norm": 0.3352976441383362, + "learning_rate": 0.00044915885708597653, + "loss": 2.8053915023803713, + "step": 42910 + }, + { + "epoch": 0.2146, + "grad_norm": 0.3319903314113617, + "learning_rate": 0.0004491348777185307, + "loss": 2.817917823791504, + "step": 42920 + }, + { + "epoch": 0.21465, + "grad_norm": 0.38007205724716187, + "learning_rate": 0.0004491108933378656, + "loss": 2.795395278930664, + "step": 42930 + }, + { + "epoch": 0.2147, + "grad_norm": 0.36327093839645386, + "learning_rate": 0.00044908690394458516, + "loss": 2.7854372024536134, + "step": 42940 + }, + { + "epoch": 0.21475, + "grad_norm": 0.31029176712036133, + "learning_rate": 0.0004490629095392932, + "loss": 2.825788879394531, + "step": 42950 + }, + { + "epoch": 0.2148, + "grad_norm": 0.38100868463516235, + "learning_rate": 0.00044903891012259385, + "loss": 2.845016098022461, + "step": 42960 + }, + { + "epoch": 0.21485, + "grad_norm": 0.3780277371406555, + "learning_rate": 0.0004490149056950912, + "loss": 2.7889453887939455, + "step": 42970 + }, + { + "epoch": 0.2149, + "grad_norm": 0.38527655601501465, + "learning_rate": 0.0004489908962573898, + "loss": 2.768167495727539, + "step": 42980 + }, + { + "epoch": 0.21495, + "grad_norm": 0.34594112634658813, + "learning_rate": 0.0004489668818100937, + "loss": 2.880566215515137, + "step": 42990 + }, + { + "epoch": 0.215, + "grad_norm": 0.3689344823360443, + "learning_rate": 0.00044894286235380775, + "loss": 2.8219200134277345, + "step": 43000 + }, + { + "epoch": 0.21505, + "grad_norm": 0.35747575759887695, + "learning_rate": 0.0004489188378891367, + "loss": 2.9166465759277345, + "step": 43010 + }, + { + "epoch": 0.2151, + "grad_norm": 0.3370111882686615, + "learning_rate": 0.0004488948084166851, + "loss": 2.788401794433594, + "step": 43020 + }, + { + "epoch": 0.21515, + "grad_norm": 0.32168713212013245, + "learning_rate": 0.000448870773937058, + "loss": 2.81167049407959, + "step": 43030 + }, + { + "epoch": 0.2152, + "grad_norm": 0.34293627738952637, + "learning_rate": 0.0004488467344508606, + "loss": 2.8034988403320313, + "step": 43040 + }, + { + "epoch": 0.21525, + "grad_norm": 0.34610456228256226, + "learning_rate": 0.0004488226899586979, + "loss": 2.787510299682617, + "step": 43050 + }, + { + "epoch": 0.2153, + "grad_norm": 0.3492356240749359, + "learning_rate": 0.00044879864046117534, + "loss": 2.843313217163086, + "step": 43060 + }, + { + "epoch": 0.21535, + "grad_norm": 0.34635522961616516, + "learning_rate": 0.0004487745859588984, + "loss": 2.8398410797119142, + "step": 43070 + }, + { + "epoch": 0.2154, + "grad_norm": 0.35296353697776794, + "learning_rate": 0.0004487505264524725, + "loss": 2.812809371948242, + "step": 43080 + }, + { + "epoch": 0.21545, + "grad_norm": 0.34774041175842285, + "learning_rate": 0.0004487264619425035, + "loss": 2.8689804077148438, + "step": 43090 + }, + { + "epoch": 0.2155, + "grad_norm": 0.35256367921829224, + "learning_rate": 0.0004487023924295971, + "loss": 2.9152042388916017, + "step": 43100 + }, + { + "epoch": 0.21555, + "grad_norm": 0.33709102869033813, + "learning_rate": 0.0004486783179143593, + "loss": 2.831477165222168, + "step": 43110 + }, + { + "epoch": 0.2156, + "grad_norm": 0.31961897015571594, + "learning_rate": 0.00044865423839739616, + "loss": 2.8016193389892576, + "step": 43120 + }, + { + "epoch": 0.21565, + "grad_norm": 0.3936953842639923, + "learning_rate": 0.00044863015387931396, + "loss": 2.856819725036621, + "step": 43130 + }, + { + "epoch": 0.2157, + "grad_norm": 0.34363797307014465, + "learning_rate": 0.00044860606436071895, + "loss": 2.8317329406738283, + "step": 43140 + }, + { + "epoch": 0.21575, + "grad_norm": 0.32562240958213806, + "learning_rate": 0.00044858196984221753, + "loss": 2.8446735382080077, + "step": 43150 + }, + { + "epoch": 0.2158, + "grad_norm": 0.33166825771331787, + "learning_rate": 0.00044855787032441636, + "loss": 2.880878448486328, + "step": 43160 + }, + { + "epoch": 0.21585, + "grad_norm": 0.32446879148483276, + "learning_rate": 0.00044853376580792216, + "loss": 2.824825668334961, + "step": 43170 + }, + { + "epoch": 0.2159, + "grad_norm": 0.36492395401000977, + "learning_rate": 0.0004485096562933417, + "loss": 2.8487234115600586, + "step": 43180 + }, + { + "epoch": 0.21595, + "grad_norm": 0.3551696538925171, + "learning_rate": 0.00044848554178128204, + "loss": 2.8306482315063475, + "step": 43190 + }, + { + "epoch": 0.216, + "grad_norm": 0.3533342182636261, + "learning_rate": 0.0004484614222723501, + "loss": 2.847543716430664, + "step": 43200 + }, + { + "epoch": 0.21605, + "grad_norm": 0.3289647102355957, + "learning_rate": 0.00044843729776715326, + "loss": 2.8298133850097655, + "step": 43210 + }, + { + "epoch": 0.2161, + "grad_norm": 0.3350399434566498, + "learning_rate": 0.0004484131682662987, + "loss": 2.857748603820801, + "step": 43220 + }, + { + "epoch": 0.21615, + "grad_norm": 0.3327401280403137, + "learning_rate": 0.000448389033770394, + "loss": 2.7943147659301757, + "step": 43230 + }, + { + "epoch": 0.2162, + "grad_norm": 0.3650300204753876, + "learning_rate": 0.00044836489428004675, + "loss": 2.8266574859619142, + "step": 43240 + }, + { + "epoch": 0.21625, + "grad_norm": 0.35996147990226746, + "learning_rate": 0.00044834074979586454, + "loss": 2.8139904022216795, + "step": 43250 + }, + { + "epoch": 0.2163, + "grad_norm": 0.41883185505867004, + "learning_rate": 0.00044831660031845535, + "loss": 2.8749549865722654, + "step": 43260 + }, + { + "epoch": 0.21635, + "grad_norm": 0.3439125716686249, + "learning_rate": 0.000448292445848427, + "loss": 2.8688039779663086, + "step": 43270 + }, + { + "epoch": 0.2164, + "grad_norm": 0.3477700352668762, + "learning_rate": 0.0004482682863863876, + "loss": 2.8501110076904297, + "step": 43280 + }, + { + "epoch": 0.21645, + "grad_norm": 0.3523405194282532, + "learning_rate": 0.0004482441219329455, + "loss": 2.8353755950927733, + "step": 43290 + }, + { + "epoch": 0.2165, + "grad_norm": 0.35361501574516296, + "learning_rate": 0.00044821995248870897, + "loss": 2.870342254638672, + "step": 43300 + }, + { + "epoch": 0.21655, + "grad_norm": 0.3578090965747833, + "learning_rate": 0.00044819577805428645, + "loss": 2.7811958312988283, + "step": 43310 + }, + { + "epoch": 0.2166, + "grad_norm": 0.38136622309684753, + "learning_rate": 0.0004481715986302865, + "loss": 2.861886215209961, + "step": 43320 + }, + { + "epoch": 0.21665, + "grad_norm": 0.33643680810928345, + "learning_rate": 0.0004481474142173179, + "loss": 2.8167110443115235, + "step": 43330 + }, + { + "epoch": 0.2167, + "grad_norm": 0.34441399574279785, + "learning_rate": 0.0004481232248159894, + "loss": 2.820071792602539, + "step": 43340 + }, + { + "epoch": 0.21675, + "grad_norm": 0.35139089822769165, + "learning_rate": 0.0004480990304269102, + "loss": 2.824982261657715, + "step": 43350 + }, + { + "epoch": 0.2168, + "grad_norm": 0.35724198818206787, + "learning_rate": 0.0004480748310506891, + "loss": 2.8610202789306642, + "step": 43360 + }, + { + "epoch": 0.21685, + "grad_norm": 0.3185963034629822, + "learning_rate": 0.0004480506266879355, + "loss": 2.8220237731933593, + "step": 43370 + }, + { + "epoch": 0.2169, + "grad_norm": 0.3472438156604767, + "learning_rate": 0.00044802641733925877, + "loss": 2.844528388977051, + "step": 43380 + }, + { + "epoch": 0.21695, + "grad_norm": 0.353866308927536, + "learning_rate": 0.00044800220300526817, + "loss": 2.8121089935302734, + "step": 43390 + }, + { + "epoch": 0.217, + "grad_norm": 0.34035417437553406, + "learning_rate": 0.0004479779836865735, + "loss": 2.8548179626464845, + "step": 43400 + }, + { + "epoch": 0.21705, + "grad_norm": 0.37453320622444153, + "learning_rate": 0.00044795375938378444, + "loss": 2.8245067596435547, + "step": 43410 + }, + { + "epoch": 0.2171, + "grad_norm": 0.3650299906730652, + "learning_rate": 0.00044792953009751077, + "loss": 2.84586238861084, + "step": 43420 + }, + { + "epoch": 0.21715, + "grad_norm": 0.4056735336780548, + "learning_rate": 0.0004479052958283625, + "loss": 2.9014816284179688, + "step": 43430 + }, + { + "epoch": 0.2172, + "grad_norm": 0.34845298528671265, + "learning_rate": 0.00044788105657694974, + "loss": 2.8649520874023438, + "step": 43440 + }, + { + "epoch": 0.21725, + "grad_norm": 0.3226182758808136, + "learning_rate": 0.00044785681234388265, + "loss": 2.8294681549072265, + "step": 43450 + }, + { + "epoch": 0.2173, + "grad_norm": 0.34070220589637756, + "learning_rate": 0.0004478325631297717, + "loss": 2.8384515762329103, + "step": 43460 + }, + { + "epoch": 0.21735, + "grad_norm": 0.3312918543815613, + "learning_rate": 0.0004478083089352273, + "loss": 2.8259422302246096, + "step": 43470 + }, + { + "epoch": 0.2174, + "grad_norm": 0.35334667563438416, + "learning_rate": 0.00044778404976086, + "loss": 2.8435497283935547, + "step": 43480 + }, + { + "epoch": 0.21745, + "grad_norm": 0.3589346408843994, + "learning_rate": 0.0004477597856072806, + "loss": 2.8302942276000977, + "step": 43490 + }, + { + "epoch": 0.2175, + "grad_norm": 0.36022287607192993, + "learning_rate": 0.00044773551647509993, + "loss": 2.7947061538696287, + "step": 43500 + }, + { + "epoch": 0.21755, + "grad_norm": 0.33927032351493835, + "learning_rate": 0.00044771124236492896, + "loss": 2.7911184310913084, + "step": 43510 + }, + { + "epoch": 0.2176, + "grad_norm": 0.3648351728916168, + "learning_rate": 0.00044768696327737874, + "loss": 2.852951431274414, + "step": 43520 + }, + { + "epoch": 0.21765, + "grad_norm": 0.32894280552864075, + "learning_rate": 0.00044766267921306056, + "loss": 2.8281270980834963, + "step": 43530 + }, + { + "epoch": 0.2177, + "grad_norm": 0.3501463830471039, + "learning_rate": 0.00044763839017258574, + "loss": 2.8339967727661133, + "step": 43540 + }, + { + "epoch": 0.21775, + "grad_norm": 0.314007043838501, + "learning_rate": 0.00044761409615656577, + "loss": 2.8120725631713865, + "step": 43550 + }, + { + "epoch": 0.2178, + "grad_norm": 0.37817370891571045, + "learning_rate": 0.00044758979716561225, + "loss": 2.8294628143310545, + "step": 43560 + }, + { + "epoch": 0.21785, + "grad_norm": 0.3371081054210663, + "learning_rate": 0.0004475654932003369, + "loss": 2.859519195556641, + "step": 43570 + }, + { + "epoch": 0.2179, + "grad_norm": 0.3626057803630829, + "learning_rate": 0.0004475411842613516, + "loss": 2.8760517120361326, + "step": 43580 + }, + { + "epoch": 0.21795, + "grad_norm": 0.321338027715683, + "learning_rate": 0.00044751687034926837, + "loss": 2.7973217010498046, + "step": 43590 + }, + { + "epoch": 0.218, + "grad_norm": 0.3584008812904358, + "learning_rate": 0.0004474925514646991, + "loss": 2.793400192260742, + "step": 43600 + }, + { + "epoch": 0.21805, + "grad_norm": 0.3300241231918335, + "learning_rate": 0.00044746822760825626, + "loss": 2.8005039215087892, + "step": 43610 + }, + { + "epoch": 0.2181, + "grad_norm": 0.37720987200737, + "learning_rate": 0.00044744389878055216, + "loss": 2.803839683532715, + "step": 43620 + }, + { + "epoch": 0.21815, + "grad_norm": 0.3543992042541504, + "learning_rate": 0.0004474195649821991, + "loss": 2.8511276245117188, + "step": 43630 + }, + { + "epoch": 0.2182, + "grad_norm": 0.33284565806388855, + "learning_rate": 0.0004473952262138099, + "loss": 2.830669403076172, + "step": 43640 + }, + { + "epoch": 0.21825, + "grad_norm": 0.3420531451702118, + "learning_rate": 0.00044737088247599716, + "loss": 2.850113868713379, + "step": 43650 + }, + { + "epoch": 0.2183, + "grad_norm": 0.3299869894981384, + "learning_rate": 0.0004473465337693738, + "loss": 2.8116504669189455, + "step": 43660 + }, + { + "epoch": 0.21835, + "grad_norm": 0.34160715341567993, + "learning_rate": 0.0004473221800945527, + "loss": 2.843792724609375, + "step": 43670 + }, + { + "epoch": 0.2184, + "grad_norm": 0.31565266847610474, + "learning_rate": 0.00044729782145214717, + "loss": 2.828138542175293, + "step": 43680 + }, + { + "epoch": 0.21845, + "grad_norm": 0.3379160761833191, + "learning_rate": 0.0004472734578427702, + "loss": 2.827197265625, + "step": 43690 + }, + { + "epoch": 0.2185, + "grad_norm": 0.35716405510902405, + "learning_rate": 0.0004472490892670352, + "loss": 2.790904998779297, + "step": 43700 + }, + { + "epoch": 0.21855, + "grad_norm": 0.3651374876499176, + "learning_rate": 0.00044722471572555583, + "loss": 2.8619075775146485, + "step": 43710 + }, + { + "epoch": 0.2186, + "grad_norm": 0.3928045630455017, + "learning_rate": 0.0004472003372189455, + "loss": 2.8384294509887695, + "step": 43720 + }, + { + "epoch": 0.21865, + "grad_norm": 0.3558192849159241, + "learning_rate": 0.00044717595374781793, + "loss": 2.81514949798584, + "step": 43730 + }, + { + "epoch": 0.2187, + "grad_norm": 0.33183395862579346, + "learning_rate": 0.0004471515653127871, + "loss": 2.818577003479004, + "step": 43740 + }, + { + "epoch": 0.21875, + "grad_norm": 0.35640230774879456, + "learning_rate": 0.000447127171914467, + "loss": 2.8188438415527344, + "step": 43750 + }, + { + "epoch": 0.2188, + "grad_norm": 0.3777505159378052, + "learning_rate": 0.0004471027735534716, + "loss": 2.8800010681152344, + "step": 43760 + }, + { + "epoch": 0.21885, + "grad_norm": 0.3445523679256439, + "learning_rate": 0.0004470783702304152, + "loss": 2.8456111907958985, + "step": 43770 + }, + { + "epoch": 0.2189, + "grad_norm": 0.35029059648513794, + "learning_rate": 0.00044705396194591213, + "loss": 2.8845027923583983, + "step": 43780 + }, + { + "epoch": 0.21895, + "grad_norm": 0.3500763177871704, + "learning_rate": 0.0004470295487005769, + "loss": 2.7872234344482423, + "step": 43790 + }, + { + "epoch": 0.219, + "grad_norm": 0.3459974229335785, + "learning_rate": 0.0004470051304950241, + "loss": 2.7718246459960936, + "step": 43800 + }, + { + "epoch": 0.21905, + "grad_norm": 0.33904197812080383, + "learning_rate": 0.0004469807073298684, + "loss": 2.806195831298828, + "step": 43810 + }, + { + "epoch": 0.2191, + "grad_norm": 0.3536299169063568, + "learning_rate": 0.00044695627920572476, + "loss": 2.8576725006103514, + "step": 43820 + }, + { + "epoch": 0.21915, + "grad_norm": 0.3922496736049652, + "learning_rate": 0.00044693184612320814, + "loss": 2.7854101181030275, + "step": 43830 + }, + { + "epoch": 0.2192, + "grad_norm": 0.3788086473941803, + "learning_rate": 0.00044690740808293357, + "loss": 2.8231639862060547, + "step": 43840 + }, + { + "epoch": 0.21925, + "grad_norm": 0.3441082537174225, + "learning_rate": 0.0004468829650855163, + "loss": 2.781816291809082, + "step": 43850 + }, + { + "epoch": 0.2193, + "grad_norm": 0.38580700755119324, + "learning_rate": 0.0004468585171315718, + "loss": 2.8541633605957033, + "step": 43860 + }, + { + "epoch": 0.21935, + "grad_norm": 0.33409586548805237, + "learning_rate": 0.00044683406422171536, + "loss": 2.8781280517578125, + "step": 43870 + }, + { + "epoch": 0.2194, + "grad_norm": 0.35280120372772217, + "learning_rate": 0.00044680960635656267, + "loss": 2.773513603210449, + "step": 43880 + }, + { + "epoch": 0.21945, + "grad_norm": 0.3788904845714569, + "learning_rate": 0.0004467851435367294, + "loss": 2.8020111083984376, + "step": 43890 + }, + { + "epoch": 0.2195, + "grad_norm": 0.3242301046848297, + "learning_rate": 0.00044676067576283153, + "loss": 2.854559898376465, + "step": 43900 + }, + { + "epoch": 0.21955, + "grad_norm": 0.3237163722515106, + "learning_rate": 0.00044673620303548495, + "loss": 2.8421005249023437, + "step": 43910 + }, + { + "epoch": 0.2196, + "grad_norm": 0.3315355181694031, + "learning_rate": 0.0004467117253553058, + "loss": 2.809035873413086, + "step": 43920 + }, + { + "epoch": 0.21965, + "grad_norm": 0.3226226568222046, + "learning_rate": 0.0004466872427229102, + "loss": 2.86633186340332, + "step": 43930 + }, + { + "epoch": 0.2197, + "grad_norm": 0.31547972559928894, + "learning_rate": 0.00044666275513891463, + "loss": 2.8583744049072264, + "step": 43940 + }, + { + "epoch": 0.21975, + "grad_norm": 0.3628997504711151, + "learning_rate": 0.0004466382626039356, + "loss": 2.8776067733764648, + "step": 43950 + }, + { + "epoch": 0.2198, + "grad_norm": 0.38782069087028503, + "learning_rate": 0.00044661376511858947, + "loss": 2.8459402084350587, + "step": 43960 + }, + { + "epoch": 0.21985, + "grad_norm": 0.3475775122642517, + "learning_rate": 0.0004465892626834932, + "loss": 2.8403045654296877, + "step": 43970 + }, + { + "epoch": 0.2199, + "grad_norm": 0.3563113510608673, + "learning_rate": 0.00044656475529926355, + "loss": 2.857710838317871, + "step": 43980 + }, + { + "epoch": 0.21995, + "grad_norm": 0.3334925174713135, + "learning_rate": 0.0004465402429665175, + "loss": 2.8523513793945314, + "step": 43990 + }, + { + "epoch": 0.22, + "grad_norm": 0.35552749037742615, + "learning_rate": 0.0004465157256858721, + "loss": 2.759809684753418, + "step": 44000 + }, + { + "epoch": 0.22005, + "grad_norm": 0.3714665472507477, + "learning_rate": 0.0004464912034579447, + "loss": 2.7991235733032225, + "step": 44010 + }, + { + "epoch": 0.2201, + "grad_norm": 0.3315805494785309, + "learning_rate": 0.00044646667628335247, + "loss": 2.840413475036621, + "step": 44020 + }, + { + "epoch": 0.22015, + "grad_norm": 0.33001142740249634, + "learning_rate": 0.00044644214416271304, + "loss": 2.8132213592529296, + "step": 44030 + }, + { + "epoch": 0.2202, + "grad_norm": 0.33635228872299194, + "learning_rate": 0.0004464176070966439, + "loss": 2.849736785888672, + "step": 44040 + }, + { + "epoch": 0.22025, + "grad_norm": 0.35710451006889343, + "learning_rate": 0.0004463930650857629, + "loss": 2.868490219116211, + "step": 44050 + }, + { + "epoch": 0.2203, + "grad_norm": 0.3404357433319092, + "learning_rate": 0.00044636851813068776, + "loss": 2.8821432113647463, + "step": 44060 + }, + { + "epoch": 0.22035, + "grad_norm": 0.3288133144378662, + "learning_rate": 0.0004463439662320364, + "loss": 2.838816833496094, + "step": 44070 + }, + { + "epoch": 0.2204, + "grad_norm": 0.3717919588088989, + "learning_rate": 0.00044631940939042714, + "loss": 2.848206329345703, + "step": 44080 + }, + { + "epoch": 0.22045, + "grad_norm": 0.3964031934738159, + "learning_rate": 0.00044629484760647797, + "loss": 2.853707695007324, + "step": 44090 + }, + { + "epoch": 0.2205, + "grad_norm": 0.3308519721031189, + "learning_rate": 0.00044627028088080735, + "loss": 2.7912769317626953, + "step": 44100 + }, + { + "epoch": 0.22055, + "grad_norm": 0.39173340797424316, + "learning_rate": 0.00044624570921403375, + "loss": 2.8617725372314453, + "step": 44110 + }, + { + "epoch": 0.2206, + "grad_norm": 0.3510302007198334, + "learning_rate": 0.00044622113260677567, + "loss": 2.8477109909057616, + "step": 44120 + }, + { + "epoch": 0.22065, + "grad_norm": 0.34022071957588196, + "learning_rate": 0.0004461965510596519, + "loss": 2.8740713119506838, + "step": 44130 + }, + { + "epoch": 0.2207, + "grad_norm": 0.347238689661026, + "learning_rate": 0.0004461719645732813, + "loss": 2.8712772369384765, + "step": 44140 + }, + { + "epoch": 0.22075, + "grad_norm": 0.37571120262145996, + "learning_rate": 0.0004461473731482828, + "loss": 2.878373908996582, + "step": 44150 + }, + { + "epoch": 0.2208, + "grad_norm": 0.35659852623939514, + "learning_rate": 0.00044612277678527545, + "loss": 2.789261817932129, + "step": 44160 + }, + { + "epoch": 0.22085, + "grad_norm": 0.36797428131103516, + "learning_rate": 0.0004460981754848785, + "loss": 2.841922378540039, + "step": 44170 + }, + { + "epoch": 0.2209, + "grad_norm": 0.3665328919887543, + "learning_rate": 0.00044607356924771137, + "loss": 2.827409553527832, + "step": 44180 + }, + { + "epoch": 0.22095, + "grad_norm": 0.338848739862442, + "learning_rate": 0.00044604895807439336, + "loss": 2.8520795822143556, + "step": 44190 + }, + { + "epoch": 0.221, + "grad_norm": 0.36393094062805176, + "learning_rate": 0.0004460243419655442, + "loss": 2.8211185455322267, + "step": 44200 + }, + { + "epoch": 0.22105, + "grad_norm": 0.32504668831825256, + "learning_rate": 0.0004459997209217835, + "loss": 2.8572158813476562, + "step": 44210 + }, + { + "epoch": 0.2211, + "grad_norm": 0.3347141742706299, + "learning_rate": 0.00044597509494373125, + "loss": 2.831799125671387, + "step": 44220 + }, + { + "epoch": 0.22115, + "grad_norm": 0.32078924775123596, + "learning_rate": 0.0004459504640320071, + "loss": 2.8410776138305662, + "step": 44230 + }, + { + "epoch": 0.2212, + "grad_norm": 0.35653796792030334, + "learning_rate": 0.0004459258281872315, + "loss": 2.8515682220458984, + "step": 44240 + }, + { + "epoch": 0.22125, + "grad_norm": 0.33930814266204834, + "learning_rate": 0.0004459011874100244, + "loss": 2.8242424011230467, + "step": 44250 + }, + { + "epoch": 0.2213, + "grad_norm": 0.31713706254959106, + "learning_rate": 0.0004458765417010062, + "loss": 2.799503707885742, + "step": 44260 + }, + { + "epoch": 0.22135, + "grad_norm": 0.35536450147628784, + "learning_rate": 0.00044585189106079745, + "loss": 2.8361812591552735, + "step": 44270 + }, + { + "epoch": 0.2214, + "grad_norm": 0.3306514620780945, + "learning_rate": 0.0004458272354900186, + "loss": 2.8455841064453127, + "step": 44280 + }, + { + "epoch": 0.22145, + "grad_norm": 0.34107959270477295, + "learning_rate": 0.00044580257498929036, + "loss": 2.813726043701172, + "step": 44290 + }, + { + "epoch": 0.2215, + "grad_norm": 0.3621120750904083, + "learning_rate": 0.0004457779095592337, + "loss": 2.8198455810546874, + "step": 44300 + }, + { + "epoch": 0.22155, + "grad_norm": 0.3472643494606018, + "learning_rate": 0.0004457532392004694, + "loss": 2.8560489654541015, + "step": 44310 + }, + { + "epoch": 0.2216, + "grad_norm": 0.32470211386680603, + "learning_rate": 0.00044572856391361865, + "loss": 2.853778076171875, + "step": 44320 + }, + { + "epoch": 0.22165, + "grad_norm": 0.3354562520980835, + "learning_rate": 0.0004457038836993026, + "loss": 2.8107332229614257, + "step": 44330 + }, + { + "epoch": 0.2217, + "grad_norm": 0.37489885091781616, + "learning_rate": 0.00044567919855814257, + "loss": 2.841092109680176, + "step": 44340 + }, + { + "epoch": 0.22175, + "grad_norm": 0.3487268388271332, + "learning_rate": 0.00044565450849076005, + "loss": 2.823917007446289, + "step": 44350 + }, + { + "epoch": 0.2218, + "grad_norm": 0.3404950499534607, + "learning_rate": 0.0004456298134977766, + "loss": 2.829069709777832, + "step": 44360 + }, + { + "epoch": 0.22185, + "grad_norm": 0.32043465971946716, + "learning_rate": 0.00044560511357981386, + "loss": 2.8230224609375, + "step": 44370 + }, + { + "epoch": 0.2219, + "grad_norm": 0.3704977035522461, + "learning_rate": 0.0004455804087374937, + "loss": 2.872623825073242, + "step": 44380 + }, + { + "epoch": 0.22195, + "grad_norm": 0.361054390668869, + "learning_rate": 0.0004455556989714381, + "loss": 2.809219741821289, + "step": 44390 + }, + { + "epoch": 0.222, + "grad_norm": 0.37762710452079773, + "learning_rate": 0.000445530984282269, + "loss": 2.8068307876586913, + "step": 44400 + }, + { + "epoch": 0.22205, + "grad_norm": 0.3567740321159363, + "learning_rate": 0.00044550626467060875, + "loss": 2.874903106689453, + "step": 44410 + }, + { + "epoch": 0.2221, + "grad_norm": 0.3491511940956116, + "learning_rate": 0.00044548154013707955, + "loss": 2.8373016357421874, + "step": 44420 + }, + { + "epoch": 0.22215, + "grad_norm": 0.3216747045516968, + "learning_rate": 0.00044545681068230395, + "loss": 2.7823745727539064, + "step": 44430 + }, + { + "epoch": 0.2222, + "grad_norm": 0.332644522190094, + "learning_rate": 0.0004454320763069044, + "loss": 2.8344036102294923, + "step": 44440 + }, + { + "epoch": 0.22225, + "grad_norm": 0.3414577841758728, + "learning_rate": 0.0004454073370115036, + "loss": 2.835733413696289, + "step": 44450 + }, + { + "epoch": 0.2223, + "grad_norm": 0.33827507495880127, + "learning_rate": 0.00044538259279672444, + "loss": 2.84066162109375, + "step": 44460 + }, + { + "epoch": 0.22235, + "grad_norm": 0.31461355090141296, + "learning_rate": 0.00044535784366318983, + "loss": 2.819081115722656, + "step": 44470 + }, + { + "epoch": 0.2224, + "grad_norm": 0.366453617811203, + "learning_rate": 0.00044533308961152275, + "loss": 2.8385387420654298, + "step": 44480 + }, + { + "epoch": 0.22245, + "grad_norm": 0.3383362889289856, + "learning_rate": 0.0004453083306423465, + "loss": 2.8293590545654297, + "step": 44490 + }, + { + "epoch": 0.2225, + "grad_norm": 0.41480451822280884, + "learning_rate": 0.00044528356675628436, + "loss": 2.814487648010254, + "step": 44500 + }, + { + "epoch": 0.22255, + "grad_norm": 0.34461864829063416, + "learning_rate": 0.0004452587979539596, + "loss": 2.8643985748291017, + "step": 44510 + }, + { + "epoch": 0.2226, + "grad_norm": 0.3341062068939209, + "learning_rate": 0.000445234024235996, + "loss": 2.8424970626831056, + "step": 44520 + }, + { + "epoch": 0.22265, + "grad_norm": 0.33653005957603455, + "learning_rate": 0.00044520924560301716, + "loss": 2.839443397521973, + "step": 44530 + }, + { + "epoch": 0.2227, + "grad_norm": 0.3607802093029022, + "learning_rate": 0.00044518446205564683, + "loss": 2.801011657714844, + "step": 44540 + }, + { + "epoch": 0.22275, + "grad_norm": 0.33390575647354126, + "learning_rate": 0.00044515967359450904, + "loss": 2.8100141525268554, + "step": 44550 + }, + { + "epoch": 0.2228, + "grad_norm": 0.35691213607788086, + "learning_rate": 0.0004451348802202276, + "loss": 2.8535533905029298, + "step": 44560 + }, + { + "epoch": 0.22285, + "grad_norm": 0.3584910035133362, + "learning_rate": 0.0004451100819334271, + "loss": 2.830525588989258, + "step": 44570 + }, + { + "epoch": 0.2229, + "grad_norm": 0.3384004533290863, + "learning_rate": 0.0004450852787347314, + "loss": 2.835072708129883, + "step": 44580 + }, + { + "epoch": 0.22295, + "grad_norm": 0.3536185920238495, + "learning_rate": 0.0004450604706247652, + "loss": 2.859430694580078, + "step": 44590 + }, + { + "epoch": 0.223, + "grad_norm": 0.33209750056266785, + "learning_rate": 0.00044503565760415286, + "loss": 2.8179174423217774, + "step": 44600 + }, + { + "epoch": 0.22305, + "grad_norm": 0.3371348977088928, + "learning_rate": 0.0004450108396735192, + "loss": 2.8307729721069337, + "step": 44610 + }, + { + "epoch": 0.2231, + "grad_norm": 0.3242034316062927, + "learning_rate": 0.000444986016833489, + "loss": 2.7870599746704103, + "step": 44620 + }, + { + "epoch": 0.22315, + "grad_norm": 0.3495365381240845, + "learning_rate": 0.0004449611890846871, + "loss": 2.8714290618896485, + "step": 44630 + }, + { + "epoch": 0.2232, + "grad_norm": 0.3547069728374481, + "learning_rate": 0.00044493635642773856, + "loss": 2.825870704650879, + "step": 44640 + }, + { + "epoch": 0.22325, + "grad_norm": 0.33796215057373047, + "learning_rate": 0.0004449115188632685, + "loss": 2.8460607528686523, + "step": 44650 + }, + { + "epoch": 0.2233, + "grad_norm": 0.3598385453224182, + "learning_rate": 0.00044488667639190236, + "loss": 2.825553321838379, + "step": 44660 + }, + { + "epoch": 0.22335, + "grad_norm": 0.3236472010612488, + "learning_rate": 0.0004448618290142654, + "loss": 2.789310646057129, + "step": 44670 + }, + { + "epoch": 0.2234, + "grad_norm": 0.34543290734291077, + "learning_rate": 0.00044483697673098314, + "loss": 2.8024150848388674, + "step": 44680 + }, + { + "epoch": 0.22345, + "grad_norm": 0.3790661096572876, + "learning_rate": 0.00044481211954268136, + "loss": 2.7941616058349608, + "step": 44690 + }, + { + "epoch": 0.2235, + "grad_norm": 0.3455739915370941, + "learning_rate": 0.0004447872574499858, + "loss": 2.834544372558594, + "step": 44700 + }, + { + "epoch": 0.22355, + "grad_norm": 0.3562193810939789, + "learning_rate": 0.00044476239045352224, + "loss": 2.871090316772461, + "step": 44710 + }, + { + "epoch": 0.2236, + "grad_norm": 0.33848997950553894, + "learning_rate": 0.0004447375185539169, + "loss": 2.8608192443847655, + "step": 44720 + }, + { + "epoch": 0.22365, + "grad_norm": 0.3345803916454315, + "learning_rate": 0.0004447126417517957, + "loss": 2.826051139831543, + "step": 44730 + }, + { + "epoch": 0.2237, + "grad_norm": 0.3600272238254547, + "learning_rate": 0.00044468776004778516, + "loss": 2.8746246337890624, + "step": 44740 + }, + { + "epoch": 0.22375, + "grad_norm": 0.3640405535697937, + "learning_rate": 0.00044466287344251153, + "loss": 2.8549692153930666, + "step": 44750 + }, + { + "epoch": 0.2238, + "grad_norm": 0.31481942534446716, + "learning_rate": 0.0004446379819366013, + "loss": 2.848249816894531, + "step": 44760 + }, + { + "epoch": 0.22385, + "grad_norm": 0.36929965019226074, + "learning_rate": 0.00044461308553068123, + "loss": 2.7930601119995115, + "step": 44770 + }, + { + "epoch": 0.2239, + "grad_norm": 0.34616705775260925, + "learning_rate": 0.00044458818422537806, + "loss": 2.840237617492676, + "step": 44780 + }, + { + "epoch": 0.22395, + "grad_norm": 0.37350088357925415, + "learning_rate": 0.0004445632780213187, + "loss": 2.8731727600097656, + "step": 44790 + }, + { + "epoch": 0.224, + "grad_norm": 0.32364651560783386, + "learning_rate": 0.00044453836691912997, + "loss": 2.860236930847168, + "step": 44800 + }, + { + "epoch": 0.22405, + "grad_norm": 0.3426031470298767, + "learning_rate": 0.00044451345091943927, + "loss": 2.853343963623047, + "step": 44810 + }, + { + "epoch": 0.2241, + "grad_norm": 0.37072959542274475, + "learning_rate": 0.0004444885300228737, + "loss": 2.856364440917969, + "step": 44820 + }, + { + "epoch": 0.22415, + "grad_norm": 0.33013230562210083, + "learning_rate": 0.00044446360423006066, + "loss": 2.8350358963012696, + "step": 44830 + }, + { + "epoch": 0.2242, + "grad_norm": 0.3275122344493866, + "learning_rate": 0.0004444386735416277, + "loss": 2.774854278564453, + "step": 44840 + }, + { + "epoch": 0.22425, + "grad_norm": 0.34939876198768616, + "learning_rate": 0.00044441373795820247, + "loss": 2.7988773345947267, + "step": 44850 + }, + { + "epoch": 0.2243, + "grad_norm": 0.3417316675186157, + "learning_rate": 0.0004443887974804126, + "loss": 2.8050559997558593, + "step": 44860 + }, + { + "epoch": 0.22435, + "grad_norm": 0.3188377618789673, + "learning_rate": 0.00044436385210888614, + "loss": 2.8231857299804686, + "step": 44870 + }, + { + "epoch": 0.2244, + "grad_norm": 0.37657973170280457, + "learning_rate": 0.000444338901844251, + "loss": 2.821158218383789, + "step": 44880 + }, + { + "epoch": 0.22445, + "grad_norm": 0.3511180579662323, + "learning_rate": 0.00044431394668713526, + "loss": 2.820878028869629, + "step": 44890 + }, + { + "epoch": 0.2245, + "grad_norm": 0.35811588168144226, + "learning_rate": 0.0004442889866381672, + "loss": 2.8364933013916014, + "step": 44900 + }, + { + "epoch": 0.22455, + "grad_norm": 0.3652416467666626, + "learning_rate": 0.00044426402169797527, + "loss": 2.860555076599121, + "step": 44910 + }, + { + "epoch": 0.2246, + "grad_norm": 0.3201248347759247, + "learning_rate": 0.0004442390518671878, + "loss": 2.8222579956054688, + "step": 44920 + }, + { + "epoch": 0.22465, + "grad_norm": 0.3870462477207184, + "learning_rate": 0.00044421407714643357, + "loss": 2.832754898071289, + "step": 44930 + }, + { + "epoch": 0.2247, + "grad_norm": 0.33719417452812195, + "learning_rate": 0.0004441890975363412, + "loss": 2.798126983642578, + "step": 44940 + }, + { + "epoch": 0.22475, + "grad_norm": 0.3648149371147156, + "learning_rate": 0.0004441641130375396, + "loss": 2.849976921081543, + "step": 44950 + }, + { + "epoch": 0.2248, + "grad_norm": 0.36008182168006897, + "learning_rate": 0.00044413912365065777, + "loss": 2.8792255401611326, + "step": 44960 + }, + { + "epoch": 0.22485, + "grad_norm": 0.39953017234802246, + "learning_rate": 0.0004441141293763248, + "loss": 2.8141590118408204, + "step": 44970 + }, + { + "epoch": 0.2249, + "grad_norm": 0.330330491065979, + "learning_rate": 0.00044408913021516997, + "loss": 2.852842330932617, + "step": 44980 + }, + { + "epoch": 0.22495, + "grad_norm": 0.35093802213668823, + "learning_rate": 0.00044406412616782254, + "loss": 2.826187515258789, + "step": 44990 + }, + { + "epoch": 0.225, + "grad_norm": 0.3344043791294098, + "learning_rate": 0.00044403911723491196, + "loss": 2.827589416503906, + "step": 45000 + }, + { + "epoch": 0.22505, + "grad_norm": 0.3304402232170105, + "learning_rate": 0.000444014103417068, + "loss": 2.807358741760254, + "step": 45010 + }, + { + "epoch": 0.2251, + "grad_norm": 0.35278603434562683, + "learning_rate": 0.0004439890847149203, + "loss": 2.8263038635253905, + "step": 45020 + }, + { + "epoch": 0.22515, + "grad_norm": 0.39286333322525024, + "learning_rate": 0.0004439640611290986, + "loss": 2.805240440368652, + "step": 45030 + }, + { + "epoch": 0.2252, + "grad_norm": 0.36931589245796204, + "learning_rate": 0.00044393903266023294, + "loss": 2.8773136138916016, + "step": 45040 + }, + { + "epoch": 0.22525, + "grad_norm": 0.37956544756889343, + "learning_rate": 0.0004439139993089535, + "loss": 2.8371496200561523, + "step": 45050 + }, + { + "epoch": 0.2253, + "grad_norm": 0.33329692482948303, + "learning_rate": 0.0004438889610758904, + "loss": 2.886583709716797, + "step": 45060 + }, + { + "epoch": 0.22535, + "grad_norm": 0.338974267244339, + "learning_rate": 0.00044386391796167403, + "loss": 2.833180236816406, + "step": 45070 + }, + { + "epoch": 0.2254, + "grad_norm": 0.34519755840301514, + "learning_rate": 0.0004438388699669348, + "loss": 2.791570472717285, + "step": 45080 + }, + { + "epoch": 0.22545, + "grad_norm": 0.3608878254890442, + "learning_rate": 0.0004438138170923033, + "loss": 2.85098819732666, + "step": 45090 + }, + { + "epoch": 0.2255, + "grad_norm": 0.3291192054748535, + "learning_rate": 0.0004437887593384102, + "loss": 2.825252151489258, + "step": 45100 + }, + { + "epoch": 0.22555, + "grad_norm": 0.3210698366165161, + "learning_rate": 0.00044376369670588646, + "loss": 2.8711681365966797, + "step": 45110 + }, + { + "epoch": 0.2256, + "grad_norm": 0.3699954152107239, + "learning_rate": 0.00044373862919536297, + "loss": 2.853860855102539, + "step": 45120 + }, + { + "epoch": 0.22565, + "grad_norm": 0.36290913820266724, + "learning_rate": 0.0004437135568074707, + "loss": 2.863733673095703, + "step": 45130 + }, + { + "epoch": 0.2257, + "grad_norm": 0.32835012674331665, + "learning_rate": 0.00044368847954284095, + "loss": 2.8561792373657227, + "step": 45140 + }, + { + "epoch": 0.22575, + "grad_norm": 0.3261044919490814, + "learning_rate": 0.0004436633974021051, + "loss": 2.821111297607422, + "step": 45150 + }, + { + "epoch": 0.2258, + "grad_norm": 0.3479663133621216, + "learning_rate": 0.0004436383103858944, + "loss": 2.826845169067383, + "step": 45160 + }, + { + "epoch": 0.22585, + "grad_norm": 0.40625154972076416, + "learning_rate": 0.0004436132184948406, + "loss": 2.8202978134155274, + "step": 45170 + }, + { + "epoch": 0.2259, + "grad_norm": 0.32904645800590515, + "learning_rate": 0.0004435881217295753, + "loss": 2.8009723663330077, + "step": 45180 + }, + { + "epoch": 0.22595, + "grad_norm": 0.31126344203948975, + "learning_rate": 0.00044356302009073036, + "loss": 2.8511983871459963, + "step": 45190 + }, + { + "epoch": 0.226, + "grad_norm": 0.3265398442745209, + "learning_rate": 0.0004435379135789377, + "loss": 2.8335926055908205, + "step": 45200 + }, + { + "epoch": 0.22605, + "grad_norm": 0.35922449827194214, + "learning_rate": 0.0004435128021948293, + "loss": 2.824930191040039, + "step": 45210 + }, + { + "epoch": 0.2261, + "grad_norm": 0.3554973602294922, + "learning_rate": 0.00044348768593903745, + "loss": 2.8837215423583986, + "step": 45220 + }, + { + "epoch": 0.22615, + "grad_norm": 0.39575469493865967, + "learning_rate": 0.00044346256481219435, + "loss": 2.828321075439453, + "step": 45230 + }, + { + "epoch": 0.2262, + "grad_norm": 0.3638007342815399, + "learning_rate": 0.0004434374388149325, + "loss": 2.815812683105469, + "step": 45240 + }, + { + "epoch": 0.22625, + "grad_norm": 0.35564130544662476, + "learning_rate": 0.0004434123079478844, + "loss": 2.8250776290893556, + "step": 45250 + }, + { + "epoch": 0.2263, + "grad_norm": 0.35863369703292847, + "learning_rate": 0.00044338717221168274, + "loss": 2.8537702560424805, + "step": 45260 + }, + { + "epoch": 0.22635, + "grad_norm": 0.3922620713710785, + "learning_rate": 0.0004433620316069603, + "loss": 2.804390716552734, + "step": 45270 + }, + { + "epoch": 0.2264, + "grad_norm": 0.34728121757507324, + "learning_rate": 0.0004433368861343501, + "loss": 2.8356109619140626, + "step": 45280 + }, + { + "epoch": 0.22645, + "grad_norm": 0.3478478491306305, + "learning_rate": 0.0004433117357944849, + "loss": 2.835322952270508, + "step": 45290 + }, + { + "epoch": 0.2265, + "grad_norm": 0.36690279841423035, + "learning_rate": 0.0004432865805879982, + "loss": 2.8364799499511717, + "step": 45300 + }, + { + "epoch": 0.22655, + "grad_norm": 0.33609116077423096, + "learning_rate": 0.0004432614205155231, + "loss": 2.8171531677246096, + "step": 45310 + }, + { + "epoch": 0.2266, + "grad_norm": 0.3475518226623535, + "learning_rate": 0.00044323625557769297, + "loss": 2.8077959060668944, + "step": 45320 + }, + { + "epoch": 0.22665, + "grad_norm": 0.36653733253479004, + "learning_rate": 0.0004432110857751415, + "loss": 2.7884805679321287, + "step": 45330 + }, + { + "epoch": 0.2267, + "grad_norm": 0.36469775438308716, + "learning_rate": 0.00044318591110850214, + "loss": 2.795069694519043, + "step": 45340 + }, + { + "epoch": 0.22675, + "grad_norm": 0.34689435362815857, + "learning_rate": 0.0004431607315784088, + "loss": 2.8243865966796875, + "step": 45350 + }, + { + "epoch": 0.2268, + "grad_norm": 0.3689448833465576, + "learning_rate": 0.0004431355471854953, + "loss": 2.790418243408203, + "step": 45360 + }, + { + "epoch": 0.22685, + "grad_norm": 0.36177858710289, + "learning_rate": 0.00044311035793039575, + "loss": 2.829158592224121, + "step": 45370 + }, + { + "epoch": 0.2269, + "grad_norm": 0.33629462122917175, + "learning_rate": 0.0004430851638137442, + "loss": 2.854570198059082, + "step": 45380 + }, + { + "epoch": 0.22695, + "grad_norm": 0.32798540592193604, + "learning_rate": 0.0004430599648361749, + "loss": 2.837279510498047, + "step": 45390 + }, + { + "epoch": 0.227, + "grad_norm": 0.35959193110466003, + "learning_rate": 0.0004430347609983223, + "loss": 2.7816076278686523, + "step": 45400 + }, + { + "epoch": 0.22705, + "grad_norm": 0.3546387255191803, + "learning_rate": 0.0004430095523008209, + "loss": 2.8273706436157227, + "step": 45410 + }, + { + "epoch": 0.2271, + "grad_norm": 0.3555634319782257, + "learning_rate": 0.00044298433874430534, + "loss": 2.7908273696899415, + "step": 45420 + }, + { + "epoch": 0.22715, + "grad_norm": 0.38323456048965454, + "learning_rate": 0.00044295912032941027, + "loss": 2.847019386291504, + "step": 45430 + }, + { + "epoch": 0.2272, + "grad_norm": 0.42165645956993103, + "learning_rate": 0.00044293389705677067, + "loss": 2.8260541915893556, + "step": 45440 + }, + { + "epoch": 0.22725, + "grad_norm": 0.36876389384269714, + "learning_rate": 0.0004429086689270215, + "loss": 2.887481117248535, + "step": 45450 + }, + { + "epoch": 0.2273, + "grad_norm": 0.31932541728019714, + "learning_rate": 0.0004428834359407979, + "loss": 2.8829484939575196, + "step": 45460 + }, + { + "epoch": 0.22735, + "grad_norm": 0.3661585748195648, + "learning_rate": 0.0004428581980987351, + "loss": 2.789453887939453, + "step": 45470 + }, + { + "epoch": 0.2274, + "grad_norm": 0.36076444387435913, + "learning_rate": 0.0004428329554014684, + "loss": 2.8656187057495117, + "step": 45480 + }, + { + "epoch": 0.22745, + "grad_norm": 0.3302929401397705, + "learning_rate": 0.0004428077078496333, + "loss": 2.8005725860595705, + "step": 45490 + }, + { + "epoch": 0.2275, + "grad_norm": 0.3387695848941803, + "learning_rate": 0.0004427824554438655, + "loss": 2.802780532836914, + "step": 45500 + }, + { + "epoch": 0.22755, + "grad_norm": 0.3546203076839447, + "learning_rate": 0.0004427571981848007, + "loss": 2.8413692474365235, + "step": 45510 + }, + { + "epoch": 0.2276, + "grad_norm": 0.34083348512649536, + "learning_rate": 0.0004427319360730747, + "loss": 2.8632999420166017, + "step": 45520 + }, + { + "epoch": 0.22765, + "grad_norm": 0.33364254236221313, + "learning_rate": 0.00044270666910932346, + "loss": 2.7783594131469727, + "step": 45530 + }, + { + "epoch": 0.2277, + "grad_norm": 0.3345731794834137, + "learning_rate": 0.0004426813972941832, + "loss": 2.8452043533325195, + "step": 45540 + }, + { + "epoch": 0.22775, + "grad_norm": 0.3281157314777374, + "learning_rate": 0.00044265612062828995, + "loss": 2.864533233642578, + "step": 45550 + }, + { + "epoch": 0.2278, + "grad_norm": 0.3440459966659546, + "learning_rate": 0.0004426308391122802, + "loss": 2.884694480895996, + "step": 45560 + }, + { + "epoch": 0.22785, + "grad_norm": 0.36085090041160583, + "learning_rate": 0.0004426055527467904, + "loss": 2.8067726135253905, + "step": 45570 + }, + { + "epoch": 0.2279, + "grad_norm": 0.34112784266471863, + "learning_rate": 0.000442580261532457, + "loss": 2.8142255783081054, + "step": 45580 + }, + { + "epoch": 0.22795, + "grad_norm": 0.3456272780895233, + "learning_rate": 0.0004425549654699169, + "loss": 2.8408977508544924, + "step": 45590 + }, + { + "epoch": 0.228, + "grad_norm": 0.37625041604042053, + "learning_rate": 0.00044252966455980675, + "loss": 2.8008468627929686, + "step": 45600 + }, + { + "epoch": 0.22805, + "grad_norm": 0.35124433040618896, + "learning_rate": 0.00044250435880276363, + "loss": 2.855885887145996, + "step": 45610 + }, + { + "epoch": 0.2281, + "grad_norm": 0.3665081262588501, + "learning_rate": 0.0004424790481994245, + "loss": 2.827336311340332, + "step": 45620 + }, + { + "epoch": 0.22815, + "grad_norm": 0.3261004686355591, + "learning_rate": 0.00044245373275042673, + "loss": 2.8178359985351564, + "step": 45630 + }, + { + "epoch": 0.2282, + "grad_norm": 0.37590786814689636, + "learning_rate": 0.0004424284124564075, + "loss": 2.7622055053710937, + "step": 45640 + }, + { + "epoch": 0.22825, + "grad_norm": 0.3417612910270691, + "learning_rate": 0.00044240308731800424, + "loss": 2.7582477569580077, + "step": 45650 + }, + { + "epoch": 0.2283, + "grad_norm": 0.34398525953292847, + "learning_rate": 0.0004423777573358545, + "loss": 2.799533462524414, + "step": 45660 + }, + { + "epoch": 0.22835, + "grad_norm": 0.34140047430992126, + "learning_rate": 0.00044235242251059603, + "loss": 2.8534671783447267, + "step": 45670 + }, + { + "epoch": 0.2284, + "grad_norm": 0.39487501978874207, + "learning_rate": 0.0004423270828428666, + "loss": 2.8446413040161134, + "step": 45680 + }, + { + "epoch": 0.22845, + "grad_norm": 0.33124294877052307, + "learning_rate": 0.00044230173833330425, + "loss": 2.843953514099121, + "step": 45690 + }, + { + "epoch": 0.2285, + "grad_norm": 0.3542773723602295, + "learning_rate": 0.0004422763889825468, + "loss": 2.870191764831543, + "step": 45700 + }, + { + "epoch": 0.22855, + "grad_norm": 0.3349216878414154, + "learning_rate": 0.00044225103479123256, + "loss": 2.775558090209961, + "step": 45710 + }, + { + "epoch": 0.2286, + "grad_norm": 0.337999165058136, + "learning_rate": 0.00044222567575999985, + "loss": 2.7892282485961912, + "step": 45720 + }, + { + "epoch": 0.22865, + "grad_norm": 0.39568376541137695, + "learning_rate": 0.000442200311889487, + "loss": 2.8137563705444335, + "step": 45730 + }, + { + "epoch": 0.2287, + "grad_norm": 0.35771510004997253, + "learning_rate": 0.0004421749431803326, + "loss": 2.791889190673828, + "step": 45740 + }, + { + "epoch": 0.22875, + "grad_norm": 0.33894094824790955, + "learning_rate": 0.00044214956963317534, + "loss": 2.8536415100097656, + "step": 45750 + }, + { + "epoch": 0.2288, + "grad_norm": 0.39464595913887024, + "learning_rate": 0.00044212419124865386, + "loss": 2.8794475555419923, + "step": 45760 + }, + { + "epoch": 0.22885, + "grad_norm": 0.4081399440765381, + "learning_rate": 0.0004420988080274072, + "loss": 2.8505605697631835, + "step": 45770 + }, + { + "epoch": 0.2289, + "grad_norm": 0.3688214123249054, + "learning_rate": 0.00044207341997007425, + "loss": 2.754367637634277, + "step": 45780 + }, + { + "epoch": 0.22895, + "grad_norm": 0.3487158417701721, + "learning_rate": 0.0004420480270772943, + "loss": 2.819036102294922, + "step": 45790 + }, + { + "epoch": 0.229, + "grad_norm": 0.37817472219467163, + "learning_rate": 0.0004420226293497066, + "loss": 2.850001335144043, + "step": 45800 + }, + { + "epoch": 0.22905, + "grad_norm": 0.3391696512699127, + "learning_rate": 0.0004419972267879504, + "loss": 2.7926759719848633, + "step": 45810 + }, + { + "epoch": 0.2291, + "grad_norm": 0.3807131350040436, + "learning_rate": 0.0004419718193926654, + "loss": 2.749459075927734, + "step": 45820 + }, + { + "epoch": 0.22915, + "grad_norm": 0.3625219464302063, + "learning_rate": 0.000441946407164491, + "loss": 2.8315109252929687, + "step": 45830 + }, + { + "epoch": 0.2292, + "grad_norm": 0.3779189884662628, + "learning_rate": 0.00044192099010406714, + "loss": 2.782154655456543, + "step": 45840 + }, + { + "epoch": 0.22925, + "grad_norm": 0.3288173973560333, + "learning_rate": 0.00044189556821203367, + "loss": 2.819871520996094, + "step": 45850 + }, + { + "epoch": 0.2293, + "grad_norm": 0.35161468386650085, + "learning_rate": 0.00044187014148903053, + "loss": 2.8556854248046877, + "step": 45860 + }, + { + "epoch": 0.22935, + "grad_norm": 0.3480682969093323, + "learning_rate": 0.00044184470993569784, + "loss": 2.7876655578613283, + "step": 45870 + }, + { + "epoch": 0.2294, + "grad_norm": 0.3651784062385559, + "learning_rate": 0.00044181927355267593, + "loss": 2.821308135986328, + "step": 45880 + }, + { + "epoch": 0.22945, + "grad_norm": 0.3391522169113159, + "learning_rate": 0.00044179383234060497, + "loss": 2.8222679138183593, + "step": 45890 + }, + { + "epoch": 0.2295, + "grad_norm": 0.3465275764465332, + "learning_rate": 0.0004417683863001257, + "loss": 2.8388513565063476, + "step": 45900 + }, + { + "epoch": 0.22955, + "grad_norm": 0.38370805978775024, + "learning_rate": 0.0004417429354318785, + "loss": 2.781410980224609, + "step": 45910 + }, + { + "epoch": 0.2296, + "grad_norm": 0.34913942217826843, + "learning_rate": 0.0004417174797365042, + "loss": 2.8618053436279296, + "step": 45920 + }, + { + "epoch": 0.22965, + "grad_norm": 0.3349117040634155, + "learning_rate": 0.00044169201921464364, + "loss": 2.792237091064453, + "step": 45930 + }, + { + "epoch": 0.2297, + "grad_norm": 0.3564988076686859, + "learning_rate": 0.0004416665538669378, + "loss": 2.837516212463379, + "step": 45940 + }, + { + "epoch": 0.22975, + "grad_norm": 0.4126798212528229, + "learning_rate": 0.00044164108369402767, + "loss": 2.892422676086426, + "step": 45950 + }, + { + "epoch": 0.2298, + "grad_norm": 0.35821646451950073, + "learning_rate": 0.0004416156086965546, + "loss": 2.793596076965332, + "step": 45960 + }, + { + "epoch": 0.22985, + "grad_norm": 0.3124660551548004, + "learning_rate": 0.0004415901288751598, + "loss": 2.753895378112793, + "step": 45970 + }, + { + "epoch": 0.2299, + "grad_norm": 0.3277362883090973, + "learning_rate": 0.0004415646442304848, + "loss": 2.8274295806884764, + "step": 45980 + }, + { + "epoch": 0.22995, + "grad_norm": 0.35911107063293457, + "learning_rate": 0.00044153915476317126, + "loss": 2.785720443725586, + "step": 45990 + }, + { + "epoch": 0.23, + "grad_norm": 0.3311454653739929, + "learning_rate": 0.00044151366047386065, + "loss": 2.8434267044067383, + "step": 46000 + }, + { + "epoch": 0.23005, + "grad_norm": 0.3217342495918274, + "learning_rate": 0.000441488161363195, + "loss": 2.8459985733032225, + "step": 46010 + }, + { + "epoch": 0.2301, + "grad_norm": 0.3406115472316742, + "learning_rate": 0.0004414626574318161, + "loss": 2.858443832397461, + "step": 46020 + }, + { + "epoch": 0.23015, + "grad_norm": 0.3435431718826294, + "learning_rate": 0.00044143714868036623, + "loss": 2.8210428237915037, + "step": 46030 + }, + { + "epoch": 0.2302, + "grad_norm": 0.3123301863670349, + "learning_rate": 0.0004414116351094873, + "loss": 2.7531299591064453, + "step": 46040 + }, + { + "epoch": 0.23025, + "grad_norm": 0.3273676633834839, + "learning_rate": 0.0004413861167198218, + "loss": 2.83780460357666, + "step": 46050 + }, + { + "epoch": 0.2303, + "grad_norm": 0.33423253893852234, + "learning_rate": 0.00044136059351201196, + "loss": 2.822166633605957, + "step": 46060 + }, + { + "epoch": 0.23035, + "grad_norm": 0.35685643553733826, + "learning_rate": 0.00044133506548670054, + "loss": 2.814803123474121, + "step": 46070 + }, + { + "epoch": 0.2304, + "grad_norm": 0.36756831407546997, + "learning_rate": 0.0004413095326445301, + "loss": 2.86636905670166, + "step": 46080 + }, + { + "epoch": 0.23045, + "grad_norm": 0.36148786544799805, + "learning_rate": 0.00044128399498614344, + "loss": 2.806817054748535, + "step": 46090 + }, + { + "epoch": 0.2305, + "grad_norm": 0.35331133008003235, + "learning_rate": 0.0004412584525121836, + "loss": 2.880113410949707, + "step": 46100 + }, + { + "epoch": 0.23055, + "grad_norm": 0.3565339744091034, + "learning_rate": 0.0004412329052232934, + "loss": 2.8061700820922852, + "step": 46110 + }, + { + "epoch": 0.2306, + "grad_norm": 0.32142701745033264, + "learning_rate": 0.00044120735312011604, + "loss": 2.801386070251465, + "step": 46120 + }, + { + "epoch": 0.23065, + "grad_norm": 0.3453209698200226, + "learning_rate": 0.00044118179620329493, + "loss": 2.8203767776489257, + "step": 46130 + }, + { + "epoch": 0.2307, + "grad_norm": 0.3476419150829315, + "learning_rate": 0.0004411562344734734, + "loss": 2.815192222595215, + "step": 46140 + }, + { + "epoch": 0.23075, + "grad_norm": 0.37376704812049866, + "learning_rate": 0.0004411306679312949, + "loss": 2.7672719955444336, + "step": 46150 + }, + { + "epoch": 0.2308, + "grad_norm": 0.31169745326042175, + "learning_rate": 0.0004411050965774031, + "loss": 2.849397659301758, + "step": 46160 + }, + { + "epoch": 0.23085, + "grad_norm": 0.3548707962036133, + "learning_rate": 0.0004410795204124418, + "loss": 2.8251110076904298, + "step": 46170 + }, + { + "epoch": 0.2309, + "grad_norm": 0.3578510284423828, + "learning_rate": 0.0004410539394370548, + "loss": 2.8197967529296877, + "step": 46180 + }, + { + "epoch": 0.23095, + "grad_norm": 0.36605581641197205, + "learning_rate": 0.0004410283536518862, + "loss": 2.849679374694824, + "step": 46190 + }, + { + "epoch": 0.231, + "grad_norm": 0.336477130651474, + "learning_rate": 0.00044100276305758003, + "loss": 2.797416114807129, + "step": 46200 + }, + { + "epoch": 0.23105, + "grad_norm": 0.41487717628479004, + "learning_rate": 0.00044097716765478063, + "loss": 2.8583652496337892, + "step": 46210 + }, + { + "epoch": 0.2311, + "grad_norm": 0.34403911232948303, + "learning_rate": 0.0004409515674441323, + "loss": 2.8396581649780273, + "step": 46220 + }, + { + "epoch": 0.23115, + "grad_norm": 0.3250097632408142, + "learning_rate": 0.00044092596242627946, + "loss": 2.850328826904297, + "step": 46230 + }, + { + "epoch": 0.2312, + "grad_norm": 0.34928324818611145, + "learning_rate": 0.0004409003526018669, + "loss": 2.8049440383911133, + "step": 46240 + }, + { + "epoch": 0.23125, + "grad_norm": 0.3793424963951111, + "learning_rate": 0.0004408747379715392, + "loss": 2.820985221862793, + "step": 46250 + }, + { + "epoch": 0.2313, + "grad_norm": 0.39207974076271057, + "learning_rate": 0.00044084911853594126, + "loss": 2.8176700592041017, + "step": 46260 + }, + { + "epoch": 0.23135, + "grad_norm": 0.347530722618103, + "learning_rate": 0.000440823494295718, + "loss": 2.8555986404418947, + "step": 46270 + }, + { + "epoch": 0.2314, + "grad_norm": 0.34722331166267395, + "learning_rate": 0.0004407978652515146, + "loss": 2.7938690185546875, + "step": 46280 + }, + { + "epoch": 0.23145, + "grad_norm": 0.35460397601127625, + "learning_rate": 0.0004407722314039761, + "loss": 2.8236515045166017, + "step": 46290 + }, + { + "epoch": 0.2315, + "grad_norm": 0.33645161986351013, + "learning_rate": 0.00044074659275374805, + "loss": 2.817551040649414, + "step": 46300 + }, + { + "epoch": 0.23155, + "grad_norm": 0.33914536237716675, + "learning_rate": 0.0004407209493014758, + "loss": 2.8606929779052734, + "step": 46310 + }, + { + "epoch": 0.2316, + "grad_norm": 0.34101998805999756, + "learning_rate": 0.00044069530104780486, + "loss": 2.879123878479004, + "step": 46320 + }, + { + "epoch": 0.23165, + "grad_norm": 0.352245032787323, + "learning_rate": 0.00044066964799338105, + "loss": 2.839212417602539, + "step": 46330 + }, + { + "epoch": 0.2317, + "grad_norm": 0.31784170866012573, + "learning_rate": 0.0004406439901388501, + "loss": 2.8932559967041014, + "step": 46340 + }, + { + "epoch": 0.23175, + "grad_norm": 0.35559648275375366, + "learning_rate": 0.00044061832748485796, + "loss": 2.8107404708862305, + "step": 46350 + }, + { + "epoch": 0.2318, + "grad_norm": 0.3266996443271637, + "learning_rate": 0.00044059266003205073, + "loss": 2.791082191467285, + "step": 46360 + }, + { + "epoch": 0.23185, + "grad_norm": 0.32323211431503296, + "learning_rate": 0.00044056698778107453, + "loss": 2.8050973892211912, + "step": 46370 + }, + { + "epoch": 0.2319, + "grad_norm": 0.3600846529006958, + "learning_rate": 0.00044054131073257576, + "loss": 2.8059274673461916, + "step": 46380 + }, + { + "epoch": 0.23195, + "grad_norm": 0.3431703746318817, + "learning_rate": 0.00044051562888720066, + "loss": 2.8110742568969727, + "step": 46390 + }, + { + "epoch": 0.232, + "grad_norm": 0.3304280936717987, + "learning_rate": 0.0004404899422455959, + "loss": 2.842738151550293, + "step": 46400 + }, + { + "epoch": 0.23205, + "grad_norm": 0.33571645617485046, + "learning_rate": 0.00044046425080840817, + "loss": 2.8654340744018554, + "step": 46410 + }, + { + "epoch": 0.2321, + "grad_norm": 0.33459043502807617, + "learning_rate": 0.0004404385545762842, + "loss": 2.8156978607177736, + "step": 46420 + }, + { + "epoch": 0.23215, + "grad_norm": 0.3539108633995056, + "learning_rate": 0.0004404128535498708, + "loss": 2.8927473068237304, + "step": 46430 + }, + { + "epoch": 0.2322, + "grad_norm": 0.34490543603897095, + "learning_rate": 0.0004403871477298151, + "loss": 2.762168312072754, + "step": 46440 + }, + { + "epoch": 0.23225, + "grad_norm": 0.35189196467399597, + "learning_rate": 0.00044036143711676434, + "loss": 2.86904296875, + "step": 46450 + }, + { + "epoch": 0.2323, + "grad_norm": 0.33516401052474976, + "learning_rate": 0.0004403357217113656, + "loss": 2.831226921081543, + "step": 46460 + }, + { + "epoch": 0.23235, + "grad_norm": 0.2971900403499603, + "learning_rate": 0.0004403100015142664, + "loss": 2.7947317123413087, + "step": 46470 + }, + { + "epoch": 0.2324, + "grad_norm": 0.3418923616409302, + "learning_rate": 0.00044028427652611406, + "loss": 2.821191596984863, + "step": 46480 + }, + { + "epoch": 0.23245, + "grad_norm": 0.33882567286491394, + "learning_rate": 0.0004402585467475564, + "loss": 2.8923467636108398, + "step": 46490 + }, + { + "epoch": 0.2325, + "grad_norm": 0.3405541479587555, + "learning_rate": 0.00044023281217924116, + "loss": 2.860665702819824, + "step": 46500 + }, + { + "epoch": 0.23255, + "grad_norm": 0.35727283358573914, + "learning_rate": 0.00044020707282181604, + "loss": 2.800685501098633, + "step": 46510 + }, + { + "epoch": 0.2326, + "grad_norm": 0.3264220058917999, + "learning_rate": 0.0004401813286759292, + "loss": 2.8198375701904297, + "step": 46520 + }, + { + "epoch": 0.23265, + "grad_norm": 0.3899862766265869, + "learning_rate": 0.00044015557974222863, + "loss": 2.8290916442871095, + "step": 46530 + }, + { + "epoch": 0.2327, + "grad_norm": 0.3372478485107422, + "learning_rate": 0.0004401298260213626, + "loss": 2.7959196090698244, + "step": 46540 + }, + { + "epoch": 0.23275, + "grad_norm": 0.31326982378959656, + "learning_rate": 0.0004401040675139795, + "loss": 2.853011131286621, + "step": 46550 + }, + { + "epoch": 0.2328, + "grad_norm": 0.3759270906448364, + "learning_rate": 0.0004400783042207278, + "loss": 2.832548141479492, + "step": 46560 + }, + { + "epoch": 0.23285, + "grad_norm": 0.33877071738243103, + "learning_rate": 0.00044005253614225605, + "loss": 2.814251518249512, + "step": 46570 + }, + { + "epoch": 0.2329, + "grad_norm": 0.3378308117389679, + "learning_rate": 0.000440026763279213, + "loss": 2.8055755615234377, + "step": 46580 + }, + { + "epoch": 0.23295, + "grad_norm": 0.340459942817688, + "learning_rate": 0.00044000098563224733, + "loss": 2.835084342956543, + "step": 46590 + }, + { + "epoch": 0.233, + "grad_norm": 0.3222876489162445, + "learning_rate": 0.0004399752032020082, + "loss": 2.8160648345947266, + "step": 46600 + }, + { + "epoch": 0.23305, + "grad_norm": 0.33802181482315063, + "learning_rate": 0.00043994941598914463, + "loss": 2.7881322860717774, + "step": 46610 + }, + { + "epoch": 0.2331, + "grad_norm": 0.3155997097492218, + "learning_rate": 0.00043992362399430577, + "loss": 2.8558631896972657, + "step": 46620 + }, + { + "epoch": 0.23315, + "grad_norm": 0.34742048382759094, + "learning_rate": 0.00043989782721814085, + "loss": 2.790632438659668, + "step": 46630 + }, + { + "epoch": 0.2332, + "grad_norm": 0.37771475315093994, + "learning_rate": 0.00043987202566129953, + "loss": 2.8606870651245115, + "step": 46640 + }, + { + "epoch": 0.23325, + "grad_norm": 0.3302416503429413, + "learning_rate": 0.00043984621932443115, + "loss": 2.8068994522094726, + "step": 46650 + }, + { + "epoch": 0.2333, + "grad_norm": 0.3444269299507141, + "learning_rate": 0.00043982040820818546, + "loss": 2.7873029708862305, + "step": 46660 + }, + { + "epoch": 0.23335, + "grad_norm": 0.3767060339450836, + "learning_rate": 0.00043979459231321234, + "loss": 2.84268741607666, + "step": 46670 + }, + { + "epoch": 0.2334, + "grad_norm": 0.35205116868019104, + "learning_rate": 0.0004397687716401616, + "loss": 2.7423355102539064, + "step": 46680 + }, + { + "epoch": 0.23345, + "grad_norm": 0.4081275463104248, + "learning_rate": 0.00043974294618968326, + "loss": 2.8458637237548827, + "step": 46690 + }, + { + "epoch": 0.2335, + "grad_norm": 0.342734158039093, + "learning_rate": 0.0004397171159624276, + "loss": 2.8025840759277343, + "step": 46700 + }, + { + "epoch": 0.23355, + "grad_norm": 0.34125056862831116, + "learning_rate": 0.0004396912809590447, + "loss": 2.8059318542480467, + "step": 46710 + }, + { + "epoch": 0.2336, + "grad_norm": 0.346599280834198, + "learning_rate": 0.00043966544118018516, + "loss": 2.8348371505737306, + "step": 46720 + }, + { + "epoch": 0.23365, + "grad_norm": 0.33484965562820435, + "learning_rate": 0.0004396395966264993, + "loss": 2.8410436630249025, + "step": 46730 + }, + { + "epoch": 0.2337, + "grad_norm": 0.3404802978038788, + "learning_rate": 0.00043961374729863795, + "loss": 2.8476919174194335, + "step": 46740 + }, + { + "epoch": 0.23375, + "grad_norm": 0.35069790482521057, + "learning_rate": 0.0004395878931972518, + "loss": 2.814188575744629, + "step": 46750 + }, + { + "epoch": 0.2338, + "grad_norm": 0.3657628297805786, + "learning_rate": 0.00043956203432299166, + "loss": 2.808117485046387, + "step": 46760 + }, + { + "epoch": 0.23385, + "grad_norm": 0.34389472007751465, + "learning_rate": 0.00043953617067650854, + "loss": 2.826603889465332, + "step": 46770 + }, + { + "epoch": 0.2339, + "grad_norm": 0.3310301899909973, + "learning_rate": 0.00043951030225845367, + "loss": 2.775644302368164, + "step": 46780 + }, + { + "epoch": 0.23395, + "grad_norm": 0.3451387882232666, + "learning_rate": 0.00043948442906947816, + "loss": 2.7982357025146483, + "step": 46790 + }, + { + "epoch": 0.234, + "grad_norm": 0.3326210379600525, + "learning_rate": 0.0004394585511102334, + "loss": 2.8146495819091797, + "step": 46800 + }, + { + "epoch": 0.23405, + "grad_norm": 0.3487163186073303, + "learning_rate": 0.00043943266838137094, + "loss": 2.8164138793945312, + "step": 46810 + }, + { + "epoch": 0.2341, + "grad_norm": 0.3326556384563446, + "learning_rate": 0.0004394067808835422, + "loss": 2.8170549392700197, + "step": 46820 + }, + { + "epoch": 0.23415, + "grad_norm": 0.33340656757354736, + "learning_rate": 0.0004393808886173992, + "loss": 2.8114959716796877, + "step": 46830 + }, + { + "epoch": 0.2342, + "grad_norm": 0.41082629561424255, + "learning_rate": 0.00043935499158359343, + "loss": 2.79259033203125, + "step": 46840 + }, + { + "epoch": 0.23425, + "grad_norm": 0.3798172175884247, + "learning_rate": 0.0004393290897827771, + "loss": 2.840772819519043, + "step": 46850 + }, + { + "epoch": 0.2343, + "grad_norm": 0.3502882719039917, + "learning_rate": 0.00043930318321560213, + "loss": 2.822649383544922, + "step": 46860 + }, + { + "epoch": 0.23435, + "grad_norm": 0.36095258593559265, + "learning_rate": 0.0004392772718827209, + "loss": 2.8167715072631836, + "step": 46870 + }, + { + "epoch": 0.2344, + "grad_norm": 0.34931308031082153, + "learning_rate": 0.00043925135578478546, + "loss": 2.831278610229492, + "step": 46880 + }, + { + "epoch": 0.23445, + "grad_norm": 0.3132226765155792, + "learning_rate": 0.0004392254349224485, + "loss": 2.8442529678344726, + "step": 46890 + }, + { + "epoch": 0.2345, + "grad_norm": 0.3251252770423889, + "learning_rate": 0.00043919950929636255, + "loss": 2.8072399139404296, + "step": 46900 + }, + { + "epoch": 0.23455, + "grad_norm": 0.3610973656177521, + "learning_rate": 0.0004391735789071801, + "loss": 2.8331415176391603, + "step": 46910 + }, + { + "epoch": 0.2346, + "grad_norm": 0.3124546706676483, + "learning_rate": 0.0004391476437555541, + "loss": 2.8157476425170898, + "step": 46920 + }, + { + "epoch": 0.23465, + "grad_norm": 0.3353816270828247, + "learning_rate": 0.00043912170384213746, + "loss": 2.809317779541016, + "step": 46930 + }, + { + "epoch": 0.2347, + "grad_norm": 0.3351481258869171, + "learning_rate": 0.00043909575916758315, + "loss": 2.816080856323242, + "step": 46940 + }, + { + "epoch": 0.23475, + "grad_norm": 0.3544747829437256, + "learning_rate": 0.0004390698097325444, + "loss": 2.843951606750488, + "step": 46950 + }, + { + "epoch": 0.2348, + "grad_norm": 0.3298306465148926, + "learning_rate": 0.0004390438555376744, + "loss": 2.8173274993896484, + "step": 46960 + }, + { + "epoch": 0.23485, + "grad_norm": 0.322655588388443, + "learning_rate": 0.00043901789658362667, + "loss": 2.8542335510253904, + "step": 46970 + }, + { + "epoch": 0.2349, + "grad_norm": 0.35526880621910095, + "learning_rate": 0.00043899193287105456, + "loss": 2.8231372833251953, + "step": 46980 + }, + { + "epoch": 0.23495, + "grad_norm": 0.3642066419124603, + "learning_rate": 0.00043896596440061187, + "loss": 2.7793857574462892, + "step": 46990 + }, + { + "epoch": 0.235, + "grad_norm": 0.333880752325058, + "learning_rate": 0.0004389399911729523, + "loss": 2.800614929199219, + "step": 47000 + }, + { + "epoch": 0.23505, + "grad_norm": 0.31829634308815, + "learning_rate": 0.0004389140131887297, + "loss": 2.764329719543457, + "step": 47010 + }, + { + "epoch": 0.2351, + "grad_norm": 0.3484399914741516, + "learning_rate": 0.0004388880304485981, + "loss": 2.801162528991699, + "step": 47020 + }, + { + "epoch": 0.23515, + "grad_norm": 0.3429854214191437, + "learning_rate": 0.0004388620429532115, + "loss": 2.8240276336669923, + "step": 47030 + }, + { + "epoch": 0.2352, + "grad_norm": 0.3144802153110504, + "learning_rate": 0.00043883605070322435, + "loss": 2.796432685852051, + "step": 47040 + }, + { + "epoch": 0.23525, + "grad_norm": 0.39365115761756897, + "learning_rate": 0.0004388100536992908, + "loss": 2.806735610961914, + "step": 47050 + }, + { + "epoch": 0.2353, + "grad_norm": 0.33106493949890137, + "learning_rate": 0.0004387840519420655, + "loss": 2.777711868286133, + "step": 47060 + }, + { + "epoch": 0.23535, + "grad_norm": 0.37855738401412964, + "learning_rate": 0.0004387580454322029, + "loss": 2.787763977050781, + "step": 47070 + }, + { + "epoch": 0.2354, + "grad_norm": 0.3485962450504303, + "learning_rate": 0.0004387320341703578, + "loss": 2.8948732376098634, + "step": 47080 + }, + { + "epoch": 0.23545, + "grad_norm": 0.33136433362960815, + "learning_rate": 0.0004387060181571849, + "loss": 2.807208442687988, + "step": 47090 + }, + { + "epoch": 0.2355, + "grad_norm": 0.37266963720321655, + "learning_rate": 0.00043867999739333934, + "loss": 2.8546421051025392, + "step": 47100 + }, + { + "epoch": 0.23555, + "grad_norm": 0.3251207172870636, + "learning_rate": 0.0004386539718794761, + "loss": 2.7593698501586914, + "step": 47110 + }, + { + "epoch": 0.2356, + "grad_norm": 0.32476145029067993, + "learning_rate": 0.0004386279416162504, + "loss": 2.8011255264282227, + "step": 47120 + }, + { + "epoch": 0.23565, + "grad_norm": 0.363925039768219, + "learning_rate": 0.0004386019066043175, + "loss": 2.826964569091797, + "step": 47130 + }, + { + "epoch": 0.2357, + "grad_norm": 0.3571464717388153, + "learning_rate": 0.0004385758668443328, + "loss": 2.780079650878906, + "step": 47140 + }, + { + "epoch": 0.23575, + "grad_norm": 0.3572477698326111, + "learning_rate": 0.000438549822336952, + "loss": 2.8564449310302735, + "step": 47150 + }, + { + "epoch": 0.2358, + "grad_norm": 0.31942540407180786, + "learning_rate": 0.0004385237730828306, + "loss": 2.854445457458496, + "step": 47160 + }, + { + "epoch": 0.23585, + "grad_norm": 0.33361002802848816, + "learning_rate": 0.0004384977190826246, + "loss": 2.8362075805664064, + "step": 47170 + }, + { + "epoch": 0.2359, + "grad_norm": 0.3486720025539398, + "learning_rate": 0.0004384716603369897, + "loss": 2.82299919128418, + "step": 47180 + }, + { + "epoch": 0.23595, + "grad_norm": 0.3337874114513397, + "learning_rate": 0.000438445596846582, + "loss": 2.8152746200561523, + "step": 47190 + }, + { + "epoch": 0.236, + "grad_norm": 0.3237634003162384, + "learning_rate": 0.00043841952861205767, + "loss": 2.824573516845703, + "step": 47200 + }, + { + "epoch": 0.23605, + "grad_norm": 0.3683710992336273, + "learning_rate": 0.00043839345563407303, + "loss": 2.8342424392700196, + "step": 47210 + }, + { + "epoch": 0.2361, + "grad_norm": 0.3658148944377899, + "learning_rate": 0.00043836737791328427, + "loss": 2.8364564895629885, + "step": 47220 + }, + { + "epoch": 0.23615, + "grad_norm": 0.3260660171508789, + "learning_rate": 0.0004383412954503482, + "loss": 2.831319236755371, + "step": 47230 + }, + { + "epoch": 0.2362, + "grad_norm": 0.4007484018802643, + "learning_rate": 0.00043831520824592115, + "loss": 2.7868831634521483, + "step": 47240 + }, + { + "epoch": 0.23625, + "grad_norm": 0.3783290684223175, + "learning_rate": 0.00043828911630066004, + "loss": 2.7652252197265623, + "step": 47250 + }, + { + "epoch": 0.2363, + "grad_norm": 0.327161580324173, + "learning_rate": 0.00043826301961522165, + "loss": 2.852010726928711, + "step": 47260 + }, + { + "epoch": 0.23635, + "grad_norm": 0.33514073491096497, + "learning_rate": 0.000438236918190263, + "loss": 2.8627702713012697, + "step": 47270 + }, + { + "epoch": 0.2364, + "grad_norm": 0.3583246171474457, + "learning_rate": 0.00043821081202644117, + "loss": 2.782851982116699, + "step": 47280 + }, + { + "epoch": 0.23645, + "grad_norm": 0.3647737205028534, + "learning_rate": 0.0004381847011244135, + "loss": 2.8542816162109377, + "step": 47290 + }, + { + "epoch": 0.2365, + "grad_norm": 0.3274952471256256, + "learning_rate": 0.0004381585854848372, + "loss": 2.8269781112670898, + "step": 47300 + }, + { + "epoch": 0.23655, + "grad_norm": 0.3432231843471527, + "learning_rate": 0.0004381324651083697, + "loss": 2.860353469848633, + "step": 47310 + }, + { + "epoch": 0.2366, + "grad_norm": 0.3574388325214386, + "learning_rate": 0.00043810633999566873, + "loss": 2.8198423385620117, + "step": 47320 + }, + { + "epoch": 0.23665, + "grad_norm": 0.38609451055526733, + "learning_rate": 0.00043808021014739187, + "loss": 2.840985107421875, + "step": 47330 + }, + { + "epoch": 0.2367, + "grad_norm": 0.41561010479927063, + "learning_rate": 0.00043805407556419694, + "loss": 2.8220775604248045, + "step": 47340 + }, + { + "epoch": 0.23675, + "grad_norm": 0.358009934425354, + "learning_rate": 0.000438027936246742, + "loss": 2.813374137878418, + "step": 47350 + }, + { + "epoch": 0.2368, + "grad_norm": 0.3621080815792084, + "learning_rate": 0.000438001792195685, + "loss": 2.8219745635986326, + "step": 47360 + }, + { + "epoch": 0.23685, + "grad_norm": 0.32724612951278687, + "learning_rate": 0.00043797564341168413, + "loss": 2.8270973205566405, + "step": 47370 + }, + { + "epoch": 0.2369, + "grad_norm": 0.34998369216918945, + "learning_rate": 0.0004379494898953976, + "loss": 2.7724159240722654, + "step": 47380 + }, + { + "epoch": 0.23695, + "grad_norm": 0.32766449451446533, + "learning_rate": 0.00043792333164748405, + "loss": 2.80936279296875, + "step": 47390 + }, + { + "epoch": 0.237, + "grad_norm": 0.3801850378513336, + "learning_rate": 0.00043789716866860184, + "loss": 2.8230573654174806, + "step": 47400 + }, + { + "epoch": 0.23705, + "grad_norm": 0.34251776337623596, + "learning_rate": 0.00043787100095940963, + "loss": 2.7508506774902344, + "step": 47410 + }, + { + "epoch": 0.2371, + "grad_norm": 0.3349398374557495, + "learning_rate": 0.0004378448285205663, + "loss": 2.813998222351074, + "step": 47420 + }, + { + "epoch": 0.23715, + "grad_norm": 0.3672226667404175, + "learning_rate": 0.00043781865135273057, + "loss": 2.796276092529297, + "step": 47430 + }, + { + "epoch": 0.2372, + "grad_norm": 0.35926148295402527, + "learning_rate": 0.00043779246945656157, + "loss": 2.8435949325561523, + "step": 47440 + }, + { + "epoch": 0.23725, + "grad_norm": 0.35680773854255676, + "learning_rate": 0.0004377662828327185, + "loss": 2.8193721771240234, + "step": 47450 + }, + { + "epoch": 0.2373, + "grad_norm": 0.3407558798789978, + "learning_rate": 0.0004377400914818604, + "loss": 2.807943916320801, + "step": 47460 + }, + { + "epoch": 0.23735, + "grad_norm": 0.3036278784275055, + "learning_rate": 0.0004377138954046468, + "loss": 2.810809516906738, + "step": 47470 + }, + { + "epoch": 0.2374, + "grad_norm": 0.31859347224235535, + "learning_rate": 0.00043768769460173716, + "loss": 2.8516555786132813, + "step": 47480 + }, + { + "epoch": 0.23745, + "grad_norm": 0.39337804913520813, + "learning_rate": 0.000437661489073791, + "loss": 2.816421127319336, + "step": 47490 + }, + { + "epoch": 0.2375, + "grad_norm": 0.3563469648361206, + "learning_rate": 0.00043763527882146816, + "loss": 2.7395374298095705, + "step": 47500 + }, + { + "epoch": 0.23755, + "grad_norm": 0.3512931168079376, + "learning_rate": 0.00043760906384542845, + "loss": 2.787601661682129, + "step": 47510 + }, + { + "epoch": 0.2376, + "grad_norm": 0.3397270739078522, + "learning_rate": 0.0004375828441463317, + "loss": 2.872249794006348, + "step": 47520 + }, + { + "epoch": 0.23765, + "grad_norm": 0.36464038491249084, + "learning_rate": 0.0004375566197248382, + "loss": 2.809745216369629, + "step": 47530 + }, + { + "epoch": 0.2377, + "grad_norm": 0.3481111526489258, + "learning_rate": 0.00043753039058160807, + "loss": 2.791069984436035, + "step": 47540 + }, + { + "epoch": 0.23775, + "grad_norm": 0.35237041115760803, + "learning_rate": 0.0004375041567173015, + "loss": 2.8600208282470705, + "step": 47550 + }, + { + "epoch": 0.2378, + "grad_norm": 0.3689851760864258, + "learning_rate": 0.0004374779181325791, + "loss": 2.779175567626953, + "step": 47560 + }, + { + "epoch": 0.23785, + "grad_norm": 0.36797112226486206, + "learning_rate": 0.00043745167482810133, + "loss": 2.867388916015625, + "step": 47570 + }, + { + "epoch": 0.2379, + "grad_norm": 0.3285422623157501, + "learning_rate": 0.00043742542680452897, + "loss": 2.7976299285888673, + "step": 47580 + }, + { + "epoch": 0.23795, + "grad_norm": 0.3546627163887024, + "learning_rate": 0.00043739917406252274, + "loss": 2.8280242919921874, + "step": 47590 + }, + { + "epoch": 0.238, + "grad_norm": 0.3385280668735504, + "learning_rate": 0.00043737291660274347, + "loss": 2.812908172607422, + "step": 47600 + }, + { + "epoch": 0.23805, + "grad_norm": 0.3422914147377014, + "learning_rate": 0.00043734665442585235, + "loss": 2.827656364440918, + "step": 47610 + }, + { + "epoch": 0.2381, + "grad_norm": 0.31226199865341187, + "learning_rate": 0.00043732038753251047, + "loss": 2.8324649810791014, + "step": 47620 + }, + { + "epoch": 0.23815, + "grad_norm": 0.3262222409248352, + "learning_rate": 0.00043729411592337904, + "loss": 2.8311901092529297, + "step": 47630 + }, + { + "epoch": 0.2382, + "grad_norm": 0.3518199920654297, + "learning_rate": 0.00043726783959911953, + "loss": 2.737621307373047, + "step": 47640 + }, + { + "epoch": 0.23825, + "grad_norm": 0.33797305822372437, + "learning_rate": 0.00043724155856039336, + "loss": 2.771009635925293, + "step": 47650 + }, + { + "epoch": 0.2383, + "grad_norm": 0.3437384366989136, + "learning_rate": 0.0004372152728078622, + "loss": 2.8357404708862304, + "step": 47660 + }, + { + "epoch": 0.23835, + "grad_norm": 0.341665118932724, + "learning_rate": 0.0004371889823421878, + "loss": 2.847621536254883, + "step": 47670 + }, + { + "epoch": 0.2384, + "grad_norm": 0.35574740171432495, + "learning_rate": 0.00043716268716403206, + "loss": 2.797610855102539, + "step": 47680 + }, + { + "epoch": 0.23845, + "grad_norm": 0.3454342484474182, + "learning_rate": 0.0004371363872740569, + "loss": 2.8182647705078123, + "step": 47690 + }, + { + "epoch": 0.2385, + "grad_norm": 0.3476695120334625, + "learning_rate": 0.0004371100826729245, + "loss": 2.86181583404541, + "step": 47700 + }, + { + "epoch": 0.23855, + "grad_norm": 0.31136417388916016, + "learning_rate": 0.00043708377336129693, + "loss": 2.803977203369141, + "step": 47710 + }, + { + "epoch": 0.2386, + "grad_norm": 0.37832170724868774, + "learning_rate": 0.00043705745933983665, + "loss": 2.7558128356933596, + "step": 47720 + }, + { + "epoch": 0.23865, + "grad_norm": 0.3380691707134247, + "learning_rate": 0.00043703114060920613, + "loss": 2.8189712524414063, + "step": 47730 + }, + { + "epoch": 0.2387, + "grad_norm": 0.3305310308933258, + "learning_rate": 0.00043700481717006784, + "loss": 2.8209461212158202, + "step": 47740 + }, + { + "epoch": 0.23875, + "grad_norm": 0.3508587181568146, + "learning_rate": 0.0004369784890230846, + "loss": 2.797454261779785, + "step": 47750 + }, + { + "epoch": 0.2388, + "grad_norm": 0.37056559324264526, + "learning_rate": 0.00043695215616891905, + "loss": 2.806288146972656, + "step": 47760 + }, + { + "epoch": 0.23885, + "grad_norm": 0.3692663013935089, + "learning_rate": 0.0004369258186082342, + "loss": 2.801869583129883, + "step": 47770 + }, + { + "epoch": 0.2389, + "grad_norm": 0.338218092918396, + "learning_rate": 0.00043689947634169323, + "loss": 2.837641143798828, + "step": 47780 + }, + { + "epoch": 0.23895, + "grad_norm": 0.37854042649269104, + "learning_rate": 0.00043687312936995914, + "loss": 2.8178939819335938, + "step": 47790 + }, + { + "epoch": 0.239, + "grad_norm": 0.3530341386795044, + "learning_rate": 0.00043684677769369526, + "loss": 2.784598731994629, + "step": 47800 + }, + { + "epoch": 0.23905, + "grad_norm": 0.32652848958969116, + "learning_rate": 0.00043682042131356496, + "loss": 2.8483476638793945, + "step": 47810 + }, + { + "epoch": 0.2391, + "grad_norm": 0.3469916582107544, + "learning_rate": 0.00043679406023023183, + "loss": 2.813518524169922, + "step": 47820 + }, + { + "epoch": 0.23915, + "grad_norm": 0.33961695432662964, + "learning_rate": 0.0004367676944443595, + "loss": 2.8155227661132813, + "step": 47830 + }, + { + "epoch": 0.2392, + "grad_norm": 0.32197079062461853, + "learning_rate": 0.0004367413239566117, + "loss": 2.857670783996582, + "step": 47840 + }, + { + "epoch": 0.23925, + "grad_norm": 0.35859376192092896, + "learning_rate": 0.0004367149487676523, + "loss": 2.8373186111450197, + "step": 47850 + }, + { + "epoch": 0.2393, + "grad_norm": 0.34297966957092285, + "learning_rate": 0.00043668856887814535, + "loss": 2.8239368438720702, + "step": 47860 + }, + { + "epoch": 0.23935, + "grad_norm": 0.32673144340515137, + "learning_rate": 0.0004366621842887549, + "loss": 2.814137268066406, + "step": 47870 + }, + { + "epoch": 0.2394, + "grad_norm": 0.3598567247390747, + "learning_rate": 0.0004366357950001453, + "loss": 2.855784034729004, + "step": 47880 + }, + { + "epoch": 0.23945, + "grad_norm": 0.3307171165943146, + "learning_rate": 0.00043660940101298066, + "loss": 2.837795448303223, + "step": 47890 + }, + { + "epoch": 0.2395, + "grad_norm": 0.3566613793373108, + "learning_rate": 0.0004365830023279257, + "loss": 2.8099996566772463, + "step": 47900 + }, + { + "epoch": 0.23955, + "grad_norm": 0.31224367022514343, + "learning_rate": 0.0004365565989456448, + "loss": 2.849606513977051, + "step": 47910 + }, + { + "epoch": 0.2396, + "grad_norm": 0.3108988404273987, + "learning_rate": 0.00043653019086680285, + "loss": 2.828246307373047, + "step": 47920 + }, + { + "epoch": 0.23965, + "grad_norm": 0.334420382976532, + "learning_rate": 0.0004365037780920646, + "loss": 2.8303693771362304, + "step": 47930 + }, + { + "epoch": 0.2397, + "grad_norm": 0.3527625799179077, + "learning_rate": 0.00043647736062209496, + "loss": 2.8194307327270507, + "step": 47940 + }, + { + "epoch": 0.23975, + "grad_norm": 0.3514990508556366, + "learning_rate": 0.000436450938457559, + "loss": 2.810409355163574, + "step": 47950 + }, + { + "epoch": 0.2398, + "grad_norm": 0.3528570532798767, + "learning_rate": 0.0004364245115991219, + "loss": 2.858650779724121, + "step": 47960 + }, + { + "epoch": 0.23985, + "grad_norm": 0.3671596050262451, + "learning_rate": 0.000436398080047449, + "loss": 2.8075305938720705, + "step": 47970 + }, + { + "epoch": 0.2399, + "grad_norm": 0.34714439511299133, + "learning_rate": 0.00043637164380320564, + "loss": 2.866752815246582, + "step": 47980 + }, + { + "epoch": 0.23995, + "grad_norm": 0.3146287202835083, + "learning_rate": 0.00043634520286705745, + "loss": 2.7718011856079103, + "step": 47990 + }, + { + "epoch": 0.24, + "grad_norm": 0.33868297934532166, + "learning_rate": 0.00043631875723966994, + "loss": 2.8405961990356445, + "step": 48000 + }, + { + "epoch": 0.24005, + "grad_norm": 0.3499642312526703, + "learning_rate": 0.0004362923069217091, + "loss": 2.814986991882324, + "step": 48010 + }, + { + "epoch": 0.2401, + "grad_norm": 0.32070159912109375, + "learning_rate": 0.0004362658519138405, + "loss": 2.816162109375, + "step": 48020 + }, + { + "epoch": 0.24015, + "grad_norm": 0.3553222119808197, + "learning_rate": 0.00043623939221673044, + "loss": 2.822132682800293, + "step": 48030 + }, + { + "epoch": 0.2402, + "grad_norm": 0.35658034682273865, + "learning_rate": 0.00043621292783104485, + "loss": 2.852560615539551, + "step": 48040 + }, + { + "epoch": 0.24025, + "grad_norm": 0.3279896676540375, + "learning_rate": 0.00043618645875745013, + "loss": 2.844436264038086, + "step": 48050 + }, + { + "epoch": 0.2403, + "grad_norm": 0.3826558589935303, + "learning_rate": 0.0004361599849966125, + "loss": 2.7970537185668944, + "step": 48060 + }, + { + "epoch": 0.24035, + "grad_norm": 0.3655315041542053, + "learning_rate": 0.00043613350654919854, + "loss": 2.8104719161987304, + "step": 48070 + }, + { + "epoch": 0.2404, + "grad_norm": 0.38456910848617554, + "learning_rate": 0.00043610702341587474, + "loss": 2.811257553100586, + "step": 48080 + }, + { + "epoch": 0.24045, + "grad_norm": 0.4625016450881958, + "learning_rate": 0.0004360805355973079, + "loss": 2.7956308364868163, + "step": 48090 + }, + { + "epoch": 0.2405, + "grad_norm": 0.4025568962097168, + "learning_rate": 0.00043605404309416486, + "loss": 2.75854434967041, + "step": 48100 + }, + { + "epoch": 0.24055, + "grad_norm": 0.34041905403137207, + "learning_rate": 0.00043602754590711245, + "loss": 2.825534439086914, + "step": 48110 + }, + { + "epoch": 0.2406, + "grad_norm": 0.323009729385376, + "learning_rate": 0.00043600104403681785, + "loss": 2.8254667282104493, + "step": 48120 + }, + { + "epoch": 0.24065, + "grad_norm": 0.32905271649360657, + "learning_rate": 0.00043597453748394824, + "loss": 2.8328128814697267, + "step": 48130 + }, + { + "epoch": 0.2407, + "grad_norm": 0.3382734954357147, + "learning_rate": 0.0004359480262491708, + "loss": 2.8106817245483398, + "step": 48140 + }, + { + "epoch": 0.24075, + "grad_norm": 0.32350027561187744, + "learning_rate": 0.00043592151033315315, + "loss": 2.791218566894531, + "step": 48150 + }, + { + "epoch": 0.2408, + "grad_norm": 0.3171815276145935, + "learning_rate": 0.00043589498973656266, + "loss": 2.7671415328979494, + "step": 48160 + }, + { + "epoch": 0.24085, + "grad_norm": 0.3325694799423218, + "learning_rate": 0.00043586846446006704, + "loss": 2.7950437545776365, + "step": 48170 + }, + { + "epoch": 0.2409, + "grad_norm": 0.3232148289680481, + "learning_rate": 0.00043584193450433407, + "loss": 2.7824745178222656, + "step": 48180 + }, + { + "epoch": 0.24095, + "grad_norm": 0.33125147223472595, + "learning_rate": 0.00043581539987003166, + "loss": 2.806507110595703, + "step": 48190 + }, + { + "epoch": 0.241, + "grad_norm": 0.367430716753006, + "learning_rate": 0.00043578886055782785, + "loss": 2.7704446792602537, + "step": 48200 + }, + { + "epoch": 0.24105, + "grad_norm": 0.31697210669517517, + "learning_rate": 0.0004357623165683906, + "loss": 2.8103986740112306, + "step": 48210 + }, + { + "epoch": 0.2411, + "grad_norm": 0.37159842252731323, + "learning_rate": 0.00043573576790238835, + "loss": 2.8413263320922852, + "step": 48220 + }, + { + "epoch": 0.24115, + "grad_norm": 0.386614590883255, + "learning_rate": 0.0004357092145604894, + "loss": 2.7920082092285154, + "step": 48230 + }, + { + "epoch": 0.2412, + "grad_norm": 0.3736366629600525, + "learning_rate": 0.00043568265654336216, + "loss": 2.7550899505615236, + "step": 48240 + }, + { + "epoch": 0.24125, + "grad_norm": 0.33361759781837463, + "learning_rate": 0.0004356560938516753, + "loss": 2.778445243835449, + "step": 48250 + }, + { + "epoch": 0.2413, + "grad_norm": 0.33140844106674194, + "learning_rate": 0.00043562952648609754, + "loss": 2.779862976074219, + "step": 48260 + }, + { + "epoch": 0.24135, + "grad_norm": 0.34529104828834534, + "learning_rate": 0.0004356029544472977, + "loss": 2.8047611236572267, + "step": 48270 + }, + { + "epoch": 0.2414, + "grad_norm": 0.30884334444999695, + "learning_rate": 0.00043557637773594465, + "loss": 2.812311363220215, + "step": 48280 + }, + { + "epoch": 0.24145, + "grad_norm": 0.2984694838523865, + "learning_rate": 0.0004355497963527076, + "loss": 2.8555131912231446, + "step": 48290 + }, + { + "epoch": 0.2415, + "grad_norm": 0.3139670789241791, + "learning_rate": 0.0004355232102982556, + "loss": 2.8526992797851562, + "step": 48300 + }, + { + "epoch": 0.24155, + "grad_norm": 0.32952257990837097, + "learning_rate": 0.0004354966195732581, + "loss": 2.764269256591797, + "step": 48310 + }, + { + "epoch": 0.2416, + "grad_norm": 0.3441306948661804, + "learning_rate": 0.0004354700241783844, + "loss": 2.8040342330932617, + "step": 48320 + }, + { + "epoch": 0.24165, + "grad_norm": 0.32683804631233215, + "learning_rate": 0.0004354434241143041, + "loss": 2.795053482055664, + "step": 48330 + }, + { + "epoch": 0.2417, + "grad_norm": 0.37470361590385437, + "learning_rate": 0.0004354168193816868, + "loss": 2.7831113815307615, + "step": 48340 + }, + { + "epoch": 0.24175, + "grad_norm": 0.40700963139533997, + "learning_rate": 0.00043539020998120237, + "loss": 2.7633270263671874, + "step": 48350 + }, + { + "epoch": 0.2418, + "grad_norm": 0.3952031433582306, + "learning_rate": 0.0004353635959135206, + "loss": 2.8063207626342774, + "step": 48360 + }, + { + "epoch": 0.24185, + "grad_norm": 0.361672580242157, + "learning_rate": 0.0004353369771793115, + "loss": 2.8235843658447264, + "step": 48370 + }, + { + "epoch": 0.2419, + "grad_norm": 0.3616844713687897, + "learning_rate": 0.0004353103537792453, + "loss": 2.795279693603516, + "step": 48380 + }, + { + "epoch": 0.24195, + "grad_norm": 0.34033337235450745, + "learning_rate": 0.0004352837257139922, + "loss": 2.839603233337402, + "step": 48390 + }, + { + "epoch": 0.242, + "grad_norm": 0.31689345836639404, + "learning_rate": 0.0004352570929842225, + "loss": 2.816489601135254, + "step": 48400 + }, + { + "epoch": 0.24205, + "grad_norm": 0.3445102870464325, + "learning_rate": 0.0004352304555906067, + "loss": 2.842007637023926, + "step": 48410 + }, + { + "epoch": 0.2421, + "grad_norm": 0.3381800949573517, + "learning_rate": 0.0004352038135338155, + "loss": 2.7729669570922852, + "step": 48420 + }, + { + "epoch": 0.24215, + "grad_norm": 0.39366215467453003, + "learning_rate": 0.00043517716681451947, + "loss": 2.8363567352294923, + "step": 48430 + }, + { + "epoch": 0.2422, + "grad_norm": 0.3543867766857147, + "learning_rate": 0.0004351505154333895, + "loss": 2.8451786041259766, + "step": 48440 + }, + { + "epoch": 0.24225, + "grad_norm": 0.34781813621520996, + "learning_rate": 0.0004351238593910966, + "loss": 2.83909912109375, + "step": 48450 + }, + { + "epoch": 0.2423, + "grad_norm": 0.3234221935272217, + "learning_rate": 0.0004350971986883117, + "loss": 2.8355031967163087, + "step": 48460 + }, + { + "epoch": 0.24235, + "grad_norm": 0.3346942365169525, + "learning_rate": 0.0004350705333257061, + "loss": 2.8770790100097656, + "step": 48470 + }, + { + "epoch": 0.2424, + "grad_norm": 0.3443107604980469, + "learning_rate": 0.0004350438633039511, + "loss": 2.8278152465820314, + "step": 48480 + }, + { + "epoch": 0.24245, + "grad_norm": 0.3284781575202942, + "learning_rate": 0.000435017188623718, + "loss": 2.8434953689575195, + "step": 48490 + }, + { + "epoch": 0.2425, + "grad_norm": 0.3543301820755005, + "learning_rate": 0.0004349905092856784, + "loss": 2.8227365493774412, + "step": 48500 + }, + { + "epoch": 0.24255, + "grad_norm": 0.33373910188674927, + "learning_rate": 0.00043496382529050403, + "loss": 2.837910842895508, + "step": 48510 + }, + { + "epoch": 0.2426, + "grad_norm": 0.39414143562316895, + "learning_rate": 0.0004349371366388666, + "loss": 2.8212863922119142, + "step": 48520 + }, + { + "epoch": 0.24265, + "grad_norm": 0.35164546966552734, + "learning_rate": 0.0004349104433314379, + "loss": 2.8198532104492187, + "step": 48530 + }, + { + "epoch": 0.2427, + "grad_norm": 0.3396400511264801, + "learning_rate": 0.00043488374536889015, + "loss": 2.804795837402344, + "step": 48540 + }, + { + "epoch": 0.24275, + "grad_norm": 0.31376928091049194, + "learning_rate": 0.0004348570427518953, + "loss": 2.8093679428100584, + "step": 48550 + }, + { + "epoch": 0.2428, + "grad_norm": 0.3333773612976074, + "learning_rate": 0.0004348303354811256, + "loss": 2.823835563659668, + "step": 48560 + }, + { + "epoch": 0.24285, + "grad_norm": 0.3453516960144043, + "learning_rate": 0.00043480362355725346, + "loss": 2.823341178894043, + "step": 48570 + }, + { + "epoch": 0.2429, + "grad_norm": 0.34246474504470825, + "learning_rate": 0.0004347769069809513, + "loss": 2.774139404296875, + "step": 48580 + }, + { + "epoch": 0.24295, + "grad_norm": 0.312188982963562, + "learning_rate": 0.00043475018575289177, + "loss": 2.8498762130737303, + "step": 48590 + }, + { + "epoch": 0.243, + "grad_norm": 0.36612311005592346, + "learning_rate": 0.00043472345987374757, + "loss": 2.832464408874512, + "step": 48600 + }, + { + "epoch": 0.24305, + "grad_norm": 0.3489023745059967, + "learning_rate": 0.0004346967293441915, + "loss": 2.8102617263793945, + "step": 48610 + }, + { + "epoch": 0.2431, + "grad_norm": 0.3291383385658264, + "learning_rate": 0.0004346699941648965, + "loss": 2.802629852294922, + "step": 48620 + }, + { + "epoch": 0.24315, + "grad_norm": 0.3316866159439087, + "learning_rate": 0.00043464325433653563, + "loss": 2.7818696975708006, + "step": 48630 + }, + { + "epoch": 0.2432, + "grad_norm": 0.3825402855873108, + "learning_rate": 0.000434616509859782, + "loss": 2.818470001220703, + "step": 48640 + }, + { + "epoch": 0.24325, + "grad_norm": 0.34143969416618347, + "learning_rate": 0.000434589760735309, + "loss": 2.818531799316406, + "step": 48650 + }, + { + "epoch": 0.2433, + "grad_norm": 0.3337445557117462, + "learning_rate": 0.00043456300696379, + "loss": 2.8061094284057617, + "step": 48660 + }, + { + "epoch": 0.24335, + "grad_norm": 0.3302741050720215, + "learning_rate": 0.00043453624854589856, + "loss": 2.812113952636719, + "step": 48670 + }, + { + "epoch": 0.2434, + "grad_norm": 0.3488406240940094, + "learning_rate": 0.00043450948548230835, + "loss": 2.8170198440551757, + "step": 48680 + }, + { + "epoch": 0.24345, + "grad_norm": 0.3446834981441498, + "learning_rate": 0.00043448271777369294, + "loss": 2.8756786346435548, + "step": 48690 + }, + { + "epoch": 0.2435, + "grad_norm": 0.3485247790813446, + "learning_rate": 0.0004344559454207265, + "loss": 2.8473526000976563, + "step": 48700 + }, + { + "epoch": 0.24355, + "grad_norm": 0.3615718483924866, + "learning_rate": 0.00043442916842408274, + "loss": 2.837722396850586, + "step": 48710 + }, + { + "epoch": 0.2436, + "grad_norm": 0.32564589381217957, + "learning_rate": 0.000434402386784436, + "loss": 2.8154001235961914, + "step": 48720 + }, + { + "epoch": 0.24365, + "grad_norm": 0.33997538685798645, + "learning_rate": 0.0004343756005024604, + "loss": 2.7808500289916993, + "step": 48730 + }, + { + "epoch": 0.2437, + "grad_norm": 0.3352994918823242, + "learning_rate": 0.00043434880957883017, + "loss": 2.814859962463379, + "step": 48740 + }, + { + "epoch": 0.24375, + "grad_norm": 0.37555789947509766, + "learning_rate": 0.00043432201401422, + "loss": 2.8054292678833006, + "step": 48750 + }, + { + "epoch": 0.2438, + "grad_norm": 0.3291011452674866, + "learning_rate": 0.00043429521380930426, + "loss": 2.768808364868164, + "step": 48760 + }, + { + "epoch": 0.24385, + "grad_norm": 0.35915902256965637, + "learning_rate": 0.0004342684089647578, + "loss": 2.8473148345947266, + "step": 48770 + }, + { + "epoch": 0.2439, + "grad_norm": 0.3566305339336395, + "learning_rate": 0.0004342415994812554, + "loss": 2.817137336730957, + "step": 48780 + }, + { + "epoch": 0.24395, + "grad_norm": 0.3127589821815491, + "learning_rate": 0.000434214785359472, + "loss": 2.8141973495483397, + "step": 48790 + }, + { + "epoch": 0.244, + "grad_norm": 0.3471936881542206, + "learning_rate": 0.0004341879666000825, + "loss": 2.818947601318359, + "step": 48800 + }, + { + "epoch": 0.24405, + "grad_norm": 0.3280351161956787, + "learning_rate": 0.0004341611432037623, + "loss": 2.813749885559082, + "step": 48810 + }, + { + "epoch": 0.2441, + "grad_norm": 0.34870052337646484, + "learning_rate": 0.0004341343151711864, + "loss": 2.8343624114990233, + "step": 48820 + }, + { + "epoch": 0.24415, + "grad_norm": 0.35461223125457764, + "learning_rate": 0.00043410748250303046, + "loss": 2.838302230834961, + "step": 48830 + }, + { + "epoch": 0.2442, + "grad_norm": 0.3292231857776642, + "learning_rate": 0.00043408064519996983, + "loss": 2.803696060180664, + "step": 48840 + }, + { + "epoch": 0.24425, + "grad_norm": 0.32314544916152954, + "learning_rate": 0.00043405380326268017, + "loss": 2.7985706329345703, + "step": 48850 + }, + { + "epoch": 0.2443, + "grad_norm": 0.34108638763427734, + "learning_rate": 0.00043402695669183736, + "loss": 2.870116424560547, + "step": 48860 + }, + { + "epoch": 0.24435, + "grad_norm": 0.33333703875541687, + "learning_rate": 0.00043400010548811697, + "loss": 2.840776062011719, + "step": 48870 + }, + { + "epoch": 0.2444, + "grad_norm": 0.3736132085323334, + "learning_rate": 0.00043397324965219533, + "loss": 2.8180803298950194, + "step": 48880 + }, + { + "epoch": 0.24445, + "grad_norm": 0.3542589545249939, + "learning_rate": 0.0004339463891847482, + "loss": 2.77585506439209, + "step": 48890 + }, + { + "epoch": 0.2445, + "grad_norm": 0.3887844979763031, + "learning_rate": 0.00043391952408645203, + "loss": 2.8055965423583986, + "step": 48900 + }, + { + "epoch": 0.24455, + "grad_norm": 0.39005231857299805, + "learning_rate": 0.000433892654357983, + "loss": 2.8469568252563477, + "step": 48910 + }, + { + "epoch": 0.2446, + "grad_norm": 0.32817381620407104, + "learning_rate": 0.0004338657800000178, + "loss": 2.819734573364258, + "step": 48920 + }, + { + "epoch": 0.24465, + "grad_norm": 0.3759654760360718, + "learning_rate": 0.00043383890101323264, + "loss": 2.8128662109375, + "step": 48930 + }, + { + "epoch": 0.2447, + "grad_norm": 0.34796828031539917, + "learning_rate": 0.0004338120173983044, + "loss": 2.842996025085449, + "step": 48940 + }, + { + "epoch": 0.24475, + "grad_norm": 0.33517998456954956, + "learning_rate": 0.00043378512915590993, + "loss": 2.8792463302612306, + "step": 48950 + }, + { + "epoch": 0.2448, + "grad_norm": 0.34909477829933167, + "learning_rate": 0.000433758236286726, + "loss": 2.792523193359375, + "step": 48960 + }, + { + "epoch": 0.24485, + "grad_norm": 0.36778154969215393, + "learning_rate": 0.0004337313387914297, + "loss": 2.7567184448242186, + "step": 48970 + }, + { + "epoch": 0.2449, + "grad_norm": 0.36400866508483887, + "learning_rate": 0.0004337044366706983, + "loss": 2.8087133407592773, + "step": 48980 + }, + { + "epoch": 0.24495, + "grad_norm": 0.36324167251586914, + "learning_rate": 0.0004336775299252087, + "loss": 2.7906747817993165, + "step": 48990 + }, + { + "epoch": 0.245, + "grad_norm": 0.3361440896987915, + "learning_rate": 0.00043365061855563867, + "loss": 2.795465850830078, + "step": 49000 + }, + { + "epoch": 0.24505, + "grad_norm": 0.32044899463653564, + "learning_rate": 0.00043362370256266555, + "loss": 2.78448486328125, + "step": 49010 + }, + { + "epoch": 0.2451, + "grad_norm": 0.33760952949523926, + "learning_rate": 0.0004335967819469669, + "loss": 2.7979522705078126, + "step": 49020 + }, + { + "epoch": 0.24515, + "grad_norm": 0.33007851243019104, + "learning_rate": 0.0004335698567092205, + "loss": 2.8401308059692383, + "step": 49030 + }, + { + "epoch": 0.2452, + "grad_norm": 0.34174349904060364, + "learning_rate": 0.00043354292685010424, + "loss": 2.8059436798095705, + "step": 49040 + }, + { + "epoch": 0.24525, + "grad_norm": 0.3681221008300781, + "learning_rate": 0.00043351599237029595, + "loss": 2.774185562133789, + "step": 49050 + }, + { + "epoch": 0.2453, + "grad_norm": 0.35944855213165283, + "learning_rate": 0.0004334890532704737, + "loss": 2.8339302062988283, + "step": 49060 + }, + { + "epoch": 0.24535, + "grad_norm": 0.32926711440086365, + "learning_rate": 0.00043346210955131595, + "loss": 2.8230167388916017, + "step": 49070 + }, + { + "epoch": 0.2454, + "grad_norm": 0.3326578438282013, + "learning_rate": 0.0004334351612135007, + "loss": 2.811781311035156, + "step": 49080 + }, + { + "epoch": 0.24545, + "grad_norm": 0.3576635718345642, + "learning_rate": 0.00043340820825770647, + "loss": 2.8158443450927733, + "step": 49090 + }, + { + "epoch": 0.2455, + "grad_norm": 0.3159853219985962, + "learning_rate": 0.0004333812506846119, + "loss": 2.7994867324829102, + "step": 49100 + }, + { + "epoch": 0.24555, + "grad_norm": 0.3416156470775604, + "learning_rate": 0.00043335428849489556, + "loss": 2.834197998046875, + "step": 49110 + }, + { + "epoch": 0.2456, + "grad_norm": 0.31884533166885376, + "learning_rate": 0.0004333273216892362, + "loss": 2.8359851837158203, + "step": 49120 + }, + { + "epoch": 0.24565, + "grad_norm": 0.3531423509120941, + "learning_rate": 0.00043330035026831274, + "loss": 2.8251110076904298, + "step": 49130 + }, + { + "epoch": 0.2457, + "grad_norm": 0.3680878281593323, + "learning_rate": 0.00043327337423280417, + "loss": 2.80633487701416, + "step": 49140 + }, + { + "epoch": 0.24575, + "grad_norm": 0.3637607991695404, + "learning_rate": 0.0004332463935833897, + "loss": 2.8307216644287108, + "step": 49150 + }, + { + "epoch": 0.2458, + "grad_norm": 0.3918369710445404, + "learning_rate": 0.0004332194083207484, + "loss": 2.8308948516845702, + "step": 49160 + }, + { + "epoch": 0.24585, + "grad_norm": 0.3457867503166199, + "learning_rate": 0.0004331924184455598, + "loss": 2.8046483993530273, + "step": 49170 + }, + { + "epoch": 0.2459, + "grad_norm": 0.3311011791229248, + "learning_rate": 0.00043316542395850324, + "loss": 2.8340192794799806, + "step": 49180 + }, + { + "epoch": 0.24595, + "grad_norm": 0.3785024583339691, + "learning_rate": 0.00043313842486025837, + "loss": 2.852337646484375, + "step": 49190 + }, + { + "epoch": 0.246, + "grad_norm": 0.3555063009262085, + "learning_rate": 0.0004331114211515049, + "loss": 2.7664180755615235, + "step": 49200 + }, + { + "epoch": 0.24605, + "grad_norm": 0.3516928255558014, + "learning_rate": 0.0004330844128329226, + "loss": 2.8692760467529297, + "step": 49210 + }, + { + "epoch": 0.2461, + "grad_norm": 0.34615573287010193, + "learning_rate": 0.00043305739990519147, + "loss": 2.835576820373535, + "step": 49220 + }, + { + "epoch": 0.24615, + "grad_norm": 0.3604532480239868, + "learning_rate": 0.0004330303823689915, + "loss": 2.854861259460449, + "step": 49230 + }, + { + "epoch": 0.2462, + "grad_norm": 0.339932382106781, + "learning_rate": 0.0004330033602250029, + "loss": 2.8346330642700197, + "step": 49240 + }, + { + "epoch": 0.24625, + "grad_norm": 0.34496352076530457, + "learning_rate": 0.00043297633347390583, + "loss": 2.7890741348266603, + "step": 49250 + }, + { + "epoch": 0.2463, + "grad_norm": 0.3582020401954651, + "learning_rate": 0.0004329493021163808, + "loss": 2.811334991455078, + "step": 49260 + }, + { + "epoch": 0.24635, + "grad_norm": 0.3669278025627136, + "learning_rate": 0.0004329222661531084, + "loss": 2.8250957489013673, + "step": 49270 + }, + { + "epoch": 0.2464, + "grad_norm": 0.35555458068847656, + "learning_rate": 0.0004328952255847691, + "loss": 2.8008115768432615, + "step": 49280 + }, + { + "epoch": 0.24645, + "grad_norm": 0.3427877724170685, + "learning_rate": 0.0004328681804120438, + "loss": 2.7877809524536135, + "step": 49290 + }, + { + "epoch": 0.2465, + "grad_norm": 0.3170597553253174, + "learning_rate": 0.0004328411306356133, + "loss": 2.843174934387207, + "step": 49300 + }, + { + "epoch": 0.24655, + "grad_norm": 0.3698081076145172, + "learning_rate": 0.0004328140762561584, + "loss": 2.842554473876953, + "step": 49310 + }, + { + "epoch": 0.2466, + "grad_norm": 0.33749136328697205, + "learning_rate": 0.0004327870172743605, + "loss": 2.7848316192626954, + "step": 49320 + }, + { + "epoch": 0.24665, + "grad_norm": 0.3387846052646637, + "learning_rate": 0.0004327599536909006, + "loss": 2.802268409729004, + "step": 49330 + }, + { + "epoch": 0.2467, + "grad_norm": 0.35204875469207764, + "learning_rate": 0.0004327328855064601, + "loss": 2.820256805419922, + "step": 49340 + }, + { + "epoch": 0.24675, + "grad_norm": 0.346565306186676, + "learning_rate": 0.00043270581272172046, + "loss": 2.7493946075439455, + "step": 49350 + }, + { + "epoch": 0.2468, + "grad_norm": 0.3398755192756653, + "learning_rate": 0.00043267873533736314, + "loss": 2.8065237045288085, + "step": 49360 + }, + { + "epoch": 0.24685, + "grad_norm": 0.36584052443504333, + "learning_rate": 0.00043265165335406994, + "loss": 2.7588712692260744, + "step": 49370 + }, + { + "epoch": 0.2469, + "grad_norm": 0.3355970084667206, + "learning_rate": 0.00043262456677252247, + "loss": 2.8341569900512695, + "step": 49380 + }, + { + "epoch": 0.24695, + "grad_norm": 0.3365468680858612, + "learning_rate": 0.0004325974755934029, + "loss": 2.792172431945801, + "step": 49390 + }, + { + "epoch": 0.247, + "grad_norm": 0.3279480040073395, + "learning_rate": 0.000432570379817393, + "loss": 2.8876033782958985, + "step": 49400 + }, + { + "epoch": 0.24705, + "grad_norm": 0.33197879791259766, + "learning_rate": 0.0004325432794451751, + "loss": 2.785018730163574, + "step": 49410 + }, + { + "epoch": 0.2471, + "grad_norm": 0.32517722249031067, + "learning_rate": 0.0004325161744774313, + "loss": 2.833082580566406, + "step": 49420 + }, + { + "epoch": 0.24715, + "grad_norm": 0.32524818181991577, + "learning_rate": 0.00043248906491484394, + "loss": 2.831659507751465, + "step": 49430 + }, + { + "epoch": 0.2472, + "grad_norm": 0.3335898816585541, + "learning_rate": 0.00043246195075809577, + "loss": 2.821773910522461, + "step": 49440 + }, + { + "epoch": 0.24725, + "grad_norm": 0.3300182819366455, + "learning_rate": 0.00043243483200786916, + "loss": 2.793663215637207, + "step": 49450 + }, + { + "epoch": 0.2473, + "grad_norm": 0.34463366866111755, + "learning_rate": 0.0004324077086648468, + "loss": 2.855737495422363, + "step": 49460 + }, + { + "epoch": 0.24735, + "grad_norm": 0.32611313462257385, + "learning_rate": 0.00043238058072971166, + "loss": 2.8629459381103515, + "step": 49470 + }, + { + "epoch": 0.2474, + "grad_norm": 0.33105576038360596, + "learning_rate": 0.00043235344820314654, + "loss": 2.7883010864257813, + "step": 49480 + }, + { + "epoch": 0.24745, + "grad_norm": 0.35682836174964905, + "learning_rate": 0.00043232631108583463, + "loss": 2.7984458923339846, + "step": 49490 + }, + { + "epoch": 0.2475, + "grad_norm": 0.3553582727909088, + "learning_rate": 0.00043229916937845904, + "loss": 2.811069679260254, + "step": 49500 + }, + { + "epoch": 0.24755, + "grad_norm": 0.3644723892211914, + "learning_rate": 0.0004322720230817031, + "loss": 2.796456146240234, + "step": 49510 + }, + { + "epoch": 0.2476, + "grad_norm": 0.34166279435157776, + "learning_rate": 0.0004322448721962502, + "loss": 2.8214141845703127, + "step": 49520 + }, + { + "epoch": 0.24765, + "grad_norm": 0.37134575843811035, + "learning_rate": 0.00043221771672278386, + "loss": 2.8195821762084963, + "step": 49530 + }, + { + "epoch": 0.2477, + "grad_norm": 0.35175782442092896, + "learning_rate": 0.0004321905566619877, + "loss": 2.87740421295166, + "step": 49540 + }, + { + "epoch": 0.24775, + "grad_norm": 0.3522413671016693, + "learning_rate": 0.0004321633920145455, + "loss": 2.8102228164672853, + "step": 49550 + }, + { + "epoch": 0.2478, + "grad_norm": 0.3576362133026123, + "learning_rate": 0.0004321362227811412, + "loss": 2.8304895401000976, + "step": 49560 + }, + { + "epoch": 0.24785, + "grad_norm": 0.32103824615478516, + "learning_rate": 0.0004321090489624586, + "loss": 2.8829484939575196, + "step": 49570 + }, + { + "epoch": 0.2479, + "grad_norm": 0.34571927785873413, + "learning_rate": 0.00043208187055918194, + "loss": 2.785100555419922, + "step": 49580 + }, + { + "epoch": 0.24795, + "grad_norm": 0.33486026525497437, + "learning_rate": 0.0004320546875719954, + "loss": 2.774957847595215, + "step": 49590 + }, + { + "epoch": 0.248, + "grad_norm": 0.3593336045742035, + "learning_rate": 0.0004320275000015833, + "loss": 2.840643310546875, + "step": 49600 + }, + { + "epoch": 0.24805, + "grad_norm": 0.3612293601036072, + "learning_rate": 0.00043200030784863014, + "loss": 2.828611373901367, + "step": 49610 + }, + { + "epoch": 0.2481, + "grad_norm": 0.36656951904296875, + "learning_rate": 0.00043197311111382045, + "loss": 2.796254348754883, + "step": 49620 + }, + { + "epoch": 0.24815, + "grad_norm": 0.38805028796195984, + "learning_rate": 0.0004319459097978389, + "loss": 2.8506677627563475, + "step": 49630 + }, + { + "epoch": 0.2482, + "grad_norm": 0.3327005207538605, + "learning_rate": 0.00043191870390137024, + "loss": 2.8805660247802733, + "step": 49640 + }, + { + "epoch": 0.24825, + "grad_norm": 0.345451295375824, + "learning_rate": 0.0004318914934250995, + "loss": 2.8165447235107424, + "step": 49650 + }, + { + "epoch": 0.2483, + "grad_norm": 0.34077683091163635, + "learning_rate": 0.0004318642783697115, + "loss": 2.8108171463012694, + "step": 49660 + }, + { + "epoch": 0.24835, + "grad_norm": 0.33974480628967285, + "learning_rate": 0.0004318370587358916, + "loss": 2.8336669921875, + "step": 49670 + }, + { + "epoch": 0.2484, + "grad_norm": 0.37030085921287537, + "learning_rate": 0.00043180983452432504, + "loss": 2.846589279174805, + "step": 49680 + }, + { + "epoch": 0.24845, + "grad_norm": 0.2915101945400238, + "learning_rate": 0.00043178260573569703, + "loss": 2.8131376266479493, + "step": 49690 + }, + { + "epoch": 0.2485, + "grad_norm": 0.33183252811431885, + "learning_rate": 0.0004317553723706931, + "loss": 2.7763877868652345, + "step": 49700 + }, + { + "epoch": 0.24855, + "grad_norm": 0.34348416328430176, + "learning_rate": 0.0004317281344299989, + "loss": 2.814225769042969, + "step": 49710 + }, + { + "epoch": 0.2486, + "grad_norm": 0.34940287470817566, + "learning_rate": 0.00043170089191430015, + "loss": 2.833464431762695, + "step": 49720 + }, + { + "epoch": 0.24865, + "grad_norm": 0.3551843464374542, + "learning_rate": 0.0004316736448242827, + "loss": 2.8043872833251955, + "step": 49730 + }, + { + "epoch": 0.2487, + "grad_norm": 0.36729520559310913, + "learning_rate": 0.0004316463931606324, + "loss": 2.801437568664551, + "step": 49740 + }, + { + "epoch": 0.24875, + "grad_norm": 0.3325875997543335, + "learning_rate": 0.0004316191369240354, + "loss": 2.8581945419311525, + "step": 49750 + }, + { + "epoch": 0.2488, + "grad_norm": 0.3370814919471741, + "learning_rate": 0.0004315918761151778, + "loss": 2.837910461425781, + "step": 49760 + }, + { + "epoch": 0.24885, + "grad_norm": 0.31870272755622864, + "learning_rate": 0.00043156461073474593, + "loss": 2.7623802185058595, + "step": 49770 + }, + { + "epoch": 0.2489, + "grad_norm": 0.36083725094795227, + "learning_rate": 0.0004315373407834263, + "loss": 2.711138916015625, + "step": 49780 + }, + { + "epoch": 0.24895, + "grad_norm": 0.3379393219947815, + "learning_rate": 0.0004315100662619052, + "loss": 2.8503684997558594, + "step": 49790 + }, + { + "epoch": 0.249, + "grad_norm": 0.3470093607902527, + "learning_rate": 0.00043148278717086946, + "loss": 2.8212814331054688, + "step": 49800 + }, + { + "epoch": 0.24905, + "grad_norm": 0.3387058675289154, + "learning_rate": 0.00043145550351100575, + "loss": 2.849533271789551, + "step": 49810 + }, + { + "epoch": 0.2491, + "grad_norm": 0.3201335370540619, + "learning_rate": 0.00043142821528300093, + "loss": 2.74930477142334, + "step": 49820 + }, + { + "epoch": 0.24915, + "grad_norm": 0.3655514717102051, + "learning_rate": 0.00043140092248754205, + "loss": 2.783322334289551, + "step": 49830 + }, + { + "epoch": 0.2492, + "grad_norm": 0.32339000701904297, + "learning_rate": 0.0004313736251253161, + "loss": 2.7799589157104494, + "step": 49840 + }, + { + "epoch": 0.24925, + "grad_norm": 0.34463056921958923, + "learning_rate": 0.0004313463231970104, + "loss": 2.7743623733520506, + "step": 49850 + }, + { + "epoch": 0.2493, + "grad_norm": 0.37399622797966003, + "learning_rate": 0.0004313190167033122, + "loss": 2.8357709884643554, + "step": 49860 + }, + { + "epoch": 0.24935, + "grad_norm": 0.3185892105102539, + "learning_rate": 0.00043129170564490903, + "loss": 2.827295684814453, + "step": 49870 + }, + { + "epoch": 0.2494, + "grad_norm": 0.3575274646282196, + "learning_rate": 0.0004312643900224883, + "loss": 2.829377555847168, + "step": 49880 + }, + { + "epoch": 0.24945, + "grad_norm": 0.3505176603794098, + "learning_rate": 0.00043123706983673785, + "loss": 2.8070974349975586, + "step": 49890 + }, + { + "epoch": 0.2495, + "grad_norm": 0.33039242029190063, + "learning_rate": 0.00043120974508834524, + "loss": 2.7794647216796875, + "step": 49900 + }, + { + "epoch": 0.24955, + "grad_norm": 0.3665037751197815, + "learning_rate": 0.0004311824157779986, + "loss": 2.8234769821166994, + "step": 49910 + }, + { + "epoch": 0.2496, + "grad_norm": 0.33681541681289673, + "learning_rate": 0.0004311550819063859, + "loss": 2.8401662826538088, + "step": 49920 + }, + { + "epoch": 0.24965, + "grad_norm": 0.356679230928421, + "learning_rate": 0.00043112774347419514, + "loss": 2.8170886993408204, + "step": 49930 + }, + { + "epoch": 0.2497, + "grad_norm": 0.34736230969429016, + "learning_rate": 0.0004311004004821146, + "loss": 2.8560203552246093, + "step": 49940 + }, + { + "epoch": 0.24975, + "grad_norm": 0.3336727023124695, + "learning_rate": 0.00043107305293083276, + "loss": 2.7729101181030273, + "step": 49950 + }, + { + "epoch": 0.2498, + "grad_norm": 0.31949013471603394, + "learning_rate": 0.00043104570082103805, + "loss": 2.838125228881836, + "step": 49960 + }, + { + "epoch": 0.24985, + "grad_norm": 0.3465896248817444, + "learning_rate": 0.000431018344153419, + "loss": 2.723069763183594, + "step": 49970 + }, + { + "epoch": 0.2499, + "grad_norm": 0.33893364667892456, + "learning_rate": 0.00043099098292866437, + "loss": 2.819744873046875, + "step": 49980 + }, + { + "epoch": 0.24995, + "grad_norm": 0.371565580368042, + "learning_rate": 0.0004309636171474629, + "loss": 2.732965087890625, + "step": 49990 + }, + { + "epoch": 0.25, + "grad_norm": 0.34269893169403076, + "learning_rate": 0.00043093624681050366, + "loss": 2.7652788162231445, + "step": 50000 + }, + { + "epoch": 0.25005, + "grad_norm": 0.34083858132362366, + "learning_rate": 0.0004309088719184755, + "loss": 2.8045373916625977, + "step": 50010 + }, + { + "epoch": 0.2501, + "grad_norm": 0.3636561632156372, + "learning_rate": 0.00043088149247206783, + "loss": 2.791265106201172, + "step": 50020 + }, + { + "epoch": 0.25015, + "grad_norm": 0.4009120762348175, + "learning_rate": 0.00043085410847196973, + "loss": 2.8179031372070313, + "step": 50030 + }, + { + "epoch": 0.2502, + "grad_norm": 0.3488523066043854, + "learning_rate": 0.00043082671991887065, + "loss": 2.780444526672363, + "step": 50040 + }, + { + "epoch": 0.25025, + "grad_norm": 0.3181380033493042, + "learning_rate": 0.00043079932681346013, + "loss": 2.8258172988891603, + "step": 50050 + }, + { + "epoch": 0.2503, + "grad_norm": 0.31111255288124084, + "learning_rate": 0.00043077192915642773, + "loss": 2.8196128845214843, + "step": 50060 + }, + { + "epoch": 0.25035, + "grad_norm": 0.3463522493839264, + "learning_rate": 0.0004307445269484632, + "loss": 2.817000961303711, + "step": 50070 + }, + { + "epoch": 0.2504, + "grad_norm": 0.31694495677948, + "learning_rate": 0.00043071712019025646, + "loss": 2.7908573150634766, + "step": 50080 + }, + { + "epoch": 0.25045, + "grad_norm": 0.3418474495410919, + "learning_rate": 0.0004306897088824974, + "loss": 2.824811553955078, + "step": 50090 + }, + { + "epoch": 0.2505, + "grad_norm": 0.34406501054763794, + "learning_rate": 0.0004306622930258762, + "loss": 2.8056636810302735, + "step": 50100 + }, + { + "epoch": 0.25055, + "grad_norm": 0.33620890974998474, + "learning_rate": 0.0004306348726210829, + "loss": 2.810621643066406, + "step": 50110 + }, + { + "epoch": 0.2506, + "grad_norm": 0.3295957148075104, + "learning_rate": 0.0004306074476688079, + "loss": 2.7990629196166994, + "step": 50120 + }, + { + "epoch": 0.25065, + "grad_norm": 0.33650869131088257, + "learning_rate": 0.00043058001816974166, + "loss": 2.8153976440429687, + "step": 50130 + }, + { + "epoch": 0.2507, + "grad_norm": 0.31780970096588135, + "learning_rate": 0.0004305525841245746, + "loss": 2.822012710571289, + "step": 50140 + }, + { + "epoch": 0.25075, + "grad_norm": 0.4035550057888031, + "learning_rate": 0.00043052514553399747, + "loss": 2.8197395324707033, + "step": 50150 + }, + { + "epoch": 0.2508, + "grad_norm": 0.37095531821250916, + "learning_rate": 0.000430497702398701, + "loss": 2.7867202758789062, + "step": 50160 + }, + { + "epoch": 0.25085, + "grad_norm": 0.3307145833969116, + "learning_rate": 0.00043047025471937614, + "loss": 2.7791894912719726, + "step": 50170 + }, + { + "epoch": 0.2509, + "grad_norm": 0.32393014430999756, + "learning_rate": 0.00043044280249671373, + "loss": 2.9209173202514647, + "step": 50180 + }, + { + "epoch": 0.25095, + "grad_norm": 0.3438321053981781, + "learning_rate": 0.00043041534573140506, + "loss": 2.8197879791259766, + "step": 50190 + }, + { + "epoch": 0.251, + "grad_norm": 0.3510034680366516, + "learning_rate": 0.00043038788442414125, + "loss": 2.8089590072631836, + "step": 50200 + }, + { + "epoch": 0.25105, + "grad_norm": 0.3281750977039337, + "learning_rate": 0.0004303604185756136, + "loss": 2.8039533615112306, + "step": 50210 + }, + { + "epoch": 0.2511, + "grad_norm": 0.39079150557518005, + "learning_rate": 0.0004303329481865136, + "loss": 2.8545385360717774, + "step": 50220 + }, + { + "epoch": 0.25115, + "grad_norm": 0.3288353681564331, + "learning_rate": 0.0004303054732575329, + "loss": 2.8261470794677734, + "step": 50230 + }, + { + "epoch": 0.2512, + "grad_norm": 0.3421873450279236, + "learning_rate": 0.00043027799378936304, + "loss": 2.8125595092773437, + "step": 50240 + }, + { + "epoch": 0.25125, + "grad_norm": 0.3378683626651764, + "learning_rate": 0.00043025050978269595, + "loss": 2.8345741271972655, + "step": 50250 + }, + { + "epoch": 0.2513, + "grad_norm": 0.34450986981391907, + "learning_rate": 0.00043022302123822356, + "loss": 2.716776466369629, + "step": 50260 + }, + { + "epoch": 0.25135, + "grad_norm": 0.3581215441226959, + "learning_rate": 0.0004301955281566377, + "loss": 2.8028932571411134, + "step": 50270 + }, + { + "epoch": 0.2514, + "grad_norm": 0.33950096368789673, + "learning_rate": 0.00043016803053863063, + "loss": 2.7479257583618164, + "step": 50280 + }, + { + "epoch": 0.25145, + "grad_norm": 0.3396832346916199, + "learning_rate": 0.0004301405283848947, + "loss": 2.821560096740723, + "step": 50290 + }, + { + "epoch": 0.2515, + "grad_norm": 0.327132910490036, + "learning_rate": 0.00043011302169612204, + "loss": 2.8094013214111326, + "step": 50300 + }, + { + "epoch": 0.25155, + "grad_norm": 0.352031409740448, + "learning_rate": 0.0004300855104730053, + "loss": 2.7659107208251954, + "step": 50310 + }, + { + "epoch": 0.2516, + "grad_norm": 0.3355056941509247, + "learning_rate": 0.0004300579947162371, + "loss": 2.821023941040039, + "step": 50320 + }, + { + "epoch": 0.25165, + "grad_norm": 0.3646096885204315, + "learning_rate": 0.00043003047442651, + "loss": 2.7868528366088867, + "step": 50330 + }, + { + "epoch": 0.2517, + "grad_norm": 0.36617979407310486, + "learning_rate": 0.00043000294960451695, + "loss": 2.769999313354492, + "step": 50340 + }, + { + "epoch": 0.25175, + "grad_norm": 0.39056968688964844, + "learning_rate": 0.0004299754202509508, + "loss": 2.753041076660156, + "step": 50350 + }, + { + "epoch": 0.2518, + "grad_norm": 0.34018218517303467, + "learning_rate": 0.00042994788636650474, + "loss": 2.818416213989258, + "step": 50360 + }, + { + "epoch": 0.25185, + "grad_norm": 0.3889447748661041, + "learning_rate": 0.0004299203479518718, + "loss": 2.7725452423095702, + "step": 50370 + }, + { + "epoch": 0.2519, + "grad_norm": 0.3281453847885132, + "learning_rate": 0.00042989280500774533, + "loss": 2.7903451919555664, + "step": 50380 + }, + { + "epoch": 0.25195, + "grad_norm": 0.3222460448741913, + "learning_rate": 0.00042986525753481864, + "loss": 2.7946441650390623, + "step": 50390 + }, + { + "epoch": 0.252, + "grad_norm": 0.32761695981025696, + "learning_rate": 0.00042983770553378534, + "loss": 2.8299564361572265, + "step": 50400 + }, + { + "epoch": 0.25205, + "grad_norm": 0.35779866576194763, + "learning_rate": 0.0004298101490053389, + "loss": 2.829323959350586, + "step": 50410 + }, + { + "epoch": 0.2521, + "grad_norm": 0.3435649573802948, + "learning_rate": 0.0004297825879501733, + "loss": 2.7758153915405273, + "step": 50420 + }, + { + "epoch": 0.25215, + "grad_norm": 0.3634573817253113, + "learning_rate": 0.0004297550223689822, + "loss": 2.7357456207275392, + "step": 50430 + }, + { + "epoch": 0.2522, + "grad_norm": 0.32486680150032043, + "learning_rate": 0.00042972745226245955, + "loss": 2.752466583251953, + "step": 50440 + }, + { + "epoch": 0.25225, + "grad_norm": 0.3376854956150055, + "learning_rate": 0.0004296998776312996, + "loss": 2.770405578613281, + "step": 50450 + }, + { + "epoch": 0.2523, + "grad_norm": 0.32690587639808655, + "learning_rate": 0.00042967229847619637, + "loss": 2.834158706665039, + "step": 50460 + }, + { + "epoch": 0.25235, + "grad_norm": 0.33874374628067017, + "learning_rate": 0.00042964471479784415, + "loss": 2.8581714630126953, + "step": 50470 + }, + { + "epoch": 0.2524, + "grad_norm": 0.3331444263458252, + "learning_rate": 0.00042961712659693755, + "loss": 2.8115518569946287, + "step": 50480 + }, + { + "epoch": 0.25245, + "grad_norm": 0.3234788775444031, + "learning_rate": 0.00042958953387417097, + "loss": 2.817202568054199, + "step": 50490 + }, + { + "epoch": 0.2525, + "grad_norm": 0.3269977867603302, + "learning_rate": 0.0004295619366302391, + "loss": 2.7999624252319335, + "step": 50500 + }, + { + "epoch": 0.25255, + "grad_norm": 0.3520045578479767, + "learning_rate": 0.0004295343348658366, + "loss": 2.779885482788086, + "step": 50510 + }, + { + "epoch": 0.2526, + "grad_norm": 0.3129745125770569, + "learning_rate": 0.00042950672858165846, + "loss": 2.793894577026367, + "step": 50520 + }, + { + "epoch": 0.25265, + "grad_norm": 0.3411996066570282, + "learning_rate": 0.00042947911777839964, + "loss": 2.827677536010742, + "step": 50530 + }, + { + "epoch": 0.2527, + "grad_norm": 0.3328293263912201, + "learning_rate": 0.0004294515024567552, + "loss": 2.8258527755737304, + "step": 50540 + }, + { + "epoch": 0.25275, + "grad_norm": 0.342388391494751, + "learning_rate": 0.0004294238826174204, + "loss": 2.842106819152832, + "step": 50550 + }, + { + "epoch": 0.2528, + "grad_norm": 0.3286673128604889, + "learning_rate": 0.00042939625826109054, + "loss": 2.8269012451171873, + "step": 50560 + }, + { + "epoch": 0.25285, + "grad_norm": 0.3383522927761078, + "learning_rate": 0.0004293686293884611, + "loss": 2.876633644104004, + "step": 50570 + }, + { + "epoch": 0.2529, + "grad_norm": 0.3471941351890564, + "learning_rate": 0.00042934099600022756, + "loss": 2.785390281677246, + "step": 50580 + }, + { + "epoch": 0.25295, + "grad_norm": 0.3545655608177185, + "learning_rate": 0.0004293133580970857, + "loss": 2.8165157318115233, + "step": 50590 + }, + { + "epoch": 0.253, + "grad_norm": 0.3529602885246277, + "learning_rate": 0.00042928571567973125, + "loss": 2.8596853256225585, + "step": 50600 + }, + { + "epoch": 0.25305, + "grad_norm": 0.3504602313041687, + "learning_rate": 0.0004292580687488601, + "loss": 2.8520586013793947, + "step": 50610 + }, + { + "epoch": 0.2531, + "grad_norm": 0.31080976128578186, + "learning_rate": 0.0004292304173051682, + "loss": 2.785508728027344, + "step": 50620 + }, + { + "epoch": 0.25315, + "grad_norm": 0.34769588708877563, + "learning_rate": 0.00042920276134935184, + "loss": 2.8386526107788086, + "step": 50630 + }, + { + "epoch": 0.2532, + "grad_norm": 0.36373165249824524, + "learning_rate": 0.0004291751008821071, + "loss": 2.7895524978637694, + "step": 50640 + }, + { + "epoch": 0.25325, + "grad_norm": 0.36389103531837463, + "learning_rate": 0.0004291474359041304, + "loss": 2.806559371948242, + "step": 50650 + }, + { + "epoch": 0.2533, + "grad_norm": 0.3182878792285919, + "learning_rate": 0.00042911976641611823, + "loss": 2.8528032302856445, + "step": 50660 + }, + { + "epoch": 0.25335, + "grad_norm": 0.35502496361732483, + "learning_rate": 0.00042909209241876714, + "loss": 2.847382736206055, + "step": 50670 + }, + { + "epoch": 0.2534, + "grad_norm": 0.3291883170604706, + "learning_rate": 0.00042906441391277374, + "loss": 2.768375587463379, + "step": 50680 + }, + { + "epoch": 0.25345, + "grad_norm": 0.34431347250938416, + "learning_rate": 0.000429036730898835, + "loss": 2.803565788269043, + "step": 50690 + }, + { + "epoch": 0.2535, + "grad_norm": 0.3357957601547241, + "learning_rate": 0.00042900904337764766, + "loss": 2.8027950286865235, + "step": 50700 + }, + { + "epoch": 0.25355, + "grad_norm": 0.3135519027709961, + "learning_rate": 0.000428981351349909, + "loss": 2.8403514862060546, + "step": 50710 + }, + { + "epoch": 0.2536, + "grad_norm": 0.3471642732620239, + "learning_rate": 0.0004289536548163159, + "loss": 2.7898216247558594, + "step": 50720 + }, + { + "epoch": 0.25365, + "grad_norm": 0.3432447910308838, + "learning_rate": 0.00042892595377756575, + "loss": 2.830305290222168, + "step": 50730 + }, + { + "epoch": 0.2537, + "grad_norm": 0.33299019932746887, + "learning_rate": 0.00042889824823435596, + "loss": 2.7998043060302735, + "step": 50740 + }, + { + "epoch": 0.25375, + "grad_norm": 0.35560205578804016, + "learning_rate": 0.00042887053818738396, + "loss": 2.796978569030762, + "step": 50750 + }, + { + "epoch": 0.2538, + "grad_norm": 0.3060586452484131, + "learning_rate": 0.00042884282363734726, + "loss": 2.8268405914306642, + "step": 50760 + }, + { + "epoch": 0.25385, + "grad_norm": 0.3573389947414398, + "learning_rate": 0.00042881510458494377, + "loss": 2.819052505493164, + "step": 50770 + }, + { + "epoch": 0.2539, + "grad_norm": 0.3341851234436035, + "learning_rate": 0.0004287873810308712, + "loss": 2.8301021575927736, + "step": 50780 + }, + { + "epoch": 0.25395, + "grad_norm": 0.3299468755722046, + "learning_rate": 0.0004287596529758275, + "loss": 2.8261993408203123, + "step": 50790 + }, + { + "epoch": 0.254, + "grad_norm": 0.3704557716846466, + "learning_rate": 0.0004287319204205107, + "loss": 2.7991302490234373, + "step": 50800 + }, + { + "epoch": 0.25405, + "grad_norm": 0.3631395995616913, + "learning_rate": 0.000428704183365619, + "loss": 2.8007007598876954, + "step": 50810 + }, + { + "epoch": 0.2541, + "grad_norm": 0.3312965929508209, + "learning_rate": 0.00042867644181185074, + "loss": 2.8464818954467774, + "step": 50820 + }, + { + "epoch": 0.25415, + "grad_norm": 0.3973892629146576, + "learning_rate": 0.00042864869575990423, + "loss": 2.8233150482177733, + "step": 50830 + }, + { + "epoch": 0.2542, + "grad_norm": 0.3181581497192383, + "learning_rate": 0.000428620945210478, + "loss": 2.8211696624755858, + "step": 50840 + }, + { + "epoch": 0.25425, + "grad_norm": 0.37942469120025635, + "learning_rate": 0.0004285931901642706, + "loss": 2.788326644897461, + "step": 50850 + }, + { + "epoch": 0.2543, + "grad_norm": 0.38046330213546753, + "learning_rate": 0.0004285654306219809, + "loss": 2.817148780822754, + "step": 50860 + }, + { + "epoch": 0.25435, + "grad_norm": 0.3316183090209961, + "learning_rate": 0.00042853766658430766, + "loss": 2.833596038818359, + "step": 50870 + }, + { + "epoch": 0.2544, + "grad_norm": 0.32730039954185486, + "learning_rate": 0.00042850989805194985, + "loss": 2.8166324615478517, + "step": 50880 + }, + { + "epoch": 0.25445, + "grad_norm": 0.3632960021495819, + "learning_rate": 0.00042848212502560656, + "loss": 2.7758563995361327, + "step": 50890 + }, + { + "epoch": 0.2545, + "grad_norm": 0.3523862659931183, + "learning_rate": 0.00042845434750597694, + "loss": 2.8271583557128905, + "step": 50900 + }, + { + "epoch": 0.25455, + "grad_norm": 0.3089194595813751, + "learning_rate": 0.00042842656549376036, + "loss": 2.8379770278930665, + "step": 50910 + }, + { + "epoch": 0.2546, + "grad_norm": 0.35963401198387146, + "learning_rate": 0.0004283987789896561, + "loss": 2.828339385986328, + "step": 50920 + }, + { + "epoch": 0.25465, + "grad_norm": 0.3688141405582428, + "learning_rate": 0.00042837098799436386, + "loss": 2.7986984252929688, + "step": 50930 + }, + { + "epoch": 0.2547, + "grad_norm": 0.33859413862228394, + "learning_rate": 0.00042834319250858316, + "loss": 2.797982406616211, + "step": 50940 + }, + { + "epoch": 0.25475, + "grad_norm": 0.3194320499897003, + "learning_rate": 0.00042831539253301376, + "loss": 2.7779680252075196, + "step": 50950 + }, + { + "epoch": 0.2548, + "grad_norm": 0.33464515209198, + "learning_rate": 0.00042828758806835553, + "loss": 2.7739261627197265, + "step": 50960 + }, + { + "epoch": 0.25485, + "grad_norm": 0.3528566360473633, + "learning_rate": 0.00042825977911530857, + "loss": 2.780558204650879, + "step": 50970 + }, + { + "epoch": 0.2549, + "grad_norm": 0.3659971356391907, + "learning_rate": 0.0004282319656745728, + "loss": 2.827273941040039, + "step": 50980 + }, + { + "epoch": 0.25495, + "grad_norm": 0.35566890239715576, + "learning_rate": 0.00042820414774684846, + "loss": 2.8365976333618166, + "step": 50990 + }, + { + "epoch": 0.255, + "grad_norm": 0.32469403743743896, + "learning_rate": 0.0004281763253328359, + "loss": 2.7555896759033205, + "step": 51000 + }, + { + "epoch": 0.25505, + "grad_norm": 0.405140221118927, + "learning_rate": 0.00042814849843323554, + "loss": 2.8373188018798827, + "step": 51010 + }, + { + "epoch": 0.2551, + "grad_norm": 0.3242659568786621, + "learning_rate": 0.0004281206670487479, + "loss": 2.790595054626465, + "step": 51020 + }, + { + "epoch": 0.25515, + "grad_norm": 0.3292941749095917, + "learning_rate": 0.0004280928311800737, + "loss": 2.787273406982422, + "step": 51030 + }, + { + "epoch": 0.2552, + "grad_norm": 0.35353267192840576, + "learning_rate": 0.0004280649908279136, + "loss": 2.806563949584961, + "step": 51040 + }, + { + "epoch": 0.25525, + "grad_norm": 0.3305346965789795, + "learning_rate": 0.0004280371459929686, + "loss": 2.7645950317382812, + "step": 51050 + }, + { + "epoch": 0.2553, + "grad_norm": 0.3248828649520874, + "learning_rate": 0.0004280092966759397, + "loss": 2.805154228210449, + "step": 51060 + }, + { + "epoch": 0.25535, + "grad_norm": 0.33314049243927, + "learning_rate": 0.0004279814428775278, + "loss": 2.826218032836914, + "step": 51070 + }, + { + "epoch": 0.2554, + "grad_norm": 0.3523172438144684, + "learning_rate": 0.00042795358459843435, + "loss": 2.8065349578857424, + "step": 51080 + }, + { + "epoch": 0.25545, + "grad_norm": 0.3895358443260193, + "learning_rate": 0.00042792572183936064, + "loss": 2.776751899719238, + "step": 51090 + }, + { + "epoch": 0.2555, + "grad_norm": 0.3498975932598114, + "learning_rate": 0.000427897854601008, + "loss": 2.8078939437866213, + "step": 51100 + }, + { + "epoch": 0.25555, + "grad_norm": 0.3214435279369354, + "learning_rate": 0.0004278699828840781, + "loss": 2.8359838485717774, + "step": 51110 + }, + { + "epoch": 0.2556, + "grad_norm": 0.3553864061832428, + "learning_rate": 0.00042784210668927254, + "loss": 2.8100540161132814, + "step": 51120 + }, + { + "epoch": 0.25565, + "grad_norm": 0.3224160373210907, + "learning_rate": 0.00042781422601729315, + "loss": 2.7979265213012696, + "step": 51130 + }, + { + "epoch": 0.2557, + "grad_norm": 0.3793036639690399, + "learning_rate": 0.00042778634086884177, + "loss": 2.8081226348876953, + "step": 51140 + }, + { + "epoch": 0.25575, + "grad_norm": 0.3487491309642792, + "learning_rate": 0.0004277584512446205, + "loss": 2.843044090270996, + "step": 51150 + }, + { + "epoch": 0.2558, + "grad_norm": 0.3380030393600464, + "learning_rate": 0.00042773055714533136, + "loss": 2.8737003326416017, + "step": 51160 + }, + { + "epoch": 0.25585, + "grad_norm": 0.3516733646392822, + "learning_rate": 0.00042770265857167665, + "loss": 2.794516754150391, + "step": 51170 + }, + { + "epoch": 0.2559, + "grad_norm": 0.39015287160873413, + "learning_rate": 0.00042767475552435874, + "loss": 2.8227750778198244, + "step": 51180 + }, + { + "epoch": 0.25595, + "grad_norm": 0.33238181471824646, + "learning_rate": 0.00042764684800408, + "loss": 2.8478977203369142, + "step": 51190 + }, + { + "epoch": 0.256, + "grad_norm": 0.3131408095359802, + "learning_rate": 0.00042761893601154306, + "loss": 2.7720972061157227, + "step": 51200 + }, + { + "epoch": 0.25605, + "grad_norm": 0.3468324840068817, + "learning_rate": 0.00042759101954745063, + "loss": 2.8273235321044923, + "step": 51210 + }, + { + "epoch": 0.2561, + "grad_norm": 0.32716605067253113, + "learning_rate": 0.0004275630986125054, + "loss": 2.7627069473266603, + "step": 51220 + }, + { + "epoch": 0.25615, + "grad_norm": 0.3299970030784607, + "learning_rate": 0.0004275351732074104, + "loss": 2.7763555526733397, + "step": 51230 + }, + { + "epoch": 0.2562, + "grad_norm": 0.3981025218963623, + "learning_rate": 0.00042750724333286854, + "loss": 2.7896421432495115, + "step": 51240 + }, + { + "epoch": 0.25625, + "grad_norm": 0.35391125082969666, + "learning_rate": 0.0004274793089895831, + "loss": 2.7747182846069336, + "step": 51250 + }, + { + "epoch": 0.2563, + "grad_norm": 0.37560200691223145, + "learning_rate": 0.0004274513701782572, + "loss": 2.7467723846435548, + "step": 51260 + }, + { + "epoch": 0.25635, + "grad_norm": 0.35709160566329956, + "learning_rate": 0.00042742342689959425, + "loss": 2.7967634201049805, + "step": 51270 + }, + { + "epoch": 0.2564, + "grad_norm": 0.375663161277771, + "learning_rate": 0.0004273954791542977, + "loss": 2.826349067687988, + "step": 51280 + }, + { + "epoch": 0.25645, + "grad_norm": 0.38453152775764465, + "learning_rate": 0.0004273675269430711, + "loss": 2.7641340255737306, + "step": 51290 + }, + { + "epoch": 0.2565, + "grad_norm": 0.3255309462547302, + "learning_rate": 0.0004273395702666183, + "loss": 2.7928945541381838, + "step": 51300 + }, + { + "epoch": 0.25655, + "grad_norm": 0.3359147012233734, + "learning_rate": 0.0004273116091256429, + "loss": 2.8156192779541014, + "step": 51310 + }, + { + "epoch": 0.2566, + "grad_norm": 0.522975742816925, + "learning_rate": 0.00042728364352084906, + "loss": 2.7897037506103515, + "step": 51320 + }, + { + "epoch": 0.25665, + "grad_norm": 0.41432228684425354, + "learning_rate": 0.0004272556734529405, + "loss": 2.82186222076416, + "step": 51330 + }, + { + "epoch": 0.2567, + "grad_norm": 0.3461954593658447, + "learning_rate": 0.0004272276989226216, + "loss": 2.8182674407958985, + "step": 51340 + }, + { + "epoch": 0.25675, + "grad_norm": 0.3519432246685028, + "learning_rate": 0.00042719971993059665, + "loss": 2.8238576889038085, + "step": 51350 + }, + { + "epoch": 0.2568, + "grad_norm": 0.3401944935321808, + "learning_rate": 0.0004271717364775699, + "loss": 2.806168556213379, + "step": 51360 + }, + { + "epoch": 0.25685, + "grad_norm": 0.33310016989707947, + "learning_rate": 0.00042714374856424585, + "loss": 2.8394210815429686, + "step": 51370 + }, + { + "epoch": 0.2569, + "grad_norm": 0.32972410321235657, + "learning_rate": 0.00042711575619132916, + "loss": 2.784037780761719, + "step": 51380 + }, + { + "epoch": 0.25695, + "grad_norm": 0.32461658120155334, + "learning_rate": 0.00042708775935952436, + "loss": 2.8214902877807617, + "step": 51390 + }, + { + "epoch": 0.257, + "grad_norm": 0.32599183917045593, + "learning_rate": 0.00042705975806953654, + "loss": 2.8123165130615235, + "step": 51400 + }, + { + "epoch": 0.25705, + "grad_norm": 0.34096580743789673, + "learning_rate": 0.0004270317523220705, + "loss": 2.797427749633789, + "step": 51410 + }, + { + "epoch": 0.2571, + "grad_norm": 0.3775445520877838, + "learning_rate": 0.0004270037421178312, + "loss": 2.835809898376465, + "step": 51420 + }, + { + "epoch": 0.25715, + "grad_norm": 0.34871652722358704, + "learning_rate": 0.00042697572745752386, + "loss": 2.8144662857055662, + "step": 51430 + }, + { + "epoch": 0.2572, + "grad_norm": 0.32548388838768005, + "learning_rate": 0.0004269477083418538, + "loss": 2.797979545593262, + "step": 51440 + }, + { + "epoch": 0.25725, + "grad_norm": 0.378066748380661, + "learning_rate": 0.00042691968477152637, + "loss": 2.854112434387207, + "step": 51450 + }, + { + "epoch": 0.2573, + "grad_norm": 0.3697924315929413, + "learning_rate": 0.00042689165674724707, + "loss": 2.8159229278564455, + "step": 51460 + }, + { + "epoch": 0.25735, + "grad_norm": 0.3361477255821228, + "learning_rate": 0.00042686362426972146, + "loss": 2.8113988876342773, + "step": 51470 + }, + { + "epoch": 0.2574, + "grad_norm": 0.33418476581573486, + "learning_rate": 0.00042683558733965535, + "loss": 2.816142463684082, + "step": 51480 + }, + { + "epoch": 0.25745, + "grad_norm": 0.34912437200546265, + "learning_rate": 0.0004268075459577544, + "loss": 2.8413166046142577, + "step": 51490 + }, + { + "epoch": 0.2575, + "grad_norm": 0.3533535301685333, + "learning_rate": 0.0004267795001247248, + "loss": 2.820369338989258, + "step": 51500 + }, + { + "epoch": 0.25755, + "grad_norm": 0.32984670996665955, + "learning_rate": 0.00042675144984127237, + "loss": 2.791619300842285, + "step": 51510 + }, + { + "epoch": 0.2576, + "grad_norm": 0.3742930591106415, + "learning_rate": 0.0004267233951081034, + "loss": 2.843683624267578, + "step": 51520 + }, + { + "epoch": 0.25765, + "grad_norm": 0.3846535384654999, + "learning_rate": 0.0004266953359259241, + "loss": 2.792150688171387, + "step": 51530 + }, + { + "epoch": 0.2577, + "grad_norm": 0.3179989457130432, + "learning_rate": 0.000426667272295441, + "loss": 2.7437765121459963, + "step": 51540 + }, + { + "epoch": 0.25775, + "grad_norm": 0.33833396434783936, + "learning_rate": 0.00042663920421736034, + "loss": 2.758404541015625, + "step": 51550 + }, + { + "epoch": 0.2578, + "grad_norm": 0.34731119871139526, + "learning_rate": 0.00042661113169238906, + "loss": 2.8036481857299806, + "step": 51560 + }, + { + "epoch": 0.25785, + "grad_norm": 0.3510054051876068, + "learning_rate": 0.0004265830547212336, + "loss": 2.7650203704833984, + "step": 51570 + }, + { + "epoch": 0.2579, + "grad_norm": 0.3881901502609253, + "learning_rate": 0.00042655497330460103, + "loss": 2.8016841888427733, + "step": 51580 + }, + { + "epoch": 0.25795, + "grad_norm": 0.34241852164268494, + "learning_rate": 0.0004265268874431981, + "loss": 2.81870174407959, + "step": 51590 + }, + { + "epoch": 0.258, + "grad_norm": 0.34204789996147156, + "learning_rate": 0.000426498797137732, + "loss": 2.8064144134521483, + "step": 51600 + }, + { + "epoch": 0.25805, + "grad_norm": 0.3404673635959625, + "learning_rate": 0.0004264707023889098, + "loss": 2.7777164459228514, + "step": 51610 + }, + { + "epoch": 0.2581, + "grad_norm": 0.33991676568984985, + "learning_rate": 0.00042644260319743896, + "loss": 2.816904830932617, + "step": 51620 + }, + { + "epoch": 0.25815, + "grad_norm": 0.3348560929298401, + "learning_rate": 0.00042641449956402666, + "loss": 2.759807586669922, + "step": 51630 + }, + { + "epoch": 0.2582, + "grad_norm": 0.3463434875011444, + "learning_rate": 0.0004263863914893805, + "loss": 2.823233795166016, + "step": 51640 + }, + { + "epoch": 0.25825, + "grad_norm": 0.3956027030944824, + "learning_rate": 0.0004263582789742082, + "loss": 2.860524559020996, + "step": 51650 + }, + { + "epoch": 0.2583, + "grad_norm": 0.3531897962093353, + "learning_rate": 0.00042633016201921737, + "loss": 2.7939279556274412, + "step": 51660 + }, + { + "epoch": 0.25835, + "grad_norm": 0.32629525661468506, + "learning_rate": 0.0004263020406251159, + "loss": 2.778402328491211, + "step": 51670 + }, + { + "epoch": 0.2584, + "grad_norm": 0.29054194688796997, + "learning_rate": 0.0004262739147926117, + "loss": 2.773983383178711, + "step": 51680 + }, + { + "epoch": 0.25845, + "grad_norm": 0.31625398993492126, + "learning_rate": 0.0004262457845224129, + "loss": 2.7878990173339844, + "step": 51690 + }, + { + "epoch": 0.2585, + "grad_norm": 0.3192029297351837, + "learning_rate": 0.00042621764981522767, + "loss": 2.830127143859863, + "step": 51700 + }, + { + "epoch": 0.25855, + "grad_norm": 0.3407951593399048, + "learning_rate": 0.00042618951067176426, + "loss": 2.7507015228271485, + "step": 51710 + }, + { + "epoch": 0.2586, + "grad_norm": 0.3291935920715332, + "learning_rate": 0.00042616136709273104, + "loss": 2.8093379974365233, + "step": 51720 + }, + { + "epoch": 0.25865, + "grad_norm": 0.3885956406593323, + "learning_rate": 0.0004261332190788366, + "loss": 2.799823760986328, + "step": 51730 + }, + { + "epoch": 0.2587, + "grad_norm": 0.3951147794723511, + "learning_rate": 0.0004261050666307896, + "loss": 2.791062927246094, + "step": 51740 + }, + { + "epoch": 0.25875, + "grad_norm": 0.3618467450141907, + "learning_rate": 0.00042607690974929874, + "loss": 2.7852266311645506, + "step": 51750 + }, + { + "epoch": 0.2588, + "grad_norm": 0.34050440788269043, + "learning_rate": 0.0004260487484350728, + "loss": 2.758548927307129, + "step": 51760 + }, + { + "epoch": 0.25885, + "grad_norm": 0.32361871004104614, + "learning_rate": 0.00042602058268882085, + "loss": 2.794818115234375, + "step": 51770 + }, + { + "epoch": 0.2589, + "grad_norm": 0.3511068522930145, + "learning_rate": 0.00042599241251125187, + "loss": 2.8467716217041015, + "step": 51780 + }, + { + "epoch": 0.25895, + "grad_norm": 0.34909820556640625, + "learning_rate": 0.0004259642379030751, + "loss": 2.8089895248413086, + "step": 51790 + }, + { + "epoch": 0.259, + "grad_norm": 0.3687528669834137, + "learning_rate": 0.00042593605886499976, + "loss": 2.802117347717285, + "step": 51800 + }, + { + "epoch": 0.25905, + "grad_norm": 0.3347025513648987, + "learning_rate": 0.00042590787539773537, + "loss": 2.8200614929199217, + "step": 51810 + }, + { + "epoch": 0.2591, + "grad_norm": 0.3219013512134552, + "learning_rate": 0.00042587968750199136, + "loss": 2.870022773742676, + "step": 51820 + }, + { + "epoch": 0.25915, + "grad_norm": 0.3581763505935669, + "learning_rate": 0.0004258514951784774, + "loss": 2.795486068725586, + "step": 51830 + }, + { + "epoch": 0.2592, + "grad_norm": 0.33550751209259033, + "learning_rate": 0.00042582329842790324, + "loss": 2.775157165527344, + "step": 51840 + }, + { + "epoch": 0.25925, + "grad_norm": 0.33584240078926086, + "learning_rate": 0.0004257950972509788, + "loss": 2.7382659912109375, + "step": 51850 + }, + { + "epoch": 0.2593, + "grad_norm": 0.3198757767677307, + "learning_rate": 0.00042576689164841374, + "loss": 2.7822185516357423, + "step": 51860 + }, + { + "epoch": 0.25935, + "grad_norm": 0.3244568407535553, + "learning_rate": 0.00042573868162091856, + "loss": 2.7448261260986326, + "step": 51870 + }, + { + "epoch": 0.2594, + "grad_norm": 0.3601987361907959, + "learning_rate": 0.00042571046716920316, + "loss": 2.8115398406982424, + "step": 51880 + }, + { + "epoch": 0.25945, + "grad_norm": 0.32239577174186707, + "learning_rate": 0.00042568224829397797, + "loss": 2.750393295288086, + "step": 51890 + }, + { + "epoch": 0.2595, + "grad_norm": 0.35688236355781555, + "learning_rate": 0.00042565402499595325, + "loss": 2.8298667907714843, + "step": 51900 + }, + { + "epoch": 0.25955, + "grad_norm": 0.325221449136734, + "learning_rate": 0.0004256257972758397, + "loss": 2.8434696197509766, + "step": 51910 + }, + { + "epoch": 0.2596, + "grad_norm": 0.3641858398914337, + "learning_rate": 0.0004255975651343479, + "loss": 2.8187204360961915, + "step": 51920 + }, + { + "epoch": 0.25965, + "grad_norm": 0.3381689786911011, + "learning_rate": 0.00042556932857218855, + "loss": 2.763795280456543, + "step": 51930 + }, + { + "epoch": 0.2597, + "grad_norm": 0.3211507499217987, + "learning_rate": 0.0004255410875900725, + "loss": 2.806276321411133, + "step": 51940 + }, + { + "epoch": 0.25975, + "grad_norm": 0.32409432530403137, + "learning_rate": 0.0004255128421887107, + "loss": 2.845358657836914, + "step": 51950 + }, + { + "epoch": 0.2598, + "grad_norm": 0.3367061913013458, + "learning_rate": 0.00042548459236881433, + "loss": 2.767511749267578, + "step": 51960 + }, + { + "epoch": 0.25985, + "grad_norm": 0.36892303824424744, + "learning_rate": 0.0004254563381310945, + "loss": 2.813341522216797, + "step": 51970 + }, + { + "epoch": 0.2599, + "grad_norm": 0.33143410086631775, + "learning_rate": 0.0004254280794762626, + "loss": 2.8055351257324217, + "step": 51980 + }, + { + "epoch": 0.25995, + "grad_norm": 0.32846584916114807, + "learning_rate": 0.00042539981640502987, + "loss": 2.805752182006836, + "step": 51990 + }, + { + "epoch": 0.26, + "grad_norm": 0.3758331537246704, + "learning_rate": 0.00042537154891810803, + "loss": 2.752425956726074, + "step": 52000 + }, + { + "epoch": 0.26005, + "grad_norm": 0.3369205594062805, + "learning_rate": 0.00042534327701620854, + "loss": 2.839817428588867, + "step": 52010 + }, + { + "epoch": 0.2601, + "grad_norm": 0.32258984446525574, + "learning_rate": 0.00042531500070004327, + "loss": 2.8213071823120117, + "step": 52020 + }, + { + "epoch": 0.26015, + "grad_norm": 0.349500834941864, + "learning_rate": 0.000425286719970324, + "loss": 2.785094451904297, + "step": 52030 + }, + { + "epoch": 0.2602, + "grad_norm": 0.3226802349090576, + "learning_rate": 0.0004252584348277628, + "loss": 2.840983772277832, + "step": 52040 + }, + { + "epoch": 0.26025, + "grad_norm": 0.35135647654533386, + "learning_rate": 0.0004252301452730716, + "loss": 2.7723409652709963, + "step": 52050 + }, + { + "epoch": 0.2603, + "grad_norm": 0.35485267639160156, + "learning_rate": 0.00042520185130696264, + "loss": 2.7829627990722656, + "step": 52060 + }, + { + "epoch": 0.26035, + "grad_norm": 0.3262607157230377, + "learning_rate": 0.0004251735529301483, + "loss": 2.807387351989746, + "step": 52070 + }, + { + "epoch": 0.2604, + "grad_norm": 0.3348061442375183, + "learning_rate": 0.0004251452501433409, + "loss": 2.8238906860351562, + "step": 52080 + }, + { + "epoch": 0.26045, + "grad_norm": 0.3480919599533081, + "learning_rate": 0.00042511694294725306, + "loss": 2.7686948776245117, + "step": 52090 + }, + { + "epoch": 0.2605, + "grad_norm": 0.31188687682151794, + "learning_rate": 0.00042508863134259724, + "loss": 2.8350879669189455, + "step": 52100 + }, + { + "epoch": 0.26055, + "grad_norm": 0.3348425328731537, + "learning_rate": 0.00042506031533008637, + "loss": 2.8303455352783202, + "step": 52110 + }, + { + "epoch": 0.2606, + "grad_norm": 0.34083157777786255, + "learning_rate": 0.00042503199491043326, + "loss": 2.872488021850586, + "step": 52120 + }, + { + "epoch": 0.26065, + "grad_norm": 0.3354133665561676, + "learning_rate": 0.0004250036700843508, + "loss": 2.7831600189208983, + "step": 52130 + }, + { + "epoch": 0.2607, + "grad_norm": 0.3800514340400696, + "learning_rate": 0.00042497534085255215, + "loss": 2.8274761199951173, + "step": 52140 + }, + { + "epoch": 0.26075, + "grad_norm": 0.35250696539878845, + "learning_rate": 0.0004249470072157504, + "loss": 2.7854055404663085, + "step": 52150 + }, + { + "epoch": 0.2608, + "grad_norm": 0.33107122778892517, + "learning_rate": 0.0004249186691746589, + "loss": 2.8188989639282225, + "step": 52160 + }, + { + "epoch": 0.26085, + "grad_norm": 0.3932209312915802, + "learning_rate": 0.0004248903267299912, + "loss": 2.761277198791504, + "step": 52170 + }, + { + "epoch": 0.2609, + "grad_norm": 0.32273373007774353, + "learning_rate": 0.0004248619798824606, + "loss": 2.7793228149414064, + "step": 52180 + }, + { + "epoch": 0.26095, + "grad_norm": 0.3604273796081543, + "learning_rate": 0.00042483362863278087, + "loss": 2.824418067932129, + "step": 52190 + }, + { + "epoch": 0.261, + "grad_norm": 0.33799463510513306, + "learning_rate": 0.00042480527298166564, + "loss": 2.8110240936279296, + "step": 52200 + }, + { + "epoch": 0.26105, + "grad_norm": 0.3303333818912506, + "learning_rate": 0.0004247769129298289, + "loss": 2.810482215881348, + "step": 52210 + }, + { + "epoch": 0.2611, + "grad_norm": 0.3444991111755371, + "learning_rate": 0.0004247485484779845, + "loss": 2.819915771484375, + "step": 52220 + }, + { + "epoch": 0.26115, + "grad_norm": 0.3540799915790558, + "learning_rate": 0.0004247201796268466, + "loss": 2.8195516586303713, + "step": 52230 + }, + { + "epoch": 0.2612, + "grad_norm": 0.36090323328971863, + "learning_rate": 0.00042469180637712934, + "loss": 2.8139204025268554, + "step": 52240 + }, + { + "epoch": 0.26125, + "grad_norm": 0.32696110010147095, + "learning_rate": 0.000424663428729547, + "loss": 2.830618667602539, + "step": 52250 + }, + { + "epoch": 0.2613, + "grad_norm": 0.3512895703315735, + "learning_rate": 0.00042463504668481403, + "loss": 2.7965038299560545, + "step": 52260 + }, + { + "epoch": 0.26135, + "grad_norm": 0.3725566864013672, + "learning_rate": 0.0004246066602436449, + "loss": 2.787379264831543, + "step": 52270 + }, + { + "epoch": 0.2614, + "grad_norm": 0.34741851687431335, + "learning_rate": 0.0004245782694067543, + "loss": 2.7848125457763673, + "step": 52280 + }, + { + "epoch": 0.26145, + "grad_norm": 0.36538055539131165, + "learning_rate": 0.00042454987417485683, + "loss": 2.781942939758301, + "step": 52290 + }, + { + "epoch": 0.2615, + "grad_norm": 0.32851457595825195, + "learning_rate": 0.00042452147454866753, + "loss": 2.804365348815918, + "step": 52300 + }, + { + "epoch": 0.26155, + "grad_norm": 0.32069289684295654, + "learning_rate": 0.0004244930705289013, + "loss": 2.7982282638549805, + "step": 52310 + }, + { + "epoch": 0.2616, + "grad_norm": 0.32947519421577454, + "learning_rate": 0.0004244646621162731, + "loss": 2.77535343170166, + "step": 52320 + }, + { + "epoch": 0.26165, + "grad_norm": 0.3433816432952881, + "learning_rate": 0.00042443624931149824, + "loss": 2.8030643463134766, + "step": 52330 + }, + { + "epoch": 0.2617, + "grad_norm": 0.32475990056991577, + "learning_rate": 0.000424407832115292, + "loss": 2.823213005065918, + "step": 52340 + }, + { + "epoch": 0.26175, + "grad_norm": 0.3388550281524658, + "learning_rate": 0.00042437941052836977, + "loss": 2.8039321899414062, + "step": 52350 + }, + { + "epoch": 0.2618, + "grad_norm": 0.31106746196746826, + "learning_rate": 0.0004243509845514469, + "loss": 2.76735782623291, + "step": 52360 + }, + { + "epoch": 0.26185, + "grad_norm": 0.36566442251205444, + "learning_rate": 0.00042432255418523935, + "loss": 2.7582149505615234, + "step": 52370 + }, + { + "epoch": 0.2619, + "grad_norm": 0.32245147228240967, + "learning_rate": 0.00042429411943046256, + "loss": 2.7725799560546873, + "step": 52380 + }, + { + "epoch": 0.26195, + "grad_norm": 0.3993712067604065, + "learning_rate": 0.0004242656802878325, + "loss": 2.784577560424805, + "step": 52390 + }, + { + "epoch": 0.262, + "grad_norm": 0.3778539001941681, + "learning_rate": 0.0004242372367580651, + "loss": 2.801395225524902, + "step": 52400 + }, + { + "epoch": 0.26205, + "grad_norm": 0.3533709645271301, + "learning_rate": 0.0004242087888418764, + "loss": 2.8051425933837892, + "step": 52410 + }, + { + "epoch": 0.2621, + "grad_norm": 0.34282559156417847, + "learning_rate": 0.00042418033653998264, + "loss": 2.8280704498291014, + "step": 52420 + }, + { + "epoch": 0.26215, + "grad_norm": 0.3338209390640259, + "learning_rate": 0.0004241518798531001, + "loss": 2.7756784439086912, + "step": 52430 + }, + { + "epoch": 0.2622, + "grad_norm": 0.3145742416381836, + "learning_rate": 0.0004241234187819451, + "loss": 2.775310516357422, + "step": 52440 + }, + { + "epoch": 0.26225, + "grad_norm": 0.31927886605262756, + "learning_rate": 0.0004240949533272342, + "loss": 2.8146490097045898, + "step": 52450 + }, + { + "epoch": 0.2623, + "grad_norm": 0.37371012568473816, + "learning_rate": 0.000424066483489684, + "loss": 2.856693458557129, + "step": 52460 + }, + { + "epoch": 0.26235, + "grad_norm": 0.35472947359085083, + "learning_rate": 0.0004240380092700113, + "loss": 2.805633544921875, + "step": 52470 + }, + { + "epoch": 0.2624, + "grad_norm": 0.39650464057922363, + "learning_rate": 0.0004240095306689328, + "loss": 2.811622428894043, + "step": 52480 + }, + { + "epoch": 0.26245, + "grad_norm": 0.34494057297706604, + "learning_rate": 0.0004239810476871656, + "loss": 2.7736412048339845, + "step": 52490 + }, + { + "epoch": 0.2625, + "grad_norm": 0.3624149262905121, + "learning_rate": 0.0004239525603254267, + "loss": 2.814092445373535, + "step": 52500 + }, + { + "epoch": 0.26255, + "grad_norm": 0.3416518568992615, + "learning_rate": 0.00042392406858443317, + "loss": 2.729833793640137, + "step": 52510 + }, + { + "epoch": 0.2626, + "grad_norm": 0.37473264336586, + "learning_rate": 0.0004238955724649024, + "loss": 2.7382234573364257, + "step": 52520 + }, + { + "epoch": 0.26265, + "grad_norm": 0.3583819568157196, + "learning_rate": 0.0004238670719675518, + "loss": 2.8178600311279296, + "step": 52530 + }, + { + "epoch": 0.2627, + "grad_norm": 0.3438234329223633, + "learning_rate": 0.00042383856709309883, + "loss": 2.822194290161133, + "step": 52540 + }, + { + "epoch": 0.26275, + "grad_norm": 0.40525469183921814, + "learning_rate": 0.00042381005784226103, + "loss": 2.823273468017578, + "step": 52550 + }, + { + "epoch": 0.2628, + "grad_norm": 0.36102625727653503, + "learning_rate": 0.0004237815442157562, + "loss": 2.7843900680541993, + "step": 52560 + }, + { + "epoch": 0.26285, + "grad_norm": 0.31220483779907227, + "learning_rate": 0.0004237530262143022, + "loss": 2.816519546508789, + "step": 52570 + }, + { + "epoch": 0.2629, + "grad_norm": 0.3472341299057007, + "learning_rate": 0.00042372450383861693, + "loss": 2.803740692138672, + "step": 52580 + }, + { + "epoch": 0.26295, + "grad_norm": 0.332547128200531, + "learning_rate": 0.0004236959770894183, + "loss": 2.7713016510009765, + "step": 52590 + }, + { + "epoch": 0.263, + "grad_norm": 0.3138367533683777, + "learning_rate": 0.0004236674459674248, + "loss": 2.8148931503295898, + "step": 52600 + }, + { + "epoch": 0.26305, + "grad_norm": 0.351303368806839, + "learning_rate": 0.0004236389104733544, + "loss": 2.8177474975585937, + "step": 52610 + }, + { + "epoch": 0.2631, + "grad_norm": 0.3520369231700897, + "learning_rate": 0.00042361037060792563, + "loss": 2.8213516235351563, + "step": 52620 + }, + { + "epoch": 0.26315, + "grad_norm": 0.33754992485046387, + "learning_rate": 0.00042358182637185687, + "loss": 2.843800926208496, + "step": 52630 + }, + { + "epoch": 0.2632, + "grad_norm": 0.33535251021385193, + "learning_rate": 0.0004235532777658669, + "loss": 2.78064022064209, + "step": 52640 + }, + { + "epoch": 0.26325, + "grad_norm": 0.41425612568855286, + "learning_rate": 0.0004235247247906742, + "loss": 2.8297119140625, + "step": 52650 + }, + { + "epoch": 0.2633, + "grad_norm": 0.3450850546360016, + "learning_rate": 0.0004234961674469978, + "loss": 2.8093210220336915, + "step": 52660 + }, + { + "epoch": 0.26335, + "grad_norm": 0.3364538252353668, + "learning_rate": 0.0004234676057355565, + "loss": 2.8031248092651366, + "step": 52670 + }, + { + "epoch": 0.2634, + "grad_norm": 0.30834972858428955, + "learning_rate": 0.0004234390396570694, + "loss": 2.812693977355957, + "step": 52680 + }, + { + "epoch": 0.26345, + "grad_norm": 0.3477528691291809, + "learning_rate": 0.0004234104692122557, + "loss": 2.8632293701171876, + "step": 52690 + }, + { + "epoch": 0.2635, + "grad_norm": 0.3383314609527588, + "learning_rate": 0.0004233818944018345, + "loss": 2.850432586669922, + "step": 52700 + }, + { + "epoch": 0.26355, + "grad_norm": 0.3490696847438812, + "learning_rate": 0.00042335331522652534, + "loss": 2.823301887512207, + "step": 52710 + }, + { + "epoch": 0.2636, + "grad_norm": 0.33350586891174316, + "learning_rate": 0.0004233247316870475, + "loss": 2.801587677001953, + "step": 52720 + }, + { + "epoch": 0.26365, + "grad_norm": 0.36078745126724243, + "learning_rate": 0.0004232961437841208, + "loss": 2.70916690826416, + "step": 52730 + }, + { + "epoch": 0.2637, + "grad_norm": 0.3598547577857971, + "learning_rate": 0.0004232675515184649, + "loss": 2.832539749145508, + "step": 52740 + }, + { + "epoch": 0.26375, + "grad_norm": 0.32483434677124023, + "learning_rate": 0.00042323895489079945, + "loss": 2.8262996673583984, + "step": 52750 + }, + { + "epoch": 0.2638, + "grad_norm": 0.3667438328266144, + "learning_rate": 0.0004232103539018445, + "loss": 2.8119174957275392, + "step": 52760 + }, + { + "epoch": 0.26385, + "grad_norm": 0.3211953639984131, + "learning_rate": 0.0004231817485523201, + "loss": 2.732596778869629, + "step": 52770 + }, + { + "epoch": 0.2639, + "grad_norm": 0.3744625449180603, + "learning_rate": 0.00042315313884294626, + "loss": 2.7749170303344726, + "step": 52780 + }, + { + "epoch": 0.26395, + "grad_norm": 0.32788923382759094, + "learning_rate": 0.0004231245247744433, + "loss": 2.7965869903564453, + "step": 52790 + }, + { + "epoch": 0.264, + "grad_norm": 0.39011502265930176, + "learning_rate": 0.00042309590634753165, + "loss": 2.8315235137939454, + "step": 52800 + }, + { + "epoch": 0.26405, + "grad_norm": 0.3551771342754364, + "learning_rate": 0.00042306728356293166, + "loss": 2.7875776290893555, + "step": 52810 + }, + { + "epoch": 0.2641, + "grad_norm": 0.31938377022743225, + "learning_rate": 0.00042303865642136397, + "loss": 2.806903266906738, + "step": 52820 + }, + { + "epoch": 0.26415, + "grad_norm": 0.31424471735954285, + "learning_rate": 0.0004230100249235492, + "loss": 2.7702713012695312, + "step": 52830 + }, + { + "epoch": 0.2642, + "grad_norm": 0.3536199629306793, + "learning_rate": 0.00042298138907020835, + "loss": 2.7988739013671875, + "step": 52840 + }, + { + "epoch": 0.26425, + "grad_norm": 0.3379589021205902, + "learning_rate": 0.0004229527488620621, + "loss": 2.8390523910522463, + "step": 52850 + }, + { + "epoch": 0.2643, + "grad_norm": 0.364846795797348, + "learning_rate": 0.00042292410429983155, + "loss": 2.758604621887207, + "step": 52860 + }, + { + "epoch": 0.26435, + "grad_norm": 0.33901411294937134, + "learning_rate": 0.0004228954553842379, + "loss": 2.777402877807617, + "step": 52870 + }, + { + "epoch": 0.2644, + "grad_norm": 0.3489440083503723, + "learning_rate": 0.00042286680211600224, + "loss": 2.795454406738281, + "step": 52880 + }, + { + "epoch": 0.26445, + "grad_norm": 0.3482224941253662, + "learning_rate": 0.000422838144495846, + "loss": 2.817069244384766, + "step": 52890 + }, + { + "epoch": 0.2645, + "grad_norm": 0.3388036787509918, + "learning_rate": 0.0004228094825244907, + "loss": 2.7852134704589844, + "step": 52900 + }, + { + "epoch": 0.26455, + "grad_norm": 0.32975903153419495, + "learning_rate": 0.00042278081620265775, + "loss": 2.801435089111328, + "step": 52910 + }, + { + "epoch": 0.2646, + "grad_norm": 0.35384082794189453, + "learning_rate": 0.0004227521455310689, + "loss": 2.8101030349731446, + "step": 52920 + }, + { + "epoch": 0.26465, + "grad_norm": 0.3182262182235718, + "learning_rate": 0.00042272347051044594, + "loss": 2.796567916870117, + "step": 52930 + }, + { + "epoch": 0.2647, + "grad_norm": 0.3364742696285248, + "learning_rate": 0.0004226947911415109, + "loss": 2.7535469055175783, + "step": 52940 + }, + { + "epoch": 0.26475, + "grad_norm": 0.37643903493881226, + "learning_rate": 0.00042266610742498545, + "loss": 2.771833610534668, + "step": 52950 + }, + { + "epoch": 0.2648, + "grad_norm": 0.35071223974227905, + "learning_rate": 0.00042263741936159207, + "loss": 2.830924415588379, + "step": 52960 + }, + { + "epoch": 0.26485, + "grad_norm": 0.39180511236190796, + "learning_rate": 0.00042260872695205263, + "loss": 2.7765630722045898, + "step": 52970 + }, + { + "epoch": 0.2649, + "grad_norm": 0.35396409034729004, + "learning_rate": 0.0004225800301970898, + "loss": 2.8068700790405274, + "step": 52980 + }, + { + "epoch": 0.26495, + "grad_norm": 0.3375014364719391, + "learning_rate": 0.00042255132909742576, + "loss": 2.839621925354004, + "step": 52990 + }, + { + "epoch": 0.265, + "grad_norm": 0.3335377871990204, + "learning_rate": 0.00042252262365378323, + "loss": 2.8109970092773438, + "step": 53000 + }, + { + "epoch": 0.26505, + "grad_norm": 0.3404606580734253, + "learning_rate": 0.0004224939138668848, + "loss": 2.7511816024780273, + "step": 53010 + }, + { + "epoch": 0.2651, + "grad_norm": 0.36444059014320374, + "learning_rate": 0.0004224651997374532, + "loss": 2.7604301452636717, + "step": 53020 + }, + { + "epoch": 0.26515, + "grad_norm": 0.36463990807533264, + "learning_rate": 0.00042243648126621145, + "loss": 2.8209423065185546, + "step": 53030 + }, + { + "epoch": 0.2652, + "grad_norm": 0.3409425914287567, + "learning_rate": 0.00042240775845388235, + "loss": 2.777011489868164, + "step": 53040 + }, + { + "epoch": 0.26525, + "grad_norm": 0.32303720712661743, + "learning_rate": 0.00042237903130118903, + "loss": 2.8104198455810545, + "step": 53050 + }, + { + "epoch": 0.2653, + "grad_norm": 0.31506016850471497, + "learning_rate": 0.0004223502998088549, + "loss": 2.8362968444824217, + "step": 53060 + }, + { + "epoch": 0.26535, + "grad_norm": 0.3265000581741333, + "learning_rate": 0.00042232156397760303, + "loss": 2.8061906814575197, + "step": 53070 + }, + { + "epoch": 0.2654, + "grad_norm": 0.3656376302242279, + "learning_rate": 0.00042229282380815695, + "loss": 2.8258031845092773, + "step": 53080 + }, + { + "epoch": 0.26545, + "grad_norm": 0.32794174551963806, + "learning_rate": 0.0004222640793012401, + "loss": 2.76973934173584, + "step": 53090 + }, + { + "epoch": 0.2655, + "grad_norm": 0.33550193905830383, + "learning_rate": 0.0004222353304575763, + "loss": 2.8062313079833983, + "step": 53100 + }, + { + "epoch": 0.26555, + "grad_norm": 0.37595781683921814, + "learning_rate": 0.0004222065772778892, + "loss": 2.8361480712890623, + "step": 53110 + }, + { + "epoch": 0.2656, + "grad_norm": 0.34190139174461365, + "learning_rate": 0.00042217781976290267, + "loss": 2.7776391983032225, + "step": 53120 + }, + { + "epoch": 0.26565, + "grad_norm": 0.3275102972984314, + "learning_rate": 0.0004221490579133407, + "loss": 2.7891618728637697, + "step": 53130 + }, + { + "epoch": 0.2657, + "grad_norm": 0.32996299862861633, + "learning_rate": 0.0004221202917299273, + "loss": 2.832809066772461, + "step": 53140 + }, + { + "epoch": 0.26575, + "grad_norm": 0.3279847204685211, + "learning_rate": 0.00042209152121338667, + "loss": 2.7884559631347656, + "step": 53150 + }, + { + "epoch": 0.2658, + "grad_norm": 0.3276940584182739, + "learning_rate": 0.0004220627463644432, + "loss": 2.7933832168579102, + "step": 53160 + }, + { + "epoch": 0.26585, + "grad_norm": 0.3469283878803253, + "learning_rate": 0.00042203396718382125, + "loss": 2.78405876159668, + "step": 53170 + }, + { + "epoch": 0.2659, + "grad_norm": 0.7579748034477234, + "learning_rate": 0.00042200518367224526, + "loss": 2.820280075073242, + "step": 53180 + }, + { + "epoch": 0.26595, + "grad_norm": 0.35426732897758484, + "learning_rate": 0.00042197639583044, + "loss": 2.8321714401245117, + "step": 53190 + }, + { + "epoch": 0.266, + "grad_norm": 0.3261466920375824, + "learning_rate": 0.0004219476036591301, + "loss": 2.8018802642822265, + "step": 53200 + }, + { + "epoch": 0.26605, + "grad_norm": 0.3485972285270691, + "learning_rate": 0.0004219188071590404, + "loss": 2.7958728790283205, + "step": 53210 + }, + { + "epoch": 0.2661, + "grad_norm": 0.32334551215171814, + "learning_rate": 0.0004218900063308959, + "loss": 2.737451934814453, + "step": 53220 + }, + { + "epoch": 0.26615, + "grad_norm": 0.33027926087379456, + "learning_rate": 0.0004218612011754216, + "loss": 2.824596405029297, + "step": 53230 + }, + { + "epoch": 0.2662, + "grad_norm": 0.3156469166278839, + "learning_rate": 0.00042183239169334277, + "loss": 2.7924192428588865, + "step": 53240 + }, + { + "epoch": 0.26625, + "grad_norm": 0.3386729955673218, + "learning_rate": 0.0004218035778853846, + "loss": 2.7766584396362304, + "step": 53250 + }, + { + "epoch": 0.2663, + "grad_norm": 0.32750749588012695, + "learning_rate": 0.0004217747597522725, + "loss": 2.8076622009277346, + "step": 53260 + }, + { + "epoch": 0.26635, + "grad_norm": 0.34066861867904663, + "learning_rate": 0.00042174593729473196, + "loss": 2.8673553466796875, + "step": 53270 + }, + { + "epoch": 0.2664, + "grad_norm": 0.3116694688796997, + "learning_rate": 0.0004217171105134886, + "loss": 2.8081792831420898, + "step": 53280 + }, + { + "epoch": 0.26645, + "grad_norm": 0.35737931728363037, + "learning_rate": 0.00042168827940926816, + "loss": 2.7845991134643553, + "step": 53290 + }, + { + "epoch": 0.2665, + "grad_norm": 0.3767668604850769, + "learning_rate": 0.00042165944398279645, + "loss": 2.811736297607422, + "step": 53300 + }, + { + "epoch": 0.26655, + "grad_norm": 0.34083110094070435, + "learning_rate": 0.0004216306042347994, + "loss": 2.786005210876465, + "step": 53310 + }, + { + "epoch": 0.2666, + "grad_norm": 0.3394831418991089, + "learning_rate": 0.0004216017601660029, + "loss": 2.772882270812988, + "step": 53320 + }, + { + "epoch": 0.26665, + "grad_norm": 0.331432580947876, + "learning_rate": 0.00042157291177713334, + "loss": 2.7989526748657227, + "step": 53330 + }, + { + "epoch": 0.2667, + "grad_norm": 0.32983553409576416, + "learning_rate": 0.00042154405906891685, + "loss": 2.772089195251465, + "step": 53340 + }, + { + "epoch": 0.26675, + "grad_norm": 0.3571232259273529, + "learning_rate": 0.0004215152020420798, + "loss": 2.7770317077636717, + "step": 53350 + }, + { + "epoch": 0.2668, + "grad_norm": 0.3664045035839081, + "learning_rate": 0.0004214863406973487, + "loss": 2.8047321319580076, + "step": 53360 + }, + { + "epoch": 0.26685, + "grad_norm": 0.32729703187942505, + "learning_rate": 0.0004214574750354502, + "loss": 2.8362579345703125, + "step": 53370 + }, + { + "epoch": 0.2669, + "grad_norm": 0.3494473695755005, + "learning_rate": 0.00042142860505711076, + "loss": 2.806278610229492, + "step": 53380 + }, + { + "epoch": 0.26695, + "grad_norm": 0.3094148337841034, + "learning_rate": 0.00042139973076305747, + "loss": 2.812583160400391, + "step": 53390 + }, + { + "epoch": 0.267, + "grad_norm": 0.34041038155555725, + "learning_rate": 0.00042137085215401707, + "loss": 2.7685724258422852, + "step": 53400 + }, + { + "epoch": 0.26705, + "grad_norm": 0.3468150198459625, + "learning_rate": 0.00042134196923071656, + "loss": 2.8092893600463866, + "step": 53410 + }, + { + "epoch": 0.2671, + "grad_norm": 0.3022952377796173, + "learning_rate": 0.00042131308199388313, + "loss": 2.809025001525879, + "step": 53420 + }, + { + "epoch": 0.26715, + "grad_norm": 0.3751043379306793, + "learning_rate": 0.00042128419044424405, + "loss": 2.772929000854492, + "step": 53430 + }, + { + "epoch": 0.2672, + "grad_norm": 0.31478968262672424, + "learning_rate": 0.00042125529458252655, + "loss": 2.793777847290039, + "step": 53440 + }, + { + "epoch": 0.26725, + "grad_norm": 0.3580355644226074, + "learning_rate": 0.0004212263944094582, + "loss": 2.815704345703125, + "step": 53450 + }, + { + "epoch": 0.2673, + "grad_norm": 0.34854361414909363, + "learning_rate": 0.0004211974899257665, + "loss": 2.8000383377075195, + "step": 53460 + }, + { + "epoch": 0.26735, + "grad_norm": 0.33095309138298035, + "learning_rate": 0.00042116858113217914, + "loss": 2.836198425292969, + "step": 53470 + }, + { + "epoch": 0.2674, + "grad_norm": 0.3079381287097931, + "learning_rate": 0.00042113966802942394, + "loss": 2.8240556716918945, + "step": 53480 + }, + { + "epoch": 0.26745, + "grad_norm": 0.3473573923110962, + "learning_rate": 0.0004211107506182287, + "loss": 2.8102079391479493, + "step": 53490 + }, + { + "epoch": 0.2675, + "grad_norm": 0.322834849357605, + "learning_rate": 0.00042108182889932143, + "loss": 2.7559463500976564, + "step": 53500 + }, + { + "epoch": 0.26755, + "grad_norm": 0.3337390124797821, + "learning_rate": 0.0004210529028734303, + "loss": 2.824360656738281, + "step": 53510 + }, + { + "epoch": 0.2676, + "grad_norm": 0.3352685272693634, + "learning_rate": 0.0004210239725412834, + "loss": 2.7762353897094725, + "step": 53520 + }, + { + "epoch": 0.26765, + "grad_norm": 0.3413403034210205, + "learning_rate": 0.00042099503790360925, + "loss": 2.748918151855469, + "step": 53530 + }, + { + "epoch": 0.2677, + "grad_norm": 0.32813605666160583, + "learning_rate": 0.00042096609896113615, + "loss": 2.814161491394043, + "step": 53540 + }, + { + "epoch": 0.26775, + "grad_norm": 0.3549848198890686, + "learning_rate": 0.00042093715571459256, + "loss": 2.7939125061035157, + "step": 53550 + }, + { + "epoch": 0.2678, + "grad_norm": 0.3545287251472473, + "learning_rate": 0.0004209082081647073, + "loss": 2.772532272338867, + "step": 53560 + }, + { + "epoch": 0.26785, + "grad_norm": 0.3381141722202301, + "learning_rate": 0.000420879256312209, + "loss": 2.8353464126586916, + "step": 53570 + }, + { + "epoch": 0.2679, + "grad_norm": 0.35633474588394165, + "learning_rate": 0.0004208503001578266, + "loss": 2.7760393142700197, + "step": 53580 + }, + { + "epoch": 0.26795, + "grad_norm": 0.32690751552581787, + "learning_rate": 0.000420821339702289, + "loss": 2.8148305892944334, + "step": 53590 + }, + { + "epoch": 0.268, + "grad_norm": 0.3575488328933716, + "learning_rate": 0.00042079237494632536, + "loss": 2.8192668914794923, + "step": 53600 + }, + { + "epoch": 0.26805, + "grad_norm": 0.3422922194004059, + "learning_rate": 0.0004207634058906648, + "loss": 2.81988410949707, + "step": 53610 + }, + { + "epoch": 0.2681, + "grad_norm": 0.3288203775882721, + "learning_rate": 0.0004207344325360366, + "loss": 2.857144355773926, + "step": 53620 + }, + { + "epoch": 0.26815, + "grad_norm": 0.35367879271507263, + "learning_rate": 0.00042070545488317025, + "loss": 2.807565116882324, + "step": 53630 + }, + { + "epoch": 0.2682, + "grad_norm": 0.3575258255004883, + "learning_rate": 0.00042067647293279523, + "loss": 2.7772453308105467, + "step": 53640 + }, + { + "epoch": 0.26825, + "grad_norm": 0.3398452401161194, + "learning_rate": 0.00042064748668564113, + "loss": 2.826671600341797, + "step": 53650 + }, + { + "epoch": 0.2683, + "grad_norm": 0.3722931146621704, + "learning_rate": 0.0004206184961424376, + "loss": 2.7537940979003905, + "step": 53660 + }, + { + "epoch": 0.26835, + "grad_norm": 0.31765106320381165, + "learning_rate": 0.00042058950130391463, + "loss": 2.7672229766845704, + "step": 53670 + }, + { + "epoch": 0.2684, + "grad_norm": 0.33410364389419556, + "learning_rate": 0.0004205605021708021, + "loss": 2.7917854309082033, + "step": 53680 + }, + { + "epoch": 0.26845, + "grad_norm": 0.3406658172607422, + "learning_rate": 0.0004205314987438301, + "loss": 2.781546211242676, + "step": 53690 + }, + { + "epoch": 0.2685, + "grad_norm": 0.33788108825683594, + "learning_rate": 0.00042050249102372874, + "loss": 2.7947200775146483, + "step": 53700 + }, + { + "epoch": 0.26855, + "grad_norm": 0.3491554260253906, + "learning_rate": 0.00042047347901122834, + "loss": 2.812271499633789, + "step": 53710 + }, + { + "epoch": 0.2686, + "grad_norm": 0.3334310054779053, + "learning_rate": 0.0004204444627070592, + "loss": 2.8165876388549806, + "step": 53720 + }, + { + "epoch": 0.26865, + "grad_norm": 0.38780632615089417, + "learning_rate": 0.0004204154421119519, + "loss": 2.7976837158203125, + "step": 53730 + }, + { + "epoch": 0.2687, + "grad_norm": 0.36768773198127747, + "learning_rate": 0.00042038641722663694, + "loss": 2.7313888549804686, + "step": 53740 + }, + { + "epoch": 0.26875, + "grad_norm": 0.344806432723999, + "learning_rate": 0.0004203573880518451, + "loss": 2.776239585876465, + "step": 53750 + }, + { + "epoch": 0.2688, + "grad_norm": 0.3397451639175415, + "learning_rate": 0.0004203283545883071, + "loss": 2.794830322265625, + "step": 53760 + }, + { + "epoch": 0.26885, + "grad_norm": 0.33318427205085754, + "learning_rate": 0.00042029931683675394, + "loss": 2.8658321380615233, + "step": 53770 + }, + { + "epoch": 0.2689, + "grad_norm": 0.3355787396430969, + "learning_rate": 0.0004202702747979166, + "loss": 2.812905502319336, + "step": 53780 + }, + { + "epoch": 0.26895, + "grad_norm": 0.36280107498168945, + "learning_rate": 0.0004202412284725263, + "loss": 2.7772195816040037, + "step": 53790 + }, + { + "epoch": 0.269, + "grad_norm": 0.3660412132740021, + "learning_rate": 0.0004202121778613141, + "loss": 2.7708301544189453, + "step": 53800 + }, + { + "epoch": 0.26905, + "grad_norm": 0.3435702323913574, + "learning_rate": 0.0004201831229650115, + "loss": 2.7852312088012696, + "step": 53810 + }, + { + "epoch": 0.2691, + "grad_norm": 0.3391478657722473, + "learning_rate": 0.00042015406378434993, + "loss": 2.8023487091064454, + "step": 53820 + }, + { + "epoch": 0.26915, + "grad_norm": 0.362068772315979, + "learning_rate": 0.000420125000320061, + "loss": 2.746112251281738, + "step": 53830 + }, + { + "epoch": 0.2692, + "grad_norm": 0.3295942544937134, + "learning_rate": 0.0004200959325728763, + "loss": 2.736598777770996, + "step": 53840 + }, + { + "epoch": 0.26925, + "grad_norm": 0.3634640574455261, + "learning_rate": 0.0004200668605435276, + "loss": 2.79073543548584, + "step": 53850 + }, + { + "epoch": 0.2693, + "grad_norm": 0.31991010904312134, + "learning_rate": 0.0004200377842327467, + "loss": 2.7997953414916994, + "step": 53860 + }, + { + "epoch": 0.26935, + "grad_norm": 0.35555407404899597, + "learning_rate": 0.000420008703641266, + "loss": 2.8037405014038086, + "step": 53870 + }, + { + "epoch": 0.2694, + "grad_norm": 0.3808906674385071, + "learning_rate": 0.000419979618769817, + "loss": 2.8531564712524413, + "step": 53880 + }, + { + "epoch": 0.26945, + "grad_norm": 0.3432667553424835, + "learning_rate": 0.0004199505296191325, + "loss": 2.8174280166625976, + "step": 53890 + }, + { + "epoch": 0.2695, + "grad_norm": 0.33299753069877625, + "learning_rate": 0.00041992143618994443, + "loss": 2.7572813034057617, + "step": 53900 + }, + { + "epoch": 0.26955, + "grad_norm": 0.33605220913887024, + "learning_rate": 0.00041989233848298534, + "loss": 2.8016109466552734, + "step": 53910 + }, + { + "epoch": 0.2696, + "grad_norm": 0.3927464187145233, + "learning_rate": 0.00041986323649898785, + "loss": 2.7780811309814455, + "step": 53920 + }, + { + "epoch": 0.26965, + "grad_norm": 0.31668198108673096, + "learning_rate": 0.0004198341302386844, + "loss": 2.7822071075439454, + "step": 53930 + }, + { + "epoch": 0.2697, + "grad_norm": 0.3560342490673065, + "learning_rate": 0.00041980501970280794, + "loss": 2.7983070373535157, + "step": 53940 + }, + { + "epoch": 0.26975, + "grad_norm": 0.32559728622436523, + "learning_rate": 0.00041977590489209116, + "loss": 2.811792755126953, + "step": 53950 + }, + { + "epoch": 0.2698, + "grad_norm": 0.33867543935775757, + "learning_rate": 0.00041974678580726716, + "loss": 2.7666677474975585, + "step": 53960 + }, + { + "epoch": 0.26985, + "grad_norm": 0.34810787439346313, + "learning_rate": 0.00041971766244906897, + "loss": 2.851176071166992, + "step": 53970 + }, + { + "epoch": 0.2699, + "grad_norm": 0.33211326599121094, + "learning_rate": 0.0004196885348182298, + "loss": 2.812287139892578, + "step": 53980 + }, + { + "epoch": 0.26995, + "grad_norm": 0.34798958897590637, + "learning_rate": 0.00041965940291548284, + "loss": 2.8071533203125, + "step": 53990 + }, + { + "epoch": 0.27, + "grad_norm": 0.36467793583869934, + "learning_rate": 0.0004196302667415615, + "loss": 2.8432695388793947, + "step": 54000 + }, + { + "epoch": 0.27005, + "grad_norm": 0.3910052180290222, + "learning_rate": 0.0004196011262971994, + "loss": 2.776417922973633, + "step": 54010 + }, + { + "epoch": 0.2701, + "grad_norm": 0.369688481092453, + "learning_rate": 0.00041957198158313003, + "loss": 2.7995847702026366, + "step": 54020 + }, + { + "epoch": 0.27015, + "grad_norm": 0.3984317183494568, + "learning_rate": 0.00041954283260008725, + "loss": 2.7767038345336914, + "step": 54030 + }, + { + "epoch": 0.2702, + "grad_norm": 0.37291431427001953, + "learning_rate": 0.0004195136793488047, + "loss": 2.7784809112548827, + "step": 54040 + }, + { + "epoch": 0.27025, + "grad_norm": 0.3544721007347107, + "learning_rate": 0.0004194845218300165, + "loss": 2.783823776245117, + "step": 54050 + }, + { + "epoch": 0.2703, + "grad_norm": 0.34652993083000183, + "learning_rate": 0.0004194553600444565, + "loss": 2.7623899459838865, + "step": 54060 + }, + { + "epoch": 0.27035, + "grad_norm": 0.36937928199768066, + "learning_rate": 0.000419426193992859, + "loss": 2.7813133239746093, + "step": 54070 + }, + { + "epoch": 0.2704, + "grad_norm": 0.34128108620643616, + "learning_rate": 0.0004193970236759581, + "loss": 2.806340980529785, + "step": 54080 + }, + { + "epoch": 0.27045, + "grad_norm": 0.3471355438232422, + "learning_rate": 0.0004193678490944884, + "loss": 2.7780855178833006, + "step": 54090 + }, + { + "epoch": 0.2705, + "grad_norm": 0.33304154872894287, + "learning_rate": 0.00041933867024918417, + "loss": 2.806402015686035, + "step": 54100 + }, + { + "epoch": 0.27055, + "grad_norm": 0.3103722035884857, + "learning_rate": 0.00041930948714078, + "loss": 2.7714122772216796, + "step": 54110 + }, + { + "epoch": 0.2706, + "grad_norm": 0.32669076323509216, + "learning_rate": 0.0004192802997700107, + "loss": 2.814716339111328, + "step": 54120 + }, + { + "epoch": 0.27065, + "grad_norm": 0.3556991517543793, + "learning_rate": 0.00041925110813761094, + "loss": 2.807015037536621, + "step": 54130 + }, + { + "epoch": 0.2707, + "grad_norm": 0.3448830842971802, + "learning_rate": 0.0004192219122443157, + "loss": 2.7984992980957033, + "step": 54140 + }, + { + "epoch": 0.27075, + "grad_norm": 0.32289907336235046, + "learning_rate": 0.0004191927120908599, + "loss": 2.800619125366211, + "step": 54150 + }, + { + "epoch": 0.2708, + "grad_norm": 0.38459956645965576, + "learning_rate": 0.0004191635076779787, + "loss": 2.7268686294555664, + "step": 54160 + }, + { + "epoch": 0.27085, + "grad_norm": 0.3421574831008911, + "learning_rate": 0.0004191342990064074, + "loss": 2.856816291809082, + "step": 54170 + }, + { + "epoch": 0.2709, + "grad_norm": 0.3139074444770813, + "learning_rate": 0.00041910508607688127, + "loss": 2.7680133819580077, + "step": 54180 + }, + { + "epoch": 0.27095, + "grad_norm": 0.33122292160987854, + "learning_rate": 0.0004190758688901356, + "loss": 2.7898818969726564, + "step": 54190 + }, + { + "epoch": 0.271, + "grad_norm": 0.35948216915130615, + "learning_rate": 0.0004190466474469061, + "loss": 2.829456329345703, + "step": 54200 + }, + { + "epoch": 0.27105, + "grad_norm": 0.37120676040649414, + "learning_rate": 0.00041901742174792845, + "loss": 2.757603073120117, + "step": 54210 + }, + { + "epoch": 0.2711, + "grad_norm": 0.35402628779411316, + "learning_rate": 0.0004189881917939383, + "loss": 2.8367143630981446, + "step": 54220 + }, + { + "epoch": 0.27115, + "grad_norm": 0.3315489590167999, + "learning_rate": 0.0004189589575856715, + "loss": 2.7785211563110352, + "step": 54230 + }, + { + "epoch": 0.2712, + "grad_norm": 0.3257052004337311, + "learning_rate": 0.00041892971912386415, + "loss": 2.77537841796875, + "step": 54240 + }, + { + "epoch": 0.27125, + "grad_norm": 0.36170732975006104, + "learning_rate": 0.0004189004764092522, + "loss": 2.864597702026367, + "step": 54250 + }, + { + "epoch": 0.2713, + "grad_norm": 0.35888251662254333, + "learning_rate": 0.0004188712294425719, + "loss": 2.7935626983642576, + "step": 54260 + }, + { + "epoch": 0.27135, + "grad_norm": 0.33552446961402893, + "learning_rate": 0.0004188419782245595, + "loss": 2.743942642211914, + "step": 54270 + }, + { + "epoch": 0.2714, + "grad_norm": 0.3320589065551758, + "learning_rate": 0.00041881272275595146, + "loss": 2.818702507019043, + "step": 54280 + }, + { + "epoch": 0.27145, + "grad_norm": 0.3565087616443634, + "learning_rate": 0.00041878346303748427, + "loss": 2.8044765472412108, + "step": 54290 + }, + { + "epoch": 0.2715, + "grad_norm": 0.3313139081001282, + "learning_rate": 0.0004187541990698945, + "loss": 2.7655670166015627, + "step": 54300 + }, + { + "epoch": 0.27155, + "grad_norm": 0.3333439826965332, + "learning_rate": 0.00041872493085391883, + "loss": 2.8236162185668947, + "step": 54310 + }, + { + "epoch": 0.2716, + "grad_norm": 0.3495479226112366, + "learning_rate": 0.00041869565839029423, + "loss": 2.8592288970947264, + "step": 54320 + }, + { + "epoch": 0.27165, + "grad_norm": 0.3370932638645172, + "learning_rate": 0.00041866638167975756, + "loss": 2.762051582336426, + "step": 54330 + }, + { + "epoch": 0.2717, + "grad_norm": 0.35683563351631165, + "learning_rate": 0.0004186371007230458, + "loss": 2.7835529327392576, + "step": 54340 + }, + { + "epoch": 0.27175, + "grad_norm": 0.3231656551361084, + "learning_rate": 0.00041860781552089624, + "loss": 2.797895050048828, + "step": 54350 + }, + { + "epoch": 0.2718, + "grad_norm": 0.374582976102829, + "learning_rate": 0.00041857852607404605, + "loss": 2.833662986755371, + "step": 54360 + }, + { + "epoch": 0.27185, + "grad_norm": 0.36041492223739624, + "learning_rate": 0.0004185492323832325, + "loss": 2.822633743286133, + "step": 54370 + }, + { + "epoch": 0.2719, + "grad_norm": 0.3380582630634308, + "learning_rate": 0.0004185199344491932, + "loss": 2.8279985427856444, + "step": 54380 + }, + { + "epoch": 0.27195, + "grad_norm": 0.3401537537574768, + "learning_rate": 0.0004184906322726657, + "loss": 2.825374984741211, + "step": 54390 + }, + { + "epoch": 0.272, + "grad_norm": 0.351784348487854, + "learning_rate": 0.00041846132585438765, + "loss": 2.8197305679321287, + "step": 54400 + }, + { + "epoch": 0.27205, + "grad_norm": 0.3695158362388611, + "learning_rate": 0.00041843201519509686, + "loss": 2.8169864654541015, + "step": 54410 + }, + { + "epoch": 0.2721, + "grad_norm": 0.31739431619644165, + "learning_rate": 0.00041840270029553116, + "loss": 2.8270984649658204, + "step": 54420 + }, + { + "epoch": 0.27215, + "grad_norm": 0.3207361698150635, + "learning_rate": 0.00041837338115642867, + "loss": 2.7990936279296874, + "step": 54430 + }, + { + "epoch": 0.2722, + "grad_norm": 0.3382663428783417, + "learning_rate": 0.0004183440577785274, + "loss": 2.7551746368408203, + "step": 54440 + }, + { + "epoch": 0.27225, + "grad_norm": 0.3150711953639984, + "learning_rate": 0.0004183147301625656, + "loss": 2.7747098922729494, + "step": 54450 + }, + { + "epoch": 0.2723, + "grad_norm": 0.35628101229667664, + "learning_rate": 0.00041828539830928156, + "loss": 2.816634750366211, + "step": 54460 + }, + { + "epoch": 0.27235, + "grad_norm": 0.3456466794013977, + "learning_rate": 0.0004182560622194138, + "loss": 2.8544593811035157, + "step": 54470 + }, + { + "epoch": 0.2724, + "grad_norm": 0.34196236729621887, + "learning_rate": 0.0004182267218937008, + "loss": 2.8544254302978516, + "step": 54480 + }, + { + "epoch": 0.27245, + "grad_norm": 0.330879807472229, + "learning_rate": 0.0004181973773328811, + "loss": 2.7360376358032226, + "step": 54490 + }, + { + "epoch": 0.2725, + "grad_norm": 0.33766093850135803, + "learning_rate": 0.00041816802853769365, + "loss": 2.778955841064453, + "step": 54500 + }, + { + "epoch": 0.27255, + "grad_norm": 0.32950258255004883, + "learning_rate": 0.00041813867550887716, + "loss": 2.79847412109375, + "step": 54510 + }, + { + "epoch": 0.2726, + "grad_norm": 0.33403587341308594, + "learning_rate": 0.0004181093182471707, + "loss": 2.8354331970214846, + "step": 54520 + }, + { + "epoch": 0.27265, + "grad_norm": 0.3501390814781189, + "learning_rate": 0.0004180799567533132, + "loss": 2.803419494628906, + "step": 54530 + }, + { + "epoch": 0.2727, + "grad_norm": 0.3207949995994568, + "learning_rate": 0.000418050591028044, + "loss": 2.769761657714844, + "step": 54540 + }, + { + "epoch": 0.27275, + "grad_norm": 0.34277620911598206, + "learning_rate": 0.0004180212210721022, + "loss": 2.749900054931641, + "step": 54550 + }, + { + "epoch": 0.2728, + "grad_norm": 0.33875131607055664, + "learning_rate": 0.00041799184688622737, + "loss": 2.78249454498291, + "step": 54560 + }, + { + "epoch": 0.27285, + "grad_norm": 0.32135480642318726, + "learning_rate": 0.0004179624684711588, + "loss": 2.8001800537109376, + "step": 54570 + }, + { + "epoch": 0.2729, + "grad_norm": 0.34092584252357483, + "learning_rate": 0.00041793308582763634, + "loss": 2.8243650436401366, + "step": 54580 + }, + { + "epoch": 0.27295, + "grad_norm": 0.35581517219543457, + "learning_rate": 0.00041790369895639945, + "loss": 2.844073677062988, + "step": 54590 + }, + { + "epoch": 0.273, + "grad_norm": 0.35373395681381226, + "learning_rate": 0.00041787430785818815, + "loss": 2.7972850799560547, + "step": 54600 + }, + { + "epoch": 0.27305, + "grad_norm": 0.3221515119075775, + "learning_rate": 0.00041784491253374226, + "loss": 2.7705047607421873, + "step": 54610 + }, + { + "epoch": 0.2731, + "grad_norm": 0.3136185109615326, + "learning_rate": 0.00041781551298380174, + "loss": 2.801368522644043, + "step": 54620 + }, + { + "epoch": 0.27315, + "grad_norm": 0.30165478587150574, + "learning_rate": 0.0004177861092091069, + "loss": 2.8252084732055662, + "step": 54630 + }, + { + "epoch": 0.2732, + "grad_norm": 0.3370544910430908, + "learning_rate": 0.00041775670121039783, + "loss": 2.7619935989379885, + "step": 54640 + }, + { + "epoch": 0.27325, + "grad_norm": 0.32397425174713135, + "learning_rate": 0.0004177272889884149, + "loss": 2.8042768478393554, + "step": 54650 + }, + { + "epoch": 0.2733, + "grad_norm": 0.32026195526123047, + "learning_rate": 0.0004176978725438987, + "loss": 2.7946760177612306, + "step": 54660 + }, + { + "epoch": 0.27335, + "grad_norm": 0.32209646701812744, + "learning_rate": 0.0004176684518775895, + "loss": 2.798611640930176, + "step": 54670 + }, + { + "epoch": 0.2734, + "grad_norm": 0.3632374703884125, + "learning_rate": 0.0004176390269902283, + "loss": 2.8023500442504883, + "step": 54680 + }, + { + "epoch": 0.27345, + "grad_norm": 0.3377440571784973, + "learning_rate": 0.00041760959788255563, + "loss": 2.851888656616211, + "step": 54690 + }, + { + "epoch": 0.2735, + "grad_norm": 0.3262783885002136, + "learning_rate": 0.00041758016455531254, + "loss": 2.785232734680176, + "step": 54700 + }, + { + "epoch": 0.27355, + "grad_norm": 0.3127083480358124, + "learning_rate": 0.00041755072700923985, + "loss": 2.8081842422485352, + "step": 54710 + }, + { + "epoch": 0.2736, + "grad_norm": 0.3278895914554596, + "learning_rate": 0.00041752128524507877, + "loss": 2.767742919921875, + "step": 54720 + }, + { + "epoch": 0.27365, + "grad_norm": 0.33334264159202576, + "learning_rate": 0.0004174918392635704, + "loss": 2.8029184341430664, + "step": 54730 + }, + { + "epoch": 0.2737, + "grad_norm": 0.3275618255138397, + "learning_rate": 0.0004174623890654562, + "loss": 2.7368370056152345, + "step": 54740 + }, + { + "epoch": 0.27375, + "grad_norm": 0.3503311276435852, + "learning_rate": 0.00041743293465147733, + "loss": 2.8081501007080076, + "step": 54750 + }, + { + "epoch": 0.2738, + "grad_norm": 0.32886171340942383, + "learning_rate": 0.00041740347602237554, + "loss": 2.7857986450195313, + "step": 54760 + }, + { + "epoch": 0.27385, + "grad_norm": 0.3614272475242615, + "learning_rate": 0.00041737401317889236, + "loss": 2.7438453674316405, + "step": 54770 + }, + { + "epoch": 0.2739, + "grad_norm": 0.3634161055088043, + "learning_rate": 0.00041734454612176955, + "loss": 2.795622444152832, + "step": 54780 + }, + { + "epoch": 0.27395, + "grad_norm": 0.33882465958595276, + "learning_rate": 0.0004173150748517488, + "loss": 2.830332946777344, + "step": 54790 + }, + { + "epoch": 0.274, + "grad_norm": 0.333222359418869, + "learning_rate": 0.00041728559936957224, + "loss": 2.786119270324707, + "step": 54800 + }, + { + "epoch": 0.27405, + "grad_norm": 0.35549259185791016, + "learning_rate": 0.00041725611967598176, + "loss": 2.814278984069824, + "step": 54810 + }, + { + "epoch": 0.2741, + "grad_norm": 0.35111701488494873, + "learning_rate": 0.0004172266357717197, + "loss": 2.7823375701904296, + "step": 54820 + }, + { + "epoch": 0.27415, + "grad_norm": 0.33398619294166565, + "learning_rate": 0.00041719714765752814, + "loss": 2.8228084564208986, + "step": 54830 + }, + { + "epoch": 0.2742, + "grad_norm": 0.35588154196739197, + "learning_rate": 0.00041716765533414944, + "loss": 2.826364517211914, + "step": 54840 + }, + { + "epoch": 0.27425, + "grad_norm": 0.41272395849227905, + "learning_rate": 0.00041713815880232627, + "loss": 2.7817575454711916, + "step": 54850 + }, + { + "epoch": 0.2743, + "grad_norm": 0.3286924660205841, + "learning_rate": 0.00041710865806280094, + "loss": 2.7742336273193358, + "step": 54860 + }, + { + "epoch": 0.27435, + "grad_norm": 0.3104027211666107, + "learning_rate": 0.00041707915311631633, + "loss": 2.828962516784668, + "step": 54870 + }, + { + "epoch": 0.2744, + "grad_norm": 0.3201368749141693, + "learning_rate": 0.0004170496439636152, + "loss": 2.790635108947754, + "step": 54880 + }, + { + "epoch": 0.27445, + "grad_norm": 0.3688223361968994, + "learning_rate": 0.0004170201306054403, + "loss": 2.7676591873168945, + "step": 54890 + }, + { + "epoch": 0.2745, + "grad_norm": 0.3537802994251251, + "learning_rate": 0.00041699061304253476, + "loss": 2.85131950378418, + "step": 54900 + }, + { + "epoch": 0.27455, + "grad_norm": 0.34153714776039124, + "learning_rate": 0.00041696109127564165, + "loss": 2.791832160949707, + "step": 54910 + }, + { + "epoch": 0.2746, + "grad_norm": 0.35516777634620667, + "learning_rate": 0.0004169315653055042, + "loss": 2.8083881378173827, + "step": 54920 + }, + { + "epoch": 0.27465, + "grad_norm": 0.37200111150741577, + "learning_rate": 0.0004169020351328657, + "loss": 2.785401153564453, + "step": 54930 + }, + { + "epoch": 0.2747, + "grad_norm": 0.37558743357658386, + "learning_rate": 0.0004168725007584696, + "loss": 2.742722511291504, + "step": 54940 + }, + { + "epoch": 0.27475, + "grad_norm": 0.30767887830734253, + "learning_rate": 0.0004168429621830594, + "loss": 2.7675233840942384, + "step": 54950 + }, + { + "epoch": 0.2748, + "grad_norm": 0.3723757565021515, + "learning_rate": 0.0004168134194073787, + "loss": 2.799302864074707, + "step": 54960 + }, + { + "epoch": 0.27485, + "grad_norm": 0.3404439687728882, + "learning_rate": 0.00041678387243217137, + "loss": 2.7976448059082033, + "step": 54970 + }, + { + "epoch": 0.2749, + "grad_norm": 0.3402661383152008, + "learning_rate": 0.0004167543212581811, + "loss": 2.751120758056641, + "step": 54980 + }, + { + "epoch": 0.27495, + "grad_norm": 0.3482840061187744, + "learning_rate": 0.00041672476588615196, + "loss": 2.7935766220092773, + "step": 54990 + }, + { + "epoch": 0.275, + "grad_norm": 0.36889347434043884, + "learning_rate": 0.0004166952063168279, + "loss": 2.767715072631836, + "step": 55000 + }, + { + "epoch": 0.27505, + "grad_norm": 0.32029175758361816, + "learning_rate": 0.0004166656425509532, + "loss": 2.824719047546387, + "step": 55010 + }, + { + "epoch": 0.2751, + "grad_norm": 0.3222712576389313, + "learning_rate": 0.00041663607458927195, + "loss": 2.758609962463379, + "step": 55020 + }, + { + "epoch": 0.27515, + "grad_norm": 0.3218227028846741, + "learning_rate": 0.0004166065024325287, + "loss": 2.8201530456542967, + "step": 55030 + }, + { + "epoch": 0.2752, + "grad_norm": 0.3757990002632141, + "learning_rate": 0.00041657692608146785, + "loss": 2.7883073806762697, + "step": 55040 + }, + { + "epoch": 0.27525, + "grad_norm": 0.3327326774597168, + "learning_rate": 0.00041654734553683403, + "loss": 2.7795698165893556, + "step": 55050 + }, + { + "epoch": 0.2753, + "grad_norm": 0.3348015248775482, + "learning_rate": 0.0004165177607993719, + "loss": 2.845808410644531, + "step": 55060 + }, + { + "epoch": 0.27535, + "grad_norm": 0.3178005516529083, + "learning_rate": 0.0004164881718698262, + "loss": 2.8154293060302735, + "step": 55070 + }, + { + "epoch": 0.2754, + "grad_norm": 0.3385442793369293, + "learning_rate": 0.00041645857874894187, + "loss": 2.778178405761719, + "step": 55080 + }, + { + "epoch": 0.27545, + "grad_norm": 0.3422952890396118, + "learning_rate": 0.00041642898143746406, + "loss": 2.8051916122436524, + "step": 55090 + }, + { + "epoch": 0.2755, + "grad_norm": 0.348384290933609, + "learning_rate": 0.00041639937993613764, + "loss": 2.816517448425293, + "step": 55100 + }, + { + "epoch": 0.27555, + "grad_norm": 0.37306874990463257, + "learning_rate": 0.000416369774245708, + "loss": 2.8191112518310546, + "step": 55110 + }, + { + "epoch": 0.2756, + "grad_norm": 0.3737117052078247, + "learning_rate": 0.00041634016436692037, + "loss": 2.7963600158691406, + "step": 55120 + }, + { + "epoch": 0.27565, + "grad_norm": 0.3485666513442993, + "learning_rate": 0.00041631055030052024, + "loss": 2.7855695724487304, + "step": 55130 + }, + { + "epoch": 0.2757, + "grad_norm": 0.3409118056297302, + "learning_rate": 0.0004162809320472531, + "loss": 2.776492118835449, + "step": 55140 + }, + { + "epoch": 0.27575, + "grad_norm": 0.3252721130847931, + "learning_rate": 0.0004162513096078646, + "loss": 2.7883960723876955, + "step": 55150 + }, + { + "epoch": 0.2758, + "grad_norm": 0.347614586353302, + "learning_rate": 0.0004162216829831005, + "loss": 2.793193817138672, + "step": 55160 + }, + { + "epoch": 0.27585, + "grad_norm": 0.3174547553062439, + "learning_rate": 0.00041619205217370667, + "loss": 2.8519535064697266, + "step": 55170 + }, + { + "epoch": 0.2759, + "grad_norm": 0.3372173309326172, + "learning_rate": 0.00041616241718042896, + "loss": 2.8044353485107423, + "step": 55180 + }, + { + "epoch": 0.27595, + "grad_norm": 0.3266725540161133, + "learning_rate": 0.00041613277800401364, + "loss": 2.7368722915649415, + "step": 55190 + }, + { + "epoch": 0.276, + "grad_norm": 0.35628607869148254, + "learning_rate": 0.00041610313464520663, + "loss": 2.7799768447875977, + "step": 55200 + }, + { + "epoch": 0.27605, + "grad_norm": 0.33318182826042175, + "learning_rate": 0.00041607348710475436, + "loss": 2.8031589508056642, + "step": 55210 + }, + { + "epoch": 0.2761, + "grad_norm": 0.3392517566680908, + "learning_rate": 0.0004160438353834031, + "loss": 2.806205177307129, + "step": 55220 + }, + { + "epoch": 0.27615, + "grad_norm": 0.3455319404602051, + "learning_rate": 0.0004160141794818995, + "loss": 2.79346923828125, + "step": 55230 + }, + { + "epoch": 0.2762, + "grad_norm": 0.33204489946365356, + "learning_rate": 0.00041598451940098993, + "loss": 2.797610855102539, + "step": 55240 + }, + { + "epoch": 0.27625, + "grad_norm": 0.30886515974998474, + "learning_rate": 0.0004159548551414213, + "loss": 2.7346075057983397, + "step": 55250 + }, + { + "epoch": 0.2763, + "grad_norm": 0.3425353765487671, + "learning_rate": 0.0004159251867039402, + "loss": 2.7604888916015624, + "step": 55260 + }, + { + "epoch": 0.27635, + "grad_norm": 0.3379358947277069, + "learning_rate": 0.0004158955140892937, + "loss": 2.8343374252319338, + "step": 55270 + }, + { + "epoch": 0.2764, + "grad_norm": 0.32594701647758484, + "learning_rate": 0.0004158658372982287, + "loss": 2.7676132202148436, + "step": 55280 + }, + { + "epoch": 0.27645, + "grad_norm": 0.34183257818222046, + "learning_rate": 0.0004158361563314924, + "loss": 2.765800666809082, + "step": 55290 + }, + { + "epoch": 0.2765, + "grad_norm": 0.3488069176673889, + "learning_rate": 0.00041580647118983195, + "loss": 2.78631649017334, + "step": 55300 + }, + { + "epoch": 0.27655, + "grad_norm": 0.4079170823097229, + "learning_rate": 0.0004157767818739947, + "loss": 2.781680679321289, + "step": 55310 + }, + { + "epoch": 0.2766, + "grad_norm": 0.31323644518852234, + "learning_rate": 0.0004157470883847281, + "loss": 2.821961212158203, + "step": 55320 + }, + { + "epoch": 0.27665, + "grad_norm": 0.324001282453537, + "learning_rate": 0.00041571739072277956, + "loss": 2.7636852264404297, + "step": 55330 + }, + { + "epoch": 0.2767, + "grad_norm": 0.3308009207248688, + "learning_rate": 0.0004156876888888969, + "loss": 2.7810760498046876, + "step": 55340 + }, + { + "epoch": 0.27675, + "grad_norm": 0.37348005175590515, + "learning_rate": 0.00041565798288382776, + "loss": 2.8100391387939454, + "step": 55350 + }, + { + "epoch": 0.2768, + "grad_norm": 0.31239867210388184, + "learning_rate": 0.00041562827270832004, + "loss": 2.8145729064941407, + "step": 55360 + }, + { + "epoch": 0.27685, + "grad_norm": 0.31011590361595154, + "learning_rate": 0.0004155985583631216, + "loss": 2.794356346130371, + "step": 55370 + }, + { + "epoch": 0.2769, + "grad_norm": 0.32485464215278625, + "learning_rate": 0.00041556883984898057, + "loss": 2.696132469177246, + "step": 55380 + }, + { + "epoch": 0.27695, + "grad_norm": 0.30816811323165894, + "learning_rate": 0.0004155391171666452, + "loss": 2.7967765808105467, + "step": 55390 + }, + { + "epoch": 0.277, + "grad_norm": 0.32836970686912537, + "learning_rate": 0.0004155093903168635, + "loss": 2.8471664428710937, + "step": 55400 + }, + { + "epoch": 0.27705, + "grad_norm": 0.32328012585639954, + "learning_rate": 0.0004154796593003841, + "loss": 2.7405105590820313, + "step": 55410 + }, + { + "epoch": 0.2771, + "grad_norm": 0.3928207755088806, + "learning_rate": 0.0004154499241179554, + "loss": 2.796983337402344, + "step": 55420 + }, + { + "epoch": 0.27715, + "grad_norm": 0.3366868197917938, + "learning_rate": 0.00041542018477032594, + "loss": 2.850372314453125, + "step": 55430 + }, + { + "epoch": 0.2772, + "grad_norm": 0.405938059091568, + "learning_rate": 0.0004153904412582444, + "loss": 2.75683708190918, + "step": 55440 + }, + { + "epoch": 0.27725, + "grad_norm": 0.3256724178791046, + "learning_rate": 0.0004153606935824596, + "loss": 2.742029571533203, + "step": 55450 + }, + { + "epoch": 0.2773, + "grad_norm": 0.31637924909591675, + "learning_rate": 0.00041533094174372055, + "loss": 2.7827205657958984, + "step": 55460 + }, + { + "epoch": 0.27735, + "grad_norm": 0.3664241135120392, + "learning_rate": 0.0004153011857427761, + "loss": 2.798270606994629, + "step": 55470 + }, + { + "epoch": 0.2774, + "grad_norm": 0.39953291416168213, + "learning_rate": 0.00041527142558037535, + "loss": 2.7657752990722657, + "step": 55480 + }, + { + "epoch": 0.27745, + "grad_norm": 0.34794002771377563, + "learning_rate": 0.0004152416612572676, + "loss": 2.7730884552001953, + "step": 55490 + }, + { + "epoch": 0.2775, + "grad_norm": 0.3611011207103729, + "learning_rate": 0.00041521189277420213, + "loss": 2.740045928955078, + "step": 55500 + }, + { + "epoch": 0.27755, + "grad_norm": 0.3605385422706604, + "learning_rate": 0.00041518212013192833, + "loss": 2.8382848739624023, + "step": 55510 + }, + { + "epoch": 0.2776, + "grad_norm": 0.35547688603401184, + "learning_rate": 0.0004151523433311958, + "loss": 2.8233259201049803, + "step": 55520 + }, + { + "epoch": 0.27765, + "grad_norm": 0.3629808723926544, + "learning_rate": 0.00041512256237275413, + "loss": 2.796675682067871, + "step": 55530 + }, + { + "epoch": 0.2777, + "grad_norm": 0.32388049364089966, + "learning_rate": 0.0004150927772573531, + "loss": 2.8237085342407227, + "step": 55540 + }, + { + "epoch": 0.27775, + "grad_norm": 0.3106570243835449, + "learning_rate": 0.00041506298798574243, + "loss": 2.7819049835205076, + "step": 55550 + }, + { + "epoch": 0.2778, + "grad_norm": 0.3310966491699219, + "learning_rate": 0.00041503319455867216, + "loss": 2.774235153198242, + "step": 55560 + }, + { + "epoch": 0.27785, + "grad_norm": 0.34881991147994995, + "learning_rate": 0.00041500339697689233, + "loss": 2.82376766204834, + "step": 55570 + }, + { + "epoch": 0.2779, + "grad_norm": 0.3560093343257904, + "learning_rate": 0.0004149735952411531, + "loss": 2.8088199615478517, + "step": 55580 + }, + { + "epoch": 0.27795, + "grad_norm": 0.3862331807613373, + "learning_rate": 0.00041494378935220467, + "loss": 2.7784021377563475, + "step": 55590 + }, + { + "epoch": 0.278, + "grad_norm": 0.36905455589294434, + "learning_rate": 0.00041491397931079743, + "loss": 2.7437070846557616, + "step": 55600 + }, + { + "epoch": 0.27805, + "grad_norm": 0.33872705698013306, + "learning_rate": 0.00041488416511768197, + "loss": 2.773875427246094, + "step": 55610 + }, + { + "epoch": 0.2781, + "grad_norm": 0.3472632169723511, + "learning_rate": 0.00041485434677360867, + "loss": 2.806808280944824, + "step": 55620 + }, + { + "epoch": 0.27815, + "grad_norm": 0.3355688452720642, + "learning_rate": 0.0004148245242793283, + "loss": 2.763914680480957, + "step": 55630 + }, + { + "epoch": 0.2782, + "grad_norm": 0.32141587138175964, + "learning_rate": 0.0004147946976355916, + "loss": 2.7929248809814453, + "step": 55640 + }, + { + "epoch": 0.27825, + "grad_norm": 0.35725435614585876, + "learning_rate": 0.0004147648668431496, + "loss": 2.8134443283081056, + "step": 55650 + }, + { + "epoch": 0.2783, + "grad_norm": 0.338199645280838, + "learning_rate": 0.000414735031902753, + "loss": 2.8208789825439453, + "step": 55660 + }, + { + "epoch": 0.27835, + "grad_norm": 0.3550020456314087, + "learning_rate": 0.00041470519281515324, + "loss": 2.784756088256836, + "step": 55670 + }, + { + "epoch": 0.2784, + "grad_norm": 0.3493032157421112, + "learning_rate": 0.00041467534958110125, + "loss": 2.6971107482910157, + "step": 55680 + }, + { + "epoch": 0.27845, + "grad_norm": 0.3527316749095917, + "learning_rate": 0.0004146455022013485, + "loss": 2.80307674407959, + "step": 55690 + }, + { + "epoch": 0.2785, + "grad_norm": 0.33378857374191284, + "learning_rate": 0.0004146156506766463, + "loss": 2.7493207931518553, + "step": 55700 + }, + { + "epoch": 0.27855, + "grad_norm": 0.3255578577518463, + "learning_rate": 0.0004145857950077462, + "loss": 2.7755962371826173, + "step": 55710 + }, + { + "epoch": 0.2786, + "grad_norm": 0.32799971103668213, + "learning_rate": 0.0004145559351953998, + "loss": 2.8145132064819336, + "step": 55720 + }, + { + "epoch": 0.27865, + "grad_norm": 0.3646393418312073, + "learning_rate": 0.00041452607124035885, + "loss": 2.781851577758789, + "step": 55730 + }, + { + "epoch": 0.2787, + "grad_norm": 0.3387882113456726, + "learning_rate": 0.0004144962031433751, + "loss": 2.723408508300781, + "step": 55740 + }, + { + "epoch": 0.27875, + "grad_norm": 0.34488382935523987, + "learning_rate": 0.00041446633090520065, + "loss": 2.7490535736083985, + "step": 55750 + }, + { + "epoch": 0.2788, + "grad_norm": 0.32681500911712646, + "learning_rate": 0.0004144364545265873, + "loss": 2.741973876953125, + "step": 55760 + }, + { + "epoch": 0.27885, + "grad_norm": 0.35520321130752563, + "learning_rate": 0.0004144065740082874, + "loss": 2.798104667663574, + "step": 55770 + }, + { + "epoch": 0.2789, + "grad_norm": 0.34358856081962585, + "learning_rate": 0.000414376689351053, + "loss": 2.817652130126953, + "step": 55780 + }, + { + "epoch": 0.27895, + "grad_norm": 0.32570379972457886, + "learning_rate": 0.00041434680055563665, + "loss": 2.7762680053710938, + "step": 55790 + }, + { + "epoch": 0.279, + "grad_norm": 0.3334934711456299, + "learning_rate": 0.00041431690762279064, + "loss": 2.771595001220703, + "step": 55800 + }, + { + "epoch": 0.27905, + "grad_norm": 0.3491438329219818, + "learning_rate": 0.00041428701055326757, + "loss": 2.844997787475586, + "step": 55810 + }, + { + "epoch": 0.2791, + "grad_norm": 0.3141382038593292, + "learning_rate": 0.00041425710934782015, + "loss": 2.8125272750854493, + "step": 55820 + }, + { + "epoch": 0.27915, + "grad_norm": 0.3376762568950653, + "learning_rate": 0.00041422720400720103, + "loss": 2.8260120391845702, + "step": 55830 + }, + { + "epoch": 0.2792, + "grad_norm": 0.3241814374923706, + "learning_rate": 0.00041419729453216314, + "loss": 2.7763298034667967, + "step": 55840 + }, + { + "epoch": 0.27925, + "grad_norm": 0.36216771602630615, + "learning_rate": 0.0004141673809234595, + "loss": 2.748993682861328, + "step": 55850 + }, + { + "epoch": 0.2793, + "grad_norm": 0.37940260767936707, + "learning_rate": 0.0004141374631818432, + "loss": 2.78934326171875, + "step": 55860 + }, + { + "epoch": 0.27935, + "grad_norm": 0.37327900528907776, + "learning_rate": 0.00041410754130806726, + "loss": 2.75130615234375, + "step": 55870 + }, + { + "epoch": 0.2794, + "grad_norm": 0.3352835178375244, + "learning_rate": 0.00041407761530288514, + "loss": 2.81448974609375, + "step": 55880 + }, + { + "epoch": 0.27945, + "grad_norm": 0.3741779923439026, + "learning_rate": 0.0004140476851670502, + "loss": 2.7004253387451174, + "step": 55890 + }, + { + "epoch": 0.2795, + "grad_norm": 0.3286983370780945, + "learning_rate": 0.00041401775090131577, + "loss": 2.7897262573242188, + "step": 55900 + }, + { + "epoch": 0.27955, + "grad_norm": 0.3246200680732727, + "learning_rate": 0.0004139878125064355, + "loss": 2.7576200485229494, + "step": 55910 + }, + { + "epoch": 0.2796, + "grad_norm": 0.32395246624946594, + "learning_rate": 0.0004139578699831633, + "loss": 2.815822219848633, + "step": 55920 + }, + { + "epoch": 0.27965, + "grad_norm": 0.32093435525894165, + "learning_rate": 0.0004139279233322527, + "loss": 2.821604537963867, + "step": 55930 + }, + { + "epoch": 0.2797, + "grad_norm": 0.2962508797645569, + "learning_rate": 0.00041389797255445785, + "loss": 2.7602161407470702, + "step": 55940 + }, + { + "epoch": 0.27975, + "grad_norm": 0.4278251826763153, + "learning_rate": 0.0004138680176505325, + "loss": 2.800942802429199, + "step": 55950 + }, + { + "epoch": 0.2798, + "grad_norm": 0.35140153765678406, + "learning_rate": 0.000413838058621231, + "loss": 2.8515251159667967, + "step": 55960 + }, + { + "epoch": 0.27985, + "grad_norm": 0.3648618161678314, + "learning_rate": 0.00041380809546730745, + "loss": 2.7470779418945312, + "step": 55970 + }, + { + "epoch": 0.2799, + "grad_norm": 0.318998783826828, + "learning_rate": 0.00041377812818951614, + "loss": 2.8245370864868162, + "step": 55980 + }, + { + "epoch": 0.27995, + "grad_norm": 0.3622021973133087, + "learning_rate": 0.00041374815678861163, + "loss": 2.828797149658203, + "step": 55990 + }, + { + "epoch": 0.28, + "grad_norm": 0.41447803378105164, + "learning_rate": 0.0004137181812653483, + "loss": 2.782292938232422, + "step": 56000 + }, + { + "epoch": 0.28005, + "grad_norm": 0.34391409158706665, + "learning_rate": 0.00041368820162048084, + "loss": 2.764924430847168, + "step": 56010 + }, + { + "epoch": 0.2801, + "grad_norm": 0.3257904648780823, + "learning_rate": 0.0004136582178547641, + "loss": 2.818672752380371, + "step": 56020 + }, + { + "epoch": 0.28015, + "grad_norm": 0.3499886393547058, + "learning_rate": 0.00041362822996895266, + "loss": 2.755257797241211, + "step": 56030 + }, + { + "epoch": 0.2802, + "grad_norm": 0.33984923362731934, + "learning_rate": 0.0004135982379638018, + "loss": 2.714219856262207, + "step": 56040 + }, + { + "epoch": 0.28025, + "grad_norm": 0.2886410057544708, + "learning_rate": 0.00041356824184006626, + "loss": 2.777745246887207, + "step": 56050 + }, + { + "epoch": 0.2803, + "grad_norm": 0.34880518913269043, + "learning_rate": 0.0004135382415985014, + "loss": 2.7729854583740234, + "step": 56060 + }, + { + "epoch": 0.28035, + "grad_norm": 0.3631017208099365, + "learning_rate": 0.0004135082372398623, + "loss": 2.757848358154297, + "step": 56070 + }, + { + "epoch": 0.2804, + "grad_norm": 0.35279956459999084, + "learning_rate": 0.00041347822876490454, + "loss": 2.753879737854004, + "step": 56080 + }, + { + "epoch": 0.28045, + "grad_norm": 0.3273608386516571, + "learning_rate": 0.00041344821617438337, + "loss": 2.7728244781494142, + "step": 56090 + }, + { + "epoch": 0.2805, + "grad_norm": 0.29933688044548035, + "learning_rate": 0.00041341819946905453, + "loss": 2.8025821685791015, + "step": 56100 + }, + { + "epoch": 0.28055, + "grad_norm": 0.3459028899669647, + "learning_rate": 0.0004133881786496736, + "loss": 2.7786487579345702, + "step": 56110 + }, + { + "epoch": 0.2806, + "grad_norm": 0.33328548073768616, + "learning_rate": 0.00041335815371699633, + "loss": 2.7957515716552734, + "step": 56120 + }, + { + "epoch": 0.28065, + "grad_norm": 0.31638145446777344, + "learning_rate": 0.0004133281246717786, + "loss": 2.7728485107421874, + "step": 56130 + }, + { + "epoch": 0.2807, + "grad_norm": 0.3631184697151184, + "learning_rate": 0.0004132980915147765, + "loss": 2.765945053100586, + "step": 56140 + }, + { + "epoch": 0.28075, + "grad_norm": 0.31975406408309937, + "learning_rate": 0.000413268054246746, + "loss": 2.750407028198242, + "step": 56150 + }, + { + "epoch": 0.2808, + "grad_norm": 0.35623180866241455, + "learning_rate": 0.0004132380128684433, + "loss": 2.7906261444091798, + "step": 56160 + }, + { + "epoch": 0.28085, + "grad_norm": 0.3478773236274719, + "learning_rate": 0.00041320796738062473, + "loss": 2.7865974426269533, + "step": 56170 + }, + { + "epoch": 0.2809, + "grad_norm": 0.34278857707977295, + "learning_rate": 0.00041317791778404666, + "loss": 2.8197748184204103, + "step": 56180 + }, + { + "epoch": 0.28095, + "grad_norm": 0.32295551896095276, + "learning_rate": 0.00041314786407946563, + "loss": 2.7811586380004885, + "step": 56190 + }, + { + "epoch": 0.281, + "grad_norm": 0.3746480941772461, + "learning_rate": 0.0004131178062676382, + "loss": 2.771931457519531, + "step": 56200 + }, + { + "epoch": 0.28105, + "grad_norm": 0.33413422107696533, + "learning_rate": 0.0004130877443493211, + "loss": 2.798667144775391, + "step": 56210 + }, + { + "epoch": 0.2811, + "grad_norm": 0.3196810483932495, + "learning_rate": 0.0004130576783252712, + "loss": 2.7924610137939454, + "step": 56220 + }, + { + "epoch": 0.28115, + "grad_norm": 0.32690665125846863, + "learning_rate": 0.0004130276081962452, + "loss": 2.8019275665283203, + "step": 56230 + }, + { + "epoch": 0.2812, + "grad_norm": 0.34079959988594055, + "learning_rate": 0.00041299753396300035, + "loss": 2.801755905151367, + "step": 56240 + }, + { + "epoch": 0.28125, + "grad_norm": 0.39078205823898315, + "learning_rate": 0.00041296745562629356, + "loss": 2.7980112075805663, + "step": 56250 + }, + { + "epoch": 0.2813, + "grad_norm": 0.3613402545452118, + "learning_rate": 0.00041293737318688227, + "loss": 2.795375442504883, + "step": 56260 + }, + { + "epoch": 0.28135, + "grad_norm": 0.41323724389076233, + "learning_rate": 0.0004129072866455236, + "loss": 2.741642189025879, + "step": 56270 + }, + { + "epoch": 0.2814, + "grad_norm": 0.33804214000701904, + "learning_rate": 0.00041287719600297516, + "loss": 2.7703643798828126, + "step": 56280 + }, + { + "epoch": 0.28145, + "grad_norm": 0.33646729588508606, + "learning_rate": 0.00041284710125999435, + "loss": 2.761273193359375, + "step": 56290 + }, + { + "epoch": 0.2815, + "grad_norm": 0.3092899024486542, + "learning_rate": 0.00041281700241733895, + "loss": 2.7791046142578124, + "step": 56300 + }, + { + "epoch": 0.28155, + "grad_norm": 0.3941107392311096, + "learning_rate": 0.00041278689947576655, + "loss": 2.7605356216430663, + "step": 56310 + }, + { + "epoch": 0.2816, + "grad_norm": 0.33383333683013916, + "learning_rate": 0.00041275679243603505, + "loss": 2.781455421447754, + "step": 56320 + }, + { + "epoch": 0.28165, + "grad_norm": 0.35439547896385193, + "learning_rate": 0.00041272668129890236, + "loss": 2.7786808013916016, + "step": 56330 + }, + { + "epoch": 0.2817, + "grad_norm": 0.33221766352653503, + "learning_rate": 0.0004126965660651265, + "loss": 2.7809822082519533, + "step": 56340 + }, + { + "epoch": 0.28175, + "grad_norm": 0.401263028383255, + "learning_rate": 0.00041266644673546574, + "loss": 2.775283432006836, + "step": 56350 + }, + { + "epoch": 0.2818, + "grad_norm": 0.327955424785614, + "learning_rate": 0.0004126363233106782, + "loss": 2.7942859649658205, + "step": 56360 + }, + { + "epoch": 0.28185, + "grad_norm": 0.3124244511127472, + "learning_rate": 0.00041260619579152247, + "loss": 2.7980056762695313, + "step": 56370 + }, + { + "epoch": 0.2819, + "grad_norm": 0.34314465522766113, + "learning_rate": 0.00041257606417875677, + "loss": 2.8218357086181642, + "step": 56380 + }, + { + "epoch": 0.28195, + "grad_norm": 0.3373999297618866, + "learning_rate": 0.00041254592847313963, + "loss": 2.8655548095703125, + "step": 56390 + }, + { + "epoch": 0.282, + "grad_norm": 0.3165168762207031, + "learning_rate": 0.00041251578867543, + "loss": 2.783711624145508, + "step": 56400 + }, + { + "epoch": 0.28205, + "grad_norm": 0.3547545075416565, + "learning_rate": 0.00041248564478638637, + "loss": 2.7652210235595702, + "step": 56410 + }, + { + "epoch": 0.2821, + "grad_norm": 0.3497694432735443, + "learning_rate": 0.00041245549680676776, + "loss": 2.8231552124023436, + "step": 56420 + }, + { + "epoch": 0.28215, + "grad_norm": 0.3177368640899658, + "learning_rate": 0.0004124253447373331, + "loss": 2.818668746948242, + "step": 56430 + }, + { + "epoch": 0.2822, + "grad_norm": 0.404225617647171, + "learning_rate": 0.0004123951885788415, + "loss": 2.791560173034668, + "step": 56440 + }, + { + "epoch": 0.28225, + "grad_norm": 0.3169245421886444, + "learning_rate": 0.000412365028332052, + "loss": 2.832415008544922, + "step": 56450 + }, + { + "epoch": 0.2823, + "grad_norm": 0.3079777657985687, + "learning_rate": 0.00041233486399772415, + "loss": 2.790848159790039, + "step": 56460 + }, + { + "epoch": 0.28235, + "grad_norm": 0.32530540227890015, + "learning_rate": 0.000412304695576617, + "loss": 2.831440734863281, + "step": 56470 + }, + { + "epoch": 0.2824, + "grad_norm": 0.3599374294281006, + "learning_rate": 0.00041227452306949035, + "loss": 2.778532028198242, + "step": 56480 + }, + { + "epoch": 0.28245, + "grad_norm": 0.3721809983253479, + "learning_rate": 0.0004122443464771036, + "loss": 2.7591388702392576, + "step": 56490 + }, + { + "epoch": 0.2825, + "grad_norm": 0.2983962595462799, + "learning_rate": 0.00041221416580021665, + "loss": 2.7840938568115234, + "step": 56500 + }, + { + "epoch": 0.28255, + "grad_norm": 0.38020071387290955, + "learning_rate": 0.00041218398103958897, + "loss": 2.8176996231079103, + "step": 56510 + }, + { + "epoch": 0.2826, + "grad_norm": 0.3139925003051758, + "learning_rate": 0.0004121537921959808, + "loss": 2.792660140991211, + "step": 56520 + }, + { + "epoch": 0.28265, + "grad_norm": 0.3738161325454712, + "learning_rate": 0.000412123599270152, + "loss": 2.8590911865234374, + "step": 56530 + }, + { + "epoch": 0.2827, + "grad_norm": 0.33545663952827454, + "learning_rate": 0.00041209340226286255, + "loss": 2.788884925842285, + "step": 56540 + }, + { + "epoch": 0.28275, + "grad_norm": 0.3362540006637573, + "learning_rate": 0.0004120632011748728, + "loss": 2.7706668853759764, + "step": 56550 + }, + { + "epoch": 0.2828, + "grad_norm": 0.38199788331985474, + "learning_rate": 0.0004120329960069431, + "loss": 2.7943954467773438, + "step": 56560 + }, + { + "epoch": 0.28285, + "grad_norm": 0.3594353497028351, + "learning_rate": 0.00041200278675983384, + "loss": 2.775797462463379, + "step": 56570 + }, + { + "epoch": 0.2829, + "grad_norm": 0.3390730917453766, + "learning_rate": 0.00041197257343430544, + "loss": 2.8467899322509767, + "step": 56580 + }, + { + "epoch": 0.28295, + "grad_norm": 0.3343215584754944, + "learning_rate": 0.00041194235603111864, + "loss": 2.8310895919799806, + "step": 56590 + }, + { + "epoch": 0.283, + "grad_norm": 0.3415638506412506, + "learning_rate": 0.00041191213455103406, + "loss": 2.825593948364258, + "step": 56600 + }, + { + "epoch": 0.28305, + "grad_norm": 0.3461700677871704, + "learning_rate": 0.0004118819089948126, + "loss": 2.748556137084961, + "step": 56610 + }, + { + "epoch": 0.2831, + "grad_norm": 0.41996699571609497, + "learning_rate": 0.0004118516793632151, + "loss": 2.781764030456543, + "step": 56620 + }, + { + "epoch": 0.28315, + "grad_norm": 0.30770692229270935, + "learning_rate": 0.00041182144565700277, + "loss": 2.788214111328125, + "step": 56630 + }, + { + "epoch": 0.2832, + "grad_norm": 0.3735395073890686, + "learning_rate": 0.0004117912078769366, + "loss": 2.810822296142578, + "step": 56640 + }, + { + "epoch": 0.28325, + "grad_norm": 0.40058788657188416, + "learning_rate": 0.00041176096602377774, + "loss": 2.8093952178955077, + "step": 56650 + }, + { + "epoch": 0.2833, + "grad_norm": 0.3792332708835602, + "learning_rate": 0.0004117307200982877, + "loss": 2.823686218261719, + "step": 56660 + }, + { + "epoch": 0.28335, + "grad_norm": 0.3341028392314911, + "learning_rate": 0.0004117004701012279, + "loss": 2.8105541229248048, + "step": 56670 + }, + { + "epoch": 0.2834, + "grad_norm": 0.34578412771224976, + "learning_rate": 0.00041167021603335987, + "loss": 2.7789573669433594, + "step": 56680 + }, + { + "epoch": 0.28345, + "grad_norm": 0.3369194567203522, + "learning_rate": 0.00041163995789544514, + "loss": 2.796636962890625, + "step": 56690 + }, + { + "epoch": 0.2835, + "grad_norm": 0.29839417338371277, + "learning_rate": 0.00041160969568824557, + "loss": 2.8107524871826173, + "step": 56700 + }, + { + "epoch": 0.28355, + "grad_norm": 0.338064581155777, + "learning_rate": 0.00041157942941252303, + "loss": 2.7604141235351562, + "step": 56710 + }, + { + "epoch": 0.2836, + "grad_norm": 0.3491729497909546, + "learning_rate": 0.00041154915906903945, + "loss": 2.7437183380126955, + "step": 56720 + }, + { + "epoch": 0.28365, + "grad_norm": 0.3199251592159271, + "learning_rate": 0.00041151888465855685, + "loss": 2.8201332092285156, + "step": 56730 + }, + { + "epoch": 0.2837, + "grad_norm": 0.34941914677619934, + "learning_rate": 0.00041148860618183737, + "loss": 2.8098272323608398, + "step": 56740 + }, + { + "epoch": 0.28375, + "grad_norm": 0.31067630648612976, + "learning_rate": 0.00041145832363964337, + "loss": 2.7720502853393554, + "step": 56750 + }, + { + "epoch": 0.2838, + "grad_norm": 0.31337013840675354, + "learning_rate": 0.0004114280370327371, + "loss": 2.7724170684814453, + "step": 56760 + }, + { + "epoch": 0.28385, + "grad_norm": 0.33364397287368774, + "learning_rate": 0.0004113977463618811, + "loss": 2.7671342849731446, + "step": 56770 + }, + { + "epoch": 0.2839, + "grad_norm": 0.3138003349304199, + "learning_rate": 0.0004113674516278379, + "loss": 2.8579566955566404, + "step": 56780 + }, + { + "epoch": 0.28395, + "grad_norm": 0.4420897364616394, + "learning_rate": 0.0004113371528313702, + "loss": 2.7545839309692384, + "step": 56790 + }, + { + "epoch": 0.284, + "grad_norm": 0.37903764843940735, + "learning_rate": 0.0004113068499732407, + "loss": 2.7733327865600588, + "step": 56800 + }, + { + "epoch": 0.28405, + "grad_norm": 0.3986048698425293, + "learning_rate": 0.00041127654305421246, + "loss": 2.8304069519042967, + "step": 56810 + }, + { + "epoch": 0.2841, + "grad_norm": 0.3254458010196686, + "learning_rate": 0.0004112462320750482, + "loss": 2.8089555740356444, + "step": 56820 + }, + { + "epoch": 0.28415, + "grad_norm": 0.33172574639320374, + "learning_rate": 0.00041121591703651117, + "loss": 2.8158899307250977, + "step": 56830 + }, + { + "epoch": 0.2842, + "grad_norm": 0.34154409170150757, + "learning_rate": 0.00041118559793936447, + "loss": 2.776771354675293, + "step": 56840 + }, + { + "epoch": 0.28425, + "grad_norm": 0.3132590353488922, + "learning_rate": 0.00041115527478437154, + "loss": 2.769377517700195, + "step": 56850 + }, + { + "epoch": 0.2843, + "grad_norm": 0.359476238489151, + "learning_rate": 0.0004111249475722955, + "loss": 2.796436309814453, + "step": 56860 + }, + { + "epoch": 0.28435, + "grad_norm": 0.3551894426345825, + "learning_rate": 0.0004110946163039001, + "loss": 2.799892044067383, + "step": 56870 + }, + { + "epoch": 0.2844, + "grad_norm": 0.3686322867870331, + "learning_rate": 0.0004110642809799487, + "loss": 2.8159357070922852, + "step": 56880 + }, + { + "epoch": 0.28445, + "grad_norm": 0.342891663312912, + "learning_rate": 0.00041103394160120515, + "loss": 2.763919639587402, + "step": 56890 + }, + { + "epoch": 0.2845, + "grad_norm": 0.32634949684143066, + "learning_rate": 0.00041100359816843325, + "loss": 2.814703941345215, + "step": 56900 + }, + { + "epoch": 0.28455, + "grad_norm": 0.35493719577789307, + "learning_rate": 0.0004109732506823968, + "loss": 2.822018051147461, + "step": 56910 + }, + { + "epoch": 0.2846, + "grad_norm": 0.3137242794036865, + "learning_rate": 0.00041094289914385976, + "loss": 2.7908992767333984, + "step": 56920 + }, + { + "epoch": 0.28465, + "grad_norm": 0.33985045552253723, + "learning_rate": 0.0004109125435535864, + "loss": 2.7526611328125, + "step": 56930 + }, + { + "epoch": 0.2847, + "grad_norm": 0.3684110641479492, + "learning_rate": 0.0004108821839123408, + "loss": 2.7873348236083983, + "step": 56940 + }, + { + "epoch": 0.28475, + "grad_norm": 0.36375313997268677, + "learning_rate": 0.00041085182022088736, + "loss": 2.823494720458984, + "step": 56950 + }, + { + "epoch": 0.2848, + "grad_norm": 0.3930044174194336, + "learning_rate": 0.00041082145247999034, + "loss": 2.8006900787353515, + "step": 56960 + }, + { + "epoch": 0.28485, + "grad_norm": 0.3455347716808319, + "learning_rate": 0.0004107910806904144, + "loss": 2.797385025024414, + "step": 56970 + }, + { + "epoch": 0.2849, + "grad_norm": 0.3203083872795105, + "learning_rate": 0.000410760704852924, + "loss": 2.7662284851074217, + "step": 56980 + }, + { + "epoch": 0.28495, + "grad_norm": 0.3458009660243988, + "learning_rate": 0.000410730324968284, + "loss": 2.8042346954345705, + "step": 56990 + }, + { + "epoch": 0.285, + "grad_norm": 0.3548101782798767, + "learning_rate": 0.00041069994103725907, + "loss": 2.85766544342041, + "step": 57000 + }, + { + "epoch": 0.28505, + "grad_norm": 0.3446357548236847, + "learning_rate": 0.00041066955306061424, + "loss": 2.755265808105469, + "step": 57010 + }, + { + "epoch": 0.2851, + "grad_norm": 0.33006560802459717, + "learning_rate": 0.00041063916103911447, + "loss": 2.7623300552368164, + "step": 57020 + }, + { + "epoch": 0.28515, + "grad_norm": 0.3265634775161743, + "learning_rate": 0.000410608764973525, + "loss": 2.756167984008789, + "step": 57030 + }, + { + "epoch": 0.2852, + "grad_norm": 0.34873485565185547, + "learning_rate": 0.00041057836486461073, + "loss": 2.7947601318359374, + "step": 57040 + }, + { + "epoch": 0.28525, + "grad_norm": 0.3245150148868561, + "learning_rate": 0.00041054796071313734, + "loss": 2.746303176879883, + "step": 57050 + }, + { + "epoch": 0.2853, + "grad_norm": 0.33779415488243103, + "learning_rate": 0.00041051755251987013, + "loss": 2.8295642852783205, + "step": 57060 + }, + { + "epoch": 0.28535, + "grad_norm": 0.30775368213653564, + "learning_rate": 0.00041048714028557457, + "loss": 2.8237384796142577, + "step": 57070 + }, + { + "epoch": 0.2854, + "grad_norm": 0.310573011636734, + "learning_rate": 0.00041045672401101625, + "loss": 2.7565919876098635, + "step": 57080 + }, + { + "epoch": 0.28545, + "grad_norm": 0.3338609039783478, + "learning_rate": 0.0004104263036969611, + "loss": 2.7810998916625977, + "step": 57090 + }, + { + "epoch": 0.2855, + "grad_norm": 0.3227618336677551, + "learning_rate": 0.00041039587934417476, + "loss": 2.7744937896728517, + "step": 57100 + }, + { + "epoch": 0.28555, + "grad_norm": 0.3342578411102295, + "learning_rate": 0.00041036545095342325, + "loss": 2.798578071594238, + "step": 57110 + }, + { + "epoch": 0.2856, + "grad_norm": 0.36066457629203796, + "learning_rate": 0.0004103350185254725, + "loss": 2.7992515563964844, + "step": 57120 + }, + { + "epoch": 0.28565, + "grad_norm": 0.3230462670326233, + "learning_rate": 0.0004103045820610889, + "loss": 2.7871795654296876, + "step": 57130 + }, + { + "epoch": 0.2857, + "grad_norm": 0.3265928030014038, + "learning_rate": 0.00041027414156103836, + "loss": 2.7714311599731447, + "step": 57140 + }, + { + "epoch": 0.28575, + "grad_norm": 0.33370503783226013, + "learning_rate": 0.0004102436970260874, + "loss": 2.7730321884155273, + "step": 57150 + }, + { + "epoch": 0.2858, + "grad_norm": 0.33112603425979614, + "learning_rate": 0.0004102132484570025, + "loss": 2.7944047927856444, + "step": 57160 + }, + { + "epoch": 0.28585, + "grad_norm": 0.35680198669433594, + "learning_rate": 0.0004101827958545502, + "loss": 2.7916154861450195, + "step": 57170 + }, + { + "epoch": 0.2859, + "grad_norm": 0.3369874954223633, + "learning_rate": 0.00041015233921949695, + "loss": 2.7618738174438477, + "step": 57180 + }, + { + "epoch": 0.28595, + "grad_norm": 0.3523711860179901, + "learning_rate": 0.00041012187855260973, + "loss": 2.821013832092285, + "step": 57190 + }, + { + "epoch": 0.286, + "grad_norm": 0.33177855610847473, + "learning_rate": 0.0004100914138546552, + "loss": 2.7922269821166994, + "step": 57200 + }, + { + "epoch": 0.28605, + "grad_norm": 0.3257180452346802, + "learning_rate": 0.00041006094512640044, + "loss": 2.7744884490966797, + "step": 57210 + }, + { + "epoch": 0.2861, + "grad_norm": 0.3337023854255676, + "learning_rate": 0.00041003047236861246, + "loss": 2.7770063400268556, + "step": 57220 + }, + { + "epoch": 0.28615, + "grad_norm": 0.33503592014312744, + "learning_rate": 0.0004099999955820584, + "loss": 2.848906898498535, + "step": 57230 + }, + { + "epoch": 0.2862, + "grad_norm": 0.3158971965312958, + "learning_rate": 0.0004099695147675055, + "loss": 2.7594329833984377, + "step": 57240 + }, + { + "epoch": 0.28625, + "grad_norm": 0.3324859142303467, + "learning_rate": 0.0004099390299257212, + "loss": 2.787159729003906, + "step": 57250 + }, + { + "epoch": 0.2863, + "grad_norm": 0.34810808300971985, + "learning_rate": 0.0004099085410574728, + "loss": 2.8108905792236327, + "step": 57260 + }, + { + "epoch": 0.28635, + "grad_norm": 0.3472435176372528, + "learning_rate": 0.00040987804816352804, + "loss": 2.7447656631469726, + "step": 57270 + }, + { + "epoch": 0.2864, + "grad_norm": 0.4021829068660736, + "learning_rate": 0.00040984755124465444, + "loss": 2.788822555541992, + "step": 57280 + }, + { + "epoch": 0.28645, + "grad_norm": 0.36305713653564453, + "learning_rate": 0.0004098170503016198, + "loss": 2.828677177429199, + "step": 57290 + }, + { + "epoch": 0.2865, + "grad_norm": 0.34236517548561096, + "learning_rate": 0.00040978654533519197, + "loss": 2.760746192932129, + "step": 57300 + }, + { + "epoch": 0.28655, + "grad_norm": 0.32044127583503723, + "learning_rate": 0.000409756036346139, + "loss": 2.811080551147461, + "step": 57310 + }, + { + "epoch": 0.2866, + "grad_norm": 0.3444381058216095, + "learning_rate": 0.00040972552333522883, + "loss": 2.801481246948242, + "step": 57320 + }, + { + "epoch": 0.28665, + "grad_norm": 0.33935585618019104, + "learning_rate": 0.00040969500630322965, + "loss": 2.8113325119018553, + "step": 57330 + }, + { + "epoch": 0.2867, + "grad_norm": 0.33532020449638367, + "learning_rate": 0.00040966448525090983, + "loss": 2.8564096450805665, + "step": 57340 + }, + { + "epoch": 0.28675, + "grad_norm": 0.34510287642478943, + "learning_rate": 0.0004096339601790376, + "loss": 2.7761648178100584, + "step": 57350 + }, + { + "epoch": 0.2868, + "grad_norm": 0.327083557844162, + "learning_rate": 0.0004096034310883815, + "loss": 2.799987030029297, + "step": 57360 + }, + { + "epoch": 0.28685, + "grad_norm": 0.34350934624671936, + "learning_rate": 0.0004095728979797101, + "loss": 2.7928531646728514, + "step": 57370 + }, + { + "epoch": 0.2869, + "grad_norm": 0.32830360531806946, + "learning_rate": 0.00040954236085379204, + "loss": 2.7396171569824217, + "step": 57380 + }, + { + "epoch": 0.28695, + "grad_norm": 0.3394298255443573, + "learning_rate": 0.00040951181971139614, + "loss": 2.7652446746826174, + "step": 57390 + }, + { + "epoch": 0.287, + "grad_norm": 0.3767000138759613, + "learning_rate": 0.0004094812745532912, + "loss": 2.8036052703857424, + "step": 57400 + }, + { + "epoch": 0.28705, + "grad_norm": 0.34162092208862305, + "learning_rate": 0.0004094507253802463, + "loss": 2.787074089050293, + "step": 57410 + }, + { + "epoch": 0.2871, + "grad_norm": 0.34404394030570984, + "learning_rate": 0.0004094201721930304, + "loss": 2.7815488815307616, + "step": 57420 + }, + { + "epoch": 0.28715, + "grad_norm": 0.44787564873695374, + "learning_rate": 0.0004093896149924128, + "loss": 2.8347469329833985, + "step": 57430 + }, + { + "epoch": 0.2872, + "grad_norm": 0.3161117136478424, + "learning_rate": 0.0004093590537791626, + "loss": 2.824811553955078, + "step": 57440 + }, + { + "epoch": 0.28725, + "grad_norm": 0.33807089924812317, + "learning_rate": 0.00040932848855404936, + "loss": 2.787318992614746, + "step": 57450 + }, + { + "epoch": 0.2873, + "grad_norm": 0.361908495426178, + "learning_rate": 0.00040929791931784243, + "loss": 2.8306413650512696, + "step": 57460 + }, + { + "epoch": 0.28735, + "grad_norm": 0.3355678617954254, + "learning_rate": 0.0004092673460713116, + "loss": 2.8439870834350587, + "step": 57470 + }, + { + "epoch": 0.2874, + "grad_norm": 0.32642701268196106, + "learning_rate": 0.0004092367688152262, + "loss": 2.814422607421875, + "step": 57480 + }, + { + "epoch": 0.28745, + "grad_norm": 0.3090766668319702, + "learning_rate": 0.00040920618755035633, + "loss": 2.7487920761108398, + "step": 57490 + }, + { + "epoch": 0.2875, + "grad_norm": 0.3342254161834717, + "learning_rate": 0.0004091756022774716, + "loss": 2.8233861923217773, + "step": 57500 + }, + { + "epoch": 0.28755, + "grad_norm": 0.3530687689781189, + "learning_rate": 0.00040914501299734243, + "loss": 2.7550607681274415, + "step": 57510 + }, + { + "epoch": 0.2876, + "grad_norm": 0.3254523277282715, + "learning_rate": 0.00040911441971073836, + "loss": 2.803701400756836, + "step": 57520 + }, + { + "epoch": 0.28765, + "grad_norm": 0.3319893479347229, + "learning_rate": 0.00040908382241842993, + "loss": 2.789906120300293, + "step": 57530 + }, + { + "epoch": 0.2877, + "grad_norm": 0.3291907012462616, + "learning_rate": 0.0004090532211211874, + "loss": 2.8016420364379884, + "step": 57540 + }, + { + "epoch": 0.28775, + "grad_norm": 0.34118765592575073, + "learning_rate": 0.00040902261581978095, + "loss": 2.84511775970459, + "step": 57550 + }, + { + "epoch": 0.2878, + "grad_norm": 0.33285173773765564, + "learning_rate": 0.0004089920065149813, + "loss": 2.7582376480102537, + "step": 57560 + }, + { + "epoch": 0.28785, + "grad_norm": 0.3317106068134308, + "learning_rate": 0.00040896139320755896, + "loss": 2.7905275344848635, + "step": 57570 + }, + { + "epoch": 0.2879, + "grad_norm": 0.36314183473587036, + "learning_rate": 0.0004089307758982846, + "loss": 2.739230155944824, + "step": 57580 + }, + { + "epoch": 0.28795, + "grad_norm": 0.3714962899684906, + "learning_rate": 0.00040890015458792896, + "loss": 2.7801019668579103, + "step": 57590 + }, + { + "epoch": 0.288, + "grad_norm": 0.3414776027202606, + "learning_rate": 0.00040886952927726304, + "loss": 2.7892276763916017, + "step": 57600 + }, + { + "epoch": 0.28805, + "grad_norm": 0.3245440423488617, + "learning_rate": 0.00040883889996705776, + "loss": 2.742414093017578, + "step": 57610 + }, + { + "epoch": 0.2881, + "grad_norm": 0.329964280128479, + "learning_rate": 0.00040880826665808427, + "loss": 2.771580696105957, + "step": 57620 + }, + { + "epoch": 0.28815, + "grad_norm": 0.37830379605293274, + "learning_rate": 0.00040877762935111374, + "loss": 2.779405212402344, + "step": 57630 + }, + { + "epoch": 0.2882, + "grad_norm": 0.3780723512172699, + "learning_rate": 0.0004087469880469173, + "loss": 2.752695846557617, + "step": 57640 + }, + { + "epoch": 0.28825, + "grad_norm": 0.36784037947654724, + "learning_rate": 0.0004087163427462667, + "loss": 2.790688133239746, + "step": 57650 + }, + { + "epoch": 0.2883, + "grad_norm": 0.32113397121429443, + "learning_rate": 0.0004086856934499331, + "loss": 2.81426944732666, + "step": 57660 + }, + { + "epoch": 0.28835, + "grad_norm": 0.33397072553634644, + "learning_rate": 0.0004086550401586883, + "loss": 2.769841194152832, + "step": 57670 + }, + { + "epoch": 0.2884, + "grad_norm": 0.3440781235694885, + "learning_rate": 0.0004086243828733039, + "loss": 2.8004129409790037, + "step": 57680 + }, + { + "epoch": 0.28845, + "grad_norm": 0.37695616483688354, + "learning_rate": 0.00040859372159455164, + "loss": 2.806120681762695, + "step": 57690 + }, + { + "epoch": 0.2885, + "grad_norm": 0.3251549005508423, + "learning_rate": 0.00040856305632320367, + "loss": 2.8632070541381838, + "step": 57700 + }, + { + "epoch": 0.28855, + "grad_norm": 0.32402077317237854, + "learning_rate": 0.00040853238706003173, + "loss": 2.761090850830078, + "step": 57710 + }, + { + "epoch": 0.2886, + "grad_norm": 0.3219938576221466, + "learning_rate": 0.000408501713805808, + "loss": 2.7496620178222657, + "step": 57720 + }, + { + "epoch": 0.28865, + "grad_norm": 0.31277602910995483, + "learning_rate": 0.00040847103656130464, + "loss": 2.773913764953613, + "step": 57730 + }, + { + "epoch": 0.2887, + "grad_norm": 0.3425081968307495, + "learning_rate": 0.00040844035532729405, + "loss": 2.796656036376953, + "step": 57740 + }, + { + "epoch": 0.28875, + "grad_norm": 0.3363751173019409, + "learning_rate": 0.00040840967010454855, + "loss": 2.788227081298828, + "step": 57750 + }, + { + "epoch": 0.2888, + "grad_norm": 0.35618019104003906, + "learning_rate": 0.0004083789808938407, + "loss": 2.787749481201172, + "step": 57760 + }, + { + "epoch": 0.28885, + "grad_norm": 0.30804792046546936, + "learning_rate": 0.000408348287695943, + "loss": 2.855157661437988, + "step": 57770 + }, + { + "epoch": 0.2889, + "grad_norm": 0.30978021025657654, + "learning_rate": 0.00040831759051162826, + "loss": 2.819750213623047, + "step": 57780 + }, + { + "epoch": 0.28895, + "grad_norm": 0.34040796756744385, + "learning_rate": 0.00040828688934166926, + "loss": 2.747323989868164, + "step": 57790 + }, + { + "epoch": 0.289, + "grad_norm": 0.3652958869934082, + "learning_rate": 0.00040825618418683887, + "loss": 2.7423431396484377, + "step": 57800 + }, + { + "epoch": 0.28905, + "grad_norm": 0.32381579279899597, + "learning_rate": 0.00040822547504791004, + "loss": 2.749323844909668, + "step": 57810 + }, + { + "epoch": 0.2891, + "grad_norm": 0.33760306239128113, + "learning_rate": 0.000408194761925656, + "loss": 2.8268741607666015, + "step": 57820 + }, + { + "epoch": 0.28915, + "grad_norm": 0.3547128736972809, + "learning_rate": 0.0004081640448208498, + "loss": 2.7711484909057615, + "step": 57830 + }, + { + "epoch": 0.2892, + "grad_norm": 0.3696802854537964, + "learning_rate": 0.0004081333237342649, + "loss": 2.8177671432495117, + "step": 57840 + }, + { + "epoch": 0.28925, + "grad_norm": 0.2920630872249603, + "learning_rate": 0.00040810259866667464, + "loss": 2.746738243103027, + "step": 57850 + }, + { + "epoch": 0.2893, + "grad_norm": 0.3369910717010498, + "learning_rate": 0.00040807186961885244, + "loss": 2.822050857543945, + "step": 57860 + }, + { + "epoch": 0.28935, + "grad_norm": 0.3299894630908966, + "learning_rate": 0.00040804113659157203, + "loss": 2.801298141479492, + "step": 57870 + }, + { + "epoch": 0.2894, + "grad_norm": 0.33925315737724304, + "learning_rate": 0.0004080103995856071, + "loss": 2.804811477661133, + "step": 57880 + }, + { + "epoch": 0.28945, + "grad_norm": 0.3188876807689667, + "learning_rate": 0.00040797965860173136, + "loss": 2.8348493576049805, + "step": 57890 + }, + { + "epoch": 0.2895, + "grad_norm": 0.3378869295120239, + "learning_rate": 0.00040794891364071884, + "loss": 2.7436656951904297, + "step": 57900 + }, + { + "epoch": 0.28955, + "grad_norm": 0.3272152543067932, + "learning_rate": 0.00040791816470334336, + "loss": 2.7580549240112306, + "step": 57910 + }, + { + "epoch": 0.2896, + "grad_norm": 0.3236615061759949, + "learning_rate": 0.00040788741179037925, + "loss": 2.8170635223388674, + "step": 57920 + }, + { + "epoch": 0.28965, + "grad_norm": 0.33186396956443787, + "learning_rate": 0.00040785665490260047, + "loss": 2.801139259338379, + "step": 57930 + }, + { + "epoch": 0.2897, + "grad_norm": 0.36549636721611023, + "learning_rate": 0.0004078258940407816, + "loss": 2.8384021759033202, + "step": 57940 + }, + { + "epoch": 0.28975, + "grad_norm": 0.32096412777900696, + "learning_rate": 0.00040779512920569684, + "loss": 2.7632143020629885, + "step": 57950 + }, + { + "epoch": 0.2898, + "grad_norm": 0.3059937655925751, + "learning_rate": 0.00040776436039812076, + "loss": 2.7852785110473635, + "step": 57960 + }, + { + "epoch": 0.28985, + "grad_norm": 0.3478514552116394, + "learning_rate": 0.00040773358761882797, + "loss": 2.784869575500488, + "step": 57970 + }, + { + "epoch": 0.2899, + "grad_norm": 0.3587123155593872, + "learning_rate": 0.00040770281086859316, + "loss": 2.79190731048584, + "step": 57980 + }, + { + "epoch": 0.28995, + "grad_norm": 0.3035874366760254, + "learning_rate": 0.0004076720301481911, + "loss": 2.7642288208007812, + "step": 57990 + }, + { + "epoch": 0.29, + "grad_norm": 0.3494371175765991, + "learning_rate": 0.0004076412454583968, + "loss": 2.8172903060913086, + "step": 58000 + }, + { + "epoch": 0.29005, + "grad_norm": 0.32719001173973083, + "learning_rate": 0.0004076104567999852, + "loss": 2.7824602127075195, + "step": 58010 + }, + { + "epoch": 0.2901, + "grad_norm": 0.3491424024105072, + "learning_rate": 0.0004075796641737314, + "loss": 2.7913486480712892, + "step": 58020 + }, + { + "epoch": 0.29015, + "grad_norm": 0.3053891062736511, + "learning_rate": 0.00040754886758041054, + "loss": 2.8180545806884765, + "step": 58030 + }, + { + "epoch": 0.2902, + "grad_norm": 0.35840630531311035, + "learning_rate": 0.00040751806702079803, + "loss": 2.7857429504394533, + "step": 58040 + }, + { + "epoch": 0.29025, + "grad_norm": 0.35019806027412415, + "learning_rate": 0.00040748726249566924, + "loss": 2.769758415222168, + "step": 58050 + }, + { + "epoch": 0.2903, + "grad_norm": 0.33357706665992737, + "learning_rate": 0.0004074564540057997, + "loss": 2.8172418594360353, + "step": 58060 + }, + { + "epoch": 0.29035, + "grad_norm": 0.3357578217983246, + "learning_rate": 0.0004074256415519649, + "loss": 2.784144401550293, + "step": 58070 + }, + { + "epoch": 0.2904, + "grad_norm": 0.3684927821159363, + "learning_rate": 0.0004073948251349407, + "loss": 2.8052928924560545, + "step": 58080 + }, + { + "epoch": 0.29045, + "grad_norm": 0.3246978223323822, + "learning_rate": 0.00040736400475550286, + "loss": 2.757457733154297, + "step": 58090 + }, + { + "epoch": 0.2905, + "grad_norm": 0.36589208245277405, + "learning_rate": 0.0004073331804144271, + "loss": 2.8119958877563476, + "step": 58100 + }, + { + "epoch": 0.29055, + "grad_norm": 0.352267861366272, + "learning_rate": 0.0004073023521124897, + "loss": 2.7669662475585937, + "step": 58110 + }, + { + "epoch": 0.2906, + "grad_norm": 0.33239513635635376, + "learning_rate": 0.0004072715198504666, + "loss": 2.8261974334716795, + "step": 58120 + }, + { + "epoch": 0.29065, + "grad_norm": 0.32253116369247437, + "learning_rate": 0.000407240683629134, + "loss": 2.812027359008789, + "step": 58130 + }, + { + "epoch": 0.2907, + "grad_norm": 0.33410337567329407, + "learning_rate": 0.0004072098434492684, + "loss": 2.798988914489746, + "step": 58140 + }, + { + "epoch": 0.29075, + "grad_norm": 0.3135933578014374, + "learning_rate": 0.0004071789993116459, + "loss": 2.794045257568359, + "step": 58150 + }, + { + "epoch": 0.2908, + "grad_norm": 0.3605506122112274, + "learning_rate": 0.0004071481512170432, + "loss": 2.808979797363281, + "step": 58160 + }, + { + "epoch": 0.29085, + "grad_norm": 0.33363184332847595, + "learning_rate": 0.00040711729916623673, + "loss": 2.778797912597656, + "step": 58170 + }, + { + "epoch": 0.2909, + "grad_norm": 0.3420080244541168, + "learning_rate": 0.0004070864431600034, + "loss": 2.7696548461914063, + "step": 58180 + }, + { + "epoch": 0.29095, + "grad_norm": 0.3160352408885956, + "learning_rate": 0.00040705558319911994, + "loss": 2.7932083129882814, + "step": 58190 + }, + { + "epoch": 0.291, + "grad_norm": 0.3376712501049042, + "learning_rate": 0.00040702471928436316, + "loss": 2.821977233886719, + "step": 58200 + }, + { + "epoch": 0.29105, + "grad_norm": 0.3736508786678314, + "learning_rate": 0.00040699385141651013, + "loss": 2.826078414916992, + "step": 58210 + }, + { + "epoch": 0.2911, + "grad_norm": 0.33028164505958557, + "learning_rate": 0.000406962979596338, + "loss": 2.7724342346191406, + "step": 58220 + }, + { + "epoch": 0.29115, + "grad_norm": 0.3474246859550476, + "learning_rate": 0.00040693210382462385, + "loss": 2.750295066833496, + "step": 58230 + }, + { + "epoch": 0.2912, + "grad_norm": 0.3219127953052521, + "learning_rate": 0.00040690122410214503, + "loss": 2.7884424209594725, + "step": 58240 + }, + { + "epoch": 0.29125, + "grad_norm": 0.3186838626861572, + "learning_rate": 0.00040687034042967895, + "loss": 2.77270393371582, + "step": 58250 + }, + { + "epoch": 0.2913, + "grad_norm": 0.31664004921913147, + "learning_rate": 0.00040683945280800316, + "loss": 2.815133285522461, + "step": 58260 + }, + { + "epoch": 0.29135, + "grad_norm": 0.3652099370956421, + "learning_rate": 0.0004068085612378951, + "loss": 2.7418144226074217, + "step": 58270 + }, + { + "epoch": 0.2914, + "grad_norm": 0.36183595657348633, + "learning_rate": 0.00040677766572013265, + "loss": 2.774628829956055, + "step": 58280 + }, + { + "epoch": 0.29145, + "grad_norm": 0.32084664702415466, + "learning_rate": 0.00040674676625549345, + "loss": 2.7897939682006836, + "step": 58290 + }, + { + "epoch": 0.2915, + "grad_norm": 0.32804054021835327, + "learning_rate": 0.0004067158628447555, + "loss": 2.7746652603149413, + "step": 58300 + }, + { + "epoch": 0.29155, + "grad_norm": 0.3410564064979553, + "learning_rate": 0.00040668495548869667, + "loss": 2.765812110900879, + "step": 58310 + }, + { + "epoch": 0.2916, + "grad_norm": 0.32370278239250183, + "learning_rate": 0.00040665404418809513, + "loss": 2.7734916687011717, + "step": 58320 + }, + { + "epoch": 0.29165, + "grad_norm": 0.4042315185070038, + "learning_rate": 0.0004066231289437292, + "loss": 2.739912414550781, + "step": 58330 + }, + { + "epoch": 0.2917, + "grad_norm": 0.33566778898239136, + "learning_rate": 0.0004065922097563769, + "loss": 2.7587203979492188, + "step": 58340 + }, + { + "epoch": 0.29175, + "grad_norm": 0.32269540429115295, + "learning_rate": 0.0004065612866268169, + "loss": 2.7870725631713866, + "step": 58350 + }, + { + "epoch": 0.2918, + "grad_norm": 0.38454172015190125, + "learning_rate": 0.0004065303595558275, + "loss": 2.770817184448242, + "step": 58360 + }, + { + "epoch": 0.29185, + "grad_norm": 0.30729493498802185, + "learning_rate": 0.00040649942854418733, + "loss": 2.717572021484375, + "step": 58370 + }, + { + "epoch": 0.2919, + "grad_norm": 0.38435500860214233, + "learning_rate": 0.0004064684935926751, + "loss": 2.7293853759765625, + "step": 58380 + }, + { + "epoch": 0.29195, + "grad_norm": 0.3105766177177429, + "learning_rate": 0.00040643755470206954, + "loss": 2.724863624572754, + "step": 58390 + }, + { + "epoch": 0.292, + "grad_norm": 0.34635257720947266, + "learning_rate": 0.00040640661187314964, + "loss": 2.7221202850341797, + "step": 58400 + }, + { + "epoch": 0.29205, + "grad_norm": 0.3295058310031891, + "learning_rate": 0.0004063756651066943, + "loss": 2.779799461364746, + "step": 58410 + }, + { + "epoch": 0.2921, + "grad_norm": 0.345778226852417, + "learning_rate": 0.0004063447144034827, + "loss": 2.793882942199707, + "step": 58420 + }, + { + "epoch": 0.29215, + "grad_norm": 0.3020574152469635, + "learning_rate": 0.00040631375976429396, + "loss": 2.817987060546875, + "step": 58430 + }, + { + "epoch": 0.2922, + "grad_norm": 0.34666192531585693, + "learning_rate": 0.0004062828011899073, + "loss": 2.785046195983887, + "step": 58440 + }, + { + "epoch": 0.29225, + "grad_norm": 0.3364968001842499, + "learning_rate": 0.00040625183868110216, + "loss": 2.7355213165283203, + "step": 58450 + }, + { + "epoch": 0.2923, + "grad_norm": 0.30530133843421936, + "learning_rate": 0.0004062208722386581, + "loss": 2.781784439086914, + "step": 58460 + }, + { + "epoch": 0.29235, + "grad_norm": 0.37033718824386597, + "learning_rate": 0.0004061899018633546, + "loss": 2.853297233581543, + "step": 58470 + }, + { + "epoch": 0.2924, + "grad_norm": 0.41855373978614807, + "learning_rate": 0.00040615892755597126, + "loss": 2.806690979003906, + "step": 58480 + }, + { + "epoch": 0.29245, + "grad_norm": 0.3623182475566864, + "learning_rate": 0.0004061279493172881, + "loss": 2.7253793716430663, + "step": 58490 + }, + { + "epoch": 0.2925, + "grad_norm": 0.36437520384788513, + "learning_rate": 0.00040609696714808476, + "loss": 2.8037479400634764, + "step": 58500 + }, + { + "epoch": 0.29255, + "grad_norm": 0.34241408109664917, + "learning_rate": 0.0004060659810491414, + "loss": 2.787900924682617, + "step": 58510 + }, + { + "epoch": 0.2926, + "grad_norm": 0.3472684323787689, + "learning_rate": 0.000406034991021238, + "loss": 2.73525390625, + "step": 58520 + }, + { + "epoch": 0.29265, + "grad_norm": 0.330415815114975, + "learning_rate": 0.00040600399706515466, + "loss": 2.826451873779297, + "step": 58530 + }, + { + "epoch": 0.2927, + "grad_norm": 0.33573204278945923, + "learning_rate": 0.0004059729991816718, + "loss": 2.836102104187012, + "step": 58540 + }, + { + "epoch": 0.29275, + "grad_norm": 0.32645806670188904, + "learning_rate": 0.0004059419973715697, + "loss": 2.80712833404541, + "step": 58550 + }, + { + "epoch": 0.2928, + "grad_norm": 0.3950437009334564, + "learning_rate": 0.00040591099163562886, + "loss": 2.829640579223633, + "step": 58560 + }, + { + "epoch": 0.29285, + "grad_norm": 0.3516944646835327, + "learning_rate": 0.00040587998197462986, + "loss": 2.7699230194091795, + "step": 58570 + }, + { + "epoch": 0.2929, + "grad_norm": 0.3270959258079529, + "learning_rate": 0.00040584896838935335, + "loss": 2.7739471435546874, + "step": 58580 + }, + { + "epoch": 0.29295, + "grad_norm": 0.3535011112689972, + "learning_rate": 0.00040581795088058016, + "loss": 2.794384765625, + "step": 58590 + }, + { + "epoch": 0.293, + "grad_norm": 0.3381982147693634, + "learning_rate": 0.00040578692944909104, + "loss": 2.795833206176758, + "step": 58600 + }, + { + "epoch": 0.29305, + "grad_norm": 0.34305399656295776, + "learning_rate": 0.000405755904095667, + "loss": 2.7881275177001954, + "step": 58610 + }, + { + "epoch": 0.2931, + "grad_norm": 0.34575650095939636, + "learning_rate": 0.0004057248748210891, + "loss": 2.773886871337891, + "step": 58620 + }, + { + "epoch": 0.29315, + "grad_norm": 0.3531855642795563, + "learning_rate": 0.00040569384162613857, + "loss": 2.776173782348633, + "step": 58630 + }, + { + "epoch": 0.2932, + "grad_norm": 0.37000271677970886, + "learning_rate": 0.0004056628045115966, + "loss": 2.791850280761719, + "step": 58640 + }, + { + "epoch": 0.29325, + "grad_norm": 0.31923314929008484, + "learning_rate": 0.0004056317634782446, + "loss": 2.790294647216797, + "step": 58650 + }, + { + "epoch": 0.2933, + "grad_norm": 0.31161999702453613, + "learning_rate": 0.0004056007185268639, + "loss": 2.8097766876220702, + "step": 58660 + }, + { + "epoch": 0.29335, + "grad_norm": 0.3260277211666107, + "learning_rate": 0.0004055696696582363, + "loss": 2.74715518951416, + "step": 58670 + }, + { + "epoch": 0.2934, + "grad_norm": 0.3410741984844208, + "learning_rate": 0.00040553861687314316, + "loss": 2.766397476196289, + "step": 58680 + }, + { + "epoch": 0.29345, + "grad_norm": 0.3629574179649353, + "learning_rate": 0.00040550756017236644, + "loss": 2.7759218215942383, + "step": 58690 + }, + { + "epoch": 0.2935, + "grad_norm": 0.3316083252429962, + "learning_rate": 0.0004054764995566879, + "loss": 2.8001453399658205, + "step": 58700 + }, + { + "epoch": 0.29355, + "grad_norm": 0.321313738822937, + "learning_rate": 0.00040544543502688956, + "loss": 2.818524932861328, + "step": 58710 + }, + { + "epoch": 0.2936, + "grad_norm": 0.32086625695228577, + "learning_rate": 0.0004054143665837534, + "loss": 2.789315414428711, + "step": 58720 + }, + { + "epoch": 0.29365, + "grad_norm": 0.3190794885158539, + "learning_rate": 0.00040538329422806166, + "loss": 2.7700939178466797, + "step": 58730 + }, + { + "epoch": 0.2937, + "grad_norm": 0.3341015577316284, + "learning_rate": 0.0004053522179605965, + "loss": 2.769862174987793, + "step": 58740 + }, + { + "epoch": 0.29375, + "grad_norm": 0.30696675181388855, + "learning_rate": 0.0004053211377821403, + "loss": 2.7332408905029295, + "step": 58750 + }, + { + "epoch": 0.2938, + "grad_norm": 0.3370659649372101, + "learning_rate": 0.00040529005369347544, + "loss": 2.8302677154541014, + "step": 58760 + }, + { + "epoch": 0.29385, + "grad_norm": 0.35062187910079956, + "learning_rate": 0.0004052589656953846, + "loss": 2.8362424850463865, + "step": 58770 + }, + { + "epoch": 0.2939, + "grad_norm": 0.3854370415210724, + "learning_rate": 0.00040522787378865024, + "loss": 2.782650184631348, + "step": 58780 + }, + { + "epoch": 0.29395, + "grad_norm": 0.33734840154647827, + "learning_rate": 0.00040519677797405527, + "loss": 2.7362592697143553, + "step": 58790 + }, + { + "epoch": 0.294, + "grad_norm": 0.34456050395965576, + "learning_rate": 0.0004051656782523824, + "loss": 2.7885967254638673, + "step": 58800 + }, + { + "epoch": 0.29405, + "grad_norm": 0.388926237821579, + "learning_rate": 0.0004051345746244146, + "loss": 2.8470062255859374, + "step": 58810 + }, + { + "epoch": 0.2941, + "grad_norm": 0.32535573840141296, + "learning_rate": 0.0004051034670909349, + "loss": 2.7599538803100585, + "step": 58820 + }, + { + "epoch": 0.29415, + "grad_norm": 0.34155192971229553, + "learning_rate": 0.0004050723556527266, + "loss": 2.7378332138061525, + "step": 58830 + }, + { + "epoch": 0.2942, + "grad_norm": 0.35258617997169495, + "learning_rate": 0.00040504124031057263, + "loss": 2.7903720855712892, + "step": 58840 + }, + { + "epoch": 0.29425, + "grad_norm": 0.32498422265052795, + "learning_rate": 0.0004050101210652566, + "loss": 2.7675485610961914, + "step": 58850 + }, + { + "epoch": 0.2943, + "grad_norm": 0.3406390845775604, + "learning_rate": 0.0004049789979175617, + "loss": 2.7931690216064453, + "step": 58860 + }, + { + "epoch": 0.29435, + "grad_norm": 0.354343444108963, + "learning_rate": 0.00040494787086827166, + "loss": 2.8208322525024414, + "step": 58870 + }, + { + "epoch": 0.2944, + "grad_norm": 0.34306296706199646, + "learning_rate": 0.00040491673991816993, + "loss": 2.7801631927490233, + "step": 58880 + }, + { + "epoch": 0.29445, + "grad_norm": 0.31120577454566956, + "learning_rate": 0.00040488560506804033, + "loss": 2.7700077056884767, + "step": 58890 + }, + { + "epoch": 0.2945, + "grad_norm": 0.35057058930397034, + "learning_rate": 0.00040485446631866663, + "loss": 2.8016313552856444, + "step": 58900 + }, + { + "epoch": 0.29455, + "grad_norm": 0.35627081990242004, + "learning_rate": 0.0004048233236708329, + "loss": 2.752054786682129, + "step": 58910 + }, + { + "epoch": 0.2946, + "grad_norm": 0.3579663634300232, + "learning_rate": 0.00040479217712532283, + "loss": 2.8045909881591795, + "step": 58920 + }, + { + "epoch": 0.29465, + "grad_norm": 0.3096756041049957, + "learning_rate": 0.00040476102668292086, + "loss": 2.7598245620727537, + "step": 58930 + }, + { + "epoch": 0.2947, + "grad_norm": 0.33064478635787964, + "learning_rate": 0.00040472987234441106, + "loss": 2.806706428527832, + "step": 58940 + }, + { + "epoch": 0.29475, + "grad_norm": 0.3530920147895813, + "learning_rate": 0.00040469871411057774, + "loss": 2.796080780029297, + "step": 58950 + }, + { + "epoch": 0.2948, + "grad_norm": 0.3142610490322113, + "learning_rate": 0.0004046675519822053, + "loss": 2.782138442993164, + "step": 58960 + }, + { + "epoch": 0.29485, + "grad_norm": 0.34552842378616333, + "learning_rate": 0.00040463638596007834, + "loss": 2.7379730224609373, + "step": 58970 + }, + { + "epoch": 0.2949, + "grad_norm": 0.3499872386455536, + "learning_rate": 0.0004046052160449814, + "loss": 2.7878559112548826, + "step": 58980 + }, + { + "epoch": 0.29495, + "grad_norm": 0.3649766743183136, + "learning_rate": 0.0004045740422376991, + "loss": 2.7872203826904296, + "step": 58990 + }, + { + "epoch": 0.295, + "grad_norm": 0.32813921570777893, + "learning_rate": 0.0004045428645390164, + "loss": 2.7599817276000977, + "step": 59000 + }, + { + "epoch": 0.29505, + "grad_norm": 0.3468402922153473, + "learning_rate": 0.0004045116829497181, + "loss": 2.795644187927246, + "step": 59010 + }, + { + "epoch": 0.2951, + "grad_norm": 0.3194868564605713, + "learning_rate": 0.0004044804974705891, + "loss": 2.765121269226074, + "step": 59020 + }, + { + "epoch": 0.29515, + "grad_norm": 0.34160521626472473, + "learning_rate": 0.00040444930810241476, + "loss": 2.795079231262207, + "step": 59030 + }, + { + "epoch": 0.2952, + "grad_norm": 0.3260653018951416, + "learning_rate": 0.00040441811484598014, + "loss": 2.806710433959961, + "step": 59040 + }, + { + "epoch": 0.29525, + "grad_norm": 0.35228535532951355, + "learning_rate": 0.0004043869177020704, + "loss": 2.784760284423828, + "step": 59050 + }, + { + "epoch": 0.2953, + "grad_norm": 0.33208587765693665, + "learning_rate": 0.0004043557166714711, + "loss": 2.801470947265625, + "step": 59060 + }, + { + "epoch": 0.29535, + "grad_norm": 0.3658336102962494, + "learning_rate": 0.0004043245117549676, + "loss": 2.804566764831543, + "step": 59070 + }, + { + "epoch": 0.2954, + "grad_norm": 0.30310940742492676, + "learning_rate": 0.0004042933029533456, + "loss": 2.81719856262207, + "step": 59080 + }, + { + "epoch": 0.29545, + "grad_norm": 0.36976006627082825, + "learning_rate": 0.0004042620902673907, + "loss": 2.8596080780029296, + "step": 59090 + }, + { + "epoch": 0.2955, + "grad_norm": 0.41278186440467834, + "learning_rate": 0.0004042308736978887, + "loss": 2.7660991668701174, + "step": 59100 + }, + { + "epoch": 0.29555, + "grad_norm": 0.32398971915245056, + "learning_rate": 0.00040419965324562556, + "loss": 2.770760345458984, + "step": 59110 + }, + { + "epoch": 0.2956, + "grad_norm": 0.336279034614563, + "learning_rate": 0.0004041684289113871, + "loss": 2.853921127319336, + "step": 59120 + }, + { + "epoch": 0.29565, + "grad_norm": 0.31730785965919495, + "learning_rate": 0.00040413720069595947, + "loss": 2.7893373489379885, + "step": 59130 + }, + { + "epoch": 0.2957, + "grad_norm": 0.3262498676776886, + "learning_rate": 0.00040410596860012896, + "loss": 2.7767467498779297, + "step": 59140 + }, + { + "epoch": 0.29575, + "grad_norm": 0.33919277787208557, + "learning_rate": 0.00040407473262468166, + "loss": 2.8174190521240234, + "step": 59150 + }, + { + "epoch": 0.2958, + "grad_norm": 0.35105088353157043, + "learning_rate": 0.00040404349277040394, + "loss": 2.724583625793457, + "step": 59160 + }, + { + "epoch": 0.29585, + "grad_norm": 0.36936795711517334, + "learning_rate": 0.0004040122490380823, + "loss": 2.7852205276489257, + "step": 59170 + }, + { + "epoch": 0.2959, + "grad_norm": 0.34558528661727905, + "learning_rate": 0.0004039810014285034, + "loss": 2.74770565032959, + "step": 59180 + }, + { + "epoch": 0.29595, + "grad_norm": 0.30977877974510193, + "learning_rate": 0.0004039497499424538, + "loss": 2.776424789428711, + "step": 59190 + }, + { + "epoch": 0.296, + "grad_norm": 0.4140177369117737, + "learning_rate": 0.00040391849458072025, + "loss": 2.8318004608154297, + "step": 59200 + }, + { + "epoch": 0.29605, + "grad_norm": 0.3673388659954071, + "learning_rate": 0.00040388723534408967, + "loss": 2.822391891479492, + "step": 59210 + }, + { + "epoch": 0.2961, + "grad_norm": 0.3580965995788574, + "learning_rate": 0.0004038559722333489, + "loss": 2.790964889526367, + "step": 59220 + }, + { + "epoch": 0.29615, + "grad_norm": 0.3587223291397095, + "learning_rate": 0.00040382470524928513, + "loss": 2.7961706161499023, + "step": 59230 + }, + { + "epoch": 0.2962, + "grad_norm": 0.33275002241134644, + "learning_rate": 0.00040379343439268534, + "loss": 2.8270524978637694, + "step": 59240 + }, + { + "epoch": 0.29625, + "grad_norm": 0.390182763338089, + "learning_rate": 0.00040376215966433693, + "loss": 2.8630510330200196, + "step": 59250 + }, + { + "epoch": 0.2963, + "grad_norm": 0.3437720537185669, + "learning_rate": 0.0004037308810650272, + "loss": 2.7919105529785155, + "step": 59260 + }, + { + "epoch": 0.29635, + "grad_norm": 0.3415510058403015, + "learning_rate": 0.00040369959859554356, + "loss": 2.821924591064453, + "step": 59270 + }, + { + "epoch": 0.2964, + "grad_norm": 0.33374133706092834, + "learning_rate": 0.0004036683122566736, + "loss": 2.8065887451171876, + "step": 59280 + }, + { + "epoch": 0.29645, + "grad_norm": 0.3277069628238678, + "learning_rate": 0.00040363702204920473, + "loss": 2.7695053100585936, + "step": 59290 + }, + { + "epoch": 0.2965, + "grad_norm": 0.32833555340766907, + "learning_rate": 0.00040360572797392503, + "loss": 2.729366493225098, + "step": 59300 + }, + { + "epoch": 0.29655, + "grad_norm": 0.32886430621147156, + "learning_rate": 0.0004035744300316221, + "loss": 2.8286476135253906, + "step": 59310 + }, + { + "epoch": 0.2966, + "grad_norm": 0.3382842242717743, + "learning_rate": 0.0004035431282230839, + "loss": 2.7763147354125977, + "step": 59320 + }, + { + "epoch": 0.29665, + "grad_norm": 0.3359784483909607, + "learning_rate": 0.00040351182254909845, + "loss": 2.7432458877563475, + "step": 59330 + }, + { + "epoch": 0.2967, + "grad_norm": 0.38096538186073303, + "learning_rate": 0.00040348051301045395, + "loss": 2.772555923461914, + "step": 59340 + }, + { + "epoch": 0.29675, + "grad_norm": 0.3494737148284912, + "learning_rate": 0.0004034491996079385, + "loss": 2.7251022338867186, + "step": 59350 + }, + { + "epoch": 0.2968, + "grad_norm": 0.32365885376930237, + "learning_rate": 0.00040341788234234057, + "loss": 2.7622995376586914, + "step": 59360 + }, + { + "epoch": 0.29685, + "grad_norm": 0.35767102241516113, + "learning_rate": 0.0004033865612144484, + "loss": 2.780147361755371, + "step": 59370 + }, + { + "epoch": 0.2969, + "grad_norm": 0.3348377048969269, + "learning_rate": 0.0004033552362250505, + "loss": 2.7788764953613283, + "step": 59380 + }, + { + "epoch": 0.29695, + "grad_norm": 0.35264620184898376, + "learning_rate": 0.00040332390737493574, + "loss": 2.791437530517578, + "step": 59390 + }, + { + "epoch": 0.297, + "grad_norm": 0.31562337279319763, + "learning_rate": 0.0004032925746648926, + "loss": 2.793313217163086, + "step": 59400 + }, + { + "epoch": 0.29705, + "grad_norm": 0.304440975189209, + "learning_rate": 0.0004032612380957098, + "loss": 2.754067039489746, + "step": 59410 + }, + { + "epoch": 0.2971, + "grad_norm": 0.33886682987213135, + "learning_rate": 0.0004032298976681764, + "loss": 2.836762237548828, + "step": 59420 + }, + { + "epoch": 0.29715, + "grad_norm": 0.30292728543281555, + "learning_rate": 0.0004031985533830813, + "loss": 2.7212802886962892, + "step": 59430 + }, + { + "epoch": 0.2972, + "grad_norm": 0.35325464606285095, + "learning_rate": 0.00040316720524121374, + "loss": 2.752246284484863, + "step": 59440 + }, + { + "epoch": 0.29725, + "grad_norm": 0.3025570809841156, + "learning_rate": 0.0004031358532433628, + "loss": 2.806364822387695, + "step": 59450 + }, + { + "epoch": 0.2973, + "grad_norm": 0.3393426239490509, + "learning_rate": 0.00040310449739031773, + "loss": 2.8193021774291993, + "step": 59460 + }, + { + "epoch": 0.29735, + "grad_norm": 0.3245916962623596, + "learning_rate": 0.00040307313768286803, + "loss": 2.832629585266113, + "step": 59470 + }, + { + "epoch": 0.2974, + "grad_norm": 0.3298146426677704, + "learning_rate": 0.0004030417741218031, + "loss": 2.831835174560547, + "step": 59480 + }, + { + "epoch": 0.29745, + "grad_norm": 0.32710978388786316, + "learning_rate": 0.00040301040670791244, + "loss": 2.813401985168457, + "step": 59490 + }, + { + "epoch": 0.2975, + "grad_norm": 0.34612858295440674, + "learning_rate": 0.0004029790354419859, + "loss": 2.7523622512817383, + "step": 59500 + }, + { + "epoch": 0.29755, + "grad_norm": 0.3329753875732422, + "learning_rate": 0.00040294766032481313, + "loss": 2.81817512512207, + "step": 59510 + }, + { + "epoch": 0.2976, + "grad_norm": 0.3514355719089508, + "learning_rate": 0.00040291628135718404, + "loss": 2.766778755187988, + "step": 59520 + }, + { + "epoch": 0.29765, + "grad_norm": 0.3286838233470917, + "learning_rate": 0.0004028848985398886, + "loss": 2.826974868774414, + "step": 59530 + }, + { + "epoch": 0.2977, + "grad_norm": 0.3472480773925781, + "learning_rate": 0.00040285351187371686, + "loss": 2.834755516052246, + "step": 59540 + }, + { + "epoch": 0.29775, + "grad_norm": 0.3372628390789032, + "learning_rate": 0.00040282212135945894, + "loss": 2.7488037109375, + "step": 59550 + }, + { + "epoch": 0.2978, + "grad_norm": 0.33014875650405884, + "learning_rate": 0.0004027907269979052, + "loss": 2.7946283340454103, + "step": 59560 + }, + { + "epoch": 0.29785, + "grad_norm": 0.33913588523864746, + "learning_rate": 0.00040275932878984586, + "loss": 2.7617387771606445, + "step": 59570 + }, + { + "epoch": 0.2979, + "grad_norm": 0.35029715299606323, + "learning_rate": 0.0004027279267360715, + "loss": 2.778788185119629, + "step": 59580 + }, + { + "epoch": 0.29795, + "grad_norm": 0.33568257093429565, + "learning_rate": 0.0004026965208373725, + "loss": 2.8233036041259765, + "step": 59590 + }, + { + "epoch": 0.298, + "grad_norm": 0.3213505744934082, + "learning_rate": 0.00040266511109453973, + "loss": 2.7566097259521483, + "step": 59600 + }, + { + "epoch": 0.29805, + "grad_norm": 0.30878564715385437, + "learning_rate": 0.00040263369750836374, + "loss": 2.783543014526367, + "step": 59610 + }, + { + "epoch": 0.2981, + "grad_norm": 0.3445400893688202, + "learning_rate": 0.0004026022800796355, + "loss": 2.8295658111572264, + "step": 59620 + }, + { + "epoch": 0.29815, + "grad_norm": 0.32394492626190186, + "learning_rate": 0.0004025708588091458, + "loss": 2.7422853469848634, + "step": 59630 + }, + { + "epoch": 0.2982, + "grad_norm": 0.32339903712272644, + "learning_rate": 0.0004025394336976859, + "loss": 2.7670358657836913, + "step": 59640 + }, + { + "epoch": 0.29825, + "grad_norm": 0.31750261783599854, + "learning_rate": 0.0004025080047460466, + "loss": 2.8190635681152343, + "step": 59650 + }, + { + "epoch": 0.2983, + "grad_norm": 0.33343467116355896, + "learning_rate": 0.0004024765719550194, + "loss": 2.7550027847290037, + "step": 59660 + }, + { + "epoch": 0.29835, + "grad_norm": 0.3501276969909668, + "learning_rate": 0.0004024451353253954, + "loss": 2.7632484436035156, + "step": 59670 + }, + { + "epoch": 0.2984, + "grad_norm": 0.3349858820438385, + "learning_rate": 0.0004024136948579663, + "loss": 2.8228271484375, + "step": 59680 + }, + { + "epoch": 0.29845, + "grad_norm": 0.32717522978782654, + "learning_rate": 0.0004023822505535233, + "loss": 2.7331138610839845, + "step": 59690 + }, + { + "epoch": 0.2985, + "grad_norm": 0.3690757155418396, + "learning_rate": 0.00040235080241285827, + "loss": 2.8124320983886717, + "step": 59700 + }, + { + "epoch": 0.29855, + "grad_norm": 0.3501994013786316, + "learning_rate": 0.0004023193504367628, + "loss": 2.8316940307617187, + "step": 59710 + }, + { + "epoch": 0.2986, + "grad_norm": 0.3295276463031769, + "learning_rate": 0.0004022878946260287, + "loss": 2.802785110473633, + "step": 59720 + }, + { + "epoch": 0.29865, + "grad_norm": 0.32683107256889343, + "learning_rate": 0.00040225643498144776, + "loss": 2.7387577056884767, + "step": 59730 + }, + { + "epoch": 0.2987, + "grad_norm": 0.33910897374153137, + "learning_rate": 0.0004022249715038122, + "loss": 2.780572700500488, + "step": 59740 + }, + { + "epoch": 0.29875, + "grad_norm": 0.32706695795059204, + "learning_rate": 0.00040219350419391396, + "loss": 2.7434669494628907, + "step": 59750 + }, + { + "epoch": 0.2988, + "grad_norm": 0.33040735125541687, + "learning_rate": 0.0004021620330525453, + "loss": 2.802803802490234, + "step": 59760 + }, + { + "epoch": 0.29885, + "grad_norm": 0.4187465310096741, + "learning_rate": 0.00040213055808049837, + "loss": 2.76477165222168, + "step": 59770 + }, + { + "epoch": 0.2989, + "grad_norm": 0.3267557919025421, + "learning_rate": 0.0004020990792785658, + "loss": 2.797444152832031, + "step": 59780 + }, + { + "epoch": 0.29895, + "grad_norm": 0.3503839373588562, + "learning_rate": 0.00040206759664753984, + "loss": 2.7680788040161133, + "step": 59790 + }, + { + "epoch": 0.299, + "grad_norm": 0.31587645411491394, + "learning_rate": 0.0004020361101882132, + "loss": 2.7709829330444338, + "step": 59800 + }, + { + "epoch": 0.29905, + "grad_norm": 0.31495654582977295, + "learning_rate": 0.0004020046199013784, + "loss": 2.7813446044921877, + "step": 59810 + }, + { + "epoch": 0.2991, + "grad_norm": 0.31562232971191406, + "learning_rate": 0.0004019731257878284, + "loss": 2.7589645385742188, + "step": 59820 + }, + { + "epoch": 0.29915, + "grad_norm": 0.3952493965625763, + "learning_rate": 0.000401941627848356, + "loss": 2.808604621887207, + "step": 59830 + }, + { + "epoch": 0.2992, + "grad_norm": 0.333525687456131, + "learning_rate": 0.0004019101260837541, + "loss": 2.82464485168457, + "step": 59840 + }, + { + "epoch": 0.29925, + "grad_norm": 0.30217891931533813, + "learning_rate": 0.00040187862049481573, + "loss": 2.796425437927246, + "step": 59850 + }, + { + "epoch": 0.2993, + "grad_norm": 0.34327611327171326, + "learning_rate": 0.0004018471110823342, + "loss": 2.781453323364258, + "step": 59860 + }, + { + "epoch": 0.29935, + "grad_norm": 0.3331916630268097, + "learning_rate": 0.0004018155978471026, + "loss": 2.777194023132324, + "step": 59870 + }, + { + "epoch": 0.2994, + "grad_norm": 0.32502150535583496, + "learning_rate": 0.00040178408078991434, + "loss": 2.782948684692383, + "step": 59880 + }, + { + "epoch": 0.29945, + "grad_norm": 0.3495626747608185, + "learning_rate": 0.00040175255991156287, + "loss": 2.7614418029785157, + "step": 59890 + }, + { + "epoch": 0.2995, + "grad_norm": 0.3659369647502899, + "learning_rate": 0.00040172103521284177, + "loss": 2.7521358489990235, + "step": 59900 + }, + { + "epoch": 0.29955, + "grad_norm": 0.3407943546772003, + "learning_rate": 0.00040168950669454455, + "loss": 2.7770198822021483, + "step": 59910 + }, + { + "epoch": 0.2996, + "grad_norm": 0.3423929512500763, + "learning_rate": 0.0004016579743574651, + "loss": 2.7794198989868164, + "step": 59920 + }, + { + "epoch": 0.29965, + "grad_norm": 0.3320602476596832, + "learning_rate": 0.00040162643820239704, + "loss": 2.8186296463012694, + "step": 59930 + }, + { + "epoch": 0.2997, + "grad_norm": 0.3450561463832855, + "learning_rate": 0.0004015948982301345, + "loss": 2.783402442932129, + "step": 59940 + }, + { + "epoch": 0.29975, + "grad_norm": 0.3671383261680603, + "learning_rate": 0.00040156335444147136, + "loss": 2.834302520751953, + "step": 59950 + }, + { + "epoch": 0.2998, + "grad_norm": 0.3640058934688568, + "learning_rate": 0.0004015318068372018, + "loss": 2.72823486328125, + "step": 59960 + }, + { + "epoch": 0.29985, + "grad_norm": 0.3513984680175781, + "learning_rate": 0.00040150025541812003, + "loss": 2.77802734375, + "step": 59970 + }, + { + "epoch": 0.2999, + "grad_norm": 0.3070821166038513, + "learning_rate": 0.00040146870018502035, + "loss": 2.7834541320800783, + "step": 59980 + }, + { + "epoch": 0.29995, + "grad_norm": 0.3467007279396057, + "learning_rate": 0.0004014371411386971, + "loss": 2.830189514160156, + "step": 59990 + }, + { + "epoch": 0.3, + "grad_norm": 0.3393097221851349, + "learning_rate": 0.00040140557827994495, + "loss": 2.7580570220947265, + "step": 60000 + }, + { + "epoch": 0.30005, + "grad_norm": 0.3609584867954254, + "learning_rate": 0.00040137401160955823, + "loss": 2.763860893249512, + "step": 60010 + }, + { + "epoch": 0.3001, + "grad_norm": 0.33606284856796265, + "learning_rate": 0.0004013424411283319, + "loss": 2.802288055419922, + "step": 60020 + }, + { + "epoch": 0.30015, + "grad_norm": 0.32552629709243774, + "learning_rate": 0.0004013108668370605, + "loss": 2.809020233154297, + "step": 60030 + }, + { + "epoch": 0.3002, + "grad_norm": 0.38583847880363464, + "learning_rate": 0.00040127928873653917, + "loss": 2.827532958984375, + "step": 60040 + }, + { + "epoch": 0.30025, + "grad_norm": 0.43065422773361206, + "learning_rate": 0.0004012477068275627, + "loss": 2.7887739181518554, + "step": 60050 + }, + { + "epoch": 0.3003, + "grad_norm": 0.3485683500766754, + "learning_rate": 0.0004012161211109262, + "loss": 2.7570068359375, + "step": 60060 + }, + { + "epoch": 0.30035, + "grad_norm": 0.3412553668022156, + "learning_rate": 0.0004011845315874249, + "loss": 2.7590215682983397, + "step": 60070 + }, + { + "epoch": 0.3004, + "grad_norm": 0.32567116618156433, + "learning_rate": 0.00040115293825785406, + "loss": 2.811531639099121, + "step": 60080 + }, + { + "epoch": 0.30045, + "grad_norm": 0.33207210898399353, + "learning_rate": 0.0004011213411230089, + "loss": 2.7448217391967775, + "step": 60090 + }, + { + "epoch": 0.3005, + "grad_norm": 0.3379744589328766, + "learning_rate": 0.00040108974018368505, + "loss": 2.836958122253418, + "step": 60100 + }, + { + "epoch": 0.30055, + "grad_norm": 0.3940136134624481, + "learning_rate": 0.000401058135440678, + "loss": 2.7456756591796876, + "step": 60110 + }, + { + "epoch": 0.3006, + "grad_norm": 0.3341921269893646, + "learning_rate": 0.0004010265268947834, + "loss": 2.7733959197998046, + "step": 60120 + }, + { + "epoch": 0.30065, + "grad_norm": 0.30995258688926697, + "learning_rate": 0.000400994914546797, + "loss": 2.7863533020019533, + "step": 60130 + }, + { + "epoch": 0.3007, + "grad_norm": 0.35586103796958923, + "learning_rate": 0.00040096329839751465, + "loss": 2.7856376647949217, + "step": 60140 + }, + { + "epoch": 0.30075, + "grad_norm": 0.34897443652153015, + "learning_rate": 0.0004009316784477323, + "loss": 2.780685043334961, + "step": 60150 + }, + { + "epoch": 0.3008, + "grad_norm": 0.36518916487693787, + "learning_rate": 0.00040090005469824587, + "loss": 2.771571731567383, + "step": 60160 + }, + { + "epoch": 0.30085, + "grad_norm": 0.35622116923332214, + "learning_rate": 0.0004008684271498516, + "loss": 2.7979307174682617, + "step": 60170 + }, + { + "epoch": 0.3009, + "grad_norm": 0.3651481866836548, + "learning_rate": 0.00040083679580334565, + "loss": 2.7593481063842775, + "step": 60180 + }, + { + "epoch": 0.30095, + "grad_norm": 0.33394691348075867, + "learning_rate": 0.00040080516065952443, + "loss": 2.79056396484375, + "step": 60190 + }, + { + "epoch": 0.301, + "grad_norm": 0.3336094617843628, + "learning_rate": 0.00040077352171918423, + "loss": 2.778608703613281, + "step": 60200 + }, + { + "epoch": 0.30105, + "grad_norm": 0.3531544506549835, + "learning_rate": 0.0004007418789831218, + "loss": 2.750142288208008, + "step": 60210 + }, + { + "epoch": 0.3011, + "grad_norm": 0.35609957575798035, + "learning_rate": 0.0004007102324521334, + "loss": 2.7708452224731444, + "step": 60220 + }, + { + "epoch": 0.30115, + "grad_norm": 0.33361902832984924, + "learning_rate": 0.000400678582127016, + "loss": 2.7514448165893555, + "step": 60230 + }, + { + "epoch": 0.3012, + "grad_norm": 0.37102600932121277, + "learning_rate": 0.00040064692800856615, + "loss": 2.8207351684570314, + "step": 60240 + }, + { + "epoch": 0.30125, + "grad_norm": 0.34005746245384216, + "learning_rate": 0.000400615270097581, + "loss": 2.7889162063598634, + "step": 60250 + }, + { + "epoch": 0.3013, + "grad_norm": 0.3322615325450897, + "learning_rate": 0.00040058360839485753, + "loss": 2.749094009399414, + "step": 60260 + }, + { + "epoch": 0.30135, + "grad_norm": 0.33889997005462646, + "learning_rate": 0.00040055194290119264, + "loss": 2.799392890930176, + "step": 60270 + }, + { + "epoch": 0.3014, + "grad_norm": 0.3589109480381012, + "learning_rate": 0.00040052027361738355, + "loss": 2.7843008041381836, + "step": 60280 + }, + { + "epoch": 0.30145, + "grad_norm": 0.34263262152671814, + "learning_rate": 0.00040048860054422763, + "loss": 2.7984609603881836, + "step": 60290 + }, + { + "epoch": 0.3015, + "grad_norm": 0.32743197679519653, + "learning_rate": 0.00040045692368252217, + "loss": 2.829288673400879, + "step": 60300 + }, + { + "epoch": 0.30155, + "grad_norm": 0.3376390337944031, + "learning_rate": 0.00040042524303306464, + "loss": 2.794875907897949, + "step": 60310 + }, + { + "epoch": 0.3016, + "grad_norm": 0.366985559463501, + "learning_rate": 0.0004003935585966527, + "loss": 2.7776159286499023, + "step": 60320 + }, + { + "epoch": 0.30165, + "grad_norm": 0.42725569009780884, + "learning_rate": 0.0004003618703740839, + "loss": 2.7966283798217773, + "step": 60330 + }, + { + "epoch": 0.3017, + "grad_norm": 0.3487706184387207, + "learning_rate": 0.00040033017836615595, + "loss": 2.7439292907714843, + "step": 60340 + }, + { + "epoch": 0.30175, + "grad_norm": 0.33414894342422485, + "learning_rate": 0.0004002984825736668, + "loss": 2.7752065658569336, + "step": 60350 + }, + { + "epoch": 0.3018, + "grad_norm": 0.3543921411037445, + "learning_rate": 0.0004002667829974144, + "loss": 2.767959976196289, + "step": 60360 + }, + { + "epoch": 0.30185, + "grad_norm": 0.33044296503067017, + "learning_rate": 0.00040023507963819673, + "loss": 2.7877445220947266, + "step": 60370 + }, + { + "epoch": 0.3019, + "grad_norm": 0.365174263715744, + "learning_rate": 0.00040020337249681184, + "loss": 2.830311584472656, + "step": 60380 + }, + { + "epoch": 0.30195, + "grad_norm": 0.3247356712818146, + "learning_rate": 0.0004001716615740582, + "loss": 2.803053092956543, + "step": 60390 + }, + { + "epoch": 0.302, + "grad_norm": 0.32706114649772644, + "learning_rate": 0.0004001399468707339, + "loss": 2.7942058563232424, + "step": 60400 + }, + { + "epoch": 0.30205, + "grad_norm": 0.3300260603427887, + "learning_rate": 0.0004001082283876374, + "loss": 2.8201835632324217, + "step": 60410 + }, + { + "epoch": 0.3021, + "grad_norm": 0.32609057426452637, + "learning_rate": 0.00040007650612556726, + "loss": 2.8101734161376952, + "step": 60420 + }, + { + "epoch": 0.30215, + "grad_norm": 0.3502173125743866, + "learning_rate": 0.0004000447800853222, + "loss": 2.776152420043945, + "step": 60430 + }, + { + "epoch": 0.3022, + "grad_norm": 0.3546614348888397, + "learning_rate": 0.0004000130502677007, + "loss": 2.8084091186523437, + "step": 60440 + }, + { + "epoch": 0.30225, + "grad_norm": 0.359266072511673, + "learning_rate": 0.0003999813166735017, + "loss": 2.8342119216918946, + "step": 60450 + }, + { + "epoch": 0.3023, + "grad_norm": 0.3541928231716156, + "learning_rate": 0.00039994957930352407, + "loss": 2.7720052719116213, + "step": 60460 + }, + { + "epoch": 0.30235, + "grad_norm": 0.35576289892196655, + "learning_rate": 0.00039991783815856676, + "loss": 2.7875476837158204, + "step": 60470 + }, + { + "epoch": 0.3024, + "grad_norm": 0.33435967564582825, + "learning_rate": 0.00039988609323942883, + "loss": 2.785282325744629, + "step": 60480 + }, + { + "epoch": 0.30245, + "grad_norm": 0.33768758177757263, + "learning_rate": 0.00039985434454690955, + "loss": 2.766219711303711, + "step": 60490 + }, + { + "epoch": 0.3025, + "grad_norm": 0.3912406861782074, + "learning_rate": 0.00039982259208180815, + "loss": 2.753566360473633, + "step": 60500 + }, + { + "epoch": 0.30255, + "grad_norm": 0.33968254923820496, + "learning_rate": 0.000399790835844924, + "loss": 2.805817222595215, + "step": 60510 + }, + { + "epoch": 0.3026, + "grad_norm": 0.3445298969745636, + "learning_rate": 0.0003997590758370565, + "loss": 2.8038398742675783, + "step": 60520 + }, + { + "epoch": 0.30265, + "grad_norm": 0.3442714512348175, + "learning_rate": 0.0003997273120590054, + "loss": 2.8173295974731447, + "step": 60530 + }, + { + "epoch": 0.3027, + "grad_norm": 0.3285183310508728, + "learning_rate": 0.0003996955445115701, + "loss": 2.7849185943603514, + "step": 60540 + }, + { + "epoch": 0.30275, + "grad_norm": 0.33008313179016113, + "learning_rate": 0.00039966377319555047, + "loss": 2.8042861938476564, + "step": 60550 + }, + { + "epoch": 0.3028, + "grad_norm": 0.32314565777778625, + "learning_rate": 0.00039963199811174637, + "loss": 2.821084403991699, + "step": 60560 + }, + { + "epoch": 0.30285, + "grad_norm": 0.33661597967147827, + "learning_rate": 0.00039960021926095776, + "loss": 2.769596481323242, + "step": 60570 + }, + { + "epoch": 0.3029, + "grad_norm": 0.3671300411224365, + "learning_rate": 0.0003995684366439846, + "loss": 2.7742910385131836, + "step": 60580 + }, + { + "epoch": 0.30295, + "grad_norm": 0.3561730980873108, + "learning_rate": 0.000399536650261627, + "loss": 2.8067720413208006, + "step": 60590 + }, + { + "epoch": 0.303, + "grad_norm": 0.3202562630176544, + "learning_rate": 0.0003995048601146853, + "loss": 2.7933414459228514, + "step": 60600 + }, + { + "epoch": 0.30305, + "grad_norm": 0.3062936067581177, + "learning_rate": 0.00039947306620395976, + "loss": 2.7839841842651367, + "step": 60610 + }, + { + "epoch": 0.3031, + "grad_norm": 0.3302069306373596, + "learning_rate": 0.00039944126853025066, + "loss": 2.7528398513793944, + "step": 60620 + }, + { + "epoch": 0.30315, + "grad_norm": 0.37712743878364563, + "learning_rate": 0.0003994094670943588, + "loss": 2.7283740997314454, + "step": 60630 + }, + { + "epoch": 0.3032, + "grad_norm": 0.30992215871810913, + "learning_rate": 0.0003993776618970844, + "loss": 2.7466968536376952, + "step": 60640 + }, + { + "epoch": 0.30325, + "grad_norm": 0.30595046281814575, + "learning_rate": 0.00039934585293922853, + "loss": 2.844677734375, + "step": 60650 + }, + { + "epoch": 0.3033, + "grad_norm": 0.3576817512512207, + "learning_rate": 0.0003993140402215917, + "loss": 2.768253135681152, + "step": 60660 + }, + { + "epoch": 0.30335, + "grad_norm": 0.3648078739643097, + "learning_rate": 0.00039928222374497496, + "loss": 2.792816734313965, + "step": 60670 + }, + { + "epoch": 0.3034, + "grad_norm": 0.33420273661613464, + "learning_rate": 0.00039925040351017924, + "loss": 2.7658790588378905, + "step": 60680 + }, + { + "epoch": 0.30345, + "grad_norm": 0.33073824644088745, + "learning_rate": 0.00039921857951800554, + "loss": 2.794977569580078, + "step": 60690 + }, + { + "epoch": 0.3035, + "grad_norm": 0.3639773726463318, + "learning_rate": 0.0003991867517692552, + "loss": 2.80856876373291, + "step": 60700 + }, + { + "epoch": 0.30355, + "grad_norm": 0.33256423473358154, + "learning_rate": 0.0003991549202647293, + "loss": 2.8179290771484373, + "step": 60710 + }, + { + "epoch": 0.3036, + "grad_norm": 0.3479187786579132, + "learning_rate": 0.0003991230850052293, + "loss": 2.7662378311157227, + "step": 60720 + }, + { + "epoch": 0.30365, + "grad_norm": 0.332622230052948, + "learning_rate": 0.0003990912459915567, + "loss": 2.835460662841797, + "step": 60730 + }, + { + "epoch": 0.3037, + "grad_norm": 0.31765061616897583, + "learning_rate": 0.0003990594032245129, + "loss": 2.730476951599121, + "step": 60740 + }, + { + "epoch": 0.30375, + "grad_norm": 0.33226799964904785, + "learning_rate": 0.00039902755670489967, + "loss": 2.7693538665771484, + "step": 60750 + }, + { + "epoch": 0.3038, + "grad_norm": 0.3149641454219818, + "learning_rate": 0.00039899570643351864, + "loss": 2.794159698486328, + "step": 60760 + }, + { + "epoch": 0.30385, + "grad_norm": 0.35178041458129883, + "learning_rate": 0.0003989638524111717, + "loss": 2.8070240020751953, + "step": 60770 + }, + { + "epoch": 0.3039, + "grad_norm": 0.31131479144096375, + "learning_rate": 0.0003989319946386608, + "loss": 2.75559139251709, + "step": 60780 + }, + { + "epoch": 0.30395, + "grad_norm": 0.3475709557533264, + "learning_rate": 0.000398900133116788, + "loss": 2.7828540802001953, + "step": 60790 + }, + { + "epoch": 0.304, + "grad_norm": 0.3287048935890198, + "learning_rate": 0.0003988682678463552, + "loss": 2.7735546112060545, + "step": 60800 + }, + { + "epoch": 0.30405, + "grad_norm": 0.420900821685791, + "learning_rate": 0.0003988363988281648, + "loss": 2.861581802368164, + "step": 60810 + }, + { + "epoch": 0.3041, + "grad_norm": 0.34318506717681885, + "learning_rate": 0.0003988045260630191, + "loss": 2.8154972076416014, + "step": 60820 + }, + { + "epoch": 0.30415, + "grad_norm": 0.3233541250228882, + "learning_rate": 0.00039877264955172044, + "loss": 2.7964900970458983, + "step": 60830 + }, + { + "epoch": 0.3042, + "grad_norm": 0.36654114723205566, + "learning_rate": 0.00039874076929507124, + "loss": 2.7596033096313475, + "step": 60840 + }, + { + "epoch": 0.30425, + "grad_norm": 0.39972636103630066, + "learning_rate": 0.0003987088852938743, + "loss": 2.7825136184692383, + "step": 60850 + }, + { + "epoch": 0.3043, + "grad_norm": 0.32912448048591614, + "learning_rate": 0.00039867699754893206, + "loss": 2.806827926635742, + "step": 60860 + }, + { + "epoch": 0.30435, + "grad_norm": 0.3026493489742279, + "learning_rate": 0.0003986451060610474, + "loss": 2.8015573501586912, + "step": 60870 + }, + { + "epoch": 0.3044, + "grad_norm": 0.35296791791915894, + "learning_rate": 0.00039861321083102324, + "loss": 2.7864776611328126, + "step": 60880 + }, + { + "epoch": 0.30445, + "grad_norm": 0.32882773876190186, + "learning_rate": 0.0003985813118596625, + "loss": 2.767013168334961, + "step": 60890 + }, + { + "epoch": 0.3045, + "grad_norm": 0.3375060260295868, + "learning_rate": 0.0003985494091477682, + "loss": 2.841303253173828, + "step": 60900 + }, + { + "epoch": 0.30455, + "grad_norm": 0.32056593894958496, + "learning_rate": 0.00039851750269614353, + "loss": 2.8062065124511717, + "step": 60910 + }, + { + "epoch": 0.3046, + "grad_norm": 0.3145381510257721, + "learning_rate": 0.0003984855925055917, + "loss": 2.766830635070801, + "step": 60920 + }, + { + "epoch": 0.30465, + "grad_norm": 0.3281804323196411, + "learning_rate": 0.0003984536785769161, + "loss": 2.7781248092651367, + "step": 60930 + }, + { + "epoch": 0.3047, + "grad_norm": 0.37007462978363037, + "learning_rate": 0.0003984217609109201, + "loss": 2.7767560958862303, + "step": 60940 + }, + { + "epoch": 0.30475, + "grad_norm": 0.3290128707885742, + "learning_rate": 0.0003983898395084073, + "loss": 2.831764793395996, + "step": 60950 + }, + { + "epoch": 0.3048, + "grad_norm": 0.340742290019989, + "learning_rate": 0.00039835791437018125, + "loss": 2.8053817749023438, + "step": 60960 + }, + { + "epoch": 0.30485, + "grad_norm": 0.3525587022304535, + "learning_rate": 0.0003983259854970457, + "loss": 2.75667724609375, + "step": 60970 + }, + { + "epoch": 0.3049, + "grad_norm": 0.340401291847229, + "learning_rate": 0.00039829405288980445, + "loss": 2.8006704330444334, + "step": 60980 + }, + { + "epoch": 0.30495, + "grad_norm": 0.3220403492450714, + "learning_rate": 0.0003982621165492615, + "loss": 2.7831932067871095, + "step": 60990 + }, + { + "epoch": 0.305, + "grad_norm": 0.3340715765953064, + "learning_rate": 0.00039823017647622064, + "loss": 2.745549201965332, + "step": 61000 + }, + { + "epoch": 0.30505, + "grad_norm": 0.33622363209724426, + "learning_rate": 0.00039819823267148613, + "loss": 2.8118659973144533, + "step": 61010 + }, + { + "epoch": 0.3051, + "grad_norm": 0.3172326982021332, + "learning_rate": 0.00039816628513586206, + "loss": 2.8044361114501952, + "step": 61020 + }, + { + "epoch": 0.30515, + "grad_norm": 0.35277730226516724, + "learning_rate": 0.0003981343338701528, + "loss": 2.771168899536133, + "step": 61030 + }, + { + "epoch": 0.3052, + "grad_norm": 0.3330965042114258, + "learning_rate": 0.00039810237887516264, + "loss": 2.7797664642333983, + "step": 61040 + }, + { + "epoch": 0.30525, + "grad_norm": 0.36064276099205017, + "learning_rate": 0.00039807042015169614, + "loss": 2.7465633392333983, + "step": 61050 + }, + { + "epoch": 0.3053, + "grad_norm": 0.35066378116607666, + "learning_rate": 0.00039803845770055777, + "loss": 2.7878896713256838, + "step": 61060 + }, + { + "epoch": 0.30535, + "grad_norm": 0.34598973393440247, + "learning_rate": 0.0003980064915225522, + "loss": 2.769889068603516, + "step": 61070 + }, + { + "epoch": 0.3054, + "grad_norm": 0.34216925501823425, + "learning_rate": 0.0003979745216184842, + "loss": 2.7996212005615235, + "step": 61080 + }, + { + "epoch": 0.30545, + "grad_norm": 0.37236303091049194, + "learning_rate": 0.0003979425479891586, + "loss": 2.781284713745117, + "step": 61090 + }, + { + "epoch": 0.3055, + "grad_norm": 0.3575686514377594, + "learning_rate": 0.0003979105706353804, + "loss": 2.749688911437988, + "step": 61100 + }, + { + "epoch": 0.30555, + "grad_norm": 0.34142830967903137, + "learning_rate": 0.00039787858955795443, + "loss": 2.790568542480469, + "step": 61110 + }, + { + "epoch": 0.3056, + "grad_norm": 0.39224377274513245, + "learning_rate": 0.0003978466047576861, + "loss": 2.8085397720336913, + "step": 61120 + }, + { + "epoch": 0.30565, + "grad_norm": 0.32795450091362, + "learning_rate": 0.0003978146162353804, + "loss": 2.779157257080078, + "step": 61130 + }, + { + "epoch": 0.3057, + "grad_norm": 0.36241447925567627, + "learning_rate": 0.0003977826239918428, + "loss": 2.7896198272705077, + "step": 61140 + }, + { + "epoch": 0.30575, + "grad_norm": 0.2960101068019867, + "learning_rate": 0.00039775062802787855, + "loss": 2.781027984619141, + "step": 61150 + }, + { + "epoch": 0.3058, + "grad_norm": 0.32648202776908875, + "learning_rate": 0.0003977186283442933, + "loss": 2.7577003479003905, + "step": 61160 + }, + { + "epoch": 0.30585, + "grad_norm": 0.3409590423107147, + "learning_rate": 0.0003976866249418925, + "loss": 2.7701187133789062, + "step": 61170 + }, + { + "epoch": 0.3059, + "grad_norm": 0.3390822410583496, + "learning_rate": 0.00039765461782148196, + "loss": 2.7651575088500975, + "step": 61180 + }, + { + "epoch": 0.30595, + "grad_norm": 0.34199559688568115, + "learning_rate": 0.00039762260698386734, + "loss": 2.761462593078613, + "step": 61190 + }, + { + "epoch": 0.306, + "grad_norm": 0.3587498962879181, + "learning_rate": 0.00039759059242985465, + "loss": 2.811970329284668, + "step": 61200 + }, + { + "epoch": 0.30605, + "grad_norm": 0.36455121636390686, + "learning_rate": 0.00039755857416024973, + "loss": 2.841235542297363, + "step": 61210 + }, + { + "epoch": 0.3061, + "grad_norm": 0.3940179944038391, + "learning_rate": 0.0003975265521758587, + "loss": 2.823501205444336, + "step": 61220 + }, + { + "epoch": 0.30615, + "grad_norm": 0.3245076537132263, + "learning_rate": 0.00039749452647748766, + "loss": 2.768631362915039, + "step": 61230 + }, + { + "epoch": 0.3062, + "grad_norm": 0.3533698320388794, + "learning_rate": 0.000397462497065943, + "loss": 2.771267890930176, + "step": 61240 + }, + { + "epoch": 0.30625, + "grad_norm": 0.3182413578033447, + "learning_rate": 0.0003974304639420309, + "loss": 2.8032905578613283, + "step": 61250 + }, + { + "epoch": 0.3063, + "grad_norm": 0.3232475221157074, + "learning_rate": 0.0003973984271065579, + "loss": 2.7522356033325197, + "step": 61260 + }, + { + "epoch": 0.30635, + "grad_norm": 0.4187970757484436, + "learning_rate": 0.0003973663865603304, + "loss": 2.7337108612060548, + "step": 61270 + }, + { + "epoch": 0.3064, + "grad_norm": 0.3656420111656189, + "learning_rate": 0.00039733434230415524, + "loss": 2.7694280624389647, + "step": 61280 + }, + { + "epoch": 0.30645, + "grad_norm": 0.32798388600349426, + "learning_rate": 0.0003973022943388389, + "loss": 2.7728309631347656, + "step": 61290 + }, + { + "epoch": 0.3065, + "grad_norm": 0.3549242615699768, + "learning_rate": 0.0003972702426651883, + "loss": 2.744997978210449, + "step": 61300 + }, + { + "epoch": 0.30655, + "grad_norm": 0.3400737941265106, + "learning_rate": 0.0003972381872840103, + "loss": 2.806566619873047, + "step": 61310 + }, + { + "epoch": 0.3066, + "grad_norm": 0.32424479722976685, + "learning_rate": 0.0003972061281961119, + "loss": 2.791777801513672, + "step": 61320 + }, + { + "epoch": 0.30665, + "grad_norm": 0.35054025053977966, + "learning_rate": 0.0003971740654023003, + "loss": 2.7753808975219725, + "step": 61330 + }, + { + "epoch": 0.3067, + "grad_norm": 0.3075977563858032, + "learning_rate": 0.00039714199890338253, + "loss": 2.7786418914794924, + "step": 61340 + }, + { + "epoch": 0.30675, + "grad_norm": 0.3264601528644562, + "learning_rate": 0.00039710992870016596, + "loss": 2.7844268798828127, + "step": 61350 + }, + { + "epoch": 0.3068, + "grad_norm": 0.3701021671295166, + "learning_rate": 0.0003970778547934578, + "loss": 2.779793930053711, + "step": 61360 + }, + { + "epoch": 0.30685, + "grad_norm": 0.36855369806289673, + "learning_rate": 0.00039704577718406575, + "loss": 2.820775604248047, + "step": 61370 + }, + { + "epoch": 0.3069, + "grad_norm": 0.3445533215999603, + "learning_rate": 0.00039701369587279725, + "loss": 2.732349967956543, + "step": 61380 + }, + { + "epoch": 0.30695, + "grad_norm": 0.3159644603729248, + "learning_rate": 0.0003969816108604598, + "loss": 2.7675952911376953, + "step": 61390 + }, + { + "epoch": 0.307, + "grad_norm": 0.3290649950504303, + "learning_rate": 0.0003969495221478614, + "loss": 2.7759323120117188, + "step": 61400 + }, + { + "epoch": 0.30705, + "grad_norm": 0.32386279106140137, + "learning_rate": 0.0003969174297358097, + "loss": 2.785898780822754, + "step": 61410 + }, + { + "epoch": 0.3071, + "grad_norm": 0.3267507255077362, + "learning_rate": 0.00039688533362511276, + "loss": 2.7765832901000977, + "step": 61420 + }, + { + "epoch": 0.30715, + "grad_norm": 0.3638741672039032, + "learning_rate": 0.0003968532338165785, + "loss": 2.7714895248413085, + "step": 61430 + }, + { + "epoch": 0.3072, + "grad_norm": 0.3585849404335022, + "learning_rate": 0.00039682113031101507, + "loss": 2.7469303131103517, + "step": 61440 + }, + { + "epoch": 0.30725, + "grad_norm": 0.33251872658729553, + "learning_rate": 0.0003967890231092306, + "loss": 2.7688501358032225, + "step": 61450 + }, + { + "epoch": 0.3073, + "grad_norm": 0.34752970933914185, + "learning_rate": 0.00039675691221203346, + "loss": 2.7878026962280273, + "step": 61460 + }, + { + "epoch": 0.30735, + "grad_norm": 0.32609444856643677, + "learning_rate": 0.0003967247976202321, + "loss": 2.791761589050293, + "step": 61470 + }, + { + "epoch": 0.3074, + "grad_norm": 0.41280654072761536, + "learning_rate": 0.0003966926793346349, + "loss": 2.78649959564209, + "step": 61480 + }, + { + "epoch": 0.30745, + "grad_norm": 0.3614339232444763, + "learning_rate": 0.0003966605573560505, + "loss": 2.749970626831055, + "step": 61490 + }, + { + "epoch": 0.3075, + "grad_norm": 0.3458935618400574, + "learning_rate": 0.00039662843168528756, + "loss": 2.807902717590332, + "step": 61500 + }, + { + "epoch": 0.30755, + "grad_norm": 0.3777947723865509, + "learning_rate": 0.0003965963023231548, + "loss": 2.8365846633911134, + "step": 61510 + }, + { + "epoch": 0.3076, + "grad_norm": 0.37080851197242737, + "learning_rate": 0.0003965641692704611, + "loss": 2.765579414367676, + "step": 61520 + }, + { + "epoch": 0.30765, + "grad_norm": 0.34659022092819214, + "learning_rate": 0.00039653203252801544, + "loss": 2.7610902786254883, + "step": 61530 + }, + { + "epoch": 0.3077, + "grad_norm": 0.3527055084705353, + "learning_rate": 0.00039649989209662685, + "loss": 2.750279426574707, + "step": 61540 + }, + { + "epoch": 0.30775, + "grad_norm": 0.3844434916973114, + "learning_rate": 0.0003964677479771044, + "loss": 2.712922286987305, + "step": 61550 + }, + { + "epoch": 0.3078, + "grad_norm": 0.33245566487312317, + "learning_rate": 0.0003964356001702575, + "loss": 2.7118125915527345, + "step": 61560 + }, + { + "epoch": 0.30785, + "grad_norm": 0.3471648693084717, + "learning_rate": 0.0003964034486768952, + "loss": 2.812482452392578, + "step": 61570 + }, + { + "epoch": 0.3079, + "grad_norm": 0.33625033497810364, + "learning_rate": 0.0003963712934978272, + "loss": 2.7856489181518556, + "step": 61580 + }, + { + "epoch": 0.30795, + "grad_norm": 0.35644933581352234, + "learning_rate": 0.0003963391346338627, + "loss": 2.8077091217041015, + "step": 61590 + }, + { + "epoch": 0.308, + "grad_norm": 0.3536539673805237, + "learning_rate": 0.00039630697208581155, + "loss": 2.8166057586669924, + "step": 61600 + }, + { + "epoch": 0.30805, + "grad_norm": 0.3125661313533783, + "learning_rate": 0.0003962748058544834, + "loss": 2.8164491653442383, + "step": 61610 + }, + { + "epoch": 0.3081, + "grad_norm": 0.3455277383327484, + "learning_rate": 0.000396242635940688, + "loss": 2.7987037658691407, + "step": 61620 + }, + { + "epoch": 0.30815, + "grad_norm": 0.3680521845817566, + "learning_rate": 0.00039621046234523504, + "loss": 2.82998046875, + "step": 61630 + }, + { + "epoch": 0.3082, + "grad_norm": 0.34640637040138245, + "learning_rate": 0.00039617828506893483, + "loss": 2.750980186462402, + "step": 61640 + }, + { + "epoch": 0.30825, + "grad_norm": 0.2867099642753601, + "learning_rate": 0.00039614610411259723, + "loss": 2.77700138092041, + "step": 61650 + }, + { + "epoch": 0.3083, + "grad_norm": 0.3338114321231842, + "learning_rate": 0.00039611391947703246, + "loss": 2.770977020263672, + "step": 61660 + }, + { + "epoch": 0.30835, + "grad_norm": 0.4117962121963501, + "learning_rate": 0.0003960817311630507, + "loss": 2.7931825637817385, + "step": 61670 + }, + { + "epoch": 0.3084, + "grad_norm": 0.3068656027317047, + "learning_rate": 0.00039604953917146235, + "loss": 2.7793935775756835, + "step": 61680 + }, + { + "epoch": 0.30845, + "grad_norm": 0.330463171005249, + "learning_rate": 0.00039601734350307784, + "loss": 2.8109230041503905, + "step": 61690 + }, + { + "epoch": 0.3085, + "grad_norm": 0.34977027773857117, + "learning_rate": 0.00039598514415870766, + "loss": 2.783591079711914, + "step": 61700 + }, + { + "epoch": 0.30855, + "grad_norm": 0.33323171734809875, + "learning_rate": 0.00039595294113916246, + "loss": 2.7537769317626952, + "step": 61710 + }, + { + "epoch": 0.3086, + "grad_norm": 0.34014713764190674, + "learning_rate": 0.00039592073444525295, + "loss": 2.8151058197021483, + "step": 61720 + }, + { + "epoch": 0.30865, + "grad_norm": 0.3820238411426544, + "learning_rate": 0.0003958885240777899, + "loss": 2.8102378845214844, + "step": 61730 + }, + { + "epoch": 0.3087, + "grad_norm": 0.37728583812713623, + "learning_rate": 0.00039585631003758424, + "loss": 2.794764518737793, + "step": 61740 + }, + { + "epoch": 0.30875, + "grad_norm": 0.3400370180606842, + "learning_rate": 0.00039582409232544695, + "loss": 2.7287317276000977, + "step": 61750 + }, + { + "epoch": 0.3088, + "grad_norm": 0.325854629278183, + "learning_rate": 0.0003957918709421892, + "loss": 2.749621772766113, + "step": 61760 + }, + { + "epoch": 0.30885, + "grad_norm": 0.3233296573162079, + "learning_rate": 0.00039575964588862197, + "loss": 2.7616998672485353, + "step": 61770 + }, + { + "epoch": 0.3089, + "grad_norm": 0.35207194089889526, + "learning_rate": 0.00039572741716555667, + "loss": 2.80284423828125, + "step": 61780 + }, + { + "epoch": 0.30895, + "grad_norm": 0.32721075415611267, + "learning_rate": 0.0003956951847738045, + "loss": 2.7884674072265625, + "step": 61790 + }, + { + "epoch": 0.309, + "grad_norm": 0.3353371024131775, + "learning_rate": 0.00039566294871417717, + "loss": 2.790659523010254, + "step": 61800 + }, + { + "epoch": 0.30905, + "grad_norm": 0.3179541528224945, + "learning_rate": 0.000395630708987486, + "loss": 2.799010467529297, + "step": 61810 + }, + { + "epoch": 0.3091, + "grad_norm": 0.3637048304080963, + "learning_rate": 0.00039559846559454284, + "loss": 2.7319896697998045, + "step": 61820 + }, + { + "epoch": 0.30915, + "grad_norm": 0.3452148139476776, + "learning_rate": 0.00039556621853615914, + "loss": 2.809540557861328, + "step": 61830 + }, + { + "epoch": 0.3092, + "grad_norm": 0.31364476680755615, + "learning_rate": 0.0003955339678131469, + "loss": 2.776096534729004, + "step": 61840 + }, + { + "epoch": 0.30925, + "grad_norm": 0.33350417017936707, + "learning_rate": 0.000395501713426318, + "loss": 2.831559753417969, + "step": 61850 + }, + { + "epoch": 0.3093, + "grad_norm": 0.3301025629043579, + "learning_rate": 0.0003954694553764844, + "loss": 2.783168983459473, + "step": 61860 + }, + { + "epoch": 0.30935, + "grad_norm": 0.3632865846157074, + "learning_rate": 0.0003954371936644583, + "loss": 2.7728607177734377, + "step": 61870 + }, + { + "epoch": 0.3094, + "grad_norm": 0.35786381363868713, + "learning_rate": 0.0003954049282910518, + "loss": 2.71961669921875, + "step": 61880 + }, + { + "epoch": 0.30945, + "grad_norm": 0.3212524652481079, + "learning_rate": 0.0003953726592570772, + "loss": 2.8641910552978516, + "step": 61890 + }, + { + "epoch": 0.3095, + "grad_norm": 0.39472368359565735, + "learning_rate": 0.0003953403865633468, + "loss": 2.770468521118164, + "step": 61900 + }, + { + "epoch": 0.30955, + "grad_norm": 0.3771386742591858, + "learning_rate": 0.00039530811021067325, + "loss": 2.7520336151123046, + "step": 61910 + }, + { + "epoch": 0.3096, + "grad_norm": 0.3455604910850525, + "learning_rate": 0.0003952758301998689, + "loss": 2.759804916381836, + "step": 61920 + }, + { + "epoch": 0.30965, + "grad_norm": 0.31295567750930786, + "learning_rate": 0.0003952435465317466, + "loss": 2.8165571212768556, + "step": 61930 + }, + { + "epoch": 0.3097, + "grad_norm": 0.3695603013038635, + "learning_rate": 0.00039521125920711884, + "loss": 2.743368721008301, + "step": 61940 + }, + { + "epoch": 0.30975, + "grad_norm": 0.33122310042381287, + "learning_rate": 0.00039517896822679875, + "loss": 2.829473304748535, + "step": 61950 + }, + { + "epoch": 0.3098, + "grad_norm": 0.3414788544178009, + "learning_rate": 0.0003951466735915991, + "loss": 2.806474304199219, + "step": 61960 + }, + { + "epoch": 0.30985, + "grad_norm": 0.3577026426792145, + "learning_rate": 0.00039511437530233284, + "loss": 2.727133941650391, + "step": 61970 + }, + { + "epoch": 0.3099, + "grad_norm": 0.3199129104614258, + "learning_rate": 0.0003950820733598132, + "loss": 2.794279098510742, + "step": 61980 + }, + { + "epoch": 0.30995, + "grad_norm": 0.42746272683143616, + "learning_rate": 0.00039504976776485323, + "loss": 2.758991241455078, + "step": 61990 + }, + { + "epoch": 0.31, + "grad_norm": 0.3103788495063782, + "learning_rate": 0.00039501745851826644, + "loss": 2.788766860961914, + "step": 62000 + }, + { + "epoch": 0.31005, + "grad_norm": 0.34706729650497437, + "learning_rate": 0.00039498514562086605, + "loss": 2.7809993743896486, + "step": 62010 + }, + { + "epoch": 0.3101, + "grad_norm": 0.3273739516735077, + "learning_rate": 0.00039495282907346554, + "loss": 2.7808435440063475, + "step": 62020 + }, + { + "epoch": 0.31015, + "grad_norm": 0.34058135747909546, + "learning_rate": 0.0003949205088768786, + "loss": 2.7534637451171875, + "step": 62030 + }, + { + "epoch": 0.3102, + "grad_norm": 0.3162517845630646, + "learning_rate": 0.0003948881850319188, + "loss": 2.7646432876586915, + "step": 62040 + }, + { + "epoch": 0.31025, + "grad_norm": 0.34425240755081177, + "learning_rate": 0.00039485585753939995, + "loss": 2.8003662109375, + "step": 62050 + }, + { + "epoch": 0.3103, + "grad_norm": 0.32048553228378296, + "learning_rate": 0.00039482352640013576, + "loss": 2.7964401245117188, + "step": 62060 + }, + { + "epoch": 0.31035, + "grad_norm": 0.3224177956581116, + "learning_rate": 0.00039479119161494035, + "loss": 2.806925392150879, + "step": 62070 + }, + { + "epoch": 0.3104, + "grad_norm": 0.32127249240875244, + "learning_rate": 0.00039475885318462765, + "loss": 2.813699722290039, + "step": 62080 + }, + { + "epoch": 0.31045, + "grad_norm": 0.3259986937046051, + "learning_rate": 0.0003947265111100117, + "loss": 2.739202880859375, + "step": 62090 + }, + { + "epoch": 0.3105, + "grad_norm": 0.3327244222164154, + "learning_rate": 0.0003946941653919069, + "loss": 2.8058223724365234, + "step": 62100 + }, + { + "epoch": 0.31055, + "grad_norm": 0.31462639570236206, + "learning_rate": 0.00039466181603112736, + "loss": 2.778119659423828, + "step": 62110 + }, + { + "epoch": 0.3106, + "grad_norm": 0.36091524362564087, + "learning_rate": 0.0003946294630284877, + "loss": 2.833635711669922, + "step": 62120 + }, + { + "epoch": 0.31065, + "grad_norm": 0.3215066194534302, + "learning_rate": 0.0003945971063848022, + "loss": 2.780780220031738, + "step": 62130 + }, + { + "epoch": 0.3107, + "grad_norm": 0.35237202048301697, + "learning_rate": 0.0003945647461008854, + "loss": 2.776190185546875, + "step": 62140 + }, + { + "epoch": 0.31075, + "grad_norm": 0.3248578906059265, + "learning_rate": 0.00039453238217755226, + "loss": 2.8558279037475587, + "step": 62150 + }, + { + "epoch": 0.3108, + "grad_norm": 0.33672744035720825, + "learning_rate": 0.0003945000146156173, + "loss": 2.7781835556030274, + "step": 62160 + }, + { + "epoch": 0.31085, + "grad_norm": 0.32643401622772217, + "learning_rate": 0.00039446764341589545, + "loss": 2.783399963378906, + "step": 62170 + }, + { + "epoch": 0.3109, + "grad_norm": 0.31201034784317017, + "learning_rate": 0.0003944352685792016, + "loss": 2.73538818359375, + "step": 62180 + }, + { + "epoch": 0.31095, + "grad_norm": 0.34061771631240845, + "learning_rate": 0.0003944028901063509, + "loss": 2.7870803833007813, + "step": 62190 + }, + { + "epoch": 0.311, + "grad_norm": 0.3587060868740082, + "learning_rate": 0.0003943705079981583, + "loss": 2.762251281738281, + "step": 62200 + }, + { + "epoch": 0.31105, + "grad_norm": 0.32545650005340576, + "learning_rate": 0.00039433812225543927, + "loss": 2.77868595123291, + "step": 62210 + }, + { + "epoch": 0.3111, + "grad_norm": 0.34221360087394714, + "learning_rate": 0.00039430573287900885, + "loss": 2.748916816711426, + "step": 62220 + }, + { + "epoch": 0.31115, + "grad_norm": 0.3527939021587372, + "learning_rate": 0.00039427333986968263, + "loss": 2.7648386001586913, + "step": 62230 + }, + { + "epoch": 0.3112, + "grad_norm": 0.33024173974990845, + "learning_rate": 0.000394240943228276, + "loss": 2.7751773834228515, + "step": 62240 + }, + { + "epoch": 0.31125, + "grad_norm": 0.3263540267944336, + "learning_rate": 0.0003942085429556046, + "loss": 2.7312038421630858, + "step": 62250 + }, + { + "epoch": 0.3113, + "grad_norm": 0.3284808397293091, + "learning_rate": 0.0003941761390524842, + "loss": 2.745637321472168, + "step": 62260 + }, + { + "epoch": 0.31135, + "grad_norm": 0.33995962142944336, + "learning_rate": 0.0003941437315197303, + "loss": 2.729678726196289, + "step": 62270 + }, + { + "epoch": 0.3114, + "grad_norm": 0.33342334628105164, + "learning_rate": 0.000394111320358159, + "loss": 2.7962825775146483, + "step": 62280 + }, + { + "epoch": 0.31145, + "grad_norm": 0.3225201964378357, + "learning_rate": 0.00039407890556858617, + "loss": 2.7448354721069337, + "step": 62290 + }, + { + "epoch": 0.3115, + "grad_norm": 0.3488818407058716, + "learning_rate": 0.0003940464871518279, + "loss": 2.7546855926513674, + "step": 62300 + }, + { + "epoch": 0.31155, + "grad_norm": 0.34978869557380676, + "learning_rate": 0.0003940140651087002, + "loss": 2.7631608963012697, + "step": 62310 + }, + { + "epoch": 0.3116, + "grad_norm": 0.3346109390258789, + "learning_rate": 0.0003939816394400194, + "loss": 2.803691864013672, + "step": 62320 + }, + { + "epoch": 0.31165, + "grad_norm": 0.3275657892227173, + "learning_rate": 0.00039394921014660177, + "loss": 2.785363960266113, + "step": 62330 + }, + { + "epoch": 0.3117, + "grad_norm": 0.3611825704574585, + "learning_rate": 0.00039391677722926376, + "loss": 2.758500671386719, + "step": 62340 + }, + { + "epoch": 0.31175, + "grad_norm": 0.35450926423072815, + "learning_rate": 0.0003938843406888218, + "loss": 2.755628204345703, + "step": 62350 + }, + { + "epoch": 0.3118, + "grad_norm": 0.3819355070590973, + "learning_rate": 0.0003938519005260927, + "loss": 2.7817039489746094, + "step": 62360 + }, + { + "epoch": 0.31185, + "grad_norm": 0.391335666179657, + "learning_rate": 0.0003938194567418928, + "loss": 2.744426727294922, + "step": 62370 + }, + { + "epoch": 0.3119, + "grad_norm": 0.3178156018257141, + "learning_rate": 0.00039378700933703905, + "loss": 2.8377353668212892, + "step": 62380 + }, + { + "epoch": 0.31195, + "grad_norm": 0.34502261877059937, + "learning_rate": 0.0003937545583123483, + "loss": 2.797121047973633, + "step": 62390 + }, + { + "epoch": 0.312, + "grad_norm": 0.3419400751590729, + "learning_rate": 0.0003937221036686375, + "loss": 2.743568992614746, + "step": 62400 + }, + { + "epoch": 0.31205, + "grad_norm": 0.32766515016555786, + "learning_rate": 0.0003936896454067238, + "loss": 2.777775764465332, + "step": 62410 + }, + { + "epoch": 0.3121, + "grad_norm": 0.342292457818985, + "learning_rate": 0.0003936571835274241, + "loss": 2.7883142471313476, + "step": 62420 + }, + { + "epoch": 0.31215, + "grad_norm": 0.30479297041893005, + "learning_rate": 0.00039362471803155584, + "loss": 2.750901794433594, + "step": 62430 + }, + { + "epoch": 0.3122, + "grad_norm": 0.3453761339187622, + "learning_rate": 0.00039359224891993627, + "loss": 2.7714004516601562, + "step": 62440 + }, + { + "epoch": 0.31225, + "grad_norm": 0.30823802947998047, + "learning_rate": 0.00039355977619338277, + "loss": 2.7465505599975586, + "step": 62450 + }, + { + "epoch": 0.3123, + "grad_norm": 0.32599499821662903, + "learning_rate": 0.0003935272998527128, + "loss": 2.7413101196289062, + "step": 62460 + }, + { + "epoch": 0.31235, + "grad_norm": 0.3074691593647003, + "learning_rate": 0.00039349481989874415, + "loss": 2.8020782470703125, + "step": 62470 + }, + { + "epoch": 0.3124, + "grad_norm": 0.32307907938957214, + "learning_rate": 0.00039346233633229434, + "loss": 2.7719776153564455, + "step": 62480 + }, + { + "epoch": 0.31245, + "grad_norm": 0.30727916955947876, + "learning_rate": 0.0003934298491541811, + "loss": 2.7546102523803713, + "step": 62490 + }, + { + "epoch": 0.3125, + "grad_norm": 0.3470028340816498, + "learning_rate": 0.0003933973583652224, + "loss": 2.79781494140625, + "step": 62500 + }, + { + "epoch": 0.31255, + "grad_norm": 0.3410707414150238, + "learning_rate": 0.00039336486396623617, + "loss": 2.7929428100585936, + "step": 62510 + }, + { + "epoch": 0.3126, + "grad_norm": 0.38475272059440613, + "learning_rate": 0.00039333236595804044, + "loss": 2.7743432998657225, + "step": 62520 + }, + { + "epoch": 0.31265, + "grad_norm": 0.3496464490890503, + "learning_rate": 0.0003932998643414534, + "loss": 2.771631050109863, + "step": 62530 + }, + { + "epoch": 0.3127, + "grad_norm": 0.3528592586517334, + "learning_rate": 0.0003932673591172932, + "loss": 2.8092180252075196, + "step": 62540 + }, + { + "epoch": 0.31275, + "grad_norm": 0.3895825147628784, + "learning_rate": 0.0003932348502863781, + "loss": 2.7835155487060548, + "step": 62550 + }, + { + "epoch": 0.3128, + "grad_norm": 0.31321337819099426, + "learning_rate": 0.0003932023378495267, + "loss": 2.768472099304199, + "step": 62560 + }, + { + "epoch": 0.31285, + "grad_norm": 0.3555309474468231, + "learning_rate": 0.0003931698218075574, + "loss": 2.7578758239746093, + "step": 62570 + }, + { + "epoch": 0.3129, + "grad_norm": 0.33757051825523376, + "learning_rate": 0.0003931373021612887, + "loss": 2.7725923538208006, + "step": 62580 + }, + { + "epoch": 0.31295, + "grad_norm": 0.33230045437812805, + "learning_rate": 0.0003931047789115395, + "loss": 2.7854957580566406, + "step": 62590 + }, + { + "epoch": 0.313, + "grad_norm": 0.32634052634239197, + "learning_rate": 0.00039307225205912846, + "loss": 2.7875732421875, + "step": 62600 + }, + { + "epoch": 0.31305, + "grad_norm": 0.318863183259964, + "learning_rate": 0.00039303972160487435, + "loss": 2.767739486694336, + "step": 62610 + }, + { + "epoch": 0.3131, + "grad_norm": 0.32424789667129517, + "learning_rate": 0.0003930071875495962, + "loss": 2.779796600341797, + "step": 62620 + }, + { + "epoch": 0.31315, + "grad_norm": 0.37541377544403076, + "learning_rate": 0.00039297464989411306, + "loss": 2.785706901550293, + "step": 62630 + }, + { + "epoch": 0.3132, + "grad_norm": 0.3207271099090576, + "learning_rate": 0.00039294210863924405, + "loss": 2.7641017913818358, + "step": 62640 + }, + { + "epoch": 0.31325, + "grad_norm": 0.30831560492515564, + "learning_rate": 0.0003929095637858085, + "loss": 2.764098358154297, + "step": 62650 + }, + { + "epoch": 0.3133, + "grad_norm": 0.32935500144958496, + "learning_rate": 0.0003928770153346256, + "loss": 2.8051126480102537, + "step": 62660 + }, + { + "epoch": 0.31335, + "grad_norm": 0.3451465666294098, + "learning_rate": 0.0003928444632865147, + "loss": 2.7897415161132812, + "step": 62670 + }, + { + "epoch": 0.3134, + "grad_norm": 0.3590226173400879, + "learning_rate": 0.0003928119076422955, + "loss": 2.797193145751953, + "step": 62680 + }, + { + "epoch": 0.31345, + "grad_norm": 0.32220083475112915, + "learning_rate": 0.00039277934840278747, + "loss": 2.7416189193725584, + "step": 62690 + }, + { + "epoch": 0.3135, + "grad_norm": 0.36356064677238464, + "learning_rate": 0.0003927467855688103, + "loss": 2.839765930175781, + "step": 62700 + }, + { + "epoch": 0.31355, + "grad_norm": 0.3531944155693054, + "learning_rate": 0.00039271421914118367, + "loss": 2.8302385330200197, + "step": 62710 + }, + { + "epoch": 0.3136, + "grad_norm": 0.37171077728271484, + "learning_rate": 0.00039268164912072763, + "loss": 2.758119583129883, + "step": 62720 + }, + { + "epoch": 0.31365, + "grad_norm": 0.36287009716033936, + "learning_rate": 0.00039264907550826193, + "loss": 2.782419204711914, + "step": 62730 + }, + { + "epoch": 0.3137, + "grad_norm": 0.3371746242046356, + "learning_rate": 0.00039261649830460676, + "loss": 2.7501909255981447, + "step": 62740 + }, + { + "epoch": 0.31375, + "grad_norm": 0.33569297194480896, + "learning_rate": 0.0003925839175105822, + "loss": 2.797886276245117, + "step": 62750 + }, + { + "epoch": 0.3138, + "grad_norm": 0.37126895785331726, + "learning_rate": 0.00039255133312700855, + "loss": 2.78182373046875, + "step": 62760 + }, + { + "epoch": 0.31385, + "grad_norm": 0.31487348675727844, + "learning_rate": 0.00039251874515470585, + "loss": 2.850747489929199, + "step": 62770 + }, + { + "epoch": 0.3139, + "grad_norm": 0.30880987644195557, + "learning_rate": 0.0003924861535944949, + "loss": 2.809254837036133, + "step": 62780 + }, + { + "epoch": 0.31395, + "grad_norm": 0.37412795424461365, + "learning_rate": 0.00039245355844719584, + "loss": 2.7464454650878904, + "step": 62790 + }, + { + "epoch": 0.314, + "grad_norm": 0.3370353579521179, + "learning_rate": 0.00039242095971362946, + "loss": 2.799096870422363, + "step": 62800 + }, + { + "epoch": 0.31405, + "grad_norm": 0.3498474657535553, + "learning_rate": 0.0003923883573946163, + "loss": 2.7261432647705077, + "step": 62810 + }, + { + "epoch": 0.3141, + "grad_norm": 0.3199339807033539, + "learning_rate": 0.0003923557514909773, + "loss": 2.787121391296387, + "step": 62820 + }, + { + "epoch": 0.31415, + "grad_norm": 0.3252459764480591, + "learning_rate": 0.00039232314200353315, + "loss": 2.844984436035156, + "step": 62830 + }, + { + "epoch": 0.3142, + "grad_norm": 0.34114736318588257, + "learning_rate": 0.0003922905289331049, + "loss": 2.763096237182617, + "step": 62840 + }, + { + "epoch": 0.31425, + "grad_norm": 0.37553349137306213, + "learning_rate": 0.0003922579122805134, + "loss": 2.7453298568725586, + "step": 62850 + }, + { + "epoch": 0.3143, + "grad_norm": 0.3306816816329956, + "learning_rate": 0.00039222529204658007, + "loss": 2.8150522232055666, + "step": 62860 + }, + { + "epoch": 0.31435, + "grad_norm": 0.370641827583313, + "learning_rate": 0.0003921926682321259, + "loss": 2.7827381134033202, + "step": 62870 + }, + { + "epoch": 0.3144, + "grad_norm": 0.31219562888145447, + "learning_rate": 0.0003921600408379723, + "loss": 2.7515279769897463, + "step": 62880 + }, + { + "epoch": 0.31445, + "grad_norm": 0.3578277826309204, + "learning_rate": 0.0003921274098649405, + "loss": 2.736435127258301, + "step": 62890 + }, + { + "epoch": 0.3145, + "grad_norm": 0.32978707551956177, + "learning_rate": 0.0003920947753138522, + "loss": 2.8324745178222654, + "step": 62900 + }, + { + "epoch": 0.31455, + "grad_norm": 0.3421213626861572, + "learning_rate": 0.0003920621371855289, + "loss": 2.775504302978516, + "step": 62910 + }, + { + "epoch": 0.3146, + "grad_norm": 0.36391544342041016, + "learning_rate": 0.0003920294954807922, + "loss": 2.8138458251953127, + "step": 62920 + }, + { + "epoch": 0.31465, + "grad_norm": 0.35615628957748413, + "learning_rate": 0.00039199685020046383, + "loss": 2.7762239456176756, + "step": 62930 + }, + { + "epoch": 0.3147, + "grad_norm": 0.31525304913520813, + "learning_rate": 0.0003919642013453658, + "loss": 2.728671073913574, + "step": 62940 + }, + { + "epoch": 0.31475, + "grad_norm": 0.3289265036582947, + "learning_rate": 0.00039193154891632, + "loss": 2.7870656967163088, + "step": 62950 + }, + { + "epoch": 0.3148, + "grad_norm": 0.33909428119659424, + "learning_rate": 0.00039189889291414834, + "loss": 2.809100532531738, + "step": 62960 + }, + { + "epoch": 0.31485, + "grad_norm": 0.3462473154067993, + "learning_rate": 0.000391866233339673, + "loss": 2.8098464965820313, + "step": 62970 + }, + { + "epoch": 0.3149, + "grad_norm": 0.35535362362861633, + "learning_rate": 0.00039183357019371615, + "loss": 2.778388786315918, + "step": 62980 + }, + { + "epoch": 0.31495, + "grad_norm": 0.31856968998908997, + "learning_rate": 0.00039180090347710014, + "loss": 2.791696548461914, + "step": 62990 + }, + { + "epoch": 0.315, + "grad_norm": 0.32545673847198486, + "learning_rate": 0.00039176823319064737, + "loss": 2.8001964569091795, + "step": 63000 + }, + { + "epoch": 0.31505, + "grad_norm": 0.3360145390033722, + "learning_rate": 0.0003917355593351802, + "loss": 2.768664741516113, + "step": 63010 + }, + { + "epoch": 0.3151, + "grad_norm": 0.318851500749588, + "learning_rate": 0.00039170288191152136, + "loss": 2.7579940795898437, + "step": 63020 + }, + { + "epoch": 0.31515, + "grad_norm": 0.3377171456813812, + "learning_rate": 0.00039167020092049335, + "loss": 2.744194221496582, + "step": 63030 + }, + { + "epoch": 0.3152, + "grad_norm": 0.35596492886543274, + "learning_rate": 0.000391637516362919, + "loss": 2.7466564178466797, + "step": 63040 + }, + { + "epoch": 0.31525, + "grad_norm": 0.3578857183456421, + "learning_rate": 0.0003916048282396211, + "loss": 2.7716217041015625, + "step": 63050 + }, + { + "epoch": 0.3153, + "grad_norm": 0.3539103865623474, + "learning_rate": 0.00039157213655142265, + "loss": 2.7864036560058594, + "step": 63060 + }, + { + "epoch": 0.31535, + "grad_norm": 0.3480004072189331, + "learning_rate": 0.0003915394412991465, + "loss": 2.769593048095703, + "step": 63070 + }, + { + "epoch": 0.3154, + "grad_norm": 0.3578735589981079, + "learning_rate": 0.000391506742483616, + "loss": 2.7642351150512696, + "step": 63080 + }, + { + "epoch": 0.31545, + "grad_norm": 0.37423208355903625, + "learning_rate": 0.00039147404010565415, + "loss": 2.8159578323364256, + "step": 63090 + }, + { + "epoch": 0.3155, + "grad_norm": 0.3344326913356781, + "learning_rate": 0.0003914413341660843, + "loss": 2.7749250411987303, + "step": 63100 + }, + { + "epoch": 0.31555, + "grad_norm": 0.3229919672012329, + "learning_rate": 0.0003914086246657298, + "loss": 2.781337547302246, + "step": 63110 + }, + { + "epoch": 0.3156, + "grad_norm": 0.31847700476646423, + "learning_rate": 0.0003913759116054141, + "loss": 2.7835594177246095, + "step": 63120 + }, + { + "epoch": 0.31565, + "grad_norm": 0.3319474756717682, + "learning_rate": 0.0003913431949859607, + "loss": 2.77145938873291, + "step": 63130 + }, + { + "epoch": 0.3157, + "grad_norm": 0.3354262113571167, + "learning_rate": 0.0003913104748081935, + "loss": 2.705592155456543, + "step": 63140 + }, + { + "epoch": 0.31575, + "grad_norm": 0.3145484924316406, + "learning_rate": 0.0003912777510729358, + "loss": 2.725447654724121, + "step": 63150 + }, + { + "epoch": 0.3158, + "grad_norm": 0.32373207807540894, + "learning_rate": 0.00039124502378101185, + "loss": 2.7206212997436525, + "step": 63160 + }, + { + "epoch": 0.31585, + "grad_norm": 0.3575333058834076, + "learning_rate": 0.0003912122929332453, + "loss": 2.7770156860351562, + "step": 63170 + }, + { + "epoch": 0.3159, + "grad_norm": 0.3022885024547577, + "learning_rate": 0.00039117955853046023, + "loss": 2.7725593566894533, + "step": 63180 + }, + { + "epoch": 0.31595, + "grad_norm": 0.3445993959903717, + "learning_rate": 0.00039114682057348073, + "loss": 2.751745414733887, + "step": 63190 + }, + { + "epoch": 0.316, + "grad_norm": 0.3850545287132263, + "learning_rate": 0.0003911140790631309, + "loss": 2.76954345703125, + "step": 63200 + }, + { + "epoch": 0.31605, + "grad_norm": 0.3268568217754364, + "learning_rate": 0.0003910813340002352, + "loss": 2.7791872024536133, + "step": 63210 + }, + { + "epoch": 0.3161, + "grad_norm": 0.37041088938713074, + "learning_rate": 0.0003910485853856177, + "loss": 2.8109277725219726, + "step": 63220 + }, + { + "epoch": 0.31615, + "grad_norm": 0.3509655296802521, + "learning_rate": 0.0003910158332201031, + "loss": 2.8028337478637697, + "step": 63230 + }, + { + "epoch": 0.3162, + "grad_norm": 0.3498767018318176, + "learning_rate": 0.00039098307750451585, + "loss": 2.825515365600586, + "step": 63240 + }, + { + "epoch": 0.31625, + "grad_norm": 0.4060635268688202, + "learning_rate": 0.00039095031823968053, + "loss": 2.784683609008789, + "step": 63250 + }, + { + "epoch": 0.3163, + "grad_norm": 0.36584463715553284, + "learning_rate": 0.00039091755542642193, + "loss": 2.809419631958008, + "step": 63260 + }, + { + "epoch": 0.31635, + "grad_norm": 0.3533365726470947, + "learning_rate": 0.0003908847890655647, + "loss": 2.814574432373047, + "step": 63270 + }, + { + "epoch": 0.3164, + "grad_norm": 0.3285982310771942, + "learning_rate": 0.000390852019157934, + "loss": 2.7787460327148437, + "step": 63280 + }, + { + "epoch": 0.31645, + "grad_norm": 0.3442291021347046, + "learning_rate": 0.0003908192457043546, + "loss": 2.7182024002075194, + "step": 63290 + }, + { + "epoch": 0.3165, + "grad_norm": 0.3498513996601105, + "learning_rate": 0.0003907864687056516, + "loss": 2.815262222290039, + "step": 63300 + }, + { + "epoch": 0.31655, + "grad_norm": 0.34886473417282104, + "learning_rate": 0.0003907536881626502, + "loss": 2.8494489669799803, + "step": 63310 + }, + { + "epoch": 0.3166, + "grad_norm": 0.3020434081554413, + "learning_rate": 0.0003907209040761757, + "loss": 2.695028305053711, + "step": 63320 + }, + { + "epoch": 0.31665, + "grad_norm": 0.31665468215942383, + "learning_rate": 0.00039068811644705326, + "loss": 2.8080663681030273, + "step": 63330 + }, + { + "epoch": 0.3167, + "grad_norm": 0.3108874261379242, + "learning_rate": 0.00039065532527610857, + "loss": 2.79910888671875, + "step": 63340 + }, + { + "epoch": 0.31675, + "grad_norm": 0.344580739736557, + "learning_rate": 0.00039062253056416685, + "loss": 2.7585506439208984, + "step": 63350 + }, + { + "epoch": 0.3168, + "grad_norm": 0.3356591463088989, + "learning_rate": 0.000390589732312054, + "loss": 2.755851936340332, + "step": 63360 + }, + { + "epoch": 0.31685, + "grad_norm": 0.3278847634792328, + "learning_rate": 0.00039055693052059555, + "loss": 2.836620330810547, + "step": 63370 + }, + { + "epoch": 0.3169, + "grad_norm": 0.31274208426475525, + "learning_rate": 0.0003905241251906173, + "loss": 2.7950042724609374, + "step": 63380 + }, + { + "epoch": 0.31695, + "grad_norm": 0.3175993859767914, + "learning_rate": 0.0003904913163229451, + "loss": 2.7779218673706056, + "step": 63390 + }, + { + "epoch": 0.317, + "grad_norm": 0.33016684651374817, + "learning_rate": 0.000390458503918405, + "loss": 2.788984489440918, + "step": 63400 + }, + { + "epoch": 0.31705, + "grad_norm": 0.3053135275840759, + "learning_rate": 0.00039042568797782295, + "loss": 2.798720359802246, + "step": 63410 + }, + { + "epoch": 0.3171, + "grad_norm": 0.3652627766132355, + "learning_rate": 0.0003903928685020252, + "loss": 2.7946355819702147, + "step": 63420 + }, + { + "epoch": 0.31715, + "grad_norm": 0.3457755446434021, + "learning_rate": 0.0003903600454918379, + "loss": 2.7703413009643554, + "step": 63430 + }, + { + "epoch": 0.3172, + "grad_norm": 0.33805400133132935, + "learning_rate": 0.0003903272189480874, + "loss": 2.7703264236450194, + "step": 63440 + }, + { + "epoch": 0.31725, + "grad_norm": 0.31478527188301086, + "learning_rate": 0.0003902943888716001, + "loss": 2.8054777145385743, + "step": 63450 + }, + { + "epoch": 0.3173, + "grad_norm": 0.347711443901062, + "learning_rate": 0.0003902615552632025, + "loss": 2.7618602752685546, + "step": 63460 + }, + { + "epoch": 0.31735, + "grad_norm": 0.3224261701107025, + "learning_rate": 0.0003902287181237212, + "loss": 2.787689971923828, + "step": 63470 + }, + { + "epoch": 0.3174, + "grad_norm": 0.3150254487991333, + "learning_rate": 0.00039019587745398276, + "loss": 2.7945728302001953, + "step": 63480 + }, + { + "epoch": 0.31745, + "grad_norm": 0.3262255787849426, + "learning_rate": 0.0003901630332548141, + "loss": 2.8142934799194337, + "step": 63490 + }, + { + "epoch": 0.3175, + "grad_norm": 0.3346201479434967, + "learning_rate": 0.00039013018552704206, + "loss": 2.7946439743041993, + "step": 63500 + }, + { + "epoch": 0.31755, + "grad_norm": 0.3244723081588745, + "learning_rate": 0.0003900973342714935, + "loss": 2.7705249786376953, + "step": 63510 + }, + { + "epoch": 0.3176, + "grad_norm": 0.34310609102249146, + "learning_rate": 0.00039006447948899546, + "loss": 2.7891828536987306, + "step": 63520 + }, + { + "epoch": 0.31765, + "grad_norm": 0.3466045558452606, + "learning_rate": 0.00039003162118037514, + "loss": 2.713962936401367, + "step": 63530 + }, + { + "epoch": 0.3177, + "grad_norm": 0.31169748306274414, + "learning_rate": 0.0003899987593464597, + "loss": 2.7897480010986326, + "step": 63540 + }, + { + "epoch": 0.31775, + "grad_norm": 0.34396687150001526, + "learning_rate": 0.0003899658939880764, + "loss": 2.768999481201172, + "step": 63550 + }, + { + "epoch": 0.3178, + "grad_norm": 0.34237727522850037, + "learning_rate": 0.0003899330251060527, + "loss": 2.8166275024414062, + "step": 63560 + }, + { + "epoch": 0.31785, + "grad_norm": 0.3627862334251404, + "learning_rate": 0.0003899001527012159, + "loss": 2.7634735107421875, + "step": 63570 + }, + { + "epoch": 0.3179, + "grad_norm": 0.31107214093208313, + "learning_rate": 0.00038986727677439384, + "loss": 2.7587989807128905, + "step": 63580 + }, + { + "epoch": 0.31795, + "grad_norm": 0.36359742283821106, + "learning_rate": 0.0003898343973264139, + "loss": 2.750432014465332, + "step": 63590 + }, + { + "epoch": 0.318, + "grad_norm": 0.35363590717315674, + "learning_rate": 0.00038980151435810396, + "loss": 2.831713104248047, + "step": 63600 + }, + { + "epoch": 0.31805, + "grad_norm": 0.335629403591156, + "learning_rate": 0.00038976862787029186, + "loss": 2.752029228210449, + "step": 63610 + }, + { + "epoch": 0.3181, + "grad_norm": 0.37435972690582275, + "learning_rate": 0.00038973573786380553, + "loss": 2.7843631744384765, + "step": 63620 + }, + { + "epoch": 0.31815, + "grad_norm": 0.36230355501174927, + "learning_rate": 0.0003897028443394729, + "loss": 2.7047039031982423, + "step": 63630 + }, + { + "epoch": 0.3182, + "grad_norm": 0.39321112632751465, + "learning_rate": 0.0003896699472981221, + "loss": 2.7169517517089843, + "step": 63640 + }, + { + "epoch": 0.31825, + "grad_norm": 0.3365134596824646, + "learning_rate": 0.0003896370467405813, + "loss": 2.8082300186157227, + "step": 63650 + }, + { + "epoch": 0.3183, + "grad_norm": 0.3568336069583893, + "learning_rate": 0.0003896041426676787, + "loss": 2.7475982666015626, + "step": 63660 + }, + { + "epoch": 0.31835, + "grad_norm": 0.37972721457481384, + "learning_rate": 0.0003895712350802429, + "loss": 2.702322006225586, + "step": 63670 + }, + { + "epoch": 0.3184, + "grad_norm": 0.33463114500045776, + "learning_rate": 0.0003895383239791021, + "loss": 2.8322404861450194, + "step": 63680 + }, + { + "epoch": 0.31845, + "grad_norm": 0.323381632566452, + "learning_rate": 0.0003895054093650849, + "loss": 2.8198890686035156, + "step": 63690 + }, + { + "epoch": 0.3185, + "grad_norm": 0.3083694577217102, + "learning_rate": 0.00038947249123902003, + "loss": 2.8038660049438477, + "step": 63700 + }, + { + "epoch": 0.31855, + "grad_norm": 0.32598876953125, + "learning_rate": 0.0003894395696017361, + "loss": 2.7593297958374023, + "step": 63710 + }, + { + "epoch": 0.3186, + "grad_norm": 0.31553754210472107, + "learning_rate": 0.0003894066444540619, + "loss": 2.770488739013672, + "step": 63720 + }, + { + "epoch": 0.31865, + "grad_norm": 0.3812231123447418, + "learning_rate": 0.0003893737157968264, + "loss": 2.7796192169189453, + "step": 63730 + }, + { + "epoch": 0.3187, + "grad_norm": 0.38749727606773376, + "learning_rate": 0.00038934078363085845, + "loss": 2.780046272277832, + "step": 63740 + }, + { + "epoch": 0.31875, + "grad_norm": 0.3503393828868866, + "learning_rate": 0.0003893078479569874, + "loss": 2.7594415664672853, + "step": 63750 + }, + { + "epoch": 0.3188, + "grad_norm": 0.3527878522872925, + "learning_rate": 0.000389274908776042, + "loss": 2.7916242599487306, + "step": 63760 + }, + { + "epoch": 0.31885, + "grad_norm": 0.31913745403289795, + "learning_rate": 0.0003892419660888518, + "loss": 2.8188446044921873, + "step": 63770 + }, + { + "epoch": 0.3189, + "grad_norm": 0.36322468519210815, + "learning_rate": 0.000389209019896246, + "loss": 2.8278121948242188, + "step": 63780 + }, + { + "epoch": 0.31895, + "grad_norm": 0.36444562673568726, + "learning_rate": 0.0003891760701990541, + "loss": 2.7717737197875976, + "step": 63790 + }, + { + "epoch": 0.319, + "grad_norm": 0.3499029278755188, + "learning_rate": 0.00038914311699810543, + "loss": 2.811095428466797, + "step": 63800 + }, + { + "epoch": 0.31905, + "grad_norm": 0.3527213931083679, + "learning_rate": 0.00038911016029422984, + "loss": 2.7109182357788084, + "step": 63810 + }, + { + "epoch": 0.3191, + "grad_norm": 0.3145475685596466, + "learning_rate": 0.00038907720008825693, + "loss": 2.7683868408203125, + "step": 63820 + }, + { + "epoch": 0.31915, + "grad_norm": 0.33611583709716797, + "learning_rate": 0.0003890442363810164, + "loss": 2.8020889282226564, + "step": 63830 + }, + { + "epoch": 0.3192, + "grad_norm": 0.370435506105423, + "learning_rate": 0.00038901126917333806, + "loss": 2.8315006256103517, + "step": 63840 + }, + { + "epoch": 0.31925, + "grad_norm": 0.29962387681007385, + "learning_rate": 0.00038897829846605206, + "loss": 2.769055938720703, + "step": 63850 + }, + { + "epoch": 0.3193, + "grad_norm": 0.33246394991874695, + "learning_rate": 0.0003889453242599883, + "loss": 2.7978782653808594, + "step": 63860 + }, + { + "epoch": 0.31935, + "grad_norm": 0.31210440397262573, + "learning_rate": 0.0003889123465559769, + "loss": 2.7732059478759767, + "step": 63870 + }, + { + "epoch": 0.3194, + "grad_norm": 0.3501077890396118, + "learning_rate": 0.0003888793653548481, + "loss": 2.8230112075805662, + "step": 63880 + }, + { + "epoch": 0.31945, + "grad_norm": 0.3602078855037689, + "learning_rate": 0.00038884638065743226, + "loss": 2.769354057312012, + "step": 63890 + }, + { + "epoch": 0.3195, + "grad_norm": 0.3529110550880432, + "learning_rate": 0.00038881339246455966, + "loss": 2.7825349807739257, + "step": 63900 + }, + { + "epoch": 0.31955, + "grad_norm": 0.32027554512023926, + "learning_rate": 0.0003887804007770608, + "loss": 2.7642723083496095, + "step": 63910 + }, + { + "epoch": 0.3196, + "grad_norm": 0.32028308510780334, + "learning_rate": 0.0003887474055957664, + "loss": 2.738302230834961, + "step": 63920 + }, + { + "epoch": 0.31965, + "grad_norm": 0.35622259974479675, + "learning_rate": 0.00038871440692150687, + "loss": 2.7914302825927733, + "step": 63930 + }, + { + "epoch": 0.3197, + "grad_norm": 0.32769766449928284, + "learning_rate": 0.00038868140475511315, + "loss": 2.7813232421875, + "step": 63940 + }, + { + "epoch": 0.31975, + "grad_norm": 0.3277485966682434, + "learning_rate": 0.00038864839909741593, + "loss": 2.7733461380004885, + "step": 63950 + }, + { + "epoch": 0.3198, + "grad_norm": 0.3572827875614166, + "learning_rate": 0.0003886153899492462, + "loss": 2.7745378494262694, + "step": 63960 + }, + { + "epoch": 0.31985, + "grad_norm": 0.33273938298225403, + "learning_rate": 0.00038858237731143497, + "loss": 2.7665565490722654, + "step": 63970 + }, + { + "epoch": 0.3199, + "grad_norm": 0.33525946736335754, + "learning_rate": 0.0003885493611848133, + "loss": 2.7242172241210936, + "step": 63980 + }, + { + "epoch": 0.31995, + "grad_norm": 0.33431437611579895, + "learning_rate": 0.0003885163415702125, + "loss": 2.7694890975952147, + "step": 63990 + }, + { + "epoch": 0.32, + "grad_norm": 0.36619845032691956, + "learning_rate": 0.0003884833184684636, + "loss": 2.8128604888916016, + "step": 64000 + }, + { + "epoch": 0.32005, + "grad_norm": 0.33840081095695496, + "learning_rate": 0.00038845029188039814, + "loss": 2.7953969955444338, + "step": 64010 + }, + { + "epoch": 0.3201, + "grad_norm": 0.316633015871048, + "learning_rate": 0.0003884172618068474, + "loss": 2.7824548721313476, + "step": 64020 + }, + { + "epoch": 0.32015, + "grad_norm": 0.3416288197040558, + "learning_rate": 0.00038838422824864316, + "loss": 2.803519058227539, + "step": 64030 + }, + { + "epoch": 0.3202, + "grad_norm": 0.3302445411682129, + "learning_rate": 0.00038835119120661683, + "loss": 2.7100151062011717, + "step": 64040 + }, + { + "epoch": 0.32025, + "grad_norm": 0.3172881603240967, + "learning_rate": 0.00038831815068160024, + "loss": 2.787219429016113, + "step": 64050 + }, + { + "epoch": 0.3203, + "grad_norm": 0.3134535253047943, + "learning_rate": 0.0003882851066744251, + "loss": 2.7856586456298826, + "step": 64060 + }, + { + "epoch": 0.32035, + "grad_norm": 0.3576169013977051, + "learning_rate": 0.0003882520591859233, + "loss": 2.7927486419677736, + "step": 64070 + }, + { + "epoch": 0.3204, + "grad_norm": 0.3277982771396637, + "learning_rate": 0.00038821900821692684, + "loss": 2.8037790298461913, + "step": 64080 + }, + { + "epoch": 0.32045, + "grad_norm": 0.44480443000793457, + "learning_rate": 0.0003881859537682678, + "loss": 2.7969038009643556, + "step": 64090 + }, + { + "epoch": 0.3205, + "grad_norm": 0.3383195698261261, + "learning_rate": 0.0003881528958407783, + "loss": 2.758960151672363, + "step": 64100 + }, + { + "epoch": 0.32055, + "grad_norm": 0.3327193558216095, + "learning_rate": 0.00038811983443529053, + "loss": 2.724446678161621, + "step": 64110 + }, + { + "epoch": 0.3206, + "grad_norm": 0.3404598534107208, + "learning_rate": 0.00038808676955263684, + "loss": 2.786115837097168, + "step": 64120 + }, + { + "epoch": 0.32065, + "grad_norm": 0.33351269364356995, + "learning_rate": 0.0003880537011936497, + "loss": 2.8018119812011717, + "step": 64130 + }, + { + "epoch": 0.3207, + "grad_norm": 0.3287157416343689, + "learning_rate": 0.0003880206293591615, + "loss": 2.808099365234375, + "step": 64140 + }, + { + "epoch": 0.32075, + "grad_norm": 0.3356676697731018, + "learning_rate": 0.00038798755405000494, + "loss": 2.7779848098754885, + "step": 64150 + }, + { + "epoch": 0.3208, + "grad_norm": 0.3701429069042206, + "learning_rate": 0.0003879544752670126, + "loss": 2.751350975036621, + "step": 64160 + }, + { + "epoch": 0.32085, + "grad_norm": 0.4178856611251831, + "learning_rate": 0.00038792139301101723, + "loss": 2.7637264251708986, + "step": 64170 + }, + { + "epoch": 0.3209, + "grad_norm": 0.35256221890449524, + "learning_rate": 0.00038788830728285175, + "loss": 2.760905647277832, + "step": 64180 + }, + { + "epoch": 0.32095, + "grad_norm": 0.34523916244506836, + "learning_rate": 0.0003878552180833491, + "loss": 2.766391944885254, + "step": 64190 + }, + { + "epoch": 0.321, + "grad_norm": 0.32481294870376587, + "learning_rate": 0.00038782212541334216, + "loss": 2.8100563049316407, + "step": 64200 + }, + { + "epoch": 0.32105, + "grad_norm": 0.3231998085975647, + "learning_rate": 0.0003877890292736642, + "loss": 2.7620792388916016, + "step": 64210 + }, + { + "epoch": 0.3211, + "grad_norm": 0.3676568269729614, + "learning_rate": 0.00038775592966514835, + "loss": 2.753952980041504, + "step": 64220 + }, + { + "epoch": 0.32115, + "grad_norm": 0.3103042244911194, + "learning_rate": 0.0003877228265886279, + "loss": 2.696093940734863, + "step": 64230 + }, + { + "epoch": 0.3212, + "grad_norm": 0.365803986787796, + "learning_rate": 0.00038768972004493615, + "loss": 2.779113006591797, + "step": 64240 + }, + { + "epoch": 0.32125, + "grad_norm": 0.3534572720527649, + "learning_rate": 0.00038765661003490665, + "loss": 2.765274238586426, + "step": 64250 + }, + { + "epoch": 0.3213, + "grad_norm": 0.33442848920822144, + "learning_rate": 0.0003876234965593729, + "loss": 2.7325918197631838, + "step": 64260 + }, + { + "epoch": 0.32135, + "grad_norm": 0.3195880651473999, + "learning_rate": 0.00038759037961916856, + "loss": 2.7791122436523437, + "step": 64270 + }, + { + "epoch": 0.3214, + "grad_norm": 0.33592090010643005, + "learning_rate": 0.0003875572592151273, + "loss": 2.759059715270996, + "step": 64280 + }, + { + "epoch": 0.32145, + "grad_norm": 0.361471563577652, + "learning_rate": 0.00038752413534808304, + "loss": 2.795364761352539, + "step": 64290 + }, + { + "epoch": 0.3215, + "grad_norm": 0.35272371768951416, + "learning_rate": 0.0003874910080188695, + "loss": 2.8151882171630858, + "step": 64300 + }, + { + "epoch": 0.32155, + "grad_norm": 0.3321019411087036, + "learning_rate": 0.0003874578772283208, + "loss": 2.8426721572875975, + "step": 64310 + }, + { + "epoch": 0.3216, + "grad_norm": 0.3397807478904724, + "learning_rate": 0.000387424742977271, + "loss": 2.757215881347656, + "step": 64320 + }, + { + "epoch": 0.32165, + "grad_norm": 0.3694498836994171, + "learning_rate": 0.00038739160526655417, + "loss": 2.72936954498291, + "step": 64330 + }, + { + "epoch": 0.3217, + "grad_norm": 0.32366684079170227, + "learning_rate": 0.00038735846409700456, + "loss": 2.7583124160766603, + "step": 64340 + }, + { + "epoch": 0.32175, + "grad_norm": 0.32947486639022827, + "learning_rate": 0.0003873253194694566, + "loss": 2.8296892166137697, + "step": 64350 + }, + { + "epoch": 0.3218, + "grad_norm": 0.3656548857688904, + "learning_rate": 0.0003872921713847446, + "loss": 2.749034118652344, + "step": 64360 + }, + { + "epoch": 0.32185, + "grad_norm": 0.3565869629383087, + "learning_rate": 0.0003872590198437031, + "loss": 2.76641902923584, + "step": 64370 + }, + { + "epoch": 0.3219, + "grad_norm": 0.34362056851387024, + "learning_rate": 0.00038722586484716664, + "loss": 2.8007171630859373, + "step": 64380 + }, + { + "epoch": 0.32195, + "grad_norm": 0.3057903051376343, + "learning_rate": 0.00038719270639597005, + "loss": 2.805160140991211, + "step": 64390 + }, + { + "epoch": 0.322, + "grad_norm": 0.36743345856666565, + "learning_rate": 0.000387159544490948, + "loss": 2.7945068359375, + "step": 64400 + }, + { + "epoch": 0.32205, + "grad_norm": 0.3428398370742798, + "learning_rate": 0.00038712637913293523, + "loss": 2.7169464111328123, + "step": 64410 + }, + { + "epoch": 0.3221, + "grad_norm": 0.3198021650314331, + "learning_rate": 0.0003870932103227669, + "loss": 2.814238929748535, + "step": 64420 + }, + { + "epoch": 0.32215, + "grad_norm": 0.3235197365283966, + "learning_rate": 0.0003870600380612778, + "loss": 2.773672103881836, + "step": 64430 + }, + { + "epoch": 0.3222, + "grad_norm": 0.3039498031139374, + "learning_rate": 0.0003870268623493033, + "loss": 2.7665216445922853, + "step": 64440 + }, + { + "epoch": 0.32225, + "grad_norm": 0.31825193762779236, + "learning_rate": 0.0003869936831876783, + "loss": 2.794822883605957, + "step": 64450 + }, + { + "epoch": 0.3223, + "grad_norm": 0.311152845621109, + "learning_rate": 0.0003869605005772383, + "loss": 2.8069229125976562, + "step": 64460 + }, + { + "epoch": 0.32235, + "grad_norm": 0.31502026319503784, + "learning_rate": 0.0003869273145188186, + "loss": 2.733692932128906, + "step": 64470 + }, + { + "epoch": 0.3224, + "grad_norm": 0.3061986565589905, + "learning_rate": 0.0003868941250132547, + "loss": 2.7771251678466795, + "step": 64480 + }, + { + "epoch": 0.32245, + "grad_norm": 0.35260632634162903, + "learning_rate": 0.00038686093206138217, + "loss": 2.746393394470215, + "step": 64490 + }, + { + "epoch": 0.3225, + "grad_norm": 0.3680865466594696, + "learning_rate": 0.0003868277356640365, + "loss": 2.744266319274902, + "step": 64500 + }, + { + "epoch": 0.32255, + "grad_norm": 0.3520456552505493, + "learning_rate": 0.00038679453582205354, + "loss": 2.7898597717285156, + "step": 64510 + }, + { + "epoch": 0.3226, + "grad_norm": 0.3939192295074463, + "learning_rate": 0.0003867613325362691, + "loss": 2.791429138183594, + "step": 64520 + }, + { + "epoch": 0.32265, + "grad_norm": 0.32327744364738464, + "learning_rate": 0.00038672812580751894, + "loss": 2.6786113739013673, + "step": 64530 + }, + { + "epoch": 0.3227, + "grad_norm": 0.351521372795105, + "learning_rate": 0.0003866949156366392, + "loss": 2.8416637420654296, + "step": 64540 + }, + { + "epoch": 0.32275, + "grad_norm": 0.3948894143104553, + "learning_rate": 0.00038666170202446574, + "loss": 2.786330223083496, + "step": 64550 + }, + { + "epoch": 0.3228, + "grad_norm": 0.31663307547569275, + "learning_rate": 0.000386628484971835, + "loss": 2.833728790283203, + "step": 64560 + }, + { + "epoch": 0.32285, + "grad_norm": 0.33457690477371216, + "learning_rate": 0.0003865952644795829, + "loss": 2.756263542175293, + "step": 64570 + }, + { + "epoch": 0.3229, + "grad_norm": 0.3631688356399536, + "learning_rate": 0.0003865620405485461, + "loss": 2.7607831954956055, + "step": 64580 + }, + { + "epoch": 0.32295, + "grad_norm": 0.3190194368362427, + "learning_rate": 0.00038652881317956077, + "loss": 2.7684675216674806, + "step": 64590 + }, + { + "epoch": 0.323, + "grad_norm": 0.352816104888916, + "learning_rate": 0.0003864955823734635, + "loss": 2.799749565124512, + "step": 64600 + }, + { + "epoch": 0.32305, + "grad_norm": 0.31599733233451843, + "learning_rate": 0.0003864623481310908, + "loss": 2.778061294555664, + "step": 64610 + }, + { + "epoch": 0.3231, + "grad_norm": 0.33751562237739563, + "learning_rate": 0.0003864291104532794, + "loss": 2.7780183792114257, + "step": 64620 + }, + { + "epoch": 0.32315, + "grad_norm": 0.31166955828666687, + "learning_rate": 0.0003863958693408661, + "loss": 2.7458372116088867, + "step": 64630 + }, + { + "epoch": 0.3232, + "grad_norm": 0.36339324712753296, + "learning_rate": 0.00038636262479468773, + "loss": 2.763485908508301, + "step": 64640 + }, + { + "epoch": 0.32325, + "grad_norm": 0.3296827971935272, + "learning_rate": 0.0003863293768155811, + "loss": 2.746242332458496, + "step": 64650 + }, + { + "epoch": 0.3233, + "grad_norm": 0.33209308981895447, + "learning_rate": 0.00038629612540438344, + "loss": 2.754735565185547, + "step": 64660 + }, + { + "epoch": 0.32335, + "grad_norm": 0.34701642394065857, + "learning_rate": 0.00038626287056193163, + "loss": 2.7255508422851564, + "step": 64670 + }, + { + "epoch": 0.3234, + "grad_norm": 0.39456385374069214, + "learning_rate": 0.00038622961228906307, + "loss": 2.823482322692871, + "step": 64680 + }, + { + "epoch": 0.32345, + "grad_norm": 0.33842733502388, + "learning_rate": 0.00038619635058661486, + "loss": 2.768875503540039, + "step": 64690 + }, + { + "epoch": 0.3235, + "grad_norm": 0.32360318303108215, + "learning_rate": 0.00038616308545542444, + "loss": 2.7518999099731447, + "step": 64700 + }, + { + "epoch": 0.32355, + "grad_norm": 0.3424636721611023, + "learning_rate": 0.0003861298168963292, + "loss": 2.7481386184692385, + "step": 64710 + }, + { + "epoch": 0.3236, + "grad_norm": 0.3400861620903015, + "learning_rate": 0.0003860965449101668, + "loss": 2.774835205078125, + "step": 64720 + }, + { + "epoch": 0.32365, + "grad_norm": 0.3345588743686676, + "learning_rate": 0.00038606326949777476, + "loss": 2.7960811614990235, + "step": 64730 + }, + { + "epoch": 0.3237, + "grad_norm": 0.3497892916202545, + "learning_rate": 0.0003860299906599909, + "loss": 2.7642143249511717, + "step": 64740 + }, + { + "epoch": 0.32375, + "grad_norm": 0.32274937629699707, + "learning_rate": 0.00038599670839765276, + "loss": 2.7824588775634767, + "step": 64750 + }, + { + "epoch": 0.3238, + "grad_norm": 0.3302581310272217, + "learning_rate": 0.00038596342271159854, + "loss": 2.7727678298950194, + "step": 64760 + }, + { + "epoch": 0.32385, + "grad_norm": 0.3420636057853699, + "learning_rate": 0.000385930133602666, + "loss": 2.767006683349609, + "step": 64770 + }, + { + "epoch": 0.3239, + "grad_norm": 0.32045260071754456, + "learning_rate": 0.0003858968410716933, + "loss": 2.797442626953125, + "step": 64780 + }, + { + "epoch": 0.32395, + "grad_norm": 0.3335878551006317, + "learning_rate": 0.0003858635451195185, + "loss": 2.742608070373535, + "step": 64790 + }, + { + "epoch": 0.324, + "grad_norm": 0.41320374608039856, + "learning_rate": 0.0003858302457469799, + "loss": 2.7634550094604493, + "step": 64800 + }, + { + "epoch": 0.32405, + "grad_norm": 0.3730308711528778, + "learning_rate": 0.00038579694295491575, + "loss": 2.760434150695801, + "step": 64810 + }, + { + "epoch": 0.3241, + "grad_norm": 0.30643409490585327, + "learning_rate": 0.00038576363674416446, + "loss": 2.777770233154297, + "step": 64820 + }, + { + "epoch": 0.32415, + "grad_norm": 0.33333954215049744, + "learning_rate": 0.00038573032711556443, + "loss": 2.7962514877319338, + "step": 64830 + }, + { + "epoch": 0.3242, + "grad_norm": 0.2996315360069275, + "learning_rate": 0.0003856970140699545, + "loss": 2.7390527725219727, + "step": 64840 + }, + { + "epoch": 0.32425, + "grad_norm": 0.303483784198761, + "learning_rate": 0.000385663697608173, + "loss": 2.7717456817626953, + "step": 64850 + }, + { + "epoch": 0.3243, + "grad_norm": 0.3089207410812378, + "learning_rate": 0.0003856303777310589, + "loss": 2.7447885513305663, + "step": 64860 + }, + { + "epoch": 0.32435, + "grad_norm": 0.3563207685947418, + "learning_rate": 0.00038559705443945094, + "loss": 2.7817804336547853, + "step": 64870 + }, + { + "epoch": 0.3244, + "grad_norm": 0.3413868546485901, + "learning_rate": 0.000385563727734188, + "loss": 2.8057342529296876, + "step": 64880 + }, + { + "epoch": 0.32445, + "grad_norm": 0.37615111470222473, + "learning_rate": 0.0003855303976161092, + "loss": 2.827165985107422, + "step": 64890 + }, + { + "epoch": 0.3245, + "grad_norm": 0.33022695779800415, + "learning_rate": 0.00038549706408605347, + "loss": 2.7483152389526366, + "step": 64900 + }, + { + "epoch": 0.32455, + "grad_norm": 0.36557406187057495, + "learning_rate": 0.00038546372714486006, + "loss": 2.7645803451538087, + "step": 64910 + }, + { + "epoch": 0.3246, + "grad_norm": 0.36303842067718506, + "learning_rate": 0.0003854303867933683, + "loss": 2.7807985305786134, + "step": 64920 + }, + { + "epoch": 0.32465, + "grad_norm": 0.3260742425918579, + "learning_rate": 0.00038539704303241736, + "loss": 2.758035659790039, + "step": 64930 + }, + { + "epoch": 0.3247, + "grad_norm": 0.32800889015197754, + "learning_rate": 0.0003853636958628468, + "loss": 2.843445587158203, + "step": 64940 + }, + { + "epoch": 0.32475, + "grad_norm": 0.3661607801914215, + "learning_rate": 0.0003853303452854961, + "loss": 2.777959442138672, + "step": 64950 + }, + { + "epoch": 0.3248, + "grad_norm": 0.31537434458732605, + "learning_rate": 0.0003852969913012049, + "loss": 2.7680376052856444, + "step": 64960 + }, + { + "epoch": 0.32485, + "grad_norm": 0.3444458544254303, + "learning_rate": 0.00038526363391081284, + "loss": 2.7554224014282225, + "step": 64970 + }, + { + "epoch": 0.3249, + "grad_norm": 0.334855318069458, + "learning_rate": 0.0003852302731151597, + "loss": 2.7575876235961916, + "step": 64980 + }, + { + "epoch": 0.32495, + "grad_norm": 0.3974142074584961, + "learning_rate": 0.00038519690891508526, + "loss": 2.798872375488281, + "step": 64990 + }, + { + "epoch": 0.325, + "grad_norm": 0.3328888416290283, + "learning_rate": 0.0003851635413114296, + "loss": 2.738381576538086, + "step": 65000 + }, + { + "epoch": 0.32505, + "grad_norm": 0.32193225622177124, + "learning_rate": 0.00038513017030503274, + "loss": 2.8414207458496095, + "step": 65010 + }, + { + "epoch": 0.3251, + "grad_norm": 0.32320600748062134, + "learning_rate": 0.0003850967958967346, + "loss": 2.75498104095459, + "step": 65020 + }, + { + "epoch": 0.32515, + "grad_norm": 0.3324619233608246, + "learning_rate": 0.00038506341808737564, + "loss": 2.7102941513061523, + "step": 65030 + }, + { + "epoch": 0.3252, + "grad_norm": 0.3363543450832367, + "learning_rate": 0.00038503003687779603, + "loss": 2.7860137939453127, + "step": 65040 + }, + { + "epoch": 0.32525, + "grad_norm": 0.33604860305786133, + "learning_rate": 0.00038499665226883614, + "loss": 2.7516592025756834, + "step": 65050 + }, + { + "epoch": 0.3253, + "grad_norm": 0.3429010212421417, + "learning_rate": 0.00038496326426133644, + "loss": 2.826551628112793, + "step": 65060 + }, + { + "epoch": 0.32535, + "grad_norm": 0.3614298403263092, + "learning_rate": 0.0003849298728561374, + "loss": 2.7904645919799806, + "step": 65070 + }, + { + "epoch": 0.3254, + "grad_norm": 0.35082489252090454, + "learning_rate": 0.0003848964780540798, + "loss": 2.724397659301758, + "step": 65080 + }, + { + "epoch": 0.32545, + "grad_norm": 0.3129132390022278, + "learning_rate": 0.00038486307985600423, + "loss": 2.7854074478149413, + "step": 65090 + }, + { + "epoch": 0.3255, + "grad_norm": 0.35317349433898926, + "learning_rate": 0.0003848296782627515, + "loss": 2.772166442871094, + "step": 65100 + }, + { + "epoch": 0.32555, + "grad_norm": 0.3338810205459595, + "learning_rate": 0.00038479627327516256, + "loss": 2.7111351013183596, + "step": 65110 + }, + { + "epoch": 0.3256, + "grad_norm": 0.32964614033699036, + "learning_rate": 0.0003847628648940783, + "loss": 2.7743602752685548, + "step": 65120 + }, + { + "epoch": 0.32565, + "grad_norm": 0.3226764500141144, + "learning_rate": 0.0003847294531203398, + "loss": 2.7297924041748045, + "step": 65130 + }, + { + "epoch": 0.3257, + "grad_norm": 0.32100290060043335, + "learning_rate": 0.00038469603795478826, + "loss": 2.713528060913086, + "step": 65140 + }, + { + "epoch": 0.32575, + "grad_norm": 0.34689414501190186, + "learning_rate": 0.00038466261939826485, + "loss": 2.7445293426513673, + "step": 65150 + }, + { + "epoch": 0.3258, + "grad_norm": 0.3077094852924347, + "learning_rate": 0.00038462919745161094, + "loss": 2.732035446166992, + "step": 65160 + }, + { + "epoch": 0.32585, + "grad_norm": 0.3193971812725067, + "learning_rate": 0.00038459577211566784, + "loss": 2.753880500793457, + "step": 65170 + }, + { + "epoch": 0.3259, + "grad_norm": 0.3734096884727478, + "learning_rate": 0.000384562343391277, + "loss": 2.764912223815918, + "step": 65180 + }, + { + "epoch": 0.32595, + "grad_norm": 0.381599098443985, + "learning_rate": 0.00038452891127928025, + "loss": 2.7564348220825194, + "step": 65190 + }, + { + "epoch": 0.326, + "grad_norm": 0.3238396644592285, + "learning_rate": 0.00038449547578051893, + "loss": 2.764679718017578, + "step": 65200 + }, + { + "epoch": 0.32605, + "grad_norm": 0.3701344132423401, + "learning_rate": 0.00038446203689583493, + "loss": 2.827960395812988, + "step": 65210 + }, + { + "epoch": 0.3261, + "grad_norm": 0.3227131962776184, + "learning_rate": 0.00038442859462607004, + "loss": 2.8021190643310545, + "step": 65220 + }, + { + "epoch": 0.32615, + "grad_norm": 0.3191877603530884, + "learning_rate": 0.0003843951489720662, + "loss": 2.7752246856689453, + "step": 65230 + }, + { + "epoch": 0.3262, + "grad_norm": 0.33469536900520325, + "learning_rate": 0.0003843616999346654, + "loss": 2.7673450469970704, + "step": 65240 + }, + { + "epoch": 0.32625, + "grad_norm": 0.35151800513267517, + "learning_rate": 0.00038432824751470965, + "loss": 2.763423538208008, + "step": 65250 + }, + { + "epoch": 0.3263, + "grad_norm": 0.30435389280319214, + "learning_rate": 0.00038429479171304113, + "loss": 2.7722091674804688, + "step": 65260 + }, + { + "epoch": 0.32635, + "grad_norm": 0.3571937680244446, + "learning_rate": 0.00038426133253050227, + "loss": 2.7964685440063475, + "step": 65270 + }, + { + "epoch": 0.3264, + "grad_norm": 0.32652702927589417, + "learning_rate": 0.00038422786996793515, + "loss": 2.794063377380371, + "step": 65280 + }, + { + "epoch": 0.32645, + "grad_norm": 0.354148805141449, + "learning_rate": 0.00038419440402618233, + "loss": 2.799833869934082, + "step": 65290 + }, + { + "epoch": 0.3265, + "grad_norm": 0.33977243304252625, + "learning_rate": 0.0003841609347060863, + "loss": 2.78237361907959, + "step": 65300 + }, + { + "epoch": 0.32655, + "grad_norm": 0.37021005153656006, + "learning_rate": 0.0003841274620084896, + "loss": 2.7971187591552735, + "step": 65310 + }, + { + "epoch": 0.3266, + "grad_norm": 0.34874090552330017, + "learning_rate": 0.0003840939859342349, + "loss": 2.797547149658203, + "step": 65320 + }, + { + "epoch": 0.32665, + "grad_norm": 0.3607078790664673, + "learning_rate": 0.00038406050648416513, + "loss": 2.784065055847168, + "step": 65330 + }, + { + "epoch": 0.3267, + "grad_norm": 0.33140841126441956, + "learning_rate": 0.0003840270236591229, + "loss": 2.7397769927978515, + "step": 65340 + }, + { + "epoch": 0.32675, + "grad_norm": 0.327230840921402, + "learning_rate": 0.00038399353745995134, + "loss": 2.739827537536621, + "step": 65350 + }, + { + "epoch": 0.3268, + "grad_norm": 0.3082130253314972, + "learning_rate": 0.00038396004788749326, + "loss": 2.7822053909301756, + "step": 65360 + }, + { + "epoch": 0.32685, + "grad_norm": 0.33485147356987, + "learning_rate": 0.0003839265549425919, + "loss": 2.7973352432250977, + "step": 65370 + }, + { + "epoch": 0.3269, + "grad_norm": 0.31974348425865173, + "learning_rate": 0.0003838930586260905, + "loss": 2.7661922454833983, + "step": 65380 + }, + { + "epoch": 0.32695, + "grad_norm": 0.3306097984313965, + "learning_rate": 0.0003838595589388322, + "loss": 2.754362678527832, + "step": 65390 + }, + { + "epoch": 0.327, + "grad_norm": 0.38620632886886597, + "learning_rate": 0.0003838260558816603, + "loss": 2.733713722229004, + "step": 65400 + }, + { + "epoch": 0.32705, + "grad_norm": 0.37392041087150574, + "learning_rate": 0.00038379254945541843, + "loss": 2.8242462158203123, + "step": 65410 + }, + { + "epoch": 0.3271, + "grad_norm": 0.31185516715049744, + "learning_rate": 0.00038375903966094994, + "loss": 2.7433259963989256, + "step": 65420 + }, + { + "epoch": 0.32715, + "grad_norm": 0.31327539682388306, + "learning_rate": 0.0003837255264990986, + "loss": 2.7677986145019533, + "step": 65430 + }, + { + "epoch": 0.3272, + "grad_norm": 0.3622407615184784, + "learning_rate": 0.00038369200997070805, + "loss": 2.7809167861938477, + "step": 65440 + }, + { + "epoch": 0.32725, + "grad_norm": 0.33992674946784973, + "learning_rate": 0.000383658490076622, + "loss": 2.7492713928222656, + "step": 65450 + }, + { + "epoch": 0.3273, + "grad_norm": 0.35646238923072815, + "learning_rate": 0.00038362496681768434, + "loss": 2.716362953186035, + "step": 65460 + }, + { + "epoch": 0.32735, + "grad_norm": 0.33589187264442444, + "learning_rate": 0.000383591440194739, + "loss": 2.7259376525878904, + "step": 65470 + }, + { + "epoch": 0.3274, + "grad_norm": 0.31121572852134705, + "learning_rate": 0.0003835579102086302, + "loss": 2.780527687072754, + "step": 65480 + }, + { + "epoch": 0.32745, + "grad_norm": 0.3160688877105713, + "learning_rate": 0.0003835243768602018, + "loss": 2.7784469604492186, + "step": 65490 + }, + { + "epoch": 0.3275, + "grad_norm": 0.32647472620010376, + "learning_rate": 0.000383490840150298, + "loss": 2.804883575439453, + "step": 65500 + }, + { + "epoch": 0.32755, + "grad_norm": 0.3904622793197632, + "learning_rate": 0.00038345730007976336, + "loss": 2.761548614501953, + "step": 65510 + }, + { + "epoch": 0.3276, + "grad_norm": 0.33672741055488586, + "learning_rate": 0.00038342375664944203, + "loss": 2.7038726806640625, + "step": 65520 + }, + { + "epoch": 0.32765, + "grad_norm": 0.3195820152759552, + "learning_rate": 0.0003833902098601785, + "loss": 2.732843017578125, + "step": 65530 + }, + { + "epoch": 0.3277, + "grad_norm": 0.33435186743736267, + "learning_rate": 0.0003833566597128173, + "loss": 2.7792186737060547, + "step": 65540 + }, + { + "epoch": 0.32775, + "grad_norm": 0.3350854814052582, + "learning_rate": 0.0003833231062082031, + "loss": 2.763810729980469, + "step": 65550 + }, + { + "epoch": 0.3278, + "grad_norm": 0.34625333547592163, + "learning_rate": 0.00038328954934718063, + "loss": 2.7909267425537108, + "step": 65560 + }, + { + "epoch": 0.32785, + "grad_norm": 0.33355769515037537, + "learning_rate": 0.0003832559891305947, + "loss": 2.7732288360595705, + "step": 65570 + }, + { + "epoch": 0.3279, + "grad_norm": 0.3729526996612549, + "learning_rate": 0.0003832224255592899, + "loss": 2.767386054992676, + "step": 65580 + }, + { + "epoch": 0.32795, + "grad_norm": 0.35713455080986023, + "learning_rate": 0.0003831888586341117, + "loss": 2.7421993255615233, + "step": 65590 + }, + { + "epoch": 0.328, + "grad_norm": 0.3565027415752411, + "learning_rate": 0.00038315528835590473, + "loss": 2.736357307434082, + "step": 65600 + }, + { + "epoch": 0.32805, + "grad_norm": 0.3522573411464691, + "learning_rate": 0.0003831217147255143, + "loss": 2.7935415267944337, + "step": 65610 + }, + { + "epoch": 0.3281, + "grad_norm": 0.38144543766975403, + "learning_rate": 0.0003830881377437856, + "loss": 2.783716583251953, + "step": 65620 + }, + { + "epoch": 0.32815, + "grad_norm": 0.3497493863105774, + "learning_rate": 0.0003830545574115639, + "loss": 2.7454044342041017, + "step": 65630 + }, + { + "epoch": 0.3282, + "grad_norm": 0.36095723509788513, + "learning_rate": 0.00038302097372969456, + "loss": 2.7260021209716796, + "step": 65640 + }, + { + "epoch": 0.32825, + "grad_norm": 0.3504488468170166, + "learning_rate": 0.0003829873866990232, + "loss": 2.7256921768188476, + "step": 65650 + }, + { + "epoch": 0.3283, + "grad_norm": 0.33529940247535706, + "learning_rate": 0.00038295379632039515, + "loss": 2.750411033630371, + "step": 65660 + }, + { + "epoch": 0.32835, + "grad_norm": 0.3410096764564514, + "learning_rate": 0.00038292020259465625, + "loss": 2.771929931640625, + "step": 65670 + }, + { + "epoch": 0.3284, + "grad_norm": 0.3844609558582306, + "learning_rate": 0.00038288660552265207, + "loss": 2.778542709350586, + "step": 65680 + }, + { + "epoch": 0.32845, + "grad_norm": 0.3321518301963806, + "learning_rate": 0.0003828530051052286, + "loss": 2.7546615600585938, + "step": 65690 + }, + { + "epoch": 0.3285, + "grad_norm": 0.3416709303855896, + "learning_rate": 0.00038281940134323155, + "loss": 2.748508262634277, + "step": 65700 + }, + { + "epoch": 0.32855, + "grad_norm": 0.33216530084609985, + "learning_rate": 0.000382785794237507, + "loss": 2.7535396575927735, + "step": 65710 + }, + { + "epoch": 0.3286, + "grad_norm": 0.3634127974510193, + "learning_rate": 0.00038275218378890086, + "loss": 2.7480083465576173, + "step": 65720 + }, + { + "epoch": 0.32865, + "grad_norm": 0.33731141686439514, + "learning_rate": 0.00038271856999825947, + "loss": 2.805653953552246, + "step": 65730 + }, + { + "epoch": 0.3287, + "grad_norm": 0.333944171667099, + "learning_rate": 0.0003826849528664289, + "loss": 2.7616445541381838, + "step": 65740 + }, + { + "epoch": 0.32875, + "grad_norm": 0.32692375779151917, + "learning_rate": 0.0003826513323942556, + "loss": 2.820975685119629, + "step": 65750 + }, + { + "epoch": 0.3288, + "grad_norm": 0.2958635985851288, + "learning_rate": 0.0003826177085825858, + "loss": 2.76717472076416, + "step": 65760 + }, + { + "epoch": 0.32885, + "grad_norm": 0.314067542552948, + "learning_rate": 0.0003825840814322662, + "loss": 2.79183349609375, + "step": 65770 + }, + { + "epoch": 0.3289, + "grad_norm": 0.3447400629520416, + "learning_rate": 0.00038255045094414314, + "loss": 2.767874336242676, + "step": 65780 + }, + { + "epoch": 0.32895, + "grad_norm": 0.35122445225715637, + "learning_rate": 0.00038251681711906345, + "loss": 2.7849267959594726, + "step": 65790 + }, + { + "epoch": 0.329, + "grad_norm": 0.3289494514465332, + "learning_rate": 0.0003824831799578736, + "loss": 2.761005973815918, + "step": 65800 + }, + { + "epoch": 0.32905, + "grad_norm": 0.3350103199481964, + "learning_rate": 0.00038244953946142083, + "loss": 2.745119094848633, + "step": 65810 + }, + { + "epoch": 0.3291, + "grad_norm": 0.3174055814743042, + "learning_rate": 0.0003824158956305516, + "loss": 2.7599584579467775, + "step": 65820 + }, + { + "epoch": 0.32915, + "grad_norm": 0.31591203808784485, + "learning_rate": 0.00038238224846611315, + "loss": 2.8056459426879883, + "step": 65830 + }, + { + "epoch": 0.3292, + "grad_norm": 0.3183386027812958, + "learning_rate": 0.0003823485979689524, + "loss": 2.7389787673950194, + "step": 65840 + }, + { + "epoch": 0.32925, + "grad_norm": 0.33982720971107483, + "learning_rate": 0.00038231494413991666, + "loss": 2.773180389404297, + "step": 65850 + }, + { + "epoch": 0.3293, + "grad_norm": 0.37457790970802307, + "learning_rate": 0.00038228128697985307, + "loss": 2.754871368408203, + "step": 65860 + }, + { + "epoch": 0.32935, + "grad_norm": 0.3488258719444275, + "learning_rate": 0.00038224762648960905, + "loss": 2.7770503997802733, + "step": 65870 + }, + { + "epoch": 0.3294, + "grad_norm": 0.359223872423172, + "learning_rate": 0.00038221396267003175, + "loss": 2.7622920989990236, + "step": 65880 + }, + { + "epoch": 0.32945, + "grad_norm": 0.3419441878795624, + "learning_rate": 0.00038218029552196896, + "loss": 2.7199541091918946, + "step": 65890 + }, + { + "epoch": 0.3295, + "grad_norm": 0.33414286375045776, + "learning_rate": 0.0003821466250462681, + "loss": 2.721760940551758, + "step": 65900 + }, + { + "epoch": 0.32955, + "grad_norm": 0.32452356815338135, + "learning_rate": 0.00038211295124377677, + "loss": 2.8015186309814455, + "step": 65910 + }, + { + "epoch": 0.3296, + "grad_norm": 0.3381940722465515, + "learning_rate": 0.00038207927411534285, + "loss": 2.842654800415039, + "step": 65920 + }, + { + "epoch": 0.32965, + "grad_norm": 0.3335150480270386, + "learning_rate": 0.00038204559366181395, + "loss": 2.717409133911133, + "step": 65930 + }, + { + "epoch": 0.3297, + "grad_norm": 0.3230106830596924, + "learning_rate": 0.0003820119098840382, + "loss": 2.756159019470215, + "step": 65940 + }, + { + "epoch": 0.32975, + "grad_norm": 0.33958011865615845, + "learning_rate": 0.00038197822278286355, + "loss": 2.7506284713745117, + "step": 65950 + }, + { + "epoch": 0.3298, + "grad_norm": 0.32457271218299866, + "learning_rate": 0.00038194453235913795, + "loss": 2.697976303100586, + "step": 65960 + }, + { + "epoch": 0.32985, + "grad_norm": 0.4098268151283264, + "learning_rate": 0.0003819108386137097, + "loss": 2.731293869018555, + "step": 65970 + }, + { + "epoch": 0.3299, + "grad_norm": 0.33788952231407166, + "learning_rate": 0.0003818771415474269, + "loss": 2.776759719848633, + "step": 65980 + }, + { + "epoch": 0.32995, + "grad_norm": 0.34169474244117737, + "learning_rate": 0.000381843441161138, + "loss": 2.7094234466552733, + "step": 65990 + }, + { + "epoch": 0.33, + "grad_norm": 0.33323609828948975, + "learning_rate": 0.00038180973745569123, + "loss": 2.8017900466918944, + "step": 66000 + }, + { + "epoch": 0.33005, + "grad_norm": 0.30861058831214905, + "learning_rate": 0.00038177603043193536, + "loss": 2.727107810974121, + "step": 66010 + }, + { + "epoch": 0.3301, + "grad_norm": 0.31691229343414307, + "learning_rate": 0.00038174232009071876, + "loss": 2.739047050476074, + "step": 66020 + }, + { + "epoch": 0.33015, + "grad_norm": 0.3021847903728485, + "learning_rate": 0.0003817086064328901, + "loss": 2.7904382705688477, + "step": 66030 + }, + { + "epoch": 0.3302, + "grad_norm": 0.3634334206581116, + "learning_rate": 0.0003816748894592982, + "loss": 2.724009323120117, + "step": 66040 + }, + { + "epoch": 0.33025, + "grad_norm": 0.3294525742530823, + "learning_rate": 0.00038164116917079174, + "loss": 2.7524065017700194, + "step": 66050 + }, + { + "epoch": 0.3303, + "grad_norm": 0.32030683755874634, + "learning_rate": 0.0003816074455682198, + "loss": 2.76763801574707, + "step": 66060 + }, + { + "epoch": 0.33035, + "grad_norm": 0.3118949830532074, + "learning_rate": 0.00038157371865243125, + "loss": 2.7654302597045897, + "step": 66070 + }, + { + "epoch": 0.3304, + "grad_norm": 0.33959725499153137, + "learning_rate": 0.00038153998842427523, + "loss": 2.7993967056274416, + "step": 66080 + }, + { + "epoch": 0.33045, + "grad_norm": 0.3438769280910492, + "learning_rate": 0.0003815062548846009, + "loss": 2.7501527786254885, + "step": 66090 + }, + { + "epoch": 0.3305, + "grad_norm": 0.324328750371933, + "learning_rate": 0.0003814725180342575, + "loss": 2.7667778015136717, + "step": 66100 + }, + { + "epoch": 0.33055, + "grad_norm": 0.328125, + "learning_rate": 0.0003814387778740943, + "loss": 2.770594024658203, + "step": 66110 + }, + { + "epoch": 0.3306, + "grad_norm": 0.3244064748287201, + "learning_rate": 0.0003814050344049608, + "loss": 2.766187858581543, + "step": 66120 + }, + { + "epoch": 0.33065, + "grad_norm": 0.3421144187450409, + "learning_rate": 0.00038137128762770635, + "loss": 2.734879493713379, + "step": 66130 + }, + { + "epoch": 0.3307, + "grad_norm": 0.37592780590057373, + "learning_rate": 0.0003813375375431807, + "loss": 2.829850769042969, + "step": 66140 + }, + { + "epoch": 0.33075, + "grad_norm": 0.31622570753097534, + "learning_rate": 0.0003813037841522333, + "loss": 2.7707889556884764, + "step": 66150 + }, + { + "epoch": 0.3308, + "grad_norm": 0.30575698614120483, + "learning_rate": 0.00038127002745571415, + "loss": 2.7724782943725588, + "step": 66160 + }, + { + "epoch": 0.33085, + "grad_norm": 0.32749491930007935, + "learning_rate": 0.0003812362674544728, + "loss": 2.7945085525512696, + "step": 66170 + }, + { + "epoch": 0.3309, + "grad_norm": 0.3500199317932129, + "learning_rate": 0.00038120250414935944, + "loss": 2.7375820159912108, + "step": 66180 + }, + { + "epoch": 0.33095, + "grad_norm": 0.3678411841392517, + "learning_rate": 0.0003811687375412238, + "loss": 2.7433916091918946, + "step": 66190 + }, + { + "epoch": 0.331, + "grad_norm": 0.30628910660743713, + "learning_rate": 0.0003811349676309161, + "loss": 2.748147392272949, + "step": 66200 + }, + { + "epoch": 0.33105, + "grad_norm": 0.33354926109313965, + "learning_rate": 0.00038110119441928645, + "loss": 2.760938262939453, + "step": 66210 + }, + { + "epoch": 0.3311, + "grad_norm": 0.3619321882724762, + "learning_rate": 0.00038106741790718516, + "loss": 2.8260780334472657, + "step": 66220 + }, + { + "epoch": 0.33115, + "grad_norm": 0.4632556736469269, + "learning_rate": 0.00038103363809546245, + "loss": 2.755206489562988, + "step": 66230 + }, + { + "epoch": 0.3312, + "grad_norm": 0.33106091618537903, + "learning_rate": 0.0003809998549849688, + "loss": 2.826741600036621, + "step": 66240 + }, + { + "epoch": 0.33125, + "grad_norm": 0.33064475655555725, + "learning_rate": 0.0003809660685765547, + "loss": 2.8046632766723634, + "step": 66250 + }, + { + "epoch": 0.3313, + "grad_norm": 0.36323046684265137, + "learning_rate": 0.00038093227887107066, + "loss": 2.7682235717773436, + "step": 66260 + }, + { + "epoch": 0.33135, + "grad_norm": 0.31096842885017395, + "learning_rate": 0.0003808984858693674, + "loss": 2.7799566268920897, + "step": 66270 + }, + { + "epoch": 0.3314, + "grad_norm": 0.3709983229637146, + "learning_rate": 0.0003808646895722957, + "loss": 2.6900089263916014, + "step": 66280 + }, + { + "epoch": 0.33145, + "grad_norm": 0.3157351613044739, + "learning_rate": 0.0003808308899807061, + "loss": 2.7532346725463865, + "step": 66290 + }, + { + "epoch": 0.3315, + "grad_norm": 0.3361470401287079, + "learning_rate": 0.00038079708709544993, + "loss": 2.769987869262695, + "step": 66300 + }, + { + "epoch": 0.33155, + "grad_norm": 0.32864221930503845, + "learning_rate": 0.0003807632809173779, + "loss": 2.760721969604492, + "step": 66310 + }, + { + "epoch": 0.3316, + "grad_norm": 0.36235785484313965, + "learning_rate": 0.00038072947144734113, + "loss": 2.717623710632324, + "step": 66320 + }, + { + "epoch": 0.33165, + "grad_norm": 0.3350370228290558, + "learning_rate": 0.0003806956586861908, + "loss": 2.7419927597045897, + "step": 66330 + }, + { + "epoch": 0.3317, + "grad_norm": 0.3299534022808075, + "learning_rate": 0.0003806618426347781, + "loss": 2.7260725021362306, + "step": 66340 + }, + { + "epoch": 0.33175, + "grad_norm": 0.3377462327480316, + "learning_rate": 0.00038062802329395443, + "loss": 2.7246849060058596, + "step": 66350 + }, + { + "epoch": 0.3318, + "grad_norm": 0.3599812686443329, + "learning_rate": 0.00038059420066457116, + "loss": 2.8469472885131837, + "step": 66360 + }, + { + "epoch": 0.33185, + "grad_norm": 0.34460964798927307, + "learning_rate": 0.0003805603747474797, + "loss": 2.785201644897461, + "step": 66370 + }, + { + "epoch": 0.3319, + "grad_norm": 0.38252949714660645, + "learning_rate": 0.00038052654554353174, + "loss": 2.7693628311157226, + "step": 66380 + }, + { + "epoch": 0.33195, + "grad_norm": 0.3350801467895508, + "learning_rate": 0.0003804927130535788, + "loss": 2.8426576614379884, + "step": 66390 + }, + { + "epoch": 0.332, + "grad_norm": 0.32731807231903076, + "learning_rate": 0.0003804588772784727, + "loss": 2.7441707611083985, + "step": 66400 + }, + { + "epoch": 0.33205, + "grad_norm": 0.37033042311668396, + "learning_rate": 0.0003804250382190653, + "loss": 2.7582494735717775, + "step": 66410 + }, + { + "epoch": 0.3321, + "grad_norm": 0.33528822660446167, + "learning_rate": 0.0003803911958762084, + "loss": 2.745726776123047, + "step": 66420 + }, + { + "epoch": 0.33215, + "grad_norm": 0.3062208592891693, + "learning_rate": 0.000380357350250754, + "loss": 2.820939826965332, + "step": 66430 + }, + { + "epoch": 0.3322, + "grad_norm": 0.3209203779697418, + "learning_rate": 0.0003803235013435542, + "loss": 2.762510871887207, + "step": 66440 + }, + { + "epoch": 0.33225, + "grad_norm": 0.35359254479408264, + "learning_rate": 0.00038028964915546107, + "loss": 2.7029888153076174, + "step": 66450 + }, + { + "epoch": 0.3323, + "grad_norm": 0.3490917682647705, + "learning_rate": 0.000380255793687327, + "loss": 2.783255958557129, + "step": 66460 + }, + { + "epoch": 0.33235, + "grad_norm": 0.34160658717155457, + "learning_rate": 0.00038022193494000414, + "loss": 2.7688421249389648, + "step": 66470 + }, + { + "epoch": 0.3324, + "grad_norm": 0.32139089703559875, + "learning_rate": 0.00038018807291434486, + "loss": 2.7528663635253907, + "step": 66480 + }, + { + "epoch": 0.33245, + "grad_norm": 0.32945001125335693, + "learning_rate": 0.0003801542076112018, + "loss": 2.794088935852051, + "step": 66490 + }, + { + "epoch": 0.3325, + "grad_norm": 0.32645490765571594, + "learning_rate": 0.0003801203390314274, + "loss": 2.773613166809082, + "step": 66500 + }, + { + "epoch": 0.33255, + "grad_norm": 0.3298453390598297, + "learning_rate": 0.00038008646717587434, + "loss": 2.788191795349121, + "step": 66510 + }, + { + "epoch": 0.3326, + "grad_norm": 0.32300329208374023, + "learning_rate": 0.0003800525920453954, + "loss": 2.7723567962646483, + "step": 66520 + }, + { + "epoch": 0.33265, + "grad_norm": 0.38799726963043213, + "learning_rate": 0.0003800187136408432, + "loss": 2.764381217956543, + "step": 66530 + }, + { + "epoch": 0.3327, + "grad_norm": 0.36572596430778503, + "learning_rate": 0.00037998483196307085, + "loss": 2.756684112548828, + "step": 66540 + }, + { + "epoch": 0.33275, + "grad_norm": 0.39156273007392883, + "learning_rate": 0.00037995094701293123, + "loss": 2.806692886352539, + "step": 66550 + }, + { + "epoch": 0.3328, + "grad_norm": 0.34786537289619446, + "learning_rate": 0.00037991705879127736, + "loss": 2.8186235427856445, + "step": 66560 + }, + { + "epoch": 0.33285, + "grad_norm": 0.33802032470703125, + "learning_rate": 0.0003798831672989623, + "loss": 2.735603141784668, + "step": 66570 + }, + { + "epoch": 0.3329, + "grad_norm": 0.31932833790779114, + "learning_rate": 0.00037984927253683953, + "loss": 2.745252227783203, + "step": 66580 + }, + { + "epoch": 0.33295, + "grad_norm": 0.3170626759529114, + "learning_rate": 0.0003798153745057621, + "loss": 2.7214305877685545, + "step": 66590 + }, + { + "epoch": 0.333, + "grad_norm": 0.3457740545272827, + "learning_rate": 0.00037978147320658355, + "loss": 2.7639373779296874, + "step": 66600 + }, + { + "epoch": 0.33305, + "grad_norm": 0.35097259283065796, + "learning_rate": 0.00037974756864015725, + "loss": 2.7766693115234373, + "step": 66610 + }, + { + "epoch": 0.3331, + "grad_norm": 0.35351020097732544, + "learning_rate": 0.00037971366080733675, + "loss": 2.730050468444824, + "step": 66620 + }, + { + "epoch": 0.33315, + "grad_norm": 0.3499346375465393, + "learning_rate": 0.00037967974970897574, + "loss": 2.7373767852783204, + "step": 66630 + }, + { + "epoch": 0.3332, + "grad_norm": 0.3785187900066376, + "learning_rate": 0.0003796458353459279, + "loss": 2.722113037109375, + "step": 66640 + }, + { + "epoch": 0.33325, + "grad_norm": 0.359997034072876, + "learning_rate": 0.00037961191771904694, + "loss": 2.7595748901367188, + "step": 66650 + }, + { + "epoch": 0.3333, + "grad_norm": 0.3323831558227539, + "learning_rate": 0.00037957799682918693, + "loss": 2.7607553482055662, + "step": 66660 + }, + { + "epoch": 0.33335, + "grad_norm": 0.30642130970954895, + "learning_rate": 0.00037954407267720166, + "loss": 2.827049446105957, + "step": 66670 + }, + { + "epoch": 0.3334, + "grad_norm": 0.36628973484039307, + "learning_rate": 0.00037951014526394524, + "loss": 2.814953422546387, + "step": 66680 + }, + { + "epoch": 0.33345, + "grad_norm": 0.3419228494167328, + "learning_rate": 0.00037947621459027183, + "loss": 2.749830627441406, + "step": 66690 + }, + { + "epoch": 0.3335, + "grad_norm": 0.3308188021183014, + "learning_rate": 0.00037944228065703544, + "loss": 2.7793996810913084, + "step": 66700 + }, + { + "epoch": 0.33355, + "grad_norm": 0.33067312836647034, + "learning_rate": 0.00037940834346509065, + "loss": 2.8290584564208983, + "step": 66710 + }, + { + "epoch": 0.3336, + "grad_norm": 0.30863815546035767, + "learning_rate": 0.0003793744030152916, + "loss": 2.763117027282715, + "step": 66720 + }, + { + "epoch": 0.33365, + "grad_norm": 0.34335851669311523, + "learning_rate": 0.0003793404593084928, + "loss": 2.7766098022460937, + "step": 66730 + }, + { + "epoch": 0.3337, + "grad_norm": 0.3083382546901703, + "learning_rate": 0.00037930651234554884, + "loss": 2.7567302703857424, + "step": 66740 + }, + { + "epoch": 0.33375, + "grad_norm": 0.3458162546157837, + "learning_rate": 0.00037927256212731426, + "loss": 2.7314994812011717, + "step": 66750 + }, + { + "epoch": 0.3338, + "grad_norm": 0.34813299775123596, + "learning_rate": 0.0003792386086546439, + "loss": 2.7449466705322267, + "step": 66760 + }, + { + "epoch": 0.33385, + "grad_norm": 0.32204148173332214, + "learning_rate": 0.00037920465192839225, + "loss": 2.7635944366455076, + "step": 66770 + }, + { + "epoch": 0.3339, + "grad_norm": 0.3531690835952759, + "learning_rate": 0.0003791706919494145, + "loss": 2.771036720275879, + "step": 66780 + }, + { + "epoch": 0.33395, + "grad_norm": 0.3723186254501343, + "learning_rate": 0.00037913672871856544, + "loss": 2.745263671875, + "step": 66790 + }, + { + "epoch": 0.334, + "grad_norm": 0.3112393915653229, + "learning_rate": 0.0003791027622367, + "loss": 2.7658016204833986, + "step": 66800 + }, + { + "epoch": 0.33405, + "grad_norm": 0.34491658210754395, + "learning_rate": 0.0003790687925046734, + "loss": 2.757438850402832, + "step": 66810 + }, + { + "epoch": 0.3341, + "grad_norm": 0.3890240788459778, + "learning_rate": 0.0003790348195233409, + "loss": 2.6997091293334963, + "step": 66820 + }, + { + "epoch": 0.33415, + "grad_norm": 0.3280429542064667, + "learning_rate": 0.00037900084329355754, + "loss": 2.7758726119995116, + "step": 66830 + }, + { + "epoch": 0.3342, + "grad_norm": 0.34446340799331665, + "learning_rate": 0.00037896686381617896, + "loss": 2.7635698318481445, + "step": 66840 + }, + { + "epoch": 0.33425, + "grad_norm": 0.3202959895133972, + "learning_rate": 0.0003789328810920604, + "loss": 2.7434112548828127, + "step": 66850 + }, + { + "epoch": 0.3343, + "grad_norm": 0.36047306656837463, + "learning_rate": 0.00037889889512205744, + "loss": 2.7296588897705076, + "step": 66860 + }, + { + "epoch": 0.33435, + "grad_norm": 0.3326652944087982, + "learning_rate": 0.0003788649059070256, + "loss": 2.821704864501953, + "step": 66870 + }, + { + "epoch": 0.3344, + "grad_norm": 0.3210088014602661, + "learning_rate": 0.0003788309134478206, + "loss": 2.735629081726074, + "step": 66880 + }, + { + "epoch": 0.33445, + "grad_norm": 0.3240206837654114, + "learning_rate": 0.00037879691774529825, + "loss": 2.7213903427124024, + "step": 66890 + }, + { + "epoch": 0.3345, + "grad_norm": 0.3187462389469147, + "learning_rate": 0.0003787629188003144, + "loss": 2.7423763275146484, + "step": 66900 + }, + { + "epoch": 0.33455, + "grad_norm": 0.35875508189201355, + "learning_rate": 0.00037872891661372486, + "loss": 2.8023553848266602, + "step": 66910 + }, + { + "epoch": 0.3346, + "grad_norm": 0.33878180384635925, + "learning_rate": 0.0003786949111863858, + "loss": 2.7749452590942383, + "step": 66920 + }, + { + "epoch": 0.33465, + "grad_norm": 0.35876336693763733, + "learning_rate": 0.00037866090251915306, + "loss": 2.770898628234863, + "step": 66930 + }, + { + "epoch": 0.3347, + "grad_norm": 0.4195384979248047, + "learning_rate": 0.00037862689061288305, + "loss": 2.7422063827514647, + "step": 66940 + }, + { + "epoch": 0.33475, + "grad_norm": 0.3367250859737396, + "learning_rate": 0.000378592875468432, + "loss": 2.7957380294799803, + "step": 66950 + }, + { + "epoch": 0.3348, + "grad_norm": 0.32300466299057007, + "learning_rate": 0.00037855885708665607, + "loss": 2.7427770614624025, + "step": 66960 + }, + { + "epoch": 0.33485, + "grad_norm": 0.3191125690937042, + "learning_rate": 0.0003785248354684118, + "loss": 2.7620624542236327, + "step": 66970 + }, + { + "epoch": 0.3349, + "grad_norm": 0.3299277424812317, + "learning_rate": 0.0003784908106145556, + "loss": 2.786029815673828, + "step": 66980 + }, + { + "epoch": 0.33495, + "grad_norm": 0.3614889979362488, + "learning_rate": 0.00037845678252594416, + "loss": 2.7481075286865235, + "step": 66990 + }, + { + "epoch": 0.335, + "grad_norm": 0.33294549584388733, + "learning_rate": 0.00037842275120343404, + "loss": 2.7665988922119142, + "step": 67000 + }, + { + "epoch": 0.33505, + "grad_norm": 0.38477402925491333, + "learning_rate": 0.0003783887166478821, + "loss": 2.730635070800781, + "step": 67010 + }, + { + "epoch": 0.3351, + "grad_norm": 0.32201606035232544, + "learning_rate": 0.000378354678860145, + "loss": 2.7869958877563477, + "step": 67020 + }, + { + "epoch": 0.33515, + "grad_norm": 0.37600526213645935, + "learning_rate": 0.00037832063784107973, + "loss": 2.761496734619141, + "step": 67030 + }, + { + "epoch": 0.3352, + "grad_norm": 0.31691840291023254, + "learning_rate": 0.00037828659359154324, + "loss": 2.7779872894287108, + "step": 67040 + }, + { + "epoch": 0.33525, + "grad_norm": 0.3631610870361328, + "learning_rate": 0.0003782525461123927, + "loss": 2.7342193603515623, + "step": 67050 + }, + { + "epoch": 0.3353, + "grad_norm": 0.3663494288921356, + "learning_rate": 0.00037821849540448506, + "loss": 2.7621952056884767, + "step": 67060 + }, + { + "epoch": 0.33535, + "grad_norm": 0.3687584698200226, + "learning_rate": 0.0003781844414686777, + "loss": 2.8061351776123047, + "step": 67070 + }, + { + "epoch": 0.3354, + "grad_norm": 0.324349582195282, + "learning_rate": 0.0003781503843058279, + "loss": 2.8048095703125, + "step": 67080 + }, + { + "epoch": 0.33545, + "grad_norm": 0.3152707517147064, + "learning_rate": 0.0003781163239167931, + "loss": 2.778451919555664, + "step": 67090 + }, + { + "epoch": 0.3355, + "grad_norm": 0.40878763794898987, + "learning_rate": 0.00037808226030243056, + "loss": 2.798727798461914, + "step": 67100 + }, + { + "epoch": 0.33555, + "grad_norm": 0.3263276219367981, + "learning_rate": 0.000378048193463598, + "loss": 2.7712425231933593, + "step": 67110 + }, + { + "epoch": 0.3356, + "grad_norm": 0.3322700262069702, + "learning_rate": 0.00037801412340115304, + "loss": 2.7108497619628906, + "step": 67120 + }, + { + "epoch": 0.33565, + "grad_norm": 0.39530766010284424, + "learning_rate": 0.00037798005011595336, + "loss": 2.6993619918823244, + "step": 67130 + }, + { + "epoch": 0.3357, + "grad_norm": 0.3256371021270752, + "learning_rate": 0.00037794597360885687, + "loss": 2.832206916809082, + "step": 67140 + }, + { + "epoch": 0.33575, + "grad_norm": 0.3529179096221924, + "learning_rate": 0.00037791189388072127, + "loss": 2.801797866821289, + "step": 67150 + }, + { + "epoch": 0.3358, + "grad_norm": 0.3380773961544037, + "learning_rate": 0.0003778778109324046, + "loss": 2.7749683380126955, + "step": 67160 + }, + { + "epoch": 0.33585, + "grad_norm": 0.32143434882164, + "learning_rate": 0.00037784372476476496, + "loss": 2.779781723022461, + "step": 67170 + }, + { + "epoch": 0.3359, + "grad_norm": 0.3505733609199524, + "learning_rate": 0.00037780963537866033, + "loss": 2.7833364486694334, + "step": 67180 + }, + { + "epoch": 0.33595, + "grad_norm": 0.36041224002838135, + "learning_rate": 0.000377775542774949, + "loss": 2.750771331787109, + "step": 67190 + }, + { + "epoch": 0.336, + "grad_norm": 0.37633877992630005, + "learning_rate": 0.0003777414469544893, + "loss": 2.7284805297851564, + "step": 67200 + }, + { + "epoch": 0.33605, + "grad_norm": 0.3318568468093872, + "learning_rate": 0.0003777073479181394, + "loss": 2.724995994567871, + "step": 67210 + }, + { + "epoch": 0.3361, + "grad_norm": 0.36349180340766907, + "learning_rate": 0.00037767324566675796, + "loss": 2.775670623779297, + "step": 67220 + }, + { + "epoch": 0.33615, + "grad_norm": 0.3749900162220001, + "learning_rate": 0.0003776391402012034, + "loss": 2.7916994094848633, + "step": 67230 + }, + { + "epoch": 0.3362, + "grad_norm": 0.3426421284675598, + "learning_rate": 0.00037760503152233426, + "loss": 2.7414926528930663, + "step": 67240 + }, + { + "epoch": 0.33625, + "grad_norm": 0.3446321189403534, + "learning_rate": 0.0003775709196310094, + "loss": 2.805167579650879, + "step": 67250 + }, + { + "epoch": 0.3363, + "grad_norm": 0.31155386567115784, + "learning_rate": 0.00037753680452808747, + "loss": 2.760938262939453, + "step": 67260 + }, + { + "epoch": 0.33635, + "grad_norm": 0.3115680515766144, + "learning_rate": 0.00037750268621442734, + "loss": 2.770128631591797, + "step": 67270 + }, + { + "epoch": 0.3364, + "grad_norm": 0.3132811486721039, + "learning_rate": 0.0003774685646908879, + "loss": 2.7576045989990234, + "step": 67280 + }, + { + "epoch": 0.33645, + "grad_norm": 0.4301375150680542, + "learning_rate": 0.0003774344399583283, + "loss": 2.7356342315673827, + "step": 67290 + }, + { + "epoch": 0.3365, + "grad_norm": 0.35370078682899475, + "learning_rate": 0.0003774003120176074, + "loss": 2.7551319122314455, + "step": 67300 + }, + { + "epoch": 0.33655, + "grad_norm": 0.3421018421649933, + "learning_rate": 0.0003773661808695846, + "loss": 2.739829444885254, + "step": 67310 + }, + { + "epoch": 0.3366, + "grad_norm": 0.33632349967956543, + "learning_rate": 0.0003773320465151189, + "loss": 2.7858230590820314, + "step": 67320 + }, + { + "epoch": 0.33665, + "grad_norm": 0.3733080327510834, + "learning_rate": 0.0003772979089550699, + "loss": 2.8142063140869142, + "step": 67330 + }, + { + "epoch": 0.3367, + "grad_norm": 0.32302242517471313, + "learning_rate": 0.0003772637681902969, + "loss": 2.804935073852539, + "step": 67340 + }, + { + "epoch": 0.33675, + "grad_norm": 0.3540862202644348, + "learning_rate": 0.0003772296242216593, + "loss": 2.777179145812988, + "step": 67350 + }, + { + "epoch": 0.3368, + "grad_norm": 0.3487662672996521, + "learning_rate": 0.0003771954770500168, + "loss": 2.7754690170288088, + "step": 67360 + }, + { + "epoch": 0.33685, + "grad_norm": 0.32664066553115845, + "learning_rate": 0.000377161326676229, + "loss": 2.775611686706543, + "step": 67370 + }, + { + "epoch": 0.3369, + "grad_norm": 0.3139122426509857, + "learning_rate": 0.00037712717310115575, + "loss": 2.737133026123047, + "step": 67380 + }, + { + "epoch": 0.33695, + "grad_norm": 0.3424839377403259, + "learning_rate": 0.00037709301632565676, + "loss": 2.7559980392456054, + "step": 67390 + }, + { + "epoch": 0.337, + "grad_norm": 0.33502963185310364, + "learning_rate": 0.0003770588563505918, + "loss": 2.7579233169555666, + "step": 67400 + }, + { + "epoch": 0.33705, + "grad_norm": 0.3563230633735657, + "learning_rate": 0.00037702469317682106, + "loss": 2.77086181640625, + "step": 67410 + }, + { + "epoch": 0.3371, + "grad_norm": 0.3482460677623749, + "learning_rate": 0.0003769905268052045, + "loss": 2.7733575820922853, + "step": 67420 + }, + { + "epoch": 0.33715, + "grad_norm": 0.3119116425514221, + "learning_rate": 0.00037695635723660224, + "loss": 2.7868074417114257, + "step": 67430 + }, + { + "epoch": 0.3372, + "grad_norm": 0.32461100816726685, + "learning_rate": 0.0003769221844718746, + "loss": 2.8119552612304686, + "step": 67440 + }, + { + "epoch": 0.33725, + "grad_norm": 0.3803332448005676, + "learning_rate": 0.00037688800851188177, + "loss": 2.769273376464844, + "step": 67450 + }, + { + "epoch": 0.3373, + "grad_norm": 0.33875125646591187, + "learning_rate": 0.0003768538293574842, + "loss": 2.8242748260498045, + "step": 67460 + }, + { + "epoch": 0.33735, + "grad_norm": 0.35179591178894043, + "learning_rate": 0.0003768196470095424, + "loss": 2.7404111862182616, + "step": 67470 + }, + { + "epoch": 0.3374, + "grad_norm": 0.3611929416656494, + "learning_rate": 0.00037678546146891666, + "loss": 2.760235595703125, + "step": 67480 + }, + { + "epoch": 0.33745, + "grad_norm": 0.35867491364479065, + "learning_rate": 0.00037675127273646794, + "loss": 2.770611572265625, + "step": 67490 + }, + { + "epoch": 0.3375, + "grad_norm": 0.3097492456436157, + "learning_rate": 0.00037671708081305665, + "loss": 2.7678354263305662, + "step": 67500 + }, + { + "epoch": 0.33755, + "grad_norm": 0.33073538541793823, + "learning_rate": 0.0003766828856995438, + "loss": 2.779329299926758, + "step": 67510 + }, + { + "epoch": 0.3376, + "grad_norm": 0.3164210617542267, + "learning_rate": 0.00037664868739679003, + "loss": 2.7562835693359373, + "step": 67520 + }, + { + "epoch": 0.33765, + "grad_norm": 0.3550221025943756, + "learning_rate": 0.00037661448590565653, + "loss": 2.745503616333008, + "step": 67530 + }, + { + "epoch": 0.3377, + "grad_norm": 0.3214140236377716, + "learning_rate": 0.00037658028122700404, + "loss": 2.7906633377075196, + "step": 67540 + }, + { + "epoch": 0.33775, + "grad_norm": 0.35355475544929504, + "learning_rate": 0.000376546073361694, + "loss": 2.72534065246582, + "step": 67550 + }, + { + "epoch": 0.3378, + "grad_norm": 0.3707701563835144, + "learning_rate": 0.0003765118623105873, + "loss": 2.7524269104003904, + "step": 67560 + }, + { + "epoch": 0.33785, + "grad_norm": 0.34544259309768677, + "learning_rate": 0.0003764776480745453, + "loss": 2.7127342224121094, + "step": 67570 + }, + { + "epoch": 0.3379, + "grad_norm": 0.3272325396537781, + "learning_rate": 0.00037644343065442945, + "loss": 2.8335813522338866, + "step": 67580 + }, + { + "epoch": 0.33795, + "grad_norm": 0.3275600075721741, + "learning_rate": 0.00037640921005110094, + "loss": 2.7946598052978517, + "step": 67590 + }, + { + "epoch": 0.338, + "grad_norm": 0.3029966950416565, + "learning_rate": 0.0003763749862654215, + "loss": 2.785457992553711, + "step": 67600 + }, + { + "epoch": 0.33805, + "grad_norm": 0.3529048562049866, + "learning_rate": 0.0003763407592982526, + "loss": 2.7480207443237306, + "step": 67610 + }, + { + "epoch": 0.3381, + "grad_norm": 0.3190729320049286, + "learning_rate": 0.00037630652915045595, + "loss": 2.753936767578125, + "step": 67620 + }, + { + "epoch": 0.33815, + "grad_norm": 0.3115578889846802, + "learning_rate": 0.0003762722958228933, + "loss": 2.7908958435058593, + "step": 67630 + }, + { + "epoch": 0.3382, + "grad_norm": 0.3820028305053711, + "learning_rate": 0.00037623805931642643, + "loss": 2.7633752822875977, + "step": 67640 + }, + { + "epoch": 0.33825, + "grad_norm": 0.33225640654563904, + "learning_rate": 0.00037620381963191716, + "loss": 2.7992652893066405, + "step": 67650 + }, + { + "epoch": 0.3383, + "grad_norm": 0.2973940372467041, + "learning_rate": 0.0003761695767702277, + "loss": 2.7626136779785155, + "step": 67660 + }, + { + "epoch": 0.33835, + "grad_norm": 0.34277933835983276, + "learning_rate": 0.0003761353307322199, + "loss": 2.7219268798828127, + "step": 67670 + }, + { + "epoch": 0.3384, + "grad_norm": 0.31647786498069763, + "learning_rate": 0.000376101081518756, + "loss": 2.788456916809082, + "step": 67680 + }, + { + "epoch": 0.33845, + "grad_norm": 0.30930304527282715, + "learning_rate": 0.00037606682913069833, + "loss": 2.8214868545532226, + "step": 67690 + }, + { + "epoch": 0.3385, + "grad_norm": 0.30383947491645813, + "learning_rate": 0.00037603257356890894, + "loss": 2.8054243087768556, + "step": 67700 + }, + { + "epoch": 0.33855, + "grad_norm": 0.3484495282173157, + "learning_rate": 0.0003759983148342504, + "loss": 2.7317808151245115, + "step": 67710 + }, + { + "epoch": 0.3386, + "grad_norm": 0.3174482583999634, + "learning_rate": 0.0003759640529275852, + "loss": 2.771683120727539, + "step": 67720 + }, + { + "epoch": 0.33865, + "grad_norm": 0.3498234152793884, + "learning_rate": 0.0003759297878497757, + "loss": 2.7256534576416014, + "step": 67730 + }, + { + "epoch": 0.3387, + "grad_norm": 0.35067474842071533, + "learning_rate": 0.00037589551960168465, + "loss": 2.7218643188476563, + "step": 67740 + }, + { + "epoch": 0.33875, + "grad_norm": 0.3397684693336487, + "learning_rate": 0.0003758612481841748, + "loss": 2.7802427291870115, + "step": 67750 + }, + { + "epoch": 0.3388, + "grad_norm": 0.3185613453388214, + "learning_rate": 0.0003758269735981088, + "loss": 2.753506088256836, + "step": 67760 + }, + { + "epoch": 0.33885, + "grad_norm": 0.33735203742980957, + "learning_rate": 0.0003757926958443496, + "loss": 2.7637399673461913, + "step": 67770 + }, + { + "epoch": 0.3389, + "grad_norm": 0.3430686891078949, + "learning_rate": 0.00037575841492376013, + "loss": 2.7537504196166993, + "step": 67780 + }, + { + "epoch": 0.33895, + "grad_norm": 0.346166729927063, + "learning_rate": 0.00037572413083720345, + "loss": 2.7608053207397463, + "step": 67790 + }, + { + "epoch": 0.339, + "grad_norm": 0.3195328414440155, + "learning_rate": 0.0003756898435855426, + "loss": 2.8034740447998048, + "step": 67800 + }, + { + "epoch": 0.33905, + "grad_norm": 0.3431428372859955, + "learning_rate": 0.0003756555531696407, + "loss": 2.8269472122192383, + "step": 67810 + }, + { + "epoch": 0.3391, + "grad_norm": 0.33285972476005554, + "learning_rate": 0.00037562125959036117, + "loss": 2.7566650390625, + "step": 67820 + }, + { + "epoch": 0.33915, + "grad_norm": 0.4072999358177185, + "learning_rate": 0.0003755869628485672, + "loss": 2.714764404296875, + "step": 67830 + }, + { + "epoch": 0.3392, + "grad_norm": 0.29008087515830994, + "learning_rate": 0.0003755526629451224, + "loss": 2.751499557495117, + "step": 67840 + }, + { + "epoch": 0.33925, + "grad_norm": 0.36585402488708496, + "learning_rate": 0.00037551835988089, + "loss": 2.808220100402832, + "step": 67850 + }, + { + "epoch": 0.3393, + "grad_norm": 0.3216208219528198, + "learning_rate": 0.00037548405365673377, + "loss": 2.7742528915405273, + "step": 67860 + }, + { + "epoch": 0.33935, + "grad_norm": 0.32297483086586, + "learning_rate": 0.0003754497442735174, + "loss": 2.802867126464844, + "step": 67870 + }, + { + "epoch": 0.3394, + "grad_norm": 0.3107060492038727, + "learning_rate": 0.0003754154317321046, + "loss": 2.7941259384155273, + "step": 67880 + }, + { + "epoch": 0.33945, + "grad_norm": 0.3538084626197815, + "learning_rate": 0.0003753811160333589, + "loss": 2.6875085830688477, + "step": 67890 + }, + { + "epoch": 0.3395, + "grad_norm": 0.34693384170532227, + "learning_rate": 0.00037534679717814466, + "loss": 2.8109127044677735, + "step": 67900 + }, + { + "epoch": 0.33955, + "grad_norm": 0.38549232482910156, + "learning_rate": 0.00037531247516732556, + "loss": 2.822133445739746, + "step": 67910 + }, + { + "epoch": 0.3396, + "grad_norm": 0.32749947905540466, + "learning_rate": 0.0003752781500017658, + "loss": 2.731892776489258, + "step": 67920 + }, + { + "epoch": 0.33965, + "grad_norm": 0.3466450273990631, + "learning_rate": 0.00037524382168232937, + "loss": 2.774514007568359, + "step": 67930 + }, + { + "epoch": 0.3397, + "grad_norm": 0.3502967357635498, + "learning_rate": 0.0003752094902098806, + "loss": 2.7372798919677734, + "step": 67940 + }, + { + "epoch": 0.33975, + "grad_norm": 0.3346320688724518, + "learning_rate": 0.0003751751555852837, + "loss": 2.733610916137695, + "step": 67950 + }, + { + "epoch": 0.3398, + "grad_norm": 0.3396095037460327, + "learning_rate": 0.00037514081780940325, + "loss": 2.7961071014404295, + "step": 67960 + }, + { + "epoch": 0.33985, + "grad_norm": 0.31989753246307373, + "learning_rate": 0.00037510647688310337, + "loss": 2.7399852752685545, + "step": 67970 + }, + { + "epoch": 0.3399, + "grad_norm": 0.3373163938522339, + "learning_rate": 0.0003750721328072488, + "loss": 2.7944869995117188, + "step": 67980 + }, + { + "epoch": 0.33995, + "grad_norm": 0.3474128544330597, + "learning_rate": 0.0003750377855827042, + "loss": 2.8039134979248046, + "step": 67990 + }, + { + "epoch": 0.34, + "grad_norm": 0.3346783220767975, + "learning_rate": 0.00037500343521033413, + "loss": 2.7988086700439454, + "step": 68000 + }, + { + "epoch": 0.34005, + "grad_norm": 0.3257092833518982, + "learning_rate": 0.0003749690816910034, + "loss": 2.7288848876953127, + "step": 68010 + }, + { + "epoch": 0.3401, + "grad_norm": 0.3549114465713501, + "learning_rate": 0.0003749347250255769, + "loss": 2.727164077758789, + "step": 68020 + }, + { + "epoch": 0.34015, + "grad_norm": 0.34038013219833374, + "learning_rate": 0.00037490036521491956, + "loss": 2.7961847305297853, + "step": 68030 + }, + { + "epoch": 0.3402, + "grad_norm": 0.33180108666419983, + "learning_rate": 0.0003748660022598964, + "loss": 2.7687637329101564, + "step": 68040 + }, + { + "epoch": 0.34025, + "grad_norm": 0.35229483246803284, + "learning_rate": 0.00037483163616137234, + "loss": 2.8322422027587892, + "step": 68050 + }, + { + "epoch": 0.3403, + "grad_norm": 0.32908034324645996, + "learning_rate": 0.0003747972669202128, + "loss": 2.7502071380615236, + "step": 68060 + }, + { + "epoch": 0.34035, + "grad_norm": 0.4010922312736511, + "learning_rate": 0.0003747628945372828, + "loss": 2.7861730575561525, + "step": 68070 + }, + { + "epoch": 0.3404, + "grad_norm": 0.3562430441379547, + "learning_rate": 0.0003747285190134478, + "loss": 2.7910350799560546, + "step": 68080 + }, + { + "epoch": 0.34045, + "grad_norm": 0.3456060588359833, + "learning_rate": 0.0003746941403495732, + "loss": 2.7458213806152343, + "step": 68090 + }, + { + "epoch": 0.3405, + "grad_norm": 0.3479112684726715, + "learning_rate": 0.0003746597585465243, + "loss": 2.7518032073974608, + "step": 68100 + }, + { + "epoch": 0.34055, + "grad_norm": 0.3641105592250824, + "learning_rate": 0.00037462537360516694, + "loss": 2.7990650177001952, + "step": 68110 + }, + { + "epoch": 0.3406, + "grad_norm": 0.305372416973114, + "learning_rate": 0.0003745909855263667, + "loss": 2.796866226196289, + "step": 68120 + }, + { + "epoch": 0.34065, + "grad_norm": 0.3403930962085724, + "learning_rate": 0.0003745565943109891, + "loss": 2.7543811798095703, + "step": 68130 + }, + { + "epoch": 0.3407, + "grad_norm": 0.3498028814792633, + "learning_rate": 0.00037452219995990016, + "loss": 2.7843868255615236, + "step": 68140 + }, + { + "epoch": 0.34075, + "grad_norm": 0.33273881673812866, + "learning_rate": 0.00037448780247396565, + "loss": 2.742630386352539, + "step": 68150 + }, + { + "epoch": 0.3408, + "grad_norm": 0.3213755488395691, + "learning_rate": 0.00037445340185405146, + "loss": 2.732284927368164, + "step": 68160 + }, + { + "epoch": 0.34085, + "grad_norm": 0.32691916823387146, + "learning_rate": 0.00037441899810102377, + "loss": 2.7632503509521484, + "step": 68170 + }, + { + "epoch": 0.3409, + "grad_norm": 0.3196250796318054, + "learning_rate": 0.0003743845912157487, + "loss": 2.8084247589111326, + "step": 68180 + }, + { + "epoch": 0.34095, + "grad_norm": 0.3533404767513275, + "learning_rate": 0.00037435018119909226, + "loss": 2.7502756118774414, + "step": 68190 + }, + { + "epoch": 0.341, + "grad_norm": 0.32598721981048584, + "learning_rate": 0.0003743157680519209, + "loss": 2.7024843215942385, + "step": 68200 + }, + { + "epoch": 0.34105, + "grad_norm": 0.4313802421092987, + "learning_rate": 0.00037428135177510093, + "loss": 2.740108299255371, + "step": 68210 + }, + { + "epoch": 0.3411, + "grad_norm": 0.3148825764656067, + "learning_rate": 0.0003742469323694987, + "loss": 2.7674896240234377, + "step": 68220 + }, + { + "epoch": 0.34115, + "grad_norm": 0.35741668939590454, + "learning_rate": 0.00037421250983598085, + "loss": 2.775520133972168, + "step": 68230 + }, + { + "epoch": 0.3412, + "grad_norm": 0.3538045287132263, + "learning_rate": 0.0003741780841754139, + "loss": 2.774412727355957, + "step": 68240 + }, + { + "epoch": 0.34125, + "grad_norm": 0.30625414848327637, + "learning_rate": 0.0003741436553886644, + "loss": 2.8001167297363283, + "step": 68250 + }, + { + "epoch": 0.3413, + "grad_norm": 0.3427485525608063, + "learning_rate": 0.00037410922347659933, + "loss": 2.8000560760498048, + "step": 68260 + }, + { + "epoch": 0.34135, + "grad_norm": 0.3414103090763092, + "learning_rate": 0.00037407478844008526, + "loss": 2.7881519317626955, + "step": 68270 + }, + { + "epoch": 0.3414, + "grad_norm": 0.30132994055747986, + "learning_rate": 0.00037404035027998933, + "loss": 2.6481470108032226, + "step": 68280 + }, + { + "epoch": 0.34145, + "grad_norm": 0.3201387822628021, + "learning_rate": 0.00037400590899717837, + "loss": 2.774854850769043, + "step": 68290 + }, + { + "epoch": 0.3415, + "grad_norm": 0.3924294710159302, + "learning_rate": 0.0003739714645925195, + "loss": 2.733637046813965, + "step": 68300 + }, + { + "epoch": 0.34155, + "grad_norm": 0.36490991711616516, + "learning_rate": 0.0003739370170668798, + "loss": 2.7961259841918946, + "step": 68310 + }, + { + "epoch": 0.3416, + "grad_norm": 0.3633970618247986, + "learning_rate": 0.00037390256642112654, + "loss": 2.751624870300293, + "step": 68320 + }, + { + "epoch": 0.34165, + "grad_norm": 0.35056889057159424, + "learning_rate": 0.00037386811265612696, + "loss": 2.792185592651367, + "step": 68330 + }, + { + "epoch": 0.3417, + "grad_norm": 0.3289625644683838, + "learning_rate": 0.00037383365577274856, + "loss": 2.77508430480957, + "step": 68340 + }, + { + "epoch": 0.34175, + "grad_norm": 0.3295919895172119, + "learning_rate": 0.0003737991957718586, + "loss": 2.8218034744262694, + "step": 68350 + }, + { + "epoch": 0.3418, + "grad_norm": 0.35465461015701294, + "learning_rate": 0.00037376473265432486, + "loss": 2.791253662109375, + "step": 68360 + }, + { + "epoch": 0.34185, + "grad_norm": 0.35343924164772034, + "learning_rate": 0.0003737302664210147, + "loss": 2.732932472229004, + "step": 68370 + }, + { + "epoch": 0.3419, + "grad_norm": 0.4306500256061554, + "learning_rate": 0.0003736957970727959, + "loss": 2.761184310913086, + "step": 68380 + }, + { + "epoch": 0.34195, + "grad_norm": 0.31477582454681396, + "learning_rate": 0.00037366132461053627, + "loss": 2.7707855224609377, + "step": 68390 + }, + { + "epoch": 0.342, + "grad_norm": 0.3419616222381592, + "learning_rate": 0.00037362684903510357, + "loss": 2.7914018630981445, + "step": 68400 + }, + { + "epoch": 0.34205, + "grad_norm": 0.3504674732685089, + "learning_rate": 0.0003735923703473658, + "loss": 2.7761072158813476, + "step": 68410 + }, + { + "epoch": 0.3421, + "grad_norm": 0.3311428725719452, + "learning_rate": 0.000373557888548191, + "loss": 2.7585290908813476, + "step": 68420 + }, + { + "epoch": 0.34215, + "grad_norm": 0.32785341143608093, + "learning_rate": 0.00037352340363844706, + "loss": 2.7289825439453126, + "step": 68430 + }, + { + "epoch": 0.3422, + "grad_norm": 0.33795687556266785, + "learning_rate": 0.00037348891561900233, + "loss": 2.7575550079345703, + "step": 68440 + }, + { + "epoch": 0.34225, + "grad_norm": 0.3669722080230713, + "learning_rate": 0.00037345442449072497, + "loss": 2.7366769790649412, + "step": 68450 + }, + { + "epoch": 0.3423, + "grad_norm": 0.3174551725387573, + "learning_rate": 0.0003734199302544833, + "loss": 2.785308265686035, + "step": 68460 + }, + { + "epoch": 0.34235, + "grad_norm": 0.3503727614879608, + "learning_rate": 0.0003733854329111457, + "loss": 2.7818716049194334, + "step": 68470 + }, + { + "epoch": 0.3424, + "grad_norm": 0.3209094703197479, + "learning_rate": 0.0003733509324615807, + "loss": 2.7954479217529298, + "step": 68480 + }, + { + "epoch": 0.34245, + "grad_norm": 0.3260677754878998, + "learning_rate": 0.00037331642890665667, + "loss": 2.7651914596557616, + "step": 68490 + }, + { + "epoch": 0.3425, + "grad_norm": 0.34573647379875183, + "learning_rate": 0.0003732819222472425, + "loss": 2.7381038665771484, + "step": 68500 + }, + { + "epoch": 0.34255, + "grad_norm": 0.381046861410141, + "learning_rate": 0.00037324741248420673, + "loss": 2.716586112976074, + "step": 68510 + }, + { + "epoch": 0.3426, + "grad_norm": 0.3164343535900116, + "learning_rate": 0.00037321289961841816, + "loss": 2.79718017578125, + "step": 68520 + }, + { + "epoch": 0.34265, + "grad_norm": 0.353329062461853, + "learning_rate": 0.0003731783836507456, + "loss": 2.8002206802368166, + "step": 68530 + }, + { + "epoch": 0.3427, + "grad_norm": 0.34242746233940125, + "learning_rate": 0.00037314386458205817, + "loss": 2.7132339477539062, + "step": 68540 + }, + { + "epoch": 0.34275, + "grad_norm": 0.29559627175331116, + "learning_rate": 0.00037310934241322467, + "loss": 2.7614564895629883, + "step": 68550 + }, + { + "epoch": 0.3428, + "grad_norm": 0.3036946952342987, + "learning_rate": 0.00037307481714511436, + "loss": 2.7809572219848633, + "step": 68560 + }, + { + "epoch": 0.34285, + "grad_norm": 0.33290088176727295, + "learning_rate": 0.0003730402887785963, + "loss": 2.7787960052490233, + "step": 68570 + }, + { + "epoch": 0.3429, + "grad_norm": 0.3735373616218567, + "learning_rate": 0.0003730057573145399, + "loss": 2.732700729370117, + "step": 68580 + }, + { + "epoch": 0.34295, + "grad_norm": 0.33238568902015686, + "learning_rate": 0.0003729712227538143, + "loss": 2.805005645751953, + "step": 68590 + }, + { + "epoch": 0.343, + "grad_norm": 0.32757747173309326, + "learning_rate": 0.000372936685097289, + "loss": 2.741978645324707, + "step": 68600 + }, + { + "epoch": 0.34305, + "grad_norm": 0.34891799092292786, + "learning_rate": 0.00037290214434583345, + "loss": 2.79614143371582, + "step": 68610 + }, + { + "epoch": 0.3431, + "grad_norm": 0.3332265019416809, + "learning_rate": 0.00037286760050031726, + "loss": 2.8061323165893555, + "step": 68620 + }, + { + "epoch": 0.34315, + "grad_norm": 0.38823196291923523, + "learning_rate": 0.00037283305356161, + "loss": 2.807211685180664, + "step": 68630 + }, + { + "epoch": 0.3432, + "grad_norm": 0.32072070240974426, + "learning_rate": 0.0003727985035305816, + "loss": 2.734702491760254, + "step": 68640 + }, + { + "epoch": 0.34325, + "grad_norm": 0.3258076012134552, + "learning_rate": 0.0003727639504081015, + "loss": 2.768449401855469, + "step": 68650 + }, + { + "epoch": 0.3433, + "grad_norm": 0.3295016884803772, + "learning_rate": 0.00037272939419503987, + "loss": 2.784697914123535, + "step": 68660 + }, + { + "epoch": 0.34335, + "grad_norm": 0.32782095670700073, + "learning_rate": 0.0003726948348922665, + "loss": 2.7391389846801757, + "step": 68670 + }, + { + "epoch": 0.3434, + "grad_norm": 0.3537690043449402, + "learning_rate": 0.0003726602725006516, + "loss": 2.7265207290649416, + "step": 68680 + }, + { + "epoch": 0.34345, + "grad_norm": 0.349139541387558, + "learning_rate": 0.0003726257070210651, + "loss": 2.7818506240844725, + "step": 68690 + }, + { + "epoch": 0.3435, + "grad_norm": 0.3148556649684906, + "learning_rate": 0.00037259113845437717, + "loss": 2.7273183822631837, + "step": 68700 + }, + { + "epoch": 0.34355, + "grad_norm": 0.31944987177848816, + "learning_rate": 0.0003725565668014582, + "loss": 2.757005310058594, + "step": 68710 + }, + { + "epoch": 0.3436, + "grad_norm": 0.3255188763141632, + "learning_rate": 0.0003725219920631785, + "loss": 2.8300748825073243, + "step": 68720 + }, + { + "epoch": 0.34365, + "grad_norm": 0.3359749913215637, + "learning_rate": 0.00037248741424040843, + "loss": 2.8425357818603514, + "step": 68730 + }, + { + "epoch": 0.3437, + "grad_norm": 0.37428316473960876, + "learning_rate": 0.0003724528333340186, + "loss": 2.802277946472168, + "step": 68740 + }, + { + "epoch": 0.34375, + "grad_norm": 0.341940313577652, + "learning_rate": 0.00037241824934487944, + "loss": 2.8071884155273437, + "step": 68750 + }, + { + "epoch": 0.3438, + "grad_norm": 0.32892927527427673, + "learning_rate": 0.0003723836622738617, + "loss": 2.818654251098633, + "step": 68760 + }, + { + "epoch": 0.34385, + "grad_norm": 0.325786828994751, + "learning_rate": 0.00037234907212183613, + "loss": 2.7730430603027343, + "step": 68770 + }, + { + "epoch": 0.3439, + "grad_norm": 0.36296531558036804, + "learning_rate": 0.00037231447888967336, + "loss": 2.732350730895996, + "step": 68780 + }, + { + "epoch": 0.34395, + "grad_norm": 0.3361922800540924, + "learning_rate": 0.0003722798825782445, + "loss": 2.7318328857421874, + "step": 68790 + }, + { + "epoch": 0.344, + "grad_norm": 0.31063613295555115, + "learning_rate": 0.0003722452831884203, + "loss": 2.703550910949707, + "step": 68800 + }, + { + "epoch": 0.34405, + "grad_norm": 0.3136613965034485, + "learning_rate": 0.000372210680721072, + "loss": 2.773543930053711, + "step": 68810 + }, + { + "epoch": 0.3441, + "grad_norm": 0.33001384139060974, + "learning_rate": 0.0003721760751770706, + "loss": 2.7216028213500976, + "step": 68820 + }, + { + "epoch": 0.34415, + "grad_norm": 0.32473549246788025, + "learning_rate": 0.0003721414665572874, + "loss": 2.770208168029785, + "step": 68830 + }, + { + "epoch": 0.3442, + "grad_norm": 0.35864412784576416, + "learning_rate": 0.0003721068548625935, + "loss": 2.7836328506469727, + "step": 68840 + }, + { + "epoch": 0.34425, + "grad_norm": 0.3118985891342163, + "learning_rate": 0.0003720722400938604, + "loss": 2.752621078491211, + "step": 68850 + }, + { + "epoch": 0.3443, + "grad_norm": 0.343216210603714, + "learning_rate": 0.00037203762225195946, + "loss": 2.7501407623291017, + "step": 68860 + }, + { + "epoch": 0.34435, + "grad_norm": 0.3809548318386078, + "learning_rate": 0.00037200300133776224, + "loss": 2.803085708618164, + "step": 68870 + }, + { + "epoch": 0.3444, + "grad_norm": 0.31848961114883423, + "learning_rate": 0.0003719683773521402, + "loss": 2.766274642944336, + "step": 68880 + }, + { + "epoch": 0.34445, + "grad_norm": 0.35603123903274536, + "learning_rate": 0.0003719337502959651, + "loss": 2.74865837097168, + "step": 68890 + }, + { + "epoch": 0.3445, + "grad_norm": 0.3211800754070282, + "learning_rate": 0.00037189912017010864, + "loss": 2.732944679260254, + "step": 68900 + }, + { + "epoch": 0.34455, + "grad_norm": 0.3231425881385803, + "learning_rate": 0.00037186448697544273, + "loss": 2.7224550247192383, + "step": 68910 + }, + { + "epoch": 0.3446, + "grad_norm": 0.32804059982299805, + "learning_rate": 0.00037182985071283905, + "loss": 2.7192638397216795, + "step": 68920 + }, + { + "epoch": 0.34465, + "grad_norm": 0.329093873500824, + "learning_rate": 0.0003717952113831697, + "loss": 2.74517822265625, + "step": 68930 + }, + { + "epoch": 0.3447, + "grad_norm": 0.33695003390312195, + "learning_rate": 0.0003717605689873068, + "loss": 2.7920415878295897, + "step": 68940 + }, + { + "epoch": 0.34475, + "grad_norm": 0.3515782356262207, + "learning_rate": 0.0003717259235261223, + "loss": 2.7691438674926756, + "step": 68950 + }, + { + "epoch": 0.3448, + "grad_norm": 0.4761485755443573, + "learning_rate": 0.0003716912750004886, + "loss": 2.7351753234863283, + "step": 68960 + }, + { + "epoch": 0.34485, + "grad_norm": 0.3728354275226593, + "learning_rate": 0.00037165662341127783, + "loss": 2.7466001510620117, + "step": 68970 + }, + { + "epoch": 0.3449, + "grad_norm": 0.3597579598426819, + "learning_rate": 0.0003716219687593623, + "loss": 2.7565126419067383, + "step": 68980 + }, + { + "epoch": 0.34495, + "grad_norm": 0.3159765899181366, + "learning_rate": 0.0003715873110456146, + "loss": 2.8102748870849608, + "step": 68990 + }, + { + "epoch": 0.345, + "grad_norm": 0.37748733162879944, + "learning_rate": 0.00037155265027090714, + "loss": 2.8024951934814455, + "step": 69000 + }, + { + "epoch": 0.34505, + "grad_norm": 0.31846529245376587, + "learning_rate": 0.0003715179864361125, + "loss": 2.776141357421875, + "step": 69010 + }, + { + "epoch": 0.3451, + "grad_norm": 0.34927117824554443, + "learning_rate": 0.00037148331954210337, + "loss": 2.8049600601196287, + "step": 69020 + }, + { + "epoch": 0.34515, + "grad_norm": 0.3074184060096741, + "learning_rate": 0.0003714486495897525, + "loss": 2.7373992919921877, + "step": 69030 + }, + { + "epoch": 0.3452, + "grad_norm": 0.38479548692703247, + "learning_rate": 0.00037141397657993266, + "loss": 2.756241035461426, + "step": 69040 + }, + { + "epoch": 0.34525, + "grad_norm": 0.34515461325645447, + "learning_rate": 0.00037137930051351676, + "loss": 2.750062942504883, + "step": 69050 + }, + { + "epoch": 0.3453, + "grad_norm": 0.33362817764282227, + "learning_rate": 0.0003713446213913778, + "loss": 2.747407913208008, + "step": 69060 + }, + { + "epoch": 0.34535, + "grad_norm": 0.31314945220947266, + "learning_rate": 0.0003713099392143889, + "loss": 2.7631725311279296, + "step": 69070 + }, + { + "epoch": 0.3454, + "grad_norm": 0.3641715347766876, + "learning_rate": 0.00037127525398342296, + "loss": 2.764191246032715, + "step": 69080 + }, + { + "epoch": 0.34545, + "grad_norm": 0.3418188691139221, + "learning_rate": 0.00037124056569935336, + "loss": 2.761445999145508, + "step": 69090 + }, + { + "epoch": 0.3455, + "grad_norm": 0.3659989535808563, + "learning_rate": 0.0003712058743630533, + "loss": 2.809832954406738, + "step": 69100 + }, + { + "epoch": 0.34555, + "grad_norm": 0.3294600248336792, + "learning_rate": 0.0003711711799753962, + "loss": 2.771337127685547, + "step": 69110 + }, + { + "epoch": 0.3456, + "grad_norm": 0.40587317943573, + "learning_rate": 0.0003711364825372555, + "loss": 2.733835983276367, + "step": 69120 + }, + { + "epoch": 0.34565, + "grad_norm": 0.3417139947414398, + "learning_rate": 0.0003711017820495046, + "loss": 2.7453731536865233, + "step": 69130 + }, + { + "epoch": 0.3457, + "grad_norm": 0.33530470728874207, + "learning_rate": 0.0003710670785130171, + "loss": 2.797086715698242, + "step": 69140 + }, + { + "epoch": 0.34575, + "grad_norm": 0.35722771286964417, + "learning_rate": 0.0003710323719286668, + "loss": 2.78188533782959, + "step": 69150 + }, + { + "epoch": 0.3458, + "grad_norm": 0.3376345634460449, + "learning_rate": 0.0003709976622973273, + "loss": 2.773956298828125, + "step": 69160 + }, + { + "epoch": 0.34585, + "grad_norm": 0.38354581594467163, + "learning_rate": 0.00037096294961987245, + "loss": 2.7406147003173826, + "step": 69170 + }, + { + "epoch": 0.3459, + "grad_norm": 0.34059107303619385, + "learning_rate": 0.0003709282338971761, + "loss": 2.7666994094848634, + "step": 69180 + }, + { + "epoch": 0.34595, + "grad_norm": 0.34653520584106445, + "learning_rate": 0.00037089351513011235, + "loss": 2.726339340209961, + "step": 69190 + }, + { + "epoch": 0.346, + "grad_norm": 0.3367816209793091, + "learning_rate": 0.0003708587933195552, + "loss": 2.710009765625, + "step": 69200 + }, + { + "epoch": 0.34605, + "grad_norm": 0.4113348126411438, + "learning_rate": 0.00037082406846637874, + "loss": 2.7551172256469725, + "step": 69210 + }, + { + "epoch": 0.3461, + "grad_norm": 0.3231448531150818, + "learning_rate": 0.000370789340571457, + "loss": 2.7540531158447266, + "step": 69220 + }, + { + "epoch": 0.34615, + "grad_norm": 0.36269134283065796, + "learning_rate": 0.00037075460963566454, + "loss": 2.7685752868652345, + "step": 69230 + }, + { + "epoch": 0.3462, + "grad_norm": 0.3883567452430725, + "learning_rate": 0.00037071987565987565, + "loss": 2.7697320938110352, + "step": 69240 + }, + { + "epoch": 0.34625, + "grad_norm": 0.35694533586502075, + "learning_rate": 0.00037068513864496453, + "loss": 2.7885313034057617, + "step": 69250 + }, + { + "epoch": 0.3463, + "grad_norm": 0.34241586923599243, + "learning_rate": 0.000370650398591806, + "loss": 2.756101608276367, + "step": 69260 + }, + { + "epoch": 0.34635, + "grad_norm": 0.31732845306396484, + "learning_rate": 0.0003706156555012744, + "loss": 2.8138254165649412, + "step": 69270 + }, + { + "epoch": 0.3464, + "grad_norm": 0.3169337213039398, + "learning_rate": 0.0003705809093742445, + "loss": 2.741836166381836, + "step": 69280 + }, + { + "epoch": 0.34645, + "grad_norm": 0.315535306930542, + "learning_rate": 0.0003705461602115911, + "loss": 2.770745277404785, + "step": 69290 + }, + { + "epoch": 0.3465, + "grad_norm": 0.3190414309501648, + "learning_rate": 0.00037051140801418875, + "loss": 2.700703239440918, + "step": 69300 + }, + { + "epoch": 0.34655, + "grad_norm": 0.29930707812309265, + "learning_rate": 0.00037047665278291264, + "loss": 2.767642021179199, + "step": 69310 + }, + { + "epoch": 0.3466, + "grad_norm": 0.36303627490997314, + "learning_rate": 0.0003704418945186376, + "loss": 2.7566909790039062, + "step": 69320 + }, + { + "epoch": 0.34665, + "grad_norm": 0.3375132083892822, + "learning_rate": 0.00037040713322223864, + "loss": 2.836109924316406, + "step": 69330 + }, + { + "epoch": 0.3467, + "grad_norm": 0.33823853731155396, + "learning_rate": 0.0003703723688945909, + "loss": 2.7450754165649416, + "step": 69340 + }, + { + "epoch": 0.34675, + "grad_norm": 0.3054700791835785, + "learning_rate": 0.00037033760153656957, + "loss": 2.8339086532592774, + "step": 69350 + }, + { + "epoch": 0.3468, + "grad_norm": 0.32551470398902893, + "learning_rate": 0.00037030283114904995, + "loss": 2.7484079360961915, + "step": 69360 + }, + { + "epoch": 0.34685, + "grad_norm": 0.361797958612442, + "learning_rate": 0.0003702680577329074, + "loss": 2.8016483306884767, + "step": 69370 + }, + { + "epoch": 0.3469, + "grad_norm": 0.3499944806098938, + "learning_rate": 0.0003702332812890171, + "loss": 2.788853073120117, + "step": 69380 + }, + { + "epoch": 0.34695, + "grad_norm": 0.38564321398735046, + "learning_rate": 0.000370198501818255, + "loss": 2.7943492889404298, + "step": 69390 + }, + { + "epoch": 0.347, + "grad_norm": 0.3376723825931549, + "learning_rate": 0.00037016371932149626, + "loss": 2.7144115447998045, + "step": 69400 + }, + { + "epoch": 0.34705, + "grad_norm": 0.32762035727500916, + "learning_rate": 0.00037012893379961674, + "loss": 2.7522838592529295, + "step": 69410 + }, + { + "epoch": 0.3471, + "grad_norm": 0.3451789617538452, + "learning_rate": 0.0003700941452534922, + "loss": 2.738846206665039, + "step": 69420 + }, + { + "epoch": 0.34715, + "grad_norm": 0.3374716341495514, + "learning_rate": 0.0003700593536839982, + "loss": 2.751092529296875, + "step": 69430 + }, + { + "epoch": 0.3472, + "grad_norm": 0.3057289123535156, + "learning_rate": 0.0003700245590920108, + "loss": 2.786903953552246, + "step": 69440 + }, + { + "epoch": 0.34725, + "grad_norm": 0.3269249200820923, + "learning_rate": 0.0003699897614784061, + "loss": 2.7920034408569334, + "step": 69450 + }, + { + "epoch": 0.3473, + "grad_norm": 0.3678661584854126, + "learning_rate": 0.0003699549608440598, + "loss": 2.7672338485717773, + "step": 69460 + }, + { + "epoch": 0.34735, + "grad_norm": 0.3531397581100464, + "learning_rate": 0.0003699201571898482, + "loss": 2.8100378036499025, + "step": 69470 + }, + { + "epoch": 0.3474, + "grad_norm": 0.3511013686656952, + "learning_rate": 0.0003698853505166474, + "loss": 2.6974071502685546, + "step": 69480 + }, + { + "epoch": 0.34745, + "grad_norm": 0.3265905976295471, + "learning_rate": 0.00036985054082533375, + "loss": 2.7927236557006836, + "step": 69490 + }, + { + "epoch": 0.3475, + "grad_norm": 0.3077230751514435, + "learning_rate": 0.0003698157281167835, + "loss": 2.7619091033935548, + "step": 69500 + }, + { + "epoch": 0.34755, + "grad_norm": 0.34797403216362, + "learning_rate": 0.00036978091239187315, + "loss": 2.713092803955078, + "step": 69510 + }, + { + "epoch": 0.3476, + "grad_norm": 0.3346079885959625, + "learning_rate": 0.0003697460936514791, + "loss": 2.7660112380981445, + "step": 69520 + }, + { + "epoch": 0.34765, + "grad_norm": 0.3176204562187195, + "learning_rate": 0.00036971127189647793, + "loss": 2.7355947494506836, + "step": 69530 + }, + { + "epoch": 0.3477, + "grad_norm": 0.3059046268463135, + "learning_rate": 0.0003696764471277463, + "loss": 2.7865081787109376, + "step": 69540 + }, + { + "epoch": 0.34775, + "grad_norm": 0.3342112600803375, + "learning_rate": 0.0003696416193461609, + "loss": 2.778111457824707, + "step": 69550 + }, + { + "epoch": 0.3478, + "grad_norm": 0.36216580867767334, + "learning_rate": 0.0003696067885525986, + "loss": 2.7628190994262694, + "step": 69560 + }, + { + "epoch": 0.34785, + "grad_norm": 0.3590550124645233, + "learning_rate": 0.0003695719547479361, + "loss": 2.766571044921875, + "step": 69570 + }, + { + "epoch": 0.3479, + "grad_norm": 0.3048454523086548, + "learning_rate": 0.0003695371179330505, + "loss": 2.7325155258178713, + "step": 69580 + }, + { + "epoch": 0.34795, + "grad_norm": 0.316730260848999, + "learning_rate": 0.00036950227810881876, + "loss": 2.7580228805541993, + "step": 69590 + }, + { + "epoch": 0.348, + "grad_norm": 0.3638244867324829, + "learning_rate": 0.00036946743527611794, + "loss": 2.746399688720703, + "step": 69600 + }, + { + "epoch": 0.34805, + "grad_norm": 0.3508071303367615, + "learning_rate": 0.0003694325894358252, + "loss": 2.729234504699707, + "step": 69610 + }, + { + "epoch": 0.3481, + "grad_norm": 0.3197428286075592, + "learning_rate": 0.00036939774058881785, + "loss": 2.7124841690063475, + "step": 69620 + }, + { + "epoch": 0.34815, + "grad_norm": 0.332974910736084, + "learning_rate": 0.0003693628887359732, + "loss": 2.716303253173828, + "step": 69630 + }, + { + "epoch": 0.3482, + "grad_norm": 0.3296947479248047, + "learning_rate": 0.00036932803387816865, + "loss": 2.823324203491211, + "step": 69640 + }, + { + "epoch": 0.34825, + "grad_norm": 0.3200066387653351, + "learning_rate": 0.00036929317601628155, + "loss": 2.7790590286254884, + "step": 69650 + }, + { + "epoch": 0.3483, + "grad_norm": 0.3063153028488159, + "learning_rate": 0.0003692583151511896, + "loss": 2.74292106628418, + "step": 69660 + }, + { + "epoch": 0.34835, + "grad_norm": 0.3247356712818146, + "learning_rate": 0.0003692234512837704, + "loss": 2.7793886184692385, + "step": 69670 + }, + { + "epoch": 0.3484, + "grad_norm": 0.36272796988487244, + "learning_rate": 0.00036918858441490157, + "loss": 2.7660106658935546, + "step": 69680 + }, + { + "epoch": 0.34845, + "grad_norm": 0.3484225869178772, + "learning_rate": 0.0003691537145454609, + "loss": 2.786085510253906, + "step": 69690 + }, + { + "epoch": 0.3485, + "grad_norm": 0.32125285267829895, + "learning_rate": 0.00036911884167632633, + "loss": 2.7568344116210937, + "step": 69700 + }, + { + "epoch": 0.34855, + "grad_norm": 0.326742559671402, + "learning_rate": 0.00036908396580837563, + "loss": 2.78907470703125, + "step": 69710 + }, + { + "epoch": 0.3486, + "grad_norm": 0.3529561161994934, + "learning_rate": 0.00036904908694248694, + "loss": 2.782762336730957, + "step": 69720 + }, + { + "epoch": 0.34865, + "grad_norm": 0.335958868265152, + "learning_rate": 0.0003690142050795383, + "loss": 2.6451364517211915, + "step": 69730 + }, + { + "epoch": 0.3487, + "grad_norm": 0.3205197751522064, + "learning_rate": 0.0003689793202204078, + "loss": 2.756363296508789, + "step": 69740 + }, + { + "epoch": 0.34875, + "grad_norm": 0.35242608189582825, + "learning_rate": 0.0003689444323659737, + "loss": 2.7014122009277344, + "step": 69750 + }, + { + "epoch": 0.3488, + "grad_norm": 0.31874343752861023, + "learning_rate": 0.0003689095415171143, + "loss": 2.7655414581298827, + "step": 69760 + }, + { + "epoch": 0.34885, + "grad_norm": 0.36028069257736206, + "learning_rate": 0.000368874647674708, + "loss": 2.7142913818359373, + "step": 69770 + }, + { + "epoch": 0.3489, + "grad_norm": 0.3205546736717224, + "learning_rate": 0.00036883975083963334, + "loss": 2.7686573028564454, + "step": 69780 + }, + { + "epoch": 0.34895, + "grad_norm": 0.3282606899738312, + "learning_rate": 0.00036880485101276864, + "loss": 2.7367109298706054, + "step": 69790 + }, + { + "epoch": 0.349, + "grad_norm": 0.32370662689208984, + "learning_rate": 0.00036876994819499265, + "loss": 2.761026382446289, + "step": 69800 + }, + { + "epoch": 0.34905, + "grad_norm": 0.30869269371032715, + "learning_rate": 0.00036873504238718396, + "loss": 2.7835840225219726, + "step": 69810 + }, + { + "epoch": 0.3491, + "grad_norm": 0.3380623161792755, + "learning_rate": 0.0003687001335902214, + "loss": 2.7880504608154295, + "step": 69820 + }, + { + "epoch": 0.34915, + "grad_norm": 0.32031378149986267, + "learning_rate": 0.0003686652218049838, + "loss": 2.823986053466797, + "step": 69830 + }, + { + "epoch": 0.3492, + "grad_norm": 0.38167905807495117, + "learning_rate": 0.00036863030703235, + "loss": 2.7098621368408202, + "step": 69840 + }, + { + "epoch": 0.34925, + "grad_norm": 0.32456088066101074, + "learning_rate": 0.000368595389273199, + "loss": 2.7518840789794923, + "step": 69850 + }, + { + "epoch": 0.3493, + "grad_norm": 0.40990203619003296, + "learning_rate": 0.00036856046852840995, + "loss": 2.729239654541016, + "step": 69860 + }, + { + "epoch": 0.34935, + "grad_norm": 0.3287733793258667, + "learning_rate": 0.0003685255447988618, + "loss": 2.6932859420776367, + "step": 69870 + }, + { + "epoch": 0.3494, + "grad_norm": 0.3034193515777588, + "learning_rate": 0.00036849061808543387, + "loss": 2.7721416473388674, + "step": 69880 + }, + { + "epoch": 0.34945, + "grad_norm": 0.41040486097335815, + "learning_rate": 0.0003684556883890055, + "loss": 2.7240428924560547, + "step": 69890 + }, + { + "epoch": 0.3495, + "grad_norm": 0.29979652166366577, + "learning_rate": 0.00036842075571045586, + "loss": 2.7977012634277343, + "step": 69900 + }, + { + "epoch": 0.34955, + "grad_norm": 0.3926505148410797, + "learning_rate": 0.0003683858200506645, + "loss": 2.769056510925293, + "step": 69910 + }, + { + "epoch": 0.3496, + "grad_norm": 0.3273223042488098, + "learning_rate": 0.00036835088141051097, + "loss": 2.72823486328125, + "step": 69920 + }, + { + "epoch": 0.34965, + "grad_norm": 0.3074716627597809, + "learning_rate": 0.00036831593979087476, + "loss": 2.804503059387207, + "step": 69930 + }, + { + "epoch": 0.3497, + "grad_norm": 0.3137149214744568, + "learning_rate": 0.00036828099519263557, + "loss": 2.7907148361206056, + "step": 69940 + }, + { + "epoch": 0.34975, + "grad_norm": 0.3501441180706024, + "learning_rate": 0.00036824604761667305, + "loss": 2.710919952392578, + "step": 69950 + }, + { + "epoch": 0.3498, + "grad_norm": 0.3435761034488678, + "learning_rate": 0.00036821109706386716, + "loss": 2.7440929412841797, + "step": 69960 + }, + { + "epoch": 0.34985, + "grad_norm": 0.3444962203502655, + "learning_rate": 0.00036817614353509766, + "loss": 2.739068603515625, + "step": 69970 + }, + { + "epoch": 0.3499, + "grad_norm": 0.34614452719688416, + "learning_rate": 0.00036814118703124455, + "loss": 2.763619613647461, + "step": 69980 + }, + { + "epoch": 0.34995, + "grad_norm": 0.3465234041213989, + "learning_rate": 0.0003681062275531878, + "loss": 2.738493728637695, + "step": 69990 + }, + { + "epoch": 0.35, + "grad_norm": 0.31707412004470825, + "learning_rate": 0.0003680712651018076, + "loss": 2.7888090133666994, + "step": 70000 + }, + { + "epoch": 0.35005, + "grad_norm": 0.39214465022087097, + "learning_rate": 0.00036803629967798406, + "loss": 2.784377098083496, + "step": 70010 + }, + { + "epoch": 0.3501, + "grad_norm": 0.3119412362575531, + "learning_rate": 0.00036800133128259756, + "loss": 2.7918121337890627, + "step": 70020 + }, + { + "epoch": 0.35015, + "grad_norm": 0.3286724090576172, + "learning_rate": 0.0003679663599165282, + "loss": 2.7788507461547853, + "step": 70030 + }, + { + "epoch": 0.3502, + "grad_norm": 0.3544609248638153, + "learning_rate": 0.00036793138558065665, + "loss": 2.7692420959472654, + "step": 70040 + }, + { + "epoch": 0.35025, + "grad_norm": 0.32117050886154175, + "learning_rate": 0.00036789640827586314, + "loss": 2.795867347717285, + "step": 70050 + }, + { + "epoch": 0.3503, + "grad_norm": 0.3354894518852234, + "learning_rate": 0.00036786142800302834, + "loss": 2.731528663635254, + "step": 70060 + }, + { + "epoch": 0.35035, + "grad_norm": 0.3139188885688782, + "learning_rate": 0.00036782644476303296, + "loss": 2.6954072952270507, + "step": 70070 + }, + { + "epoch": 0.3504, + "grad_norm": 0.33032935857772827, + "learning_rate": 0.00036779145855675763, + "loss": 2.7933488845825196, + "step": 70080 + }, + { + "epoch": 0.35045, + "grad_norm": 0.3549361526966095, + "learning_rate": 0.000367756469385083, + "loss": 2.774633026123047, + "step": 70090 + }, + { + "epoch": 0.3505, + "grad_norm": 0.3239253759384155, + "learning_rate": 0.00036772147724889015, + "loss": 2.7886051177978515, + "step": 70100 + }, + { + "epoch": 0.35055, + "grad_norm": 0.3378758728504181, + "learning_rate": 0.00036768648214905987, + "loss": 2.7415517807006835, + "step": 70110 + }, + { + "epoch": 0.3506, + "grad_norm": 0.316053569316864, + "learning_rate": 0.00036765148408647314, + "loss": 2.715938186645508, + "step": 70120 + }, + { + "epoch": 0.35065, + "grad_norm": 0.34594735503196716, + "learning_rate": 0.0003676164830620111, + "loss": 2.745365524291992, + "step": 70130 + }, + { + "epoch": 0.3507, + "grad_norm": 0.31351277232170105, + "learning_rate": 0.0003675814790765549, + "loss": 2.726608657836914, + "step": 70140 + }, + { + "epoch": 0.35075, + "grad_norm": 0.3676711320877075, + "learning_rate": 0.00036754647213098574, + "loss": 2.751791572570801, + "step": 70150 + }, + { + "epoch": 0.3508, + "grad_norm": 0.34682267904281616, + "learning_rate": 0.00036751146222618495, + "loss": 2.7438573837280273, + "step": 70160 + }, + { + "epoch": 0.35085, + "grad_norm": 0.3663967251777649, + "learning_rate": 0.00036747644936303386, + "loss": 2.7700826644897463, + "step": 70170 + }, + { + "epoch": 0.3509, + "grad_norm": 0.30866920948028564, + "learning_rate": 0.00036744143354241397, + "loss": 2.7264936447143553, + "step": 70180 + }, + { + "epoch": 0.35095, + "grad_norm": 0.3364311158657074, + "learning_rate": 0.0003674064147652068, + "loss": 2.762658882141113, + "step": 70190 + }, + { + "epoch": 0.351, + "grad_norm": 0.3480393886566162, + "learning_rate": 0.0003673713930322939, + "loss": 2.7220109939575194, + "step": 70200 + }, + { + "epoch": 0.35105, + "grad_norm": 0.34755489230155945, + "learning_rate": 0.0003673363683445569, + "loss": 2.7423164367675783, + "step": 70210 + }, + { + "epoch": 0.3511, + "grad_norm": 0.3175233006477356, + "learning_rate": 0.0003673013407028777, + "loss": 2.7341424942016603, + "step": 70220 + }, + { + "epoch": 0.35115, + "grad_norm": 0.31624916195869446, + "learning_rate": 0.00036726631010813794, + "loss": 2.7669578552246095, + "step": 70230 + }, + { + "epoch": 0.3512, + "grad_norm": 0.3404744863510132, + "learning_rate": 0.0003672312765612197, + "loss": 2.697296905517578, + "step": 70240 + }, + { + "epoch": 0.35125, + "grad_norm": 0.3668581247329712, + "learning_rate": 0.00036719624006300477, + "loss": 2.782074737548828, + "step": 70250 + }, + { + "epoch": 0.3513, + "grad_norm": 0.35021325945854187, + "learning_rate": 0.00036716120061437534, + "loss": 2.7447519302368164, + "step": 70260 + }, + { + "epoch": 0.35135, + "grad_norm": 0.3386077880859375, + "learning_rate": 0.00036712615821621347, + "loss": 2.7269355773925783, + "step": 70270 + }, + { + "epoch": 0.3514, + "grad_norm": 0.3240203857421875, + "learning_rate": 0.0003670911128694013, + "loss": 2.78558349609375, + "step": 70280 + }, + { + "epoch": 0.35145, + "grad_norm": 0.31445184350013733, + "learning_rate": 0.00036705606457482123, + "loss": 2.793785285949707, + "step": 70290 + }, + { + "epoch": 0.3515, + "grad_norm": 0.3382551372051239, + "learning_rate": 0.00036702101333335545, + "loss": 2.768193244934082, + "step": 70300 + }, + { + "epoch": 0.35155, + "grad_norm": 0.35016536712646484, + "learning_rate": 0.0003669859591458864, + "loss": 2.721543312072754, + "step": 70310 + }, + { + "epoch": 0.3516, + "grad_norm": 0.3351561725139618, + "learning_rate": 0.00036695090201329674, + "loss": 2.7821670532226563, + "step": 70320 + }, + { + "epoch": 0.35165, + "grad_norm": 0.3334018290042877, + "learning_rate": 0.0003669158419364688, + "loss": 2.752473831176758, + "step": 70330 + }, + { + "epoch": 0.3517, + "grad_norm": 0.3472733497619629, + "learning_rate": 0.00036688077891628535, + "loss": 2.7444372177124023, + "step": 70340 + }, + { + "epoch": 0.35175, + "grad_norm": 0.42084169387817383, + "learning_rate": 0.0003668457129536291, + "loss": 2.7498111724853516, + "step": 70350 + }, + { + "epoch": 0.3518, + "grad_norm": 0.34016257524490356, + "learning_rate": 0.00036681064404938275, + "loss": 2.752754783630371, + "step": 70360 + }, + { + "epoch": 0.35185, + "grad_norm": 0.339155912399292, + "learning_rate": 0.0003667755722044293, + "loss": 2.732756423950195, + "step": 70370 + }, + { + "epoch": 0.3519, + "grad_norm": 0.34871459007263184, + "learning_rate": 0.00036674049741965153, + "loss": 2.7872440338134767, + "step": 70380 + }, + { + "epoch": 0.35195, + "grad_norm": 0.3884022831916809, + "learning_rate": 0.0003667054196959325, + "loss": 2.7569555282592773, + "step": 70390 + }, + { + "epoch": 0.352, + "grad_norm": 0.4435460567474365, + "learning_rate": 0.0003666703390341554, + "loss": 2.7646888732910155, + "step": 70400 + }, + { + "epoch": 0.35205, + "grad_norm": 0.3225022256374359, + "learning_rate": 0.0003666352554352032, + "loss": 2.7564687728881836, + "step": 70410 + }, + { + "epoch": 0.3521, + "grad_norm": 0.3177005350589752, + "learning_rate": 0.00036660016889995925, + "loss": 2.7221309661865236, + "step": 70420 + }, + { + "epoch": 0.35215, + "grad_norm": 0.31869736313819885, + "learning_rate": 0.0003665650794293069, + "loss": 2.7797311782836913, + "step": 70430 + }, + { + "epoch": 0.3522, + "grad_norm": 0.35326993465423584, + "learning_rate": 0.0003665299870241293, + "loss": 2.7627431869506838, + "step": 70440 + }, + { + "epoch": 0.35225, + "grad_norm": 0.35871657729148865, + "learning_rate": 0.0003664948916853102, + "loss": 2.732851982116699, + "step": 70450 + }, + { + "epoch": 0.3523, + "grad_norm": 0.3646734058856964, + "learning_rate": 0.0003664597934137329, + "loss": 2.7891977310180662, + "step": 70460 + }, + { + "epoch": 0.35235, + "grad_norm": 0.3222251534461975, + "learning_rate": 0.00036642469221028113, + "loss": 2.7475025177001955, + "step": 70470 + }, + { + "epoch": 0.3524, + "grad_norm": 0.32870298624038696, + "learning_rate": 0.00036638958807583846, + "loss": 2.6800796508789064, + "step": 70480 + }, + { + "epoch": 0.35245, + "grad_norm": 0.3481248617172241, + "learning_rate": 0.0003663544810112887, + "loss": 2.733022689819336, + "step": 70490 + }, + { + "epoch": 0.3525, + "grad_norm": 0.3207465708255768, + "learning_rate": 0.0003663193710175157, + "loss": 2.7857284545898438, + "step": 70500 + }, + { + "epoch": 0.35255, + "grad_norm": 0.37359029054641724, + "learning_rate": 0.0003662842580954033, + "loss": 2.7479293823242186, + "step": 70510 + }, + { + "epoch": 0.3526, + "grad_norm": 0.34283655881881714, + "learning_rate": 0.00036624914224583547, + "loss": 2.729780578613281, + "step": 70520 + }, + { + "epoch": 0.35265, + "grad_norm": 0.3388737440109253, + "learning_rate": 0.00036621402346969625, + "loss": 2.750274658203125, + "step": 70530 + }, + { + "epoch": 0.3527, + "grad_norm": 0.32990002632141113, + "learning_rate": 0.0003661789017678698, + "loss": 2.755856513977051, + "step": 70540 + }, + { + "epoch": 0.35275, + "grad_norm": 0.3266100287437439, + "learning_rate": 0.0003661437771412402, + "loss": 2.7503679275512694, + "step": 70550 + }, + { + "epoch": 0.3528, + "grad_norm": 0.35048380494117737, + "learning_rate": 0.00036610864959069186, + "loss": 2.7272060394287108, + "step": 70560 + }, + { + "epoch": 0.35285, + "grad_norm": 0.36395302414894104, + "learning_rate": 0.00036607351911710896, + "loss": 2.7315227508544924, + "step": 70570 + }, + { + "epoch": 0.3529, + "grad_norm": 0.34737640619277954, + "learning_rate": 0.000366038385721376, + "loss": 2.757488250732422, + "step": 70580 + }, + { + "epoch": 0.35295, + "grad_norm": 0.30010899901390076, + "learning_rate": 0.0003660032494043776, + "loss": 2.6999950408935547, + "step": 70590 + }, + { + "epoch": 0.353, + "grad_norm": 0.34384679794311523, + "learning_rate": 0.00036596811016699807, + "loss": 2.7356021881103514, + "step": 70600 + }, + { + "epoch": 0.35305, + "grad_norm": 0.3385128974914551, + "learning_rate": 0.0003659329680101221, + "loss": 2.7601943969726563, + "step": 70610 + }, + { + "epoch": 0.3531, + "grad_norm": 0.3168568015098572, + "learning_rate": 0.0003658978229346345, + "loss": 2.709530830383301, + "step": 70620 + }, + { + "epoch": 0.35315, + "grad_norm": 0.4646669030189514, + "learning_rate": 0.00036586267494142, + "loss": 2.782592010498047, + "step": 70630 + }, + { + "epoch": 0.3532, + "grad_norm": 0.335753858089447, + "learning_rate": 0.0003658275240313633, + "loss": 2.7983335494995116, + "step": 70640 + }, + { + "epoch": 0.35325, + "grad_norm": 0.32005223631858826, + "learning_rate": 0.0003657923702053496, + "loss": 2.783604621887207, + "step": 70650 + }, + { + "epoch": 0.3533, + "grad_norm": 0.32149869203567505, + "learning_rate": 0.0003657572134642637, + "loss": 2.7945295333862306, + "step": 70660 + }, + { + "epoch": 0.35335, + "grad_norm": 0.33053451776504517, + "learning_rate": 0.00036572205380899076, + "loss": 2.72943229675293, + "step": 70670 + }, + { + "epoch": 0.3534, + "grad_norm": 0.31608664989471436, + "learning_rate": 0.0003656868912404159, + "loss": 2.731455612182617, + "step": 70680 + }, + { + "epoch": 0.35345, + "grad_norm": 0.33534765243530273, + "learning_rate": 0.0003656517257594242, + "loss": 2.772097587585449, + "step": 70690 + }, + { + "epoch": 0.3535, + "grad_norm": 0.3429161012172699, + "learning_rate": 0.0003656165573669012, + "loss": 2.757258987426758, + "step": 70700 + }, + { + "epoch": 0.35355, + "grad_norm": 0.3968760073184967, + "learning_rate": 0.0003655813860637321, + "loss": 2.7558746337890625, + "step": 70710 + }, + { + "epoch": 0.3536, + "grad_norm": 0.3959134519100189, + "learning_rate": 0.0003655462118508023, + "loss": 2.7649518966674806, + "step": 70720 + }, + { + "epoch": 0.35365, + "grad_norm": 0.32380831241607666, + "learning_rate": 0.00036551103472899757, + "loss": 2.790749740600586, + "step": 70730 + }, + { + "epoch": 0.3537, + "grad_norm": 0.37987592816352844, + "learning_rate": 0.00036547585469920316, + "loss": 2.7379810333251955, + "step": 70740 + }, + { + "epoch": 0.35375, + "grad_norm": 0.36415356397628784, + "learning_rate": 0.00036544067176230493, + "loss": 2.750294494628906, + "step": 70750 + }, + { + "epoch": 0.3538, + "grad_norm": 0.301055908203125, + "learning_rate": 0.0003654054859191886, + "loss": 2.783456230163574, + "step": 70760 + }, + { + "epoch": 0.35385, + "grad_norm": 0.3193199932575226, + "learning_rate": 0.0003653702971707399, + "loss": 2.7983156204223634, + "step": 70770 + }, + { + "epoch": 0.3539, + "grad_norm": 0.3200811743736267, + "learning_rate": 0.0003653351055178447, + "loss": 2.7896888732910154, + "step": 70780 + }, + { + "epoch": 0.35395, + "grad_norm": 0.36859703063964844, + "learning_rate": 0.00036529991096138903, + "loss": 2.770461082458496, + "step": 70790 + }, + { + "epoch": 0.354, + "grad_norm": 0.33074048161506653, + "learning_rate": 0.0003652647135022588, + "loss": 2.7550086975097656, + "step": 70800 + }, + { + "epoch": 0.35405, + "grad_norm": 0.3307930529117584, + "learning_rate": 0.00036522951314134036, + "loss": 2.6824995040893556, + "step": 70810 + }, + { + "epoch": 0.3541, + "grad_norm": 0.3399643003940582, + "learning_rate": 0.00036519430987951945, + "loss": 2.7221837997436524, + "step": 70820 + }, + { + "epoch": 0.35415, + "grad_norm": 0.3481927216053009, + "learning_rate": 0.00036515910371768276, + "loss": 2.7187017440795898, + "step": 70830 + }, + { + "epoch": 0.3542, + "grad_norm": 0.33921483159065247, + "learning_rate": 0.0003651238946567163, + "loss": 2.7711002349853517, + "step": 70840 + }, + { + "epoch": 0.35425, + "grad_norm": 0.3667307198047638, + "learning_rate": 0.0003650886826975066, + "loss": 2.7526206970214844, + "step": 70850 + }, + { + "epoch": 0.3543, + "grad_norm": 0.30539295077323914, + "learning_rate": 0.0003650534678409401, + "loss": 2.766703987121582, + "step": 70860 + }, + { + "epoch": 0.35435, + "grad_norm": 0.35058483481407166, + "learning_rate": 0.00036501825008790323, + "loss": 2.7601158142089846, + "step": 70870 + }, + { + "epoch": 0.3544, + "grad_norm": 0.3602858781814575, + "learning_rate": 0.00036498302943928276, + "loss": 2.735763931274414, + "step": 70880 + }, + { + "epoch": 0.35445, + "grad_norm": 0.34089091420173645, + "learning_rate": 0.0003649478058959653, + "loss": 2.812076187133789, + "step": 70890 + }, + { + "epoch": 0.3545, + "grad_norm": 0.3347007632255554, + "learning_rate": 0.00036491257945883747, + "loss": 2.8098636627197267, + "step": 70900 + }, + { + "epoch": 0.35455, + "grad_norm": 0.33201736211776733, + "learning_rate": 0.0003648773501287864, + "loss": 2.745606231689453, + "step": 70910 + }, + { + "epoch": 0.3546, + "grad_norm": 0.35025790333747864, + "learning_rate": 0.00036484211790669877, + "loss": 2.7327136993408203, + "step": 70920 + }, + { + "epoch": 0.35465, + "grad_norm": 0.3681603670120239, + "learning_rate": 0.0003648068827934615, + "loss": 2.765797996520996, + "step": 70930 + }, + { + "epoch": 0.3547, + "grad_norm": 0.32344940304756165, + "learning_rate": 0.0003647716447899618, + "loss": 2.7227582931518555, + "step": 70940 + }, + { + "epoch": 0.35475, + "grad_norm": 0.35130852460861206, + "learning_rate": 0.0003647364038970867, + "loss": 2.760018157958984, + "step": 70950 + }, + { + "epoch": 0.3548, + "grad_norm": 0.4083612263202667, + "learning_rate": 0.0003647011601157233, + "loss": 2.7763721466064455, + "step": 70960 + }, + { + "epoch": 0.35485, + "grad_norm": 0.401496022939682, + "learning_rate": 0.0003646659134467591, + "loss": 2.732193374633789, + "step": 70970 + }, + { + "epoch": 0.3549, + "grad_norm": 0.32320213317871094, + "learning_rate": 0.0003646306638910812, + "loss": 2.769613265991211, + "step": 70980 + }, + { + "epoch": 0.35495, + "grad_norm": 0.35264697670936584, + "learning_rate": 0.00036459541144957716, + "loss": 2.7769454956054687, + "step": 70990 + }, + { + "epoch": 0.355, + "grad_norm": 0.32625454664230347, + "learning_rate": 0.0003645601561231344, + "loss": 2.7758852005004884, + "step": 71000 + }, + { + "epoch": 0.35505, + "grad_norm": 0.3604283928871155, + "learning_rate": 0.0003645248979126404, + "loss": 2.746672248840332, + "step": 71010 + }, + { + "epoch": 0.3551, + "grad_norm": 0.368644118309021, + "learning_rate": 0.00036448963681898287, + "loss": 2.746267318725586, + "step": 71020 + }, + { + "epoch": 0.35515, + "grad_norm": 0.3426849842071533, + "learning_rate": 0.0003644543728430496, + "loss": 2.7584129333496095, + "step": 71030 + }, + { + "epoch": 0.3552, + "grad_norm": 0.3401089906692505, + "learning_rate": 0.00036441910598572816, + "loss": 2.7463634490966795, + "step": 71040 + }, + { + "epoch": 0.35525, + "grad_norm": 0.3056396245956421, + "learning_rate": 0.00036438383624790656, + "loss": 2.677216720581055, + "step": 71050 + }, + { + "epoch": 0.3553, + "grad_norm": 0.3289943337440491, + "learning_rate": 0.00036434856363047266, + "loss": 2.780274772644043, + "step": 71060 + }, + { + "epoch": 0.35535, + "grad_norm": 0.35639476776123047, + "learning_rate": 0.0003643132881343144, + "loss": 2.765736198425293, + "step": 71070 + }, + { + "epoch": 0.3554, + "grad_norm": 0.32040145993232727, + "learning_rate": 0.0003642780097603198, + "loss": 2.7802043914794923, + "step": 71080 + }, + { + "epoch": 0.35545, + "grad_norm": 0.3174695670604706, + "learning_rate": 0.00036424272850937715, + "loss": 2.7427268981933595, + "step": 71090 + }, + { + "epoch": 0.3555, + "grad_norm": 0.32470229268074036, + "learning_rate": 0.0003642074443823745, + "loss": 2.792214775085449, + "step": 71100 + }, + { + "epoch": 0.35555, + "grad_norm": 0.3197057843208313, + "learning_rate": 0.0003641721573802003, + "loss": 2.751507377624512, + "step": 71110 + }, + { + "epoch": 0.3556, + "grad_norm": 0.3587433099746704, + "learning_rate": 0.00036413686750374276, + "loss": 2.794012260437012, + "step": 71120 + }, + { + "epoch": 0.35565, + "grad_norm": 0.35790345072746277, + "learning_rate": 0.0003641015747538904, + "loss": 2.7572704315185548, + "step": 71130 + }, + { + "epoch": 0.3557, + "grad_norm": 0.35185399651527405, + "learning_rate": 0.00036406627913153155, + "loss": 2.8022859573364256, + "step": 71140 + }, + { + "epoch": 0.35575, + "grad_norm": 0.3923646807670593, + "learning_rate": 0.00036403098063755493, + "loss": 2.717852020263672, + "step": 71150 + }, + { + "epoch": 0.3558, + "grad_norm": 0.33697745203971863, + "learning_rate": 0.00036399567927284916, + "loss": 2.7203090667724608, + "step": 71160 + }, + { + "epoch": 0.35585, + "grad_norm": 0.3418007791042328, + "learning_rate": 0.00036396037503830285, + "loss": 2.751076889038086, + "step": 71170 + }, + { + "epoch": 0.3559, + "grad_norm": 0.30751511454582214, + "learning_rate": 0.00036392506793480494, + "loss": 2.740552520751953, + "step": 71180 + }, + { + "epoch": 0.35595, + "grad_norm": 0.40071025490760803, + "learning_rate": 0.0003638897579632442, + "loss": 2.7356731414794924, + "step": 71190 + }, + { + "epoch": 0.356, + "grad_norm": 0.33282971382141113, + "learning_rate": 0.0003638544451245095, + "loss": 2.7323856353759766, + "step": 71200 + }, + { + "epoch": 0.35605, + "grad_norm": 0.3427298963069916, + "learning_rate": 0.00036381912941949, + "loss": 2.7580360412597655, + "step": 71210 + }, + { + "epoch": 0.3561, + "grad_norm": 0.34284672141075134, + "learning_rate": 0.00036378381084907463, + "loss": 2.7506072998046873, + "step": 71220 + }, + { + "epoch": 0.35615, + "grad_norm": 0.3850841820240021, + "learning_rate": 0.00036374848941415254, + "loss": 2.728155517578125, + "step": 71230 + }, + { + "epoch": 0.3562, + "grad_norm": 0.3247060477733612, + "learning_rate": 0.0003637131651156131, + "loss": 2.774913024902344, + "step": 71240 + }, + { + "epoch": 0.35625, + "grad_norm": 0.3307344615459442, + "learning_rate": 0.00036367783795434536, + "loss": 2.7489667892456056, + "step": 71250 + }, + { + "epoch": 0.3563, + "grad_norm": 0.34614458680152893, + "learning_rate": 0.0003636425079312389, + "loss": 2.7592758178710937, + "step": 71260 + }, + { + "epoch": 0.35635, + "grad_norm": 0.37329089641571045, + "learning_rate": 0.000363607175047183, + "loss": 2.7067705154418946, + "step": 71270 + }, + { + "epoch": 0.3564, + "grad_norm": 0.38123008608818054, + "learning_rate": 0.0003635718393030673, + "loss": 2.7492799758911133, + "step": 71280 + }, + { + "epoch": 0.35645, + "grad_norm": 0.34621748328208923, + "learning_rate": 0.0003635365006997813, + "loss": 2.747458267211914, + "step": 71290 + }, + { + "epoch": 0.3565, + "grad_norm": 0.35490503907203674, + "learning_rate": 0.0003635011592382146, + "loss": 2.699990081787109, + "step": 71300 + }, + { + "epoch": 0.35655, + "grad_norm": 0.32806846499443054, + "learning_rate": 0.00036346581491925696, + "loss": 2.7433780670166015, + "step": 71310 + }, + { + "epoch": 0.3566, + "grad_norm": 0.3472064137458801, + "learning_rate": 0.0003634304677437982, + "loss": 2.775835418701172, + "step": 71320 + }, + { + "epoch": 0.35665, + "grad_norm": 0.35511305928230286, + "learning_rate": 0.0003633951177127282, + "loss": 2.716047668457031, + "step": 71330 + }, + { + "epoch": 0.3567, + "grad_norm": 0.36005622148513794, + "learning_rate": 0.00036335976482693687, + "loss": 2.7758785247802735, + "step": 71340 + }, + { + "epoch": 0.35675, + "grad_norm": 0.3221493363380432, + "learning_rate": 0.00036332440908731426, + "loss": 2.6965932846069336, + "step": 71350 + }, + { + "epoch": 0.3568, + "grad_norm": 0.32202908396720886, + "learning_rate": 0.0003632890504947503, + "loss": 2.7826786041259766, + "step": 71360 + }, + { + "epoch": 0.35685, + "grad_norm": 0.3024974465370178, + "learning_rate": 0.00036325368905013527, + "loss": 2.7758188247680664, + "step": 71370 + }, + { + "epoch": 0.3569, + "grad_norm": 0.34028998017311096, + "learning_rate": 0.0003632183247543595, + "loss": 2.7770097732543944, + "step": 71380 + }, + { + "epoch": 0.35695, + "grad_norm": 0.3147260546684265, + "learning_rate": 0.000363182957608313, + "loss": 2.7470659255981444, + "step": 71390 + }, + { + "epoch": 0.357, + "grad_norm": 0.3119431436061859, + "learning_rate": 0.00036314758761288643, + "loss": 2.719472312927246, + "step": 71400 + }, + { + "epoch": 0.35705, + "grad_norm": 0.30277225375175476, + "learning_rate": 0.00036311221476897, + "loss": 2.7939218521118163, + "step": 71410 + }, + { + "epoch": 0.3571, + "grad_norm": 0.35718923807144165, + "learning_rate": 0.0003630768390774544, + "loss": 2.709640121459961, + "step": 71420 + }, + { + "epoch": 0.35715, + "grad_norm": 0.3125021159648895, + "learning_rate": 0.00036304146053923005, + "loss": 2.7338422775268554, + "step": 71430 + }, + { + "epoch": 0.3572, + "grad_norm": 0.3499600291252136, + "learning_rate": 0.0003630060791551877, + "loss": 2.7684112548828126, + "step": 71440 + }, + { + "epoch": 0.35725, + "grad_norm": 0.37053927779197693, + "learning_rate": 0.0003629706949262181, + "loss": 2.744318199157715, + "step": 71450 + }, + { + "epoch": 0.3573, + "grad_norm": 0.36461347341537476, + "learning_rate": 0.00036293530785321205, + "loss": 2.710402488708496, + "step": 71460 + }, + { + "epoch": 0.35735, + "grad_norm": 0.32800713181495667, + "learning_rate": 0.0003628999179370602, + "loss": 2.803122138977051, + "step": 71470 + }, + { + "epoch": 0.3574, + "grad_norm": 0.33357828855514526, + "learning_rate": 0.0003628645251786539, + "loss": 2.8278059005737304, + "step": 71480 + }, + { + "epoch": 0.35745, + "grad_norm": 0.388764888048172, + "learning_rate": 0.0003628291295788838, + "loss": 2.7551816940307616, + "step": 71490 + }, + { + "epoch": 0.3575, + "grad_norm": 0.35090285539627075, + "learning_rate": 0.0003627937311386411, + "loss": 2.724920463562012, + "step": 71500 + }, + { + "epoch": 0.35755, + "grad_norm": 0.3704383075237274, + "learning_rate": 0.00036275832985881704, + "loss": 2.7301483154296875, + "step": 71510 + }, + { + "epoch": 0.3576, + "grad_norm": 0.3658599853515625, + "learning_rate": 0.0003627229257403027, + "loss": 2.791736602783203, + "step": 71520 + }, + { + "epoch": 0.35765, + "grad_norm": 0.34973448514938354, + "learning_rate": 0.0003626875187839896, + "loss": 2.738301086425781, + "step": 71530 + }, + { + "epoch": 0.3577, + "grad_norm": 0.37771883606910706, + "learning_rate": 0.00036265210899076887, + "loss": 2.7487892150878905, + "step": 71540 + }, + { + "epoch": 0.35775, + "grad_norm": 0.3159221112728119, + "learning_rate": 0.00036261669636153204, + "loss": 2.694995307922363, + "step": 71550 + }, + { + "epoch": 0.3578, + "grad_norm": 0.30931738018989563, + "learning_rate": 0.0003625812808971707, + "loss": 2.7752933502197266, + "step": 71560 + }, + { + "epoch": 0.35785, + "grad_norm": 0.2927280366420746, + "learning_rate": 0.0003625458625985763, + "loss": 2.7597625732421873, + "step": 71570 + }, + { + "epoch": 0.3579, + "grad_norm": 0.3565196692943573, + "learning_rate": 0.00036251044146664056, + "loss": 2.766142654418945, + "step": 71580 + }, + { + "epoch": 0.35795, + "grad_norm": 0.33332812786102295, + "learning_rate": 0.00036247501750225525, + "loss": 2.7384002685546873, + "step": 71590 + }, + { + "epoch": 0.358, + "grad_norm": 0.3518783748149872, + "learning_rate": 0.00036243959070631214, + "loss": 2.8227041244506834, + "step": 71600 + }, + { + "epoch": 0.35805, + "grad_norm": 0.32977235317230225, + "learning_rate": 0.000362404161079703, + "loss": 2.7923652648925783, + "step": 71610 + }, + { + "epoch": 0.3581, + "grad_norm": 0.3554164171218872, + "learning_rate": 0.00036236872862331995, + "loss": 2.7476062774658203, + "step": 71620 + }, + { + "epoch": 0.35815, + "grad_norm": 0.3496153652667999, + "learning_rate": 0.00036233329333805486, + "loss": 2.7232105255126955, + "step": 71630 + }, + { + "epoch": 0.3582, + "grad_norm": 0.35526084899902344, + "learning_rate": 0.00036229785522479986, + "loss": 2.8122392654418946, + "step": 71640 + }, + { + "epoch": 0.35825, + "grad_norm": 0.3349860906600952, + "learning_rate": 0.0003622624142844471, + "loss": 2.773759460449219, + "step": 71650 + }, + { + "epoch": 0.3583, + "grad_norm": 0.3466451168060303, + "learning_rate": 0.00036222697051788876, + "loss": 2.736077308654785, + "step": 71660 + }, + { + "epoch": 0.35835, + "grad_norm": 0.33633556962013245, + "learning_rate": 0.0003621915239260173, + "loss": 2.753680419921875, + "step": 71670 + }, + { + "epoch": 0.3584, + "grad_norm": 0.34547969698905945, + "learning_rate": 0.000362156074509725, + "loss": 2.7521419525146484, + "step": 71680 + }, + { + "epoch": 0.35845, + "grad_norm": 0.3245192766189575, + "learning_rate": 0.0003621206222699041, + "loss": 2.748539924621582, + "step": 71690 + }, + { + "epoch": 0.3585, + "grad_norm": 0.33878615498542786, + "learning_rate": 0.00036208516720744743, + "loss": 2.727352523803711, + "step": 71700 + }, + { + "epoch": 0.35855, + "grad_norm": 0.334840327501297, + "learning_rate": 0.0003620497093232473, + "loss": 2.724610710144043, + "step": 71710 + }, + { + "epoch": 0.3586, + "grad_norm": 0.3645494878292084, + "learning_rate": 0.0003620142486181966, + "loss": 2.784885215759277, + "step": 71720 + }, + { + "epoch": 0.35865, + "grad_norm": 0.33373966813087463, + "learning_rate": 0.00036197878509318794, + "loss": 2.762895202636719, + "step": 71730 + }, + { + "epoch": 0.3587, + "grad_norm": 0.377268522977829, + "learning_rate": 0.000361943318749114, + "loss": 2.7401411056518556, + "step": 71740 + }, + { + "epoch": 0.35875, + "grad_norm": 0.3363821506500244, + "learning_rate": 0.00036190784958686783, + "loss": 2.734913635253906, + "step": 71750 + }, + { + "epoch": 0.3588, + "grad_norm": 0.3521995544433594, + "learning_rate": 0.0003618723776073424, + "loss": 2.7581939697265625, + "step": 71760 + }, + { + "epoch": 0.35885, + "grad_norm": 0.33411651849746704, + "learning_rate": 0.00036183690281143047, + "loss": 2.736798858642578, + "step": 71770 + }, + { + "epoch": 0.3589, + "grad_norm": 0.3015058636665344, + "learning_rate": 0.0003618014252000253, + "loss": 2.7577545166015627, + "step": 71780 + }, + { + "epoch": 0.35895, + "grad_norm": 0.3524395227432251, + "learning_rate": 0.00036176594477402, + "loss": 2.7150150299072267, + "step": 71790 + }, + { + "epoch": 0.359, + "grad_norm": 0.3307517468929291, + "learning_rate": 0.0003617304615343078, + "loss": 2.7213743209838865, + "step": 71800 + }, + { + "epoch": 0.35905, + "grad_norm": 0.3387511074542999, + "learning_rate": 0.00036169497548178185, + "loss": 2.7288217544555664, + "step": 71810 + }, + { + "epoch": 0.3591, + "grad_norm": 0.31720253825187683, + "learning_rate": 0.0003616594866173358, + "loss": 2.7369207382202148, + "step": 71820 + }, + { + "epoch": 0.35915, + "grad_norm": 0.3450586795806885, + "learning_rate": 0.00036162399494186283, + "loss": 2.724440574645996, + "step": 71830 + }, + { + "epoch": 0.3592, + "grad_norm": 0.3265326917171478, + "learning_rate": 0.0003615885004562566, + "loss": 2.755583572387695, + "step": 71840 + }, + { + "epoch": 0.35925, + "grad_norm": 0.32408514618873596, + "learning_rate": 0.00036155300316141056, + "loss": 2.7176336288452148, + "step": 71850 + }, + { + "epoch": 0.3593, + "grad_norm": 0.3321201503276825, + "learning_rate": 0.0003615175030582185, + "loss": 2.7477443695068358, + "step": 71860 + }, + { + "epoch": 0.35935, + "grad_norm": 0.3758310079574585, + "learning_rate": 0.00036148200014757395, + "loss": 2.762008476257324, + "step": 71870 + }, + { + "epoch": 0.3594, + "grad_norm": 0.30603206157684326, + "learning_rate": 0.0003614464944303708, + "loss": 2.8139680862426757, + "step": 71880 + }, + { + "epoch": 0.35945, + "grad_norm": 0.3177337646484375, + "learning_rate": 0.00036141098590750293, + "loss": 2.7474809646606446, + "step": 71890 + }, + { + "epoch": 0.3595, + "grad_norm": 0.32505881786346436, + "learning_rate": 0.00036137547457986433, + "loss": 2.7748859405517576, + "step": 71900 + }, + { + "epoch": 0.35955, + "grad_norm": 0.37296435236930847, + "learning_rate": 0.0003613399604483487, + "loss": 2.707189750671387, + "step": 71910 + }, + { + "epoch": 0.3596, + "grad_norm": 0.34679025411605835, + "learning_rate": 0.00036130444351385044, + "loss": 2.7051734924316406, + "step": 71920 + }, + { + "epoch": 0.35965, + "grad_norm": 0.3369842767715454, + "learning_rate": 0.0003612689237772635, + "loss": 2.7152395248413086, + "step": 71930 + }, + { + "epoch": 0.3597, + "grad_norm": 0.42479777336120605, + "learning_rate": 0.0003612334012394822, + "loss": 2.726994514465332, + "step": 71940 + }, + { + "epoch": 0.35975, + "grad_norm": 0.3556099534034729, + "learning_rate": 0.00036119787590140073, + "loss": 2.7270254135131835, + "step": 71950 + }, + { + "epoch": 0.3598, + "grad_norm": 0.3290288746356964, + "learning_rate": 0.0003611623477639135, + "loss": 2.7574234008789062, + "step": 71960 + }, + { + "epoch": 0.35985, + "grad_norm": 0.3476720154285431, + "learning_rate": 0.00036112681682791485, + "loss": 2.771531677246094, + "step": 71970 + }, + { + "epoch": 0.3599, + "grad_norm": 0.34996798634529114, + "learning_rate": 0.00036109128309429943, + "loss": 2.7861726760864256, + "step": 71980 + }, + { + "epoch": 0.35995, + "grad_norm": 0.3302905559539795, + "learning_rate": 0.00036105574656396163, + "loss": 2.789584159851074, + "step": 71990 + }, + { + "epoch": 0.36, + "grad_norm": 0.3297533690929413, + "learning_rate": 0.0003610202072377962, + "loss": 2.7360095977783203, + "step": 72000 + }, + { + "epoch": 0.36005, + "grad_norm": 0.3567659854888916, + "learning_rate": 0.00036098466511669773, + "loss": 2.730657768249512, + "step": 72010 + }, + { + "epoch": 0.3601, + "grad_norm": 0.3859923183917999, + "learning_rate": 0.0003609491202015611, + "loss": 2.7162622451782226, + "step": 72020 + }, + { + "epoch": 0.36015, + "grad_norm": 0.3312918543815613, + "learning_rate": 0.00036091357249328106, + "loss": 2.7643476486206056, + "step": 72030 + }, + { + "epoch": 0.3602, + "grad_norm": 0.3238314986228943, + "learning_rate": 0.00036087802199275264, + "loss": 2.7050554275512697, + "step": 72040 + }, + { + "epoch": 0.36025, + "grad_norm": 0.35726553201675415, + "learning_rate": 0.0003608424687008707, + "loss": 2.7009883880615235, + "step": 72050 + }, + { + "epoch": 0.3603, + "grad_norm": 0.325511634349823, + "learning_rate": 0.0003608069126185304, + "loss": 2.738974952697754, + "step": 72060 + }, + { + "epoch": 0.36035, + "grad_norm": 0.3761571943759918, + "learning_rate": 0.00036077135374662685, + "loss": 2.7425432205200195, + "step": 72070 + }, + { + "epoch": 0.3604, + "grad_norm": 0.31709983944892883, + "learning_rate": 0.0003607357920860551, + "loss": 2.7326358795166015, + "step": 72080 + }, + { + "epoch": 0.36045, + "grad_norm": 0.348528653383255, + "learning_rate": 0.0003607002276377106, + "loss": 2.7334806442260744, + "step": 72090 + }, + { + "epoch": 0.3605, + "grad_norm": 0.3499636948108673, + "learning_rate": 0.00036066466040248857, + "loss": 2.7630378723144533, + "step": 72100 + }, + { + "epoch": 0.36055, + "grad_norm": 0.3523537516593933, + "learning_rate": 0.00036062909038128455, + "loss": 2.75030574798584, + "step": 72110 + }, + { + "epoch": 0.3606, + "grad_norm": 0.3172338902950287, + "learning_rate": 0.00036059351757499375, + "loss": 2.747047233581543, + "step": 72120 + }, + { + "epoch": 0.36065, + "grad_norm": 0.3095782399177551, + "learning_rate": 0.00036055794198451205, + "loss": 2.8143802642822267, + "step": 72130 + }, + { + "epoch": 0.3607, + "grad_norm": 0.33380475640296936, + "learning_rate": 0.00036052236361073484, + "loss": 2.7492744445800783, + "step": 72140 + }, + { + "epoch": 0.36075, + "grad_norm": 0.3014073073863983, + "learning_rate": 0.0003604867824545578, + "loss": 2.726824188232422, + "step": 72150 + }, + { + "epoch": 0.3608, + "grad_norm": 0.32749301195144653, + "learning_rate": 0.0003604511985168768, + "loss": 2.7269521713256837, + "step": 72160 + }, + { + "epoch": 0.36085, + "grad_norm": 0.3214188814163208, + "learning_rate": 0.0003604156117985876, + "loss": 2.7463544845581054, + "step": 72170 + }, + { + "epoch": 0.3609, + "grad_norm": 0.3433184325695038, + "learning_rate": 0.0003603800223005861, + "loss": 2.6935989379882814, + "step": 72180 + }, + { + "epoch": 0.36095, + "grad_norm": 0.33339470624923706, + "learning_rate": 0.0003603444300237684, + "loss": 2.763571548461914, + "step": 72190 + }, + { + "epoch": 0.361, + "grad_norm": 0.34734416007995605, + "learning_rate": 0.0003603088349690302, + "loss": 2.7078567504882813, + "step": 72200 + }, + { + "epoch": 0.36105, + "grad_norm": 0.32714951038360596, + "learning_rate": 0.00036027323713726803, + "loss": 2.7192779541015626, + "step": 72210 + }, + { + "epoch": 0.3611, + "grad_norm": 0.2972794473171234, + "learning_rate": 0.0003602376365293777, + "loss": 2.7760108947753905, + "step": 72220 + }, + { + "epoch": 0.36115, + "grad_norm": 0.37808316946029663, + "learning_rate": 0.00036020203314625566, + "loss": 2.747393798828125, + "step": 72230 + }, + { + "epoch": 0.3612, + "grad_norm": 0.34346652030944824, + "learning_rate": 0.0003601664269887982, + "loss": 2.694933319091797, + "step": 72240 + }, + { + "epoch": 0.36125, + "grad_norm": 0.3375234305858612, + "learning_rate": 0.0003601308180579017, + "loss": 2.797066307067871, + "step": 72250 + }, + { + "epoch": 0.3613, + "grad_norm": 0.3645692765712738, + "learning_rate": 0.0003600952063544625, + "loss": 2.7507740020751954, + "step": 72260 + }, + { + "epoch": 0.36135, + "grad_norm": 0.3303275406360626, + "learning_rate": 0.00036005959187937725, + "loss": 2.6541215896606447, + "step": 72270 + }, + { + "epoch": 0.3614, + "grad_norm": 0.33371615409851074, + "learning_rate": 0.00036002397463354255, + "loss": 2.812285041809082, + "step": 72280 + }, + { + "epoch": 0.36145, + "grad_norm": 0.30936554074287415, + "learning_rate": 0.00035998835461785493, + "loss": 2.7455022811889647, + "step": 72290 + }, + { + "epoch": 0.3615, + "grad_norm": 0.3256843686103821, + "learning_rate": 0.0003599527318332113, + "loss": 2.6804716110229494, + "step": 72300 + }, + { + "epoch": 0.36155, + "grad_norm": 0.35876619815826416, + "learning_rate": 0.00035991710628050834, + "loss": 2.735245704650879, + "step": 72310 + }, + { + "epoch": 0.3616, + "grad_norm": 0.3476503789424896, + "learning_rate": 0.00035988147796064295, + "loss": 2.7804677963256834, + "step": 72320 + }, + { + "epoch": 0.36165, + "grad_norm": 0.3169437050819397, + "learning_rate": 0.00035984584687451216, + "loss": 2.731311798095703, + "step": 72330 + }, + { + "epoch": 0.3617, + "grad_norm": 0.3603537380695343, + "learning_rate": 0.00035981021302301287, + "loss": 2.796003723144531, + "step": 72340 + }, + { + "epoch": 0.36175, + "grad_norm": 0.32913780212402344, + "learning_rate": 0.0003597745764070422, + "loss": 2.7334827423095702, + "step": 72350 + }, + { + "epoch": 0.3618, + "grad_norm": 0.328495591878891, + "learning_rate": 0.0003597389370274973, + "loss": 2.708582878112793, + "step": 72360 + }, + { + "epoch": 0.36185, + "grad_norm": 0.3293418288230896, + "learning_rate": 0.0003597032948852754, + "loss": 2.7899734497070314, + "step": 72370 + }, + { + "epoch": 0.3619, + "grad_norm": 0.3379918932914734, + "learning_rate": 0.00035966764998127376, + "loss": 2.792612075805664, + "step": 72380 + }, + { + "epoch": 0.36195, + "grad_norm": 0.41337597370147705, + "learning_rate": 0.0003596320023163898, + "loss": 2.7733741760253907, + "step": 72390 + }, + { + "epoch": 0.362, + "grad_norm": 0.334108829498291, + "learning_rate": 0.00035959635189152086, + "loss": 2.7452558517456054, + "step": 72400 + }, + { + "epoch": 0.36205, + "grad_norm": 0.37027138471603394, + "learning_rate": 0.0003595606987075646, + "loss": 2.735258865356445, + "step": 72410 + }, + { + "epoch": 0.3621, + "grad_norm": 0.39547818899154663, + "learning_rate": 0.0003595250427654183, + "loss": 2.7961605072021483, + "step": 72420 + }, + { + "epoch": 0.36215, + "grad_norm": 0.33604463934898376, + "learning_rate": 0.00035948938406598, + "loss": 2.725718688964844, + "step": 72430 + }, + { + "epoch": 0.3622, + "grad_norm": 0.33567050099372864, + "learning_rate": 0.00035945372261014704, + "loss": 2.6753124237060546, + "step": 72440 + }, + { + "epoch": 0.36225, + "grad_norm": 0.3412356972694397, + "learning_rate": 0.0003594180583988174, + "loss": 2.7924999237060546, + "step": 72450 + }, + { + "epoch": 0.3623, + "grad_norm": 0.34830403327941895, + "learning_rate": 0.0003593823914328888, + "loss": 2.7986562728881834, + "step": 72460 + }, + { + "epoch": 0.36235, + "grad_norm": 0.3341367244720459, + "learning_rate": 0.00035934672171325933, + "loss": 2.8174020767211916, + "step": 72470 + }, + { + "epoch": 0.3624, + "grad_norm": 0.33194059133529663, + "learning_rate": 0.0003593110492408268, + "loss": 2.7562515258789064, + "step": 72480 + }, + { + "epoch": 0.36245, + "grad_norm": 0.3514642119407654, + "learning_rate": 0.00035927537401648943, + "loss": 2.7296493530273436, + "step": 72490 + }, + { + "epoch": 0.3625, + "grad_norm": 0.32343313097953796, + "learning_rate": 0.0003592396960411451, + "loss": 2.8000232696533205, + "step": 72500 + }, + { + "epoch": 0.36255, + "grad_norm": 0.33183085918426514, + "learning_rate": 0.00035920401531569225, + "loss": 2.7956106185913088, + "step": 72510 + }, + { + "epoch": 0.3626, + "grad_norm": 0.34244632720947266, + "learning_rate": 0.00035916833184102904, + "loss": 2.744618797302246, + "step": 72520 + }, + { + "epoch": 0.36265, + "grad_norm": 0.3569324314594269, + "learning_rate": 0.0003591326456180538, + "loss": 2.733755683898926, + "step": 72530 + }, + { + "epoch": 0.3627, + "grad_norm": 0.3186372220516205, + "learning_rate": 0.0003590969566476649, + "loss": 2.7159690856933594, + "step": 72540 + }, + { + "epoch": 0.36275, + "grad_norm": 0.32251688838005066, + "learning_rate": 0.0003590612649307609, + "loss": 2.7290889739990236, + "step": 72550 + }, + { + "epoch": 0.3628, + "grad_norm": 0.35661542415618896, + "learning_rate": 0.0003590255704682402, + "loss": 2.7564878463745117, + "step": 72560 + }, + { + "epoch": 0.36285, + "grad_norm": 0.3452918529510498, + "learning_rate": 0.00035898987326100164, + "loss": 2.789171600341797, + "step": 72570 + }, + { + "epoch": 0.3629, + "grad_norm": 0.319518506526947, + "learning_rate": 0.00035895417330994363, + "loss": 2.7932201385498048, + "step": 72580 + }, + { + "epoch": 0.36295, + "grad_norm": 0.36717602610588074, + "learning_rate": 0.00035891847061596504, + "loss": 2.7824417114257813, + "step": 72590 + }, + { + "epoch": 0.363, + "grad_norm": 0.34522756934165955, + "learning_rate": 0.00035888276517996473, + "loss": 2.758173942565918, + "step": 72600 + }, + { + "epoch": 0.36305, + "grad_norm": 0.3314715325832367, + "learning_rate": 0.0003588470570028415, + "loss": 2.755493927001953, + "step": 72610 + }, + { + "epoch": 0.3631, + "grad_norm": 0.32702189683914185, + "learning_rate": 0.0003588113460854943, + "loss": 2.7637266159057616, + "step": 72620 + }, + { + "epoch": 0.36315, + "grad_norm": 0.3259526491165161, + "learning_rate": 0.0003587756324288223, + "loss": 2.7032684326171874, + "step": 72630 + }, + { + "epoch": 0.3632, + "grad_norm": 0.36377158761024475, + "learning_rate": 0.00035873991603372436, + "loss": 2.7806961059570314, + "step": 72640 + }, + { + "epoch": 0.36325, + "grad_norm": 0.3325626850128174, + "learning_rate": 0.00035870419690109986, + "loss": 2.793687629699707, + "step": 72650 + }, + { + "epoch": 0.3633, + "grad_norm": 0.31972411274909973, + "learning_rate": 0.00035866847503184787, + "loss": 2.75400390625, + "step": 72660 + }, + { + "epoch": 0.36335, + "grad_norm": 0.2730211913585663, + "learning_rate": 0.00035863275042686773, + "loss": 2.731527328491211, + "step": 72670 + }, + { + "epoch": 0.3634, + "grad_norm": 0.3456284999847412, + "learning_rate": 0.00035859702308705884, + "loss": 2.766158676147461, + "step": 72680 + }, + { + "epoch": 0.36345, + "grad_norm": 0.3218432366847992, + "learning_rate": 0.0003585612930133206, + "loss": 2.774829864501953, + "step": 72690 + }, + { + "epoch": 0.3635, + "grad_norm": 0.3456466794013977, + "learning_rate": 0.0003585255602065526, + "loss": 2.79268741607666, + "step": 72700 + }, + { + "epoch": 0.36355, + "grad_norm": 0.3402145802974701, + "learning_rate": 0.00035848982466765434, + "loss": 2.721563148498535, + "step": 72710 + }, + { + "epoch": 0.3636, + "grad_norm": 0.34449172019958496, + "learning_rate": 0.00035845408639752544, + "loss": 2.807883071899414, + "step": 72720 + }, + { + "epoch": 0.36365, + "grad_norm": 0.32356196641921997, + "learning_rate": 0.00035841834539706565, + "loss": 2.782817077636719, + "step": 72730 + }, + { + "epoch": 0.3637, + "grad_norm": 0.34144219756126404, + "learning_rate": 0.0003583826016671747, + "loss": 2.7674163818359374, + "step": 72740 + }, + { + "epoch": 0.36375, + "grad_norm": 0.3357720971107483, + "learning_rate": 0.00035834685520875256, + "loss": 2.764908218383789, + "step": 72750 + }, + { + "epoch": 0.3638, + "grad_norm": 0.3471919894218445, + "learning_rate": 0.00035831110602269897, + "loss": 2.7775283813476563, + "step": 72760 + }, + { + "epoch": 0.36385, + "grad_norm": 0.32112494111061096, + "learning_rate": 0.000358275354109914, + "loss": 2.762226867675781, + "step": 72770 + }, + { + "epoch": 0.3639, + "grad_norm": 0.33042457699775696, + "learning_rate": 0.0003582395994712978, + "loss": 2.7704326629638674, + "step": 72780 + }, + { + "epoch": 0.36395, + "grad_norm": 0.3191540539264679, + "learning_rate": 0.0003582038421077504, + "loss": 2.749299240112305, + "step": 72790 + }, + { + "epoch": 0.364, + "grad_norm": 0.3380267024040222, + "learning_rate": 0.0003581680820201719, + "loss": 2.799186134338379, + "step": 72800 + }, + { + "epoch": 0.36405, + "grad_norm": 0.3204616904258728, + "learning_rate": 0.00035813231920946284, + "loss": 2.7390729904174806, + "step": 72810 + }, + { + "epoch": 0.3641, + "grad_norm": 0.34375059604644775, + "learning_rate": 0.0003580965536765232, + "loss": 2.746826171875, + "step": 72820 + }, + { + "epoch": 0.36415, + "grad_norm": 0.38391393423080444, + "learning_rate": 0.0003580607854222536, + "loss": 2.710319709777832, + "step": 72830 + }, + { + "epoch": 0.3642, + "grad_norm": 0.3346812129020691, + "learning_rate": 0.0003580250144475544, + "loss": 2.749270439147949, + "step": 72840 + }, + { + "epoch": 0.36425, + "grad_norm": 0.39679020643234253, + "learning_rate": 0.00035798924075332636, + "loss": 2.759602928161621, + "step": 72850 + }, + { + "epoch": 0.3643, + "grad_norm": 0.35796231031417847, + "learning_rate": 0.0003579534643404697, + "loss": 2.7879570007324217, + "step": 72860 + }, + { + "epoch": 0.36435, + "grad_norm": 0.3745433986186981, + "learning_rate": 0.0003579176852098854, + "loss": 2.753549575805664, + "step": 72870 + }, + { + "epoch": 0.3644, + "grad_norm": 0.33317145705223083, + "learning_rate": 0.00035788190336247407, + "loss": 2.7244134902954102, + "step": 72880 + }, + { + "epoch": 0.36445, + "grad_norm": 0.34367889165878296, + "learning_rate": 0.00035784611879913653, + "loss": 2.7824731826782227, + "step": 72890 + }, + { + "epoch": 0.3645, + "grad_norm": 0.32909175753593445, + "learning_rate": 0.0003578103315207737, + "loss": 2.7545761108398437, + "step": 72900 + }, + { + "epoch": 0.36455, + "grad_norm": 0.36609476804733276, + "learning_rate": 0.0003577745415282864, + "loss": 2.7576475143432617, + "step": 72910 + }, + { + "epoch": 0.3646, + "grad_norm": 0.35270461440086365, + "learning_rate": 0.00035773874882257586, + "loss": 2.758392333984375, + "step": 72920 + }, + { + "epoch": 0.36465, + "grad_norm": 0.3663882613182068, + "learning_rate": 0.00035770295340454303, + "loss": 2.7328845977783205, + "step": 72930 + }, + { + "epoch": 0.3647, + "grad_norm": 0.3520002067089081, + "learning_rate": 0.00035766715527508905, + "loss": 2.7794464111328123, + "step": 72940 + }, + { + "epoch": 0.36475, + "grad_norm": 0.3719000816345215, + "learning_rate": 0.00035763135443511507, + "loss": 2.729887771606445, + "step": 72950 + }, + { + "epoch": 0.3648, + "grad_norm": 0.39009296894073486, + "learning_rate": 0.0003575955508855225, + "loss": 2.7345399856567383, + "step": 72960 + }, + { + "epoch": 0.36485, + "grad_norm": 0.3451180160045624, + "learning_rate": 0.0003575597446272127, + "loss": 2.731899642944336, + "step": 72970 + }, + { + "epoch": 0.3649, + "grad_norm": 0.3134165108203888, + "learning_rate": 0.0003575239356610871, + "loss": 2.722496223449707, + "step": 72980 + }, + { + "epoch": 0.36495, + "grad_norm": 0.34307417273521423, + "learning_rate": 0.000357488123988047, + "loss": 2.7758228302001955, + "step": 72990 + }, + { + "epoch": 0.365, + "grad_norm": 0.3488752543926239, + "learning_rate": 0.0003574523096089942, + "loss": 2.7138479232788084, + "step": 73000 + }, + { + "epoch": 0.36505, + "grad_norm": 0.330469012260437, + "learning_rate": 0.0003574164925248302, + "loss": 2.729911231994629, + "step": 73010 + }, + { + "epoch": 0.3651, + "grad_norm": 0.327625572681427, + "learning_rate": 0.0003573806727364567, + "loss": 2.7300437927246093, + "step": 73020 + }, + { + "epoch": 0.36515, + "grad_norm": 0.3426269292831421, + "learning_rate": 0.00035734485024477546, + "loss": 2.7526439666748046, + "step": 73030 + }, + { + "epoch": 0.3652, + "grad_norm": 0.40119874477386475, + "learning_rate": 0.0003573090250506884, + "loss": 2.6690095901489257, + "step": 73040 + }, + { + "epoch": 0.36525, + "grad_norm": 0.3436996042728424, + "learning_rate": 0.0003572731971550973, + "loss": 2.796294403076172, + "step": 73050 + }, + { + "epoch": 0.3653, + "grad_norm": 0.33782222867012024, + "learning_rate": 0.0003572373665589042, + "loss": 2.7292057037353517, + "step": 73060 + }, + { + "epoch": 0.36535, + "grad_norm": 0.36354300379753113, + "learning_rate": 0.0003572015332630111, + "loss": 2.761050796508789, + "step": 73070 + }, + { + "epoch": 0.3654, + "grad_norm": 0.3233734667301178, + "learning_rate": 0.0003571656972683202, + "loss": 2.7605659484863283, + "step": 73080 + }, + { + "epoch": 0.36545, + "grad_norm": 0.3156599700450897, + "learning_rate": 0.00035712985857573343, + "loss": 2.749865913391113, + "step": 73090 + }, + { + "epoch": 0.3655, + "grad_norm": 0.31065112352371216, + "learning_rate": 0.00035709401718615324, + "loss": 2.736884117126465, + "step": 73100 + }, + { + "epoch": 0.36555, + "grad_norm": 0.3548067510128021, + "learning_rate": 0.00035705817310048197, + "loss": 2.7640911102294923, + "step": 73110 + }, + { + "epoch": 0.3656, + "grad_norm": 0.31530699133872986, + "learning_rate": 0.0003570223263196218, + "loss": 2.7453166961669924, + "step": 73120 + }, + { + "epoch": 0.36565, + "grad_norm": 0.327047735452652, + "learning_rate": 0.00035698647684447526, + "loss": 2.7320526123046873, + "step": 73130 + }, + { + "epoch": 0.3657, + "grad_norm": 0.3349829316139221, + "learning_rate": 0.000356950624675945, + "loss": 2.7702165603637696, + "step": 73140 + }, + { + "epoch": 0.36575, + "grad_norm": 0.3458349406719208, + "learning_rate": 0.00035691476981493343, + "loss": 2.7813724517822265, + "step": 73150 + }, + { + "epoch": 0.3658, + "grad_norm": 0.30955272912979126, + "learning_rate": 0.0003568789122623433, + "loss": 2.711921501159668, + "step": 73160 + }, + { + "epoch": 0.36585, + "grad_norm": 0.3029145300388336, + "learning_rate": 0.0003568430520190771, + "loss": 2.757832908630371, + "step": 73170 + }, + { + "epoch": 0.3659, + "grad_norm": 0.32149747014045715, + "learning_rate": 0.00035680718908603795, + "loss": 2.7021955490112304, + "step": 73180 + }, + { + "epoch": 0.36595, + "grad_norm": 0.30334970355033875, + "learning_rate": 0.00035677132346412845, + "loss": 2.7528364181518556, + "step": 73190 + }, + { + "epoch": 0.366, + "grad_norm": 0.3433597683906555, + "learning_rate": 0.00035673545515425164, + "loss": 2.7291112899780274, + "step": 73200 + }, + { + "epoch": 0.36605, + "grad_norm": 0.347989559173584, + "learning_rate": 0.00035669958415731034, + "loss": 2.7786838531494142, + "step": 73210 + }, + { + "epoch": 0.3661, + "grad_norm": 0.320700466632843, + "learning_rate": 0.00035666371047420785, + "loss": 2.783252143859863, + "step": 73220 + }, + { + "epoch": 0.36615, + "grad_norm": 0.32804667949676514, + "learning_rate": 0.00035662783410584715, + "loss": 2.727341651916504, + "step": 73230 + }, + { + "epoch": 0.3662, + "grad_norm": 0.36363333463668823, + "learning_rate": 0.00035659195505313143, + "loss": 2.743922233581543, + "step": 73240 + }, + { + "epoch": 0.36625, + "grad_norm": 0.34641483426094055, + "learning_rate": 0.0003565560733169639, + "loss": 2.7324169158935545, + "step": 73250 + }, + { + "epoch": 0.3663, + "grad_norm": 0.3429851531982422, + "learning_rate": 0.00035652018889824794, + "loss": 2.7416240692138674, + "step": 73260 + }, + { + "epoch": 0.36635, + "grad_norm": 0.33236393332481384, + "learning_rate": 0.000356484301797887, + "loss": 2.7267627716064453, + "step": 73270 + }, + { + "epoch": 0.3664, + "grad_norm": 0.35570088028907776, + "learning_rate": 0.00035644841201678445, + "loss": 2.7687299728393553, + "step": 73280 + }, + { + "epoch": 0.36645, + "grad_norm": 0.31407392024993896, + "learning_rate": 0.0003564125195558438, + "loss": 2.7329996109008787, + "step": 73290 + }, + { + "epoch": 0.3665, + "grad_norm": 0.35891714692115784, + "learning_rate": 0.00035637662441596874, + "loss": 2.7811939239501955, + "step": 73300 + }, + { + "epoch": 0.36655, + "grad_norm": 0.33315610885620117, + "learning_rate": 0.0003563407265980628, + "loss": 2.762859916687012, + "step": 73310 + }, + { + "epoch": 0.3666, + "grad_norm": 0.33329057693481445, + "learning_rate": 0.00035630482610302984, + "loss": 2.7574758529663086, + "step": 73320 + }, + { + "epoch": 0.36665, + "grad_norm": 0.36714595556259155, + "learning_rate": 0.0003562689229317736, + "loss": 2.734388542175293, + "step": 73330 + }, + { + "epoch": 0.3667, + "grad_norm": 0.33029937744140625, + "learning_rate": 0.0003562330170851979, + "loss": 2.7415374755859374, + "step": 73340 + }, + { + "epoch": 0.36675, + "grad_norm": 0.32199811935424805, + "learning_rate": 0.00035619710856420666, + "loss": 2.769732666015625, + "step": 73350 + }, + { + "epoch": 0.3668, + "grad_norm": 0.31853923201560974, + "learning_rate": 0.00035616119736970404, + "loss": 2.7502838134765626, + "step": 73360 + }, + { + "epoch": 0.36685, + "grad_norm": 0.3422892391681671, + "learning_rate": 0.0003561252835025939, + "loss": 2.7870336532592774, + "step": 73370 + }, + { + "epoch": 0.3669, + "grad_norm": 0.32076287269592285, + "learning_rate": 0.00035608936696378046, + "loss": 2.772100830078125, + "step": 73380 + }, + { + "epoch": 0.36695, + "grad_norm": 0.4206850826740265, + "learning_rate": 0.000356053447754168, + "loss": 2.7806371688842773, + "step": 73390 + }, + { + "epoch": 0.367, + "grad_norm": 0.3164833188056946, + "learning_rate": 0.0003560175258746606, + "loss": 2.7052585601806642, + "step": 73400 + }, + { + "epoch": 0.36705, + "grad_norm": 0.3549223244190216, + "learning_rate": 0.0003559816013261627, + "loss": 2.6907272338867188, + "step": 73410 + }, + { + "epoch": 0.3671, + "grad_norm": 0.32567375898361206, + "learning_rate": 0.00035594567410957874, + "loss": 2.707926559448242, + "step": 73420 + }, + { + "epoch": 0.36715, + "grad_norm": 0.3313060700893402, + "learning_rate": 0.00035590974422581315, + "loss": 2.7532075881958007, + "step": 73430 + }, + { + "epoch": 0.3672, + "grad_norm": 0.37517738342285156, + "learning_rate": 0.0003558738116757704, + "loss": 2.7406278610229493, + "step": 73440 + }, + { + "epoch": 0.36725, + "grad_norm": 0.30997565388679504, + "learning_rate": 0.00035583787646035515, + "loss": 2.740657997131348, + "step": 73450 + }, + { + "epoch": 0.3673, + "grad_norm": 0.33128929138183594, + "learning_rate": 0.0003558019385804722, + "loss": 2.720555877685547, + "step": 73460 + }, + { + "epoch": 0.36735, + "grad_norm": 0.37298691272735596, + "learning_rate": 0.00035576599803702604, + "loss": 2.7461565017700194, + "step": 73470 + }, + { + "epoch": 0.3674, + "grad_norm": 0.3350296914577484, + "learning_rate": 0.00035573005483092155, + "loss": 2.7865583419799806, + "step": 73480 + }, + { + "epoch": 0.36745, + "grad_norm": 0.3198257386684418, + "learning_rate": 0.00035569410896306366, + "loss": 2.7863473892211914, + "step": 73490 + }, + { + "epoch": 0.3675, + "grad_norm": 0.33629533648490906, + "learning_rate": 0.0003556581604343574, + "loss": 2.7101772308349608, + "step": 73500 + }, + { + "epoch": 0.36755, + "grad_norm": 0.2977299392223358, + "learning_rate": 0.0003556222092457075, + "loss": 2.768904113769531, + "step": 73510 + }, + { + "epoch": 0.3676, + "grad_norm": 0.34316152334213257, + "learning_rate": 0.00035558625539801927, + "loss": 2.7595447540283202, + "step": 73520 + }, + { + "epoch": 0.36765, + "grad_norm": 0.3502255976200104, + "learning_rate": 0.00035555029889219777, + "loss": 2.734531593322754, + "step": 73530 + }, + { + "epoch": 0.3677, + "grad_norm": 0.33936062455177307, + "learning_rate": 0.00035551433972914816, + "loss": 2.736397361755371, + "step": 73540 + }, + { + "epoch": 0.36775, + "grad_norm": 0.34132513403892517, + "learning_rate": 0.0003554783779097758, + "loss": 2.75348014831543, + "step": 73550 + }, + { + "epoch": 0.3678, + "grad_norm": 0.4000563621520996, + "learning_rate": 0.0003554424134349859, + "loss": 2.7798877716064454, + "step": 73560 + }, + { + "epoch": 0.36785, + "grad_norm": 0.3619842231273651, + "learning_rate": 0.000355406446305684, + "loss": 2.765119743347168, + "step": 73570 + }, + { + "epoch": 0.3679, + "grad_norm": 0.30955028533935547, + "learning_rate": 0.00035537047652277554, + "loss": 2.7640865325927733, + "step": 73580 + }, + { + "epoch": 0.36795, + "grad_norm": 0.34135138988494873, + "learning_rate": 0.00035533450408716595, + "loss": 2.7633783340454103, + "step": 73590 + }, + { + "epoch": 0.368, + "grad_norm": 0.33162036538124084, + "learning_rate": 0.0003552985289997611, + "loss": 2.754544258117676, + "step": 73600 + }, + { + "epoch": 0.36805, + "grad_norm": 0.3807028830051422, + "learning_rate": 0.00035526255126146634, + "loss": 2.7458906173706055, + "step": 73610 + }, + { + "epoch": 0.3681, + "grad_norm": 0.3279631435871124, + "learning_rate": 0.0003552265708731875, + "loss": 2.769580078125, + "step": 73620 + }, + { + "epoch": 0.36815, + "grad_norm": 0.35813337564468384, + "learning_rate": 0.0003551905878358306, + "loss": 2.710239791870117, + "step": 73630 + }, + { + "epoch": 0.3682, + "grad_norm": 0.42102935910224915, + "learning_rate": 0.00035515460215030124, + "loss": 2.7013322830200197, + "step": 73640 + }, + { + "epoch": 0.36825, + "grad_norm": 0.32969239354133606, + "learning_rate": 0.00035511861381750553, + "loss": 2.6980051040649413, + "step": 73650 + }, + { + "epoch": 0.3683, + "grad_norm": 0.32395821809768677, + "learning_rate": 0.00035508262283834947, + "loss": 2.710295295715332, + "step": 73660 + }, + { + "epoch": 0.36835, + "grad_norm": 0.32029861211776733, + "learning_rate": 0.000355046629213739, + "loss": 2.765619659423828, + "step": 73670 + }, + { + "epoch": 0.3684, + "grad_norm": 0.3320925831794739, + "learning_rate": 0.0003550106329445804, + "loss": 2.7148365020751952, + "step": 73680 + }, + { + "epoch": 0.36845, + "grad_norm": 0.3440917134284973, + "learning_rate": 0.0003549746340317797, + "loss": 2.7383148193359377, + "step": 73690 + }, + { + "epoch": 0.3685, + "grad_norm": 0.35279881954193115, + "learning_rate": 0.00035493863247624337, + "loss": 2.7688512802124023, + "step": 73700 + }, + { + "epoch": 0.36855, + "grad_norm": 0.3175922930240631, + "learning_rate": 0.00035490262827887764, + "loss": 2.7015153884887697, + "step": 73710 + }, + { + "epoch": 0.3686, + "grad_norm": 0.31075698137283325, + "learning_rate": 0.000354866621440589, + "loss": 2.777872657775879, + "step": 73720 + }, + { + "epoch": 0.36865, + "grad_norm": 0.39659595489501953, + "learning_rate": 0.00035483061196228384, + "loss": 2.76175594329834, + "step": 73730 + }, + { + "epoch": 0.3687, + "grad_norm": 0.3240712285041809, + "learning_rate": 0.00035479459984486873, + "loss": 2.7324197769165037, + "step": 73740 + }, + { + "epoch": 0.36875, + "grad_norm": 0.3155030906200409, + "learning_rate": 0.00035475858508925026, + "loss": 2.810909461975098, + "step": 73750 + }, + { + "epoch": 0.3688, + "grad_norm": 0.3221328556537628, + "learning_rate": 0.00035472256769633517, + "loss": 2.7235528945922853, + "step": 73760 + }, + { + "epoch": 0.36885, + "grad_norm": 0.32620498538017273, + "learning_rate": 0.00035468654766703, + "loss": 2.7392047882080077, + "step": 73770 + }, + { + "epoch": 0.3689, + "grad_norm": 0.3878863751888275, + "learning_rate": 0.0003546505250022418, + "loss": 2.7433515548706056, + "step": 73780 + }, + { + "epoch": 0.36895, + "grad_norm": 0.36651042103767395, + "learning_rate": 0.00035461449970287733, + "loss": 2.781142807006836, + "step": 73790 + }, + { + "epoch": 0.369, + "grad_norm": 0.34655946493148804, + "learning_rate": 0.0003545784717698435, + "loss": 2.764843559265137, + "step": 73800 + }, + { + "epoch": 0.36905, + "grad_norm": 0.3401358127593994, + "learning_rate": 0.0003545424412040473, + "loss": 2.7720956802368164, + "step": 73810 + }, + { + "epoch": 0.3691, + "grad_norm": 0.3333001136779785, + "learning_rate": 0.00035450640800639596, + "loss": 2.7224651336669923, + "step": 73820 + }, + { + "epoch": 0.36915, + "grad_norm": 0.3684108853340149, + "learning_rate": 0.00035447037217779635, + "loss": 2.718475341796875, + "step": 73830 + }, + { + "epoch": 0.3692, + "grad_norm": 0.3717421293258667, + "learning_rate": 0.0003544343337191559, + "loss": 2.6863203048706055, + "step": 73840 + }, + { + "epoch": 0.36925, + "grad_norm": 0.3795083463191986, + "learning_rate": 0.00035439829263138187, + "loss": 2.7471549987792967, + "step": 73850 + }, + { + "epoch": 0.3693, + "grad_norm": 0.3315773606300354, + "learning_rate": 0.00035436224891538136, + "loss": 2.7263307571411133, + "step": 73860 + }, + { + "epoch": 0.36935, + "grad_norm": 0.3450697958469391, + "learning_rate": 0.00035432620257206207, + "loss": 2.7315155029296876, + "step": 73870 + }, + { + "epoch": 0.3694, + "grad_norm": 0.35449841618537903, + "learning_rate": 0.00035429015360233126, + "loss": 2.7263776779174806, + "step": 73880 + }, + { + "epoch": 0.36945, + "grad_norm": 0.34673410654067993, + "learning_rate": 0.00035425410200709654, + "loss": 2.8043996810913088, + "step": 73890 + }, + { + "epoch": 0.3695, + "grad_norm": 0.34746116399765015, + "learning_rate": 0.00035421804778726547, + "loss": 2.775592231750488, + "step": 73900 + }, + { + "epoch": 0.36955, + "grad_norm": 0.3411979377269745, + "learning_rate": 0.00035418199094374575, + "loss": 2.7216705322265624, + "step": 73910 + }, + { + "epoch": 0.3696, + "grad_norm": 0.3431650400161743, + "learning_rate": 0.00035414593147744514, + "loss": 2.7637115478515626, + "step": 73920 + }, + { + "epoch": 0.36965, + "grad_norm": 0.3255371153354645, + "learning_rate": 0.0003541098693892714, + "loss": 2.7503072738647463, + "step": 73930 + }, + { + "epoch": 0.3697, + "grad_norm": 0.3678377568721771, + "learning_rate": 0.00035407380468013236, + "loss": 2.6947023391723635, + "step": 73940 + }, + { + "epoch": 0.36975, + "grad_norm": 0.36472055315971375, + "learning_rate": 0.00035403773735093603, + "loss": 2.7143299102783205, + "step": 73950 + }, + { + "epoch": 0.3698, + "grad_norm": 0.3566860556602478, + "learning_rate": 0.00035400166740259033, + "loss": 2.684691047668457, + "step": 73960 + }, + { + "epoch": 0.36985, + "grad_norm": 0.33121412992477417, + "learning_rate": 0.00035396559483600335, + "loss": 2.769158363342285, + "step": 73970 + }, + { + "epoch": 0.3699, + "grad_norm": 0.31576433777809143, + "learning_rate": 0.00035392951965208314, + "loss": 2.767110824584961, + "step": 73980 + }, + { + "epoch": 0.36995, + "grad_norm": 0.3051203787326813, + "learning_rate": 0.00035389344185173807, + "loss": 2.788186264038086, + "step": 73990 + }, + { + "epoch": 0.37, + "grad_norm": 0.35185596346855164, + "learning_rate": 0.0003538573614358762, + "loss": 2.760975646972656, + "step": 74000 + }, + { + "epoch": 0.37005, + "grad_norm": 0.2975808084011078, + "learning_rate": 0.00035382127840540614, + "loss": 2.75350341796875, + "step": 74010 + }, + { + "epoch": 0.3701, + "grad_norm": 0.30999553203582764, + "learning_rate": 0.00035378519276123586, + "loss": 2.7364845275878906, + "step": 74020 + }, + { + "epoch": 0.37015, + "grad_norm": 0.3269255459308624, + "learning_rate": 0.0003537491045042742, + "loss": 2.7613994598388674, + "step": 74030 + }, + { + "epoch": 0.3702, + "grad_norm": 0.32954084873199463, + "learning_rate": 0.00035371301363542945, + "loss": 2.6960475921630858, + "step": 74040 + }, + { + "epoch": 0.37025, + "grad_norm": 0.33463314175605774, + "learning_rate": 0.0003536769201556103, + "loss": 2.7520235061645506, + "step": 74050 + }, + { + "epoch": 0.3703, + "grad_norm": 0.3342912495136261, + "learning_rate": 0.0003536408240657254, + "loss": 2.7095571517944337, + "step": 74060 + }, + { + "epoch": 0.37035, + "grad_norm": 0.38778156042099, + "learning_rate": 0.0003536047253666834, + "loss": 2.7748327255249023, + "step": 74070 + }, + { + "epoch": 0.3704, + "grad_norm": 0.32055220007896423, + "learning_rate": 0.00035356862405939315, + "loss": 2.729847526550293, + "step": 74080 + }, + { + "epoch": 0.37045, + "grad_norm": 0.36046266555786133, + "learning_rate": 0.00035353252014476357, + "loss": 2.786068344116211, + "step": 74090 + }, + { + "epoch": 0.3705, + "grad_norm": 0.3276388347148895, + "learning_rate": 0.00035349641362370334, + "loss": 2.771139907836914, + "step": 74100 + }, + { + "epoch": 0.37055, + "grad_norm": 0.35038766264915466, + "learning_rate": 0.00035346030449712163, + "loss": 2.7813629150390624, + "step": 74110 + }, + { + "epoch": 0.3706, + "grad_norm": 0.32199832797050476, + "learning_rate": 0.00035342419276592754, + "loss": 2.733054351806641, + "step": 74120 + }, + { + "epoch": 0.37065, + "grad_norm": 0.373421847820282, + "learning_rate": 0.00035338807843102995, + "loss": 2.749858856201172, + "step": 74130 + }, + { + "epoch": 0.3707, + "grad_norm": 0.35618191957473755, + "learning_rate": 0.00035335196149333825, + "loss": 2.7558212280273438, + "step": 74140 + }, + { + "epoch": 0.37075, + "grad_norm": 0.3320159912109375, + "learning_rate": 0.00035331584195376163, + "loss": 2.743123245239258, + "step": 74150 + }, + { + "epoch": 0.3708, + "grad_norm": 0.3871246576309204, + "learning_rate": 0.0003532797198132093, + "loss": 2.7790151596069337, + "step": 74160 + }, + { + "epoch": 0.37085, + "grad_norm": 0.3087109923362732, + "learning_rate": 0.00035324359507259086, + "loss": 2.800386428833008, + "step": 74170 + }, + { + "epoch": 0.3709, + "grad_norm": 0.3424051105976105, + "learning_rate": 0.0003532074677328155, + "loss": 2.690110206604004, + "step": 74180 + }, + { + "epoch": 0.37095, + "grad_norm": 0.3262627422809601, + "learning_rate": 0.00035317133779479284, + "loss": 2.720518112182617, + "step": 74190 + }, + { + "epoch": 0.371, + "grad_norm": 0.3614110052585602, + "learning_rate": 0.00035313520525943236, + "loss": 2.747167778015137, + "step": 74200 + }, + { + "epoch": 0.37105, + "grad_norm": 0.3379867672920227, + "learning_rate": 0.0003530990701276439, + "loss": 2.771078872680664, + "step": 74210 + }, + { + "epoch": 0.3711, + "grad_norm": 0.30792513489723206, + "learning_rate": 0.00035306293240033697, + "loss": 2.6976806640625, + "step": 74220 + }, + { + "epoch": 0.37115, + "grad_norm": 0.33707159757614136, + "learning_rate": 0.00035302679207842145, + "loss": 2.7610200881958007, + "step": 74230 + }, + { + "epoch": 0.3712, + "grad_norm": 0.3440256416797638, + "learning_rate": 0.0003529906491628071, + "loss": 2.7050094604492188, + "step": 74240 + }, + { + "epoch": 0.37125, + "grad_norm": 0.3198845684528351, + "learning_rate": 0.0003529545036544038, + "loss": 2.7357627868652346, + "step": 74250 + }, + { + "epoch": 0.3713, + "grad_norm": 0.3098306655883789, + "learning_rate": 0.00035291835555412165, + "loss": 2.7412561416625976, + "step": 74260 + }, + { + "epoch": 0.37135, + "grad_norm": 0.3769911229610443, + "learning_rate": 0.0003528822048628706, + "loss": 2.7567541122436525, + "step": 74270 + }, + { + "epoch": 0.3714, + "grad_norm": 0.3588760495185852, + "learning_rate": 0.0003528460515815606, + "loss": 2.7442001342773437, + "step": 74280 + }, + { + "epoch": 0.37145, + "grad_norm": 0.39431118965148926, + "learning_rate": 0.00035280989571110207, + "loss": 2.6821041107177734, + "step": 74290 + }, + { + "epoch": 0.3715, + "grad_norm": 0.3755283057689667, + "learning_rate": 0.00035277373725240503, + "loss": 2.7673622131347657, + "step": 74300 + }, + { + "epoch": 0.37155, + "grad_norm": 0.3226281404495239, + "learning_rate": 0.0003527375762063799, + "loss": 2.739105224609375, + "step": 74310 + }, + { + "epoch": 0.3716, + "grad_norm": 0.3853510618209839, + "learning_rate": 0.0003527014125739369, + "loss": 2.7642047882080076, + "step": 74320 + }, + { + "epoch": 0.37165, + "grad_norm": 0.3839632272720337, + "learning_rate": 0.0003526652463559866, + "loss": 2.768701171875, + "step": 74330 + }, + { + "epoch": 0.3717, + "grad_norm": 0.32274091243743896, + "learning_rate": 0.0003526290775534394, + "loss": 2.7162849426269533, + "step": 74340 + }, + { + "epoch": 0.37175, + "grad_norm": 0.3405095040798187, + "learning_rate": 0.0003525929061672058, + "loss": 2.7692527770996094, + "step": 74350 + }, + { + "epoch": 0.3718, + "grad_norm": 0.3939390778541565, + "learning_rate": 0.0003525567321981965, + "loss": 2.775654602050781, + "step": 74360 + }, + { + "epoch": 0.37185, + "grad_norm": 0.35357028245925903, + "learning_rate": 0.0003525205556473221, + "loss": 2.7849966049194337, + "step": 74370 + }, + { + "epoch": 0.3719, + "grad_norm": 0.32986488938331604, + "learning_rate": 0.0003524843765154935, + "loss": 2.7496219635009767, + "step": 74380 + }, + { + "epoch": 0.37195, + "grad_norm": 0.4125044643878937, + "learning_rate": 0.0003524481948036213, + "loss": 2.7662069320678713, + "step": 74390 + }, + { + "epoch": 0.372, + "grad_norm": 0.42025381326675415, + "learning_rate": 0.0003524120105126166, + "loss": 2.758590507507324, + "step": 74400 + }, + { + "epoch": 0.37205, + "grad_norm": 0.30588334798812866, + "learning_rate": 0.0003523758236433902, + "loss": 2.7897804260253904, + "step": 74410 + }, + { + "epoch": 0.3721, + "grad_norm": 0.28681665658950806, + "learning_rate": 0.000352339634196853, + "loss": 2.7491302490234375, + "step": 74420 + }, + { + "epoch": 0.37215, + "grad_norm": 0.34155985713005066, + "learning_rate": 0.0003523034421739163, + "loss": 2.7393110275268553, + "step": 74430 + }, + { + "epoch": 0.3722, + "grad_norm": 0.36636340618133545, + "learning_rate": 0.0003522672475754911, + "loss": 2.779221534729004, + "step": 74440 + }, + { + "epoch": 0.37225, + "grad_norm": 0.330992192029953, + "learning_rate": 0.00035223105040248866, + "loss": 2.709442138671875, + "step": 74450 + }, + { + "epoch": 0.3723, + "grad_norm": 0.3283140957355499, + "learning_rate": 0.00035219485065582015, + "loss": 2.737521743774414, + "step": 74460 + }, + { + "epoch": 0.37235, + "grad_norm": 0.3454437255859375, + "learning_rate": 0.000352158648336397, + "loss": 2.7901517868041994, + "step": 74470 + }, + { + "epoch": 0.3724, + "grad_norm": 0.35412389039993286, + "learning_rate": 0.00035212244344513057, + "loss": 2.691701126098633, + "step": 74480 + }, + { + "epoch": 0.37245, + "grad_norm": 0.325249046087265, + "learning_rate": 0.0003520862359829322, + "loss": 2.6380653381347656, + "step": 74490 + }, + { + "epoch": 0.3725, + "grad_norm": 0.511508047580719, + "learning_rate": 0.0003520500259507136, + "loss": 2.7081110000610353, + "step": 74500 + }, + { + "epoch": 0.37255, + "grad_norm": 0.3534014821052551, + "learning_rate": 0.0003520138133493863, + "loss": 2.7477678298950194, + "step": 74510 + }, + { + "epoch": 0.3726, + "grad_norm": 0.3938809335231781, + "learning_rate": 0.00035197759817986193, + "loss": 2.738829231262207, + "step": 74520 + }, + { + "epoch": 0.37265, + "grad_norm": 0.3277965188026428, + "learning_rate": 0.00035194138044305226, + "loss": 2.743724822998047, + "step": 74530 + }, + { + "epoch": 0.3727, + "grad_norm": 0.34371453523635864, + "learning_rate": 0.0003519051601398688, + "loss": 2.7787477493286135, + "step": 74540 + }, + { + "epoch": 0.37275, + "grad_norm": 0.33715271949768066, + "learning_rate": 0.0003518689372712238, + "loss": 2.7234098434448244, + "step": 74550 + }, + { + "epoch": 0.3728, + "grad_norm": 0.3434193730354309, + "learning_rate": 0.000351832711838029, + "loss": 2.7717580795288086, + "step": 74560 + }, + { + "epoch": 0.37285, + "grad_norm": 0.34814053773880005, + "learning_rate": 0.0003517964838411963, + "loss": 2.7695028305053713, + "step": 74570 + }, + { + "epoch": 0.3729, + "grad_norm": 0.3290565311908722, + "learning_rate": 0.0003517602532816378, + "loss": 2.723565673828125, + "step": 74580 + }, + { + "epoch": 0.37295, + "grad_norm": 0.3498433828353882, + "learning_rate": 0.00035172402016026564, + "loss": 2.7136722564697267, + "step": 74590 + }, + { + "epoch": 0.373, + "grad_norm": 0.3576316237449646, + "learning_rate": 0.0003516877844779919, + "loss": 2.790173149108887, + "step": 74600 + }, + { + "epoch": 0.37305, + "grad_norm": 0.3422233760356903, + "learning_rate": 0.00035165154623572893, + "loss": 2.711020088195801, + "step": 74610 + }, + { + "epoch": 0.3731, + "grad_norm": 0.33832964301109314, + "learning_rate": 0.0003516153054343889, + "loss": 2.787586975097656, + "step": 74620 + }, + { + "epoch": 0.37315, + "grad_norm": 0.3471378684043884, + "learning_rate": 0.0003515790620748842, + "loss": 2.7534250259399413, + "step": 74630 + }, + { + "epoch": 0.3732, + "grad_norm": 0.310350239276886, + "learning_rate": 0.0003515428161581274, + "loss": 2.760318374633789, + "step": 74640 + }, + { + "epoch": 0.37325, + "grad_norm": 0.31632891297340393, + "learning_rate": 0.0003515065676850308, + "loss": 2.742532730102539, + "step": 74650 + }, + { + "epoch": 0.3733, + "grad_norm": 0.3106931149959564, + "learning_rate": 0.0003514703166565071, + "loss": 2.7029232025146483, + "step": 74660 + }, + { + "epoch": 0.37335, + "grad_norm": 0.31509262323379517, + "learning_rate": 0.00035143406307346883, + "loss": 2.729236030578613, + "step": 74670 + }, + { + "epoch": 0.3734, + "grad_norm": 0.35832908749580383, + "learning_rate": 0.00035139780693682865, + "loss": 2.706273078918457, + "step": 74680 + }, + { + "epoch": 0.37345, + "grad_norm": 0.3548688292503357, + "learning_rate": 0.00035136154824749944, + "loss": 2.7310506820678713, + "step": 74690 + }, + { + "epoch": 0.3735, + "grad_norm": 0.34799283742904663, + "learning_rate": 0.0003513252870063939, + "loss": 2.7268404006958007, + "step": 74700 + }, + { + "epoch": 0.37355, + "grad_norm": 0.29600369930267334, + "learning_rate": 0.0003512890232144249, + "loss": 2.750954437255859, + "step": 74710 + }, + { + "epoch": 0.3736, + "grad_norm": 0.3266894817352295, + "learning_rate": 0.00035125275687250546, + "loss": 2.7517780303955077, + "step": 74720 + }, + { + "epoch": 0.37365, + "grad_norm": 0.3786882162094116, + "learning_rate": 0.00035121648798154846, + "loss": 2.791667175292969, + "step": 74730 + }, + { + "epoch": 0.3737, + "grad_norm": 0.32314223051071167, + "learning_rate": 0.00035118021654246715, + "loss": 2.7324617385864256, + "step": 74740 + }, + { + "epoch": 0.37375, + "grad_norm": 0.37998926639556885, + "learning_rate": 0.00035114394255617454, + "loss": 2.7517717361450194, + "step": 74750 + }, + { + "epoch": 0.3738, + "grad_norm": 0.3323594629764557, + "learning_rate": 0.00035110766602358386, + "loss": 2.7305763244628904, + "step": 74760 + }, + { + "epoch": 0.37385, + "grad_norm": 0.3286076486110687, + "learning_rate": 0.00035107138694560837, + "loss": 2.7711057662963867, + "step": 74770 + }, + { + "epoch": 0.3739, + "grad_norm": 0.33694830536842346, + "learning_rate": 0.00035103510532316135, + "loss": 2.7247257232666016, + "step": 74780 + }, + { + "epoch": 0.37395, + "grad_norm": 0.32508137822151184, + "learning_rate": 0.0003509988211571562, + "loss": 2.7591951370239256, + "step": 74790 + }, + { + "epoch": 0.374, + "grad_norm": 0.30612364411354065, + "learning_rate": 0.0003509625344485065, + "loss": 2.6922746658325196, + "step": 74800 + }, + { + "epoch": 0.37405, + "grad_norm": 0.3196461498737335, + "learning_rate": 0.0003509262451981255, + "loss": 2.742932891845703, + "step": 74810 + }, + { + "epoch": 0.3741, + "grad_norm": 0.31244874000549316, + "learning_rate": 0.0003508899534069272, + "loss": 2.7353374481201174, + "step": 74820 + }, + { + "epoch": 0.37415, + "grad_norm": 0.3459746837615967, + "learning_rate": 0.0003508536590758248, + "loss": 2.764387321472168, + "step": 74830 + }, + { + "epoch": 0.3742, + "grad_norm": 0.33335861563682556, + "learning_rate": 0.00035081736220573227, + "loss": 2.7350683212280273, + "step": 74840 + }, + { + "epoch": 0.37425, + "grad_norm": 0.342383474111557, + "learning_rate": 0.0003507810627975634, + "loss": 2.793722724914551, + "step": 74850 + }, + { + "epoch": 0.3743, + "grad_norm": 0.35646191239356995, + "learning_rate": 0.00035074476085223186, + "loss": 2.7322513580322267, + "step": 74860 + }, + { + "epoch": 0.37435, + "grad_norm": 0.3479102551937103, + "learning_rate": 0.00035070845637065164, + "loss": 2.7916221618652344, + "step": 74870 + }, + { + "epoch": 0.3744, + "grad_norm": 0.32004567980766296, + "learning_rate": 0.0003506721493537367, + "loss": 2.738429641723633, + "step": 74880 + }, + { + "epoch": 0.37445, + "grad_norm": 0.3479127287864685, + "learning_rate": 0.00035063583980240103, + "loss": 2.741348075866699, + "step": 74890 + }, + { + "epoch": 0.3745, + "grad_norm": 0.32580846548080444, + "learning_rate": 0.0003505995277175589, + "loss": 2.79879207611084, + "step": 74900 + }, + { + "epoch": 0.37455, + "grad_norm": 0.36258408427238464, + "learning_rate": 0.0003505632131001242, + "loss": 2.7653076171875, + "step": 74910 + }, + { + "epoch": 0.3746, + "grad_norm": 0.3511681854724884, + "learning_rate": 0.00035052689595101134, + "loss": 2.7578285217285154, + "step": 74920 + }, + { + "epoch": 0.37465, + "grad_norm": 0.3200025260448456, + "learning_rate": 0.00035049057627113456, + "loss": 2.748368835449219, + "step": 74930 + }, + { + "epoch": 0.3747, + "grad_norm": 0.324084997177124, + "learning_rate": 0.0003504542540614082, + "loss": 2.7829010009765627, + "step": 74940 + }, + { + "epoch": 0.37475, + "grad_norm": 0.32855483889579773, + "learning_rate": 0.0003504179293227466, + "loss": 2.775753402709961, + "step": 74950 + }, + { + "epoch": 0.3748, + "grad_norm": 0.3210921287536621, + "learning_rate": 0.0003503816020560644, + "loss": 2.714329147338867, + "step": 74960 + }, + { + "epoch": 0.37485, + "grad_norm": 0.31918928027153015, + "learning_rate": 0.00035034527226227594, + "loss": 2.772370529174805, + "step": 74970 + }, + { + "epoch": 0.3749, + "grad_norm": 0.35244813561439514, + "learning_rate": 0.00035030893994229595, + "loss": 2.763414001464844, + "step": 74980 + }, + { + "epoch": 0.37495, + "grad_norm": 0.3596433103084564, + "learning_rate": 0.00035027260509703904, + "loss": 2.7869190216064452, + "step": 74990 + }, + { + "epoch": 0.375, + "grad_norm": 0.3018629252910614, + "learning_rate": 0.00035023626772742, + "loss": 2.7085405349731446, + "step": 75000 + }, + { + "epoch": 0.37505, + "grad_norm": 0.392232209444046, + "learning_rate": 0.0003501999278343536, + "loss": 2.7157310485839843, + "step": 75010 + }, + { + "epoch": 0.3751, + "grad_norm": 0.333279550075531, + "learning_rate": 0.0003501635854187548, + "loss": 2.728981781005859, + "step": 75020 + }, + { + "epoch": 0.37515, + "grad_norm": 0.33400216698646545, + "learning_rate": 0.0003501272404815382, + "loss": 2.6941951751708983, + "step": 75030 + }, + { + "epoch": 0.3752, + "grad_norm": 0.34683758020401, + "learning_rate": 0.0003500908930236191, + "loss": 2.7358728408813477, + "step": 75040 + }, + { + "epoch": 0.37525, + "grad_norm": 0.3420390486717224, + "learning_rate": 0.00035005454304591247, + "loss": 2.769448089599609, + "step": 75050 + }, + { + "epoch": 0.3753, + "grad_norm": 0.35066863894462585, + "learning_rate": 0.00035001819054933327, + "loss": 2.707656478881836, + "step": 75060 + }, + { + "epoch": 0.37535, + "grad_norm": 0.32788076996803284, + "learning_rate": 0.00034998183553479686, + "loss": 2.6860107421875, + "step": 75070 + }, + { + "epoch": 0.3754, + "grad_norm": 0.3497314751148224, + "learning_rate": 0.0003499454780032184, + "loss": 2.707647705078125, + "step": 75080 + }, + { + "epoch": 0.37545, + "grad_norm": 0.3336949050426483, + "learning_rate": 0.00034990911795551317, + "loss": 2.747800827026367, + "step": 75090 + }, + { + "epoch": 0.3755, + "grad_norm": 0.3407692313194275, + "learning_rate": 0.00034987275539259666, + "loss": 2.73494815826416, + "step": 75100 + }, + { + "epoch": 0.37555, + "grad_norm": 0.3172776401042938, + "learning_rate": 0.00034983639031538403, + "loss": 2.7194778442382814, + "step": 75110 + }, + { + "epoch": 0.3756, + "grad_norm": 0.3538026511669159, + "learning_rate": 0.0003498000227247911, + "loss": 2.7189544677734374, + "step": 75120 + }, + { + "epoch": 0.37565, + "grad_norm": 0.3364328145980835, + "learning_rate": 0.0003497636526217332, + "loss": 2.772370719909668, + "step": 75130 + }, + { + "epoch": 0.3757, + "grad_norm": 0.34231632947921753, + "learning_rate": 0.0003497272800071261, + "loss": 2.7753427505493162, + "step": 75140 + }, + { + "epoch": 0.37575, + "grad_norm": 0.35137563943862915, + "learning_rate": 0.00034969090488188524, + "loss": 2.77486515045166, + "step": 75150 + }, + { + "epoch": 0.3758, + "grad_norm": 0.3237961530685425, + "learning_rate": 0.0003496545272469266, + "loss": 2.763633155822754, + "step": 75160 + }, + { + "epoch": 0.37585, + "grad_norm": 0.3240990936756134, + "learning_rate": 0.00034961814710316586, + "loss": 2.74767951965332, + "step": 75170 + }, + { + "epoch": 0.3759, + "grad_norm": 0.3358598053455353, + "learning_rate": 0.000349581764451519, + "loss": 2.778069496154785, + "step": 75180 + }, + { + "epoch": 0.37595, + "grad_norm": 0.3515474200248718, + "learning_rate": 0.0003495453792929018, + "loss": 2.7542205810546876, + "step": 75190 + }, + { + "epoch": 0.376, + "grad_norm": 0.3168518841266632, + "learning_rate": 0.0003495089916282305, + "loss": 2.781446838378906, + "step": 75200 + }, + { + "epoch": 0.37605, + "grad_norm": 0.33697837591171265, + "learning_rate": 0.00034947260145842086, + "loss": 2.7314443588256836, + "step": 75210 + }, + { + "epoch": 0.3761, + "grad_norm": 0.35159993171691895, + "learning_rate": 0.00034943620878438914, + "loss": 2.748832893371582, + "step": 75220 + }, + { + "epoch": 0.37615, + "grad_norm": 0.2974920868873596, + "learning_rate": 0.00034939981360705154, + "loss": 2.7112802505493163, + "step": 75230 + }, + { + "epoch": 0.3762, + "grad_norm": 0.3289106786251068, + "learning_rate": 0.00034936341592732437, + "loss": 2.697266960144043, + "step": 75240 + }, + { + "epoch": 0.37625, + "grad_norm": 0.3625704050064087, + "learning_rate": 0.0003493270157461237, + "loss": 2.7608064651489257, + "step": 75250 + }, + { + "epoch": 0.3763, + "grad_norm": 0.30960211157798767, + "learning_rate": 0.0003492906130643663, + "loss": 2.7603708267211915, + "step": 75260 + }, + { + "epoch": 0.37635, + "grad_norm": 0.3568331003189087, + "learning_rate": 0.0003492542078829682, + "loss": 2.7319995880126955, + "step": 75270 + }, + { + "epoch": 0.3764, + "grad_norm": 0.3665531575679779, + "learning_rate": 0.00034921780020284623, + "loss": 2.751388740539551, + "step": 75280 + }, + { + "epoch": 0.37645, + "grad_norm": 0.34579265117645264, + "learning_rate": 0.00034918139002491666, + "loss": 2.7342891693115234, + "step": 75290 + }, + { + "epoch": 0.3765, + "grad_norm": 0.331009179353714, + "learning_rate": 0.0003491449773500963, + "loss": 2.7112476348876955, + "step": 75300 + }, + { + "epoch": 0.37655, + "grad_norm": 0.37366777658462524, + "learning_rate": 0.00034910856217930185, + "loss": 2.735085868835449, + "step": 75310 + }, + { + "epoch": 0.3766, + "grad_norm": 0.32463183999061584, + "learning_rate": 0.00034907214451345, + "loss": 2.723371696472168, + "step": 75320 + }, + { + "epoch": 0.37665, + "grad_norm": 0.333740234375, + "learning_rate": 0.0003490357243534575, + "loss": 2.720694160461426, + "step": 75330 + }, + { + "epoch": 0.3767, + "grad_norm": 0.342163622379303, + "learning_rate": 0.0003489993017002414, + "loss": 2.7576189041137695, + "step": 75340 + }, + { + "epoch": 0.37675, + "grad_norm": 0.34138762950897217, + "learning_rate": 0.0003489628765547185, + "loss": 2.7449893951416016, + "step": 75350 + }, + { + "epoch": 0.3768, + "grad_norm": 0.316820353269577, + "learning_rate": 0.0003489264489178058, + "loss": 2.7183666229248047, + "step": 75360 + }, + { + "epoch": 0.37685, + "grad_norm": 0.3652818500995636, + "learning_rate": 0.0003488900187904205, + "loss": 2.7612445831298826, + "step": 75370 + }, + { + "epoch": 0.3769, + "grad_norm": 0.35349562764167786, + "learning_rate": 0.0003488535861734796, + "loss": 2.7443275451660156, + "step": 75380 + }, + { + "epoch": 0.37695, + "grad_norm": 0.351416677236557, + "learning_rate": 0.00034881715106790023, + "loss": 2.7299509048461914, + "step": 75390 + }, + { + "epoch": 0.377, + "grad_norm": 0.3486122488975525, + "learning_rate": 0.0003487807134745999, + "loss": 2.7273122787475588, + "step": 75400 + }, + { + "epoch": 0.37705, + "grad_norm": 0.3550716042518616, + "learning_rate": 0.00034874427339449563, + "loss": 2.7684606552124023, + "step": 75410 + }, + { + "epoch": 0.3771, + "grad_norm": 0.32404911518096924, + "learning_rate": 0.00034870783082850503, + "loss": 2.7756038665771485, + "step": 75420 + }, + { + "epoch": 0.37715, + "grad_norm": 0.3291057050228119, + "learning_rate": 0.00034867138577754543, + "loss": 2.7169305801391603, + "step": 75430 + }, + { + "epoch": 0.3772, + "grad_norm": 0.3476495146751404, + "learning_rate": 0.00034863493824253435, + "loss": 2.7428348541259764, + "step": 75440 + }, + { + "epoch": 0.37725, + "grad_norm": 0.2979298233985901, + "learning_rate": 0.00034859848822438934, + "loss": 2.7686933517456054, + "step": 75450 + }, + { + "epoch": 0.3773, + "grad_norm": 0.31790196895599365, + "learning_rate": 0.0003485620357240281, + "loss": 2.742257308959961, + "step": 75460 + }, + { + "epoch": 0.37735, + "grad_norm": 0.3204340636730194, + "learning_rate": 0.0003485255807423682, + "loss": 2.7463891983032225, + "step": 75470 + }, + { + "epoch": 0.3774, + "grad_norm": 0.328606516122818, + "learning_rate": 0.0003484891232803275, + "loss": 2.706715202331543, + "step": 75480 + }, + { + "epoch": 0.37745, + "grad_norm": 0.34504932165145874, + "learning_rate": 0.0003484526633388238, + "loss": 2.716644859313965, + "step": 75490 + }, + { + "epoch": 0.3775, + "grad_norm": 0.3316860795021057, + "learning_rate": 0.0003484162009187749, + "loss": 2.7216970443725588, + "step": 75500 + }, + { + "epoch": 0.37755, + "grad_norm": 0.31801143288612366, + "learning_rate": 0.0003483797360210988, + "loss": 2.776106071472168, + "step": 75510 + }, + { + "epoch": 0.3776, + "grad_norm": 0.37149178981781006, + "learning_rate": 0.0003483432686467135, + "loss": 2.743829536437988, + "step": 75520 + }, + { + "epoch": 0.37765, + "grad_norm": 0.3311257064342499, + "learning_rate": 0.00034830679879653716, + "loss": 2.7320266723632813, + "step": 75530 + }, + { + "epoch": 0.3777, + "grad_norm": 0.32828831672668457, + "learning_rate": 0.0003482703264714877, + "loss": 2.7381217956542967, + "step": 75540 + }, + { + "epoch": 0.37775, + "grad_norm": 0.33992481231689453, + "learning_rate": 0.00034823385167248355, + "loss": 2.749689292907715, + "step": 75550 + }, + { + "epoch": 0.3778, + "grad_norm": 0.3455585539340973, + "learning_rate": 0.00034819737440044274, + "loss": 2.741040802001953, + "step": 75560 + }, + { + "epoch": 0.37785, + "grad_norm": 0.3319137394428253, + "learning_rate": 0.00034816089465628376, + "loss": 2.738943862915039, + "step": 75570 + }, + { + "epoch": 0.3779, + "grad_norm": 0.371208518743515, + "learning_rate": 0.00034812441244092484, + "loss": 2.709202766418457, + "step": 75580 + }, + { + "epoch": 0.37795, + "grad_norm": 0.3252973258495331, + "learning_rate": 0.0003480879277552845, + "loss": 2.7077442169189454, + "step": 75590 + }, + { + "epoch": 0.378, + "grad_norm": 0.3501824140548706, + "learning_rate": 0.00034805144060028124, + "loss": 2.7492439270019533, + "step": 75600 + }, + { + "epoch": 0.37805, + "grad_norm": 0.3210947811603546, + "learning_rate": 0.00034801495097683374, + "loss": 2.7453065872192384, + "step": 75610 + }, + { + "epoch": 0.3781, + "grad_norm": 0.30786237120628357, + "learning_rate": 0.00034797845888586034, + "loss": 2.727623176574707, + "step": 75620 + }, + { + "epoch": 0.37815, + "grad_norm": 0.37634992599487305, + "learning_rate": 0.00034794196432828004, + "loss": 2.7096698760986326, + "step": 75630 + }, + { + "epoch": 0.3782, + "grad_norm": 0.32391437888145447, + "learning_rate": 0.00034790546730501133, + "loss": 2.7031558990478515, + "step": 75640 + }, + { + "epoch": 0.37825, + "grad_norm": 0.3250219523906708, + "learning_rate": 0.00034786896781697314, + "loss": 2.713270378112793, + "step": 75650 + }, + { + "epoch": 0.3783, + "grad_norm": 0.36097297072410583, + "learning_rate": 0.0003478324658650843, + "loss": 2.73233642578125, + "step": 75660 + }, + { + "epoch": 0.37835, + "grad_norm": 0.34953197836875916, + "learning_rate": 0.00034779596145026394, + "loss": 2.740534782409668, + "step": 75670 + }, + { + "epoch": 0.3784, + "grad_norm": 0.35681354999542236, + "learning_rate": 0.0003477594545734307, + "loss": 2.7448715209960937, + "step": 75680 + }, + { + "epoch": 0.37845, + "grad_norm": 0.3191402852535248, + "learning_rate": 0.0003477229452355041, + "loss": 2.7315303802490236, + "step": 75690 + }, + { + "epoch": 0.3785, + "grad_norm": 0.38553616404533386, + "learning_rate": 0.0003476864334374028, + "loss": 2.7640771865844727, + "step": 75700 + }, + { + "epoch": 0.37855, + "grad_norm": 0.3334280848503113, + "learning_rate": 0.00034764991918004626, + "loss": 2.751972961425781, + "step": 75710 + }, + { + "epoch": 0.3786, + "grad_norm": 0.3423268795013428, + "learning_rate": 0.00034761340246435365, + "loss": 2.781558036804199, + "step": 75720 + }, + { + "epoch": 0.37865, + "grad_norm": 0.36324918270111084, + "learning_rate": 0.0003475768832912443, + "loss": 2.699707794189453, + "step": 75730 + }, + { + "epoch": 0.3787, + "grad_norm": 0.3548745810985565, + "learning_rate": 0.0003475403616616375, + "loss": 2.7445775985717775, + "step": 75740 + }, + { + "epoch": 0.37875, + "grad_norm": 0.31652092933654785, + "learning_rate": 0.0003475038375764529, + "loss": 2.7267129898071287, + "step": 75750 + }, + { + "epoch": 0.3788, + "grad_norm": 0.38737693428993225, + "learning_rate": 0.0003474673110366097, + "loss": 2.765634536743164, + "step": 75760 + }, + { + "epoch": 0.37885, + "grad_norm": 0.3576512336730957, + "learning_rate": 0.0003474307820430276, + "loss": 2.7183242797851563, + "step": 75770 + }, + { + "epoch": 0.3789, + "grad_norm": 0.41781488060951233, + "learning_rate": 0.0003473942505966263, + "loss": 2.7891262054443358, + "step": 75780 + }, + { + "epoch": 0.37895, + "grad_norm": 0.3712616562843323, + "learning_rate": 0.00034735771669832527, + "loss": 2.7713296890258787, + "step": 75790 + }, + { + "epoch": 0.379, + "grad_norm": 0.33434417843818665, + "learning_rate": 0.0003473211803490445, + "loss": 2.79184513092041, + "step": 75800 + }, + { + "epoch": 0.37905, + "grad_norm": 0.36510175466537476, + "learning_rate": 0.00034728464154970356, + "loss": 2.7607349395751952, + "step": 75810 + }, + { + "epoch": 0.3791, + "grad_norm": 0.3669196367263794, + "learning_rate": 0.00034724810030122244, + "loss": 2.722483825683594, + "step": 75820 + }, + { + "epoch": 0.37915, + "grad_norm": 0.3350026309490204, + "learning_rate": 0.00034721155660452115, + "loss": 2.822025680541992, + "step": 75830 + }, + { + "epoch": 0.3792, + "grad_norm": 0.3096131980419159, + "learning_rate": 0.0003471750104605194, + "loss": 2.733211326599121, + "step": 75840 + }, + { + "epoch": 0.37925, + "grad_norm": 0.32274243235588074, + "learning_rate": 0.00034713846187013755, + "loss": 2.741160202026367, + "step": 75850 + }, + { + "epoch": 0.3793, + "grad_norm": 0.3558662533760071, + "learning_rate": 0.00034710191083429557, + "loss": 2.7196514129638674, + "step": 75860 + }, + { + "epoch": 0.37935, + "grad_norm": 0.35486483573913574, + "learning_rate": 0.0003470653573539135, + "loss": 2.718338394165039, + "step": 75870 + }, + { + "epoch": 0.3794, + "grad_norm": 0.3389519453048706, + "learning_rate": 0.00034702880142991184, + "loss": 2.759715461730957, + "step": 75880 + }, + { + "epoch": 0.37945, + "grad_norm": 0.36875641345977783, + "learning_rate": 0.0003469922430632107, + "loss": 2.7825750350952148, + "step": 75890 + }, + { + "epoch": 0.3795, + "grad_norm": 0.35809001326560974, + "learning_rate": 0.0003469556822547305, + "loss": 2.6522138595581053, + "step": 75900 + }, + { + "epoch": 0.37955, + "grad_norm": 0.31802189350128174, + "learning_rate": 0.0003469191190053917, + "loss": 2.7283645629882813, + "step": 75910 + }, + { + "epoch": 0.3796, + "grad_norm": 0.32643628120422363, + "learning_rate": 0.0003468825533161146, + "loss": 2.723804473876953, + "step": 75920 + }, + { + "epoch": 0.37965, + "grad_norm": 0.32261425256729126, + "learning_rate": 0.00034684598518781997, + "loss": 2.7336973190307616, + "step": 75930 + }, + { + "epoch": 0.3797, + "grad_norm": 0.2966252863407135, + "learning_rate": 0.0003468094146214283, + "loss": 2.752411460876465, + "step": 75940 + }, + { + "epoch": 0.37975, + "grad_norm": 0.34544143080711365, + "learning_rate": 0.0003467728416178602, + "loss": 2.7026519775390625, + "step": 75950 + }, + { + "epoch": 0.3798, + "grad_norm": 0.3212791383266449, + "learning_rate": 0.0003467362661780365, + "loss": 2.764943504333496, + "step": 75960 + }, + { + "epoch": 0.37985, + "grad_norm": 0.32320424914360046, + "learning_rate": 0.00034669968830287803, + "loss": 2.7512886047363283, + "step": 75970 + }, + { + "epoch": 0.3799, + "grad_norm": 0.3458441197872162, + "learning_rate": 0.00034666310799330536, + "loss": 2.7340871810913088, + "step": 75980 + }, + { + "epoch": 0.37995, + "grad_norm": 0.35487473011016846, + "learning_rate": 0.00034662652525023973, + "loss": 2.7497095108032226, + "step": 75990 + }, + { + "epoch": 0.38, + "grad_norm": 0.3127692639827728, + "learning_rate": 0.000346589940074602, + "loss": 2.725162124633789, + "step": 76000 + }, + { + "epoch": 0.38005, + "grad_norm": 0.36053070425987244, + "learning_rate": 0.00034655335246731314, + "loss": 2.74404296875, + "step": 76010 + }, + { + "epoch": 0.3801, + "grad_norm": 0.34977200627326965, + "learning_rate": 0.0003465167624292942, + "loss": 2.729772186279297, + "step": 76020 + }, + { + "epoch": 0.38015, + "grad_norm": 0.36187997460365295, + "learning_rate": 0.0003464801699614665, + "loss": 2.714883232116699, + "step": 76030 + }, + { + "epoch": 0.3802, + "grad_norm": 0.3323005735874176, + "learning_rate": 0.0003464435750647511, + "loss": 2.7948844909667967, + "step": 76040 + }, + { + "epoch": 0.38025, + "grad_norm": 0.3176027536392212, + "learning_rate": 0.0003464069777400694, + "loss": 2.759117317199707, + "step": 76050 + }, + { + "epoch": 0.3803, + "grad_norm": 0.39930933713912964, + "learning_rate": 0.0003463703779883426, + "loss": 2.745214653015137, + "step": 76060 + }, + { + "epoch": 0.38035, + "grad_norm": 0.3420693576335907, + "learning_rate": 0.0003463337758104923, + "loss": 2.7547664642333984, + "step": 76070 + }, + { + "epoch": 0.3804, + "grad_norm": 0.34276723861694336, + "learning_rate": 0.00034629717120743974, + "loss": 2.754214859008789, + "step": 76080 + }, + { + "epoch": 0.38045, + "grad_norm": 0.33199986815452576, + "learning_rate": 0.00034626056418010653, + "loss": 2.6956270217895506, + "step": 76090 + }, + { + "epoch": 0.3805, + "grad_norm": 0.3495066165924072, + "learning_rate": 0.00034622395472941434, + "loss": 2.804239845275879, + "step": 76100 + }, + { + "epoch": 0.38055, + "grad_norm": 0.32912495732307434, + "learning_rate": 0.0003461873428562847, + "loss": 2.747456359863281, + "step": 76110 + }, + { + "epoch": 0.3806, + "grad_norm": 0.3309055268764496, + "learning_rate": 0.00034615072856163933, + "loss": 2.753783416748047, + "step": 76120 + }, + { + "epoch": 0.38065, + "grad_norm": 0.32973712682724, + "learning_rate": 0.0003461141118464, + "loss": 2.697393226623535, + "step": 76130 + }, + { + "epoch": 0.3807, + "grad_norm": 0.34830132126808167, + "learning_rate": 0.00034607749271148855, + "loss": 2.6974878311157227, + "step": 76140 + }, + { + "epoch": 0.38075, + "grad_norm": 0.34889060258865356, + "learning_rate": 0.0003460408711578269, + "loss": 2.7369667053222657, + "step": 76150 + }, + { + "epoch": 0.3808, + "grad_norm": 0.41397470235824585, + "learning_rate": 0.00034600424718633695, + "loss": 2.7763505935668946, + "step": 76160 + }, + { + "epoch": 0.38085, + "grad_norm": 0.3300183117389679, + "learning_rate": 0.0003459676207979407, + "loss": 2.775400161743164, + "step": 76170 + }, + { + "epoch": 0.3809, + "grad_norm": 0.3623397648334503, + "learning_rate": 0.0003459309919935602, + "loss": 2.7746292114257813, + "step": 76180 + }, + { + "epoch": 0.38095, + "grad_norm": 0.3707458972930908, + "learning_rate": 0.0003458943607741177, + "loss": 2.7195695877075194, + "step": 76190 + }, + { + "epoch": 0.381, + "grad_norm": 0.3149195611476898, + "learning_rate": 0.00034585772714053526, + "loss": 2.7623931884765627, + "step": 76200 + }, + { + "epoch": 0.38105, + "grad_norm": 0.340715616941452, + "learning_rate": 0.0003458210910937352, + "loss": 2.7467565536499023, + "step": 76210 + }, + { + "epoch": 0.3811, + "grad_norm": 0.3536602556705475, + "learning_rate": 0.00034578445263463975, + "loss": 2.6655031204223634, + "step": 76220 + }, + { + "epoch": 0.38115, + "grad_norm": 0.3435898721218109, + "learning_rate": 0.0003457478117641715, + "loss": 2.750849723815918, + "step": 76230 + }, + { + "epoch": 0.3812, + "grad_norm": 0.34342554211616516, + "learning_rate": 0.0003457111684832526, + "loss": 2.7647720336914063, + "step": 76240 + }, + { + "epoch": 0.38125, + "grad_norm": 0.3234880864620209, + "learning_rate": 0.00034567452279280565, + "loss": 2.759039115905762, + "step": 76250 + }, + { + "epoch": 0.3813, + "grad_norm": 0.32859089970588684, + "learning_rate": 0.0003456378746937533, + "loss": 2.7425670623779297, + "step": 76260 + }, + { + "epoch": 0.38135, + "grad_norm": 0.33725741505622864, + "learning_rate": 0.0003456012241870181, + "loss": 2.779242515563965, + "step": 76270 + }, + { + "epoch": 0.3814, + "grad_norm": 0.3214852511882782, + "learning_rate": 0.00034556457127352266, + "loss": 2.696158027648926, + "step": 76280 + }, + { + "epoch": 0.38145, + "grad_norm": 0.3271154463291168, + "learning_rate": 0.0003455279159541898, + "loss": 2.7210840225219726, + "step": 76290 + }, + { + "epoch": 0.3815, + "grad_norm": 0.3496933579444885, + "learning_rate": 0.0003454912582299423, + "loss": 2.764784812927246, + "step": 76300 + }, + { + "epoch": 0.38155, + "grad_norm": 0.38135871291160583, + "learning_rate": 0.00034545459810170307, + "loss": 2.7383167266845705, + "step": 76310 + }, + { + "epoch": 0.3816, + "grad_norm": 0.34362274408340454, + "learning_rate": 0.00034541793557039485, + "loss": 2.7545461654663086, + "step": 76320 + }, + { + "epoch": 0.38165, + "grad_norm": 0.33481451869010925, + "learning_rate": 0.00034538127063694083, + "loss": 2.7173250198364256, + "step": 76330 + }, + { + "epoch": 0.3817, + "grad_norm": 0.36940351128578186, + "learning_rate": 0.000345344603302264, + "loss": 2.8374868392944337, + "step": 76340 + }, + { + "epoch": 0.38175, + "grad_norm": 0.35382935404777527, + "learning_rate": 0.00034530793356728727, + "loss": 2.773042678833008, + "step": 76350 + }, + { + "epoch": 0.3818, + "grad_norm": 0.34317341446876526, + "learning_rate": 0.0003452712614329341, + "loss": 2.764485168457031, + "step": 76360 + }, + { + "epoch": 0.38185, + "grad_norm": 0.3691813051700592, + "learning_rate": 0.00034523458690012744, + "loss": 2.726452445983887, + "step": 76370 + }, + { + "epoch": 0.3819, + "grad_norm": 0.3330724537372589, + "learning_rate": 0.0003451979099697907, + "loss": 2.715706443786621, + "step": 76380 + }, + { + "epoch": 0.38195, + "grad_norm": 0.32346436381340027, + "learning_rate": 0.0003451612306428473, + "loss": 2.7179595947265627, + "step": 76390 + }, + { + "epoch": 0.382, + "grad_norm": 0.33125194907188416, + "learning_rate": 0.00034512454892022057, + "loss": 2.746341514587402, + "step": 76400 + }, + { + "epoch": 0.38205, + "grad_norm": 0.3596917688846588, + "learning_rate": 0.00034508786480283383, + "loss": 2.680157470703125, + "step": 76410 + }, + { + "epoch": 0.3821, + "grad_norm": 0.4039789140224457, + "learning_rate": 0.00034505117829161083, + "loss": 2.7404306411743162, + "step": 76420 + }, + { + "epoch": 0.38215, + "grad_norm": 0.34001612663269043, + "learning_rate": 0.000345014489387475, + "loss": 2.7456741333007812, + "step": 76430 + }, + { + "epoch": 0.3822, + "grad_norm": 0.319555401802063, + "learning_rate": 0.00034497779809135006, + "loss": 2.6727724075317383, + "step": 76440 + }, + { + "epoch": 0.38225, + "grad_norm": 0.3136324882507324, + "learning_rate": 0.00034494110440415966, + "loss": 2.7740535736083984, + "step": 76450 + }, + { + "epoch": 0.3823, + "grad_norm": 0.34377387166023254, + "learning_rate": 0.0003449044083268276, + "loss": 2.7482532501220702, + "step": 76460 + }, + { + "epoch": 0.38235, + "grad_norm": 0.37261345982551575, + "learning_rate": 0.0003448677098602777, + "loss": 2.71408634185791, + "step": 76470 + }, + { + "epoch": 0.3824, + "grad_norm": 0.3642290532588959, + "learning_rate": 0.00034483100900543386, + "loss": 2.763450813293457, + "step": 76480 + }, + { + "epoch": 0.38245, + "grad_norm": 0.4139556884765625, + "learning_rate": 0.0003447943057632199, + "loss": 2.734159469604492, + "step": 76490 + }, + { + "epoch": 0.3825, + "grad_norm": 0.34828656911849976, + "learning_rate": 0.0003447576001345601, + "loss": 2.6919025421142577, + "step": 76500 + }, + { + "epoch": 0.38255, + "grad_norm": 0.3549612760543823, + "learning_rate": 0.0003447208921203783, + "loss": 2.7287824630737303, + "step": 76510 + }, + { + "epoch": 0.3826, + "grad_norm": 0.3354656398296356, + "learning_rate": 0.00034468418172159854, + "loss": 2.7535654067993165, + "step": 76520 + }, + { + "epoch": 0.38265, + "grad_norm": 0.3219509422779083, + "learning_rate": 0.00034464746893914523, + "loss": 2.7570432662963866, + "step": 76530 + }, + { + "epoch": 0.3827, + "grad_norm": 0.3292123079299927, + "learning_rate": 0.00034461075377394254, + "loss": 2.7573877334594727, + "step": 76540 + }, + { + "epoch": 0.38275, + "grad_norm": 0.32999712228775024, + "learning_rate": 0.00034457403622691475, + "loss": 2.7256036758422852, + "step": 76550 + }, + { + "epoch": 0.3828, + "grad_norm": 0.3399330675601959, + "learning_rate": 0.0003445373162989862, + "loss": 2.7134174346923827, + "step": 76560 + }, + { + "epoch": 0.38285, + "grad_norm": 0.31129661202430725, + "learning_rate": 0.00034450059399108134, + "loss": 2.7316951751708984, + "step": 76570 + }, + { + "epoch": 0.3829, + "grad_norm": 0.312796950340271, + "learning_rate": 0.00034446386930412476, + "loss": 2.7230607986450197, + "step": 76580 + }, + { + "epoch": 0.38295, + "grad_norm": 0.34156158566474915, + "learning_rate": 0.0003444271422390408, + "loss": 2.72308235168457, + "step": 76590 + }, + { + "epoch": 0.383, + "grad_norm": 0.3865812420845032, + "learning_rate": 0.00034439041279675416, + "loss": 2.750191879272461, + "step": 76600 + }, + { + "epoch": 0.38305, + "grad_norm": 0.2926737368106842, + "learning_rate": 0.0003443536809781895, + "loss": 2.736955451965332, + "step": 76610 + }, + { + "epoch": 0.3831, + "grad_norm": 0.324631929397583, + "learning_rate": 0.00034431694678427164, + "loss": 2.7106367111206056, + "step": 76620 + }, + { + "epoch": 0.38315, + "grad_norm": 0.35456401109695435, + "learning_rate": 0.0003442802102159252, + "loss": 2.7543636322021485, + "step": 76630 + }, + { + "epoch": 0.3832, + "grad_norm": 0.38037338852882385, + "learning_rate": 0.0003442434712740752, + "loss": 2.779117965698242, + "step": 76640 + }, + { + "epoch": 0.38325, + "grad_norm": 0.35709843039512634, + "learning_rate": 0.0003442067299596463, + "loss": 2.7610435485839844, + "step": 76650 + }, + { + "epoch": 0.3833, + "grad_norm": 0.3316778540611267, + "learning_rate": 0.00034416998627356365, + "loss": 2.773544502258301, + "step": 76660 + }, + { + "epoch": 0.38335, + "grad_norm": 0.3390791416168213, + "learning_rate": 0.00034413324021675224, + "loss": 2.7611223220825196, + "step": 76670 + }, + { + "epoch": 0.3834, + "grad_norm": 0.34977465867996216, + "learning_rate": 0.00034409649179013716, + "loss": 2.7218122482299805, + "step": 76680 + }, + { + "epoch": 0.38345, + "grad_norm": 0.3632274568080902, + "learning_rate": 0.0003440597409946435, + "loss": 2.7195043563842773, + "step": 76690 + }, + { + "epoch": 0.3835, + "grad_norm": 0.3119291365146637, + "learning_rate": 0.0003440229878311965, + "loss": 2.7809738159179687, + "step": 76700 + }, + { + "epoch": 0.38355, + "grad_norm": 0.35925382375717163, + "learning_rate": 0.00034398623230072137, + "loss": 2.759941291809082, + "step": 76710 + }, + { + "epoch": 0.3836, + "grad_norm": 0.3531816303730011, + "learning_rate": 0.00034394947440414354, + "loss": 2.750679588317871, + "step": 76720 + }, + { + "epoch": 0.38365, + "grad_norm": 0.3377598226070404, + "learning_rate": 0.0003439127141423883, + "loss": 2.7842098236083985, + "step": 76730 + }, + { + "epoch": 0.3837, + "grad_norm": 0.33008310198783875, + "learning_rate": 0.0003438759515163811, + "loss": 2.7177862167358398, + "step": 76740 + }, + { + "epoch": 0.38375, + "grad_norm": 0.34067484736442566, + "learning_rate": 0.00034383918652704736, + "loss": 2.7739620208740234, + "step": 76750 + }, + { + "epoch": 0.3838, + "grad_norm": 0.3454793393611908, + "learning_rate": 0.00034380241917531277, + "loss": 2.7613079071044924, + "step": 76760 + }, + { + "epoch": 0.38385, + "grad_norm": 0.32197415828704834, + "learning_rate": 0.00034376564946210297, + "loss": 2.7287445068359375, + "step": 76770 + }, + { + "epoch": 0.3839, + "grad_norm": 0.3465311527252197, + "learning_rate": 0.0003437288773883436, + "loss": 2.768386459350586, + "step": 76780 + }, + { + "epoch": 0.38395, + "grad_norm": 0.3440542221069336, + "learning_rate": 0.00034369210295496025, + "loss": 2.773102378845215, + "step": 76790 + }, + { + "epoch": 0.384, + "grad_norm": 0.3581540882587433, + "learning_rate": 0.00034365532616287893, + "loss": 2.6599878311157226, + "step": 76800 + }, + { + "epoch": 0.38405, + "grad_norm": 0.3749864101409912, + "learning_rate": 0.0003436185470130254, + "loss": 2.7406118392944334, + "step": 76810 + }, + { + "epoch": 0.3841, + "grad_norm": 0.301746666431427, + "learning_rate": 0.00034358176550632546, + "loss": 2.7220035552978517, + "step": 76820 + }, + { + "epoch": 0.38415, + "grad_norm": 0.2758578062057495, + "learning_rate": 0.0003435449816437053, + "loss": 2.7316120147705076, + "step": 76830 + }, + { + "epoch": 0.3842, + "grad_norm": 0.3756757378578186, + "learning_rate": 0.0003435081954260908, + "loss": 2.7503923416137694, + "step": 76840 + }, + { + "epoch": 0.38425, + "grad_norm": 0.33851858973503113, + "learning_rate": 0.00034347140685440816, + "loss": 2.746083068847656, + "step": 76850 + }, + { + "epoch": 0.3843, + "grad_norm": 0.30065932869911194, + "learning_rate": 0.0003434346159295835, + "loss": 2.6717647552490233, + "step": 76860 + }, + { + "epoch": 0.38435, + "grad_norm": 0.34483182430267334, + "learning_rate": 0.0003433978226525429, + "loss": 2.748577117919922, + "step": 76870 + }, + { + "epoch": 0.3844, + "grad_norm": 0.37645408511161804, + "learning_rate": 0.00034336102702421276, + "loss": 2.735739898681641, + "step": 76880 + }, + { + "epoch": 0.38445, + "grad_norm": 0.3501030504703522, + "learning_rate": 0.00034332422904551947, + "loss": 2.7448219299316405, + "step": 76890 + }, + { + "epoch": 0.3845, + "grad_norm": 0.3404141366481781, + "learning_rate": 0.0003432874287173893, + "loss": 2.7461503982543944, + "step": 76900 + }, + { + "epoch": 0.38455, + "grad_norm": 0.3238210082054138, + "learning_rate": 0.00034325062604074867, + "loss": 2.702819061279297, + "step": 76910 + }, + { + "epoch": 0.3846, + "grad_norm": 0.3083793520927429, + "learning_rate": 0.0003432138210165243, + "loss": 2.7333555221557617, + "step": 76920 + }, + { + "epoch": 0.38465, + "grad_norm": 0.3416577875614166, + "learning_rate": 0.00034317701364564246, + "loss": 2.718702507019043, + "step": 76930 + }, + { + "epoch": 0.3847, + "grad_norm": 0.3666040897369385, + "learning_rate": 0.00034314020392903, + "loss": 2.7033809661865233, + "step": 76940 + }, + { + "epoch": 0.38475, + "grad_norm": 0.32346832752227783, + "learning_rate": 0.00034310339186761354, + "loss": 2.743411636352539, + "step": 76950 + }, + { + "epoch": 0.3848, + "grad_norm": 0.36163151264190674, + "learning_rate": 0.0003430665774623197, + "loss": 2.717450714111328, + "step": 76960 + }, + { + "epoch": 0.38485, + "grad_norm": 0.32756298780441284, + "learning_rate": 0.00034302976071407557, + "loss": 2.713779258728027, + "step": 76970 + }, + { + "epoch": 0.3849, + "grad_norm": 0.3057890236377716, + "learning_rate": 0.00034299294162380765, + "loss": 2.730538177490234, + "step": 76980 + }, + { + "epoch": 0.38495, + "grad_norm": 0.3290538489818573, + "learning_rate": 0.00034295612019244313, + "loss": 2.7099224090576173, + "step": 76990 + }, + { + "epoch": 0.385, + "grad_norm": 0.36099177598953247, + "learning_rate": 0.00034291929642090905, + "loss": 2.7692371368408204, + "step": 77000 + }, + { + "epoch": 0.38505, + "grad_norm": 0.3236372172832489, + "learning_rate": 0.000342882470310132, + "loss": 2.7453792572021483, + "step": 77010 + }, + { + "epoch": 0.3851, + "grad_norm": 0.32944291830062866, + "learning_rate": 0.0003428456418610396, + "loss": 2.6887186050415037, + "step": 77020 + }, + { + "epoch": 0.38515, + "grad_norm": 0.3582788407802582, + "learning_rate": 0.00034280881107455875, + "loss": 2.725630187988281, + "step": 77030 + }, + { + "epoch": 0.3852, + "grad_norm": 0.40550002455711365, + "learning_rate": 0.0003427719779516167, + "loss": 2.78402099609375, + "step": 77040 + }, + { + "epoch": 0.38525, + "grad_norm": 0.3496074676513672, + "learning_rate": 0.0003427351424931407, + "loss": 2.7295307159423827, + "step": 77050 + }, + { + "epoch": 0.3853, + "grad_norm": 0.3609922230243683, + "learning_rate": 0.00034269830470005816, + "loss": 2.738515853881836, + "step": 77060 + }, + { + "epoch": 0.38535, + "grad_norm": 0.34482693672180176, + "learning_rate": 0.00034266146457329635, + "loss": 2.7459234237670898, + "step": 77070 + }, + { + "epoch": 0.3854, + "grad_norm": 0.3568064570426941, + "learning_rate": 0.00034262462211378283, + "loss": 2.7101680755615236, + "step": 77080 + }, + { + "epoch": 0.38545, + "grad_norm": 0.33320462703704834, + "learning_rate": 0.000342587777322445, + "loss": 2.7658639907836915, + "step": 77090 + }, + { + "epoch": 0.3855, + "grad_norm": 0.3166840076446533, + "learning_rate": 0.00034255093020021054, + "loss": 2.719900894165039, + "step": 77100 + }, + { + "epoch": 0.38555, + "grad_norm": 0.3388938903808594, + "learning_rate": 0.00034251408074800706, + "loss": 2.724796104431152, + "step": 77110 + }, + { + "epoch": 0.3856, + "grad_norm": 0.3475969135761261, + "learning_rate": 0.0003424772289667622, + "loss": 2.7474912643432616, + "step": 77120 + }, + { + "epoch": 0.38565, + "grad_norm": 0.31751322746276855, + "learning_rate": 0.0003424403748574037, + "loss": 2.7206501007080077, + "step": 77130 + }, + { + "epoch": 0.3857, + "grad_norm": 0.33483731746673584, + "learning_rate": 0.0003424035184208594, + "loss": 2.746212196350098, + "step": 77140 + }, + { + "epoch": 0.38575, + "grad_norm": 0.31462669372558594, + "learning_rate": 0.0003423666596580571, + "loss": 2.7149612426757814, + "step": 77150 + }, + { + "epoch": 0.3858, + "grad_norm": 0.3533337116241455, + "learning_rate": 0.00034232979856992486, + "loss": 2.7140165328979493, + "step": 77160 + }, + { + "epoch": 0.38585, + "grad_norm": 0.38385969400405884, + "learning_rate": 0.00034229293515739044, + "loss": 2.699262809753418, + "step": 77170 + }, + { + "epoch": 0.3859, + "grad_norm": 0.3316398859024048, + "learning_rate": 0.0003422560694213821, + "loss": 2.751161575317383, + "step": 77180 + }, + { + "epoch": 0.38595, + "grad_norm": 0.36703231930732727, + "learning_rate": 0.0003422192013628277, + "loss": 2.7819774627685545, + "step": 77190 + }, + { + "epoch": 0.386, + "grad_norm": 0.32941457629203796, + "learning_rate": 0.00034218233098265563, + "loss": 2.7276399612426756, + "step": 77200 + }, + { + "epoch": 0.38605, + "grad_norm": 0.34426507353782654, + "learning_rate": 0.000342145458281794, + "loss": 2.706533432006836, + "step": 77210 + }, + { + "epoch": 0.3861, + "grad_norm": 0.36684292554855347, + "learning_rate": 0.000342108583261171, + "loss": 2.757851791381836, + "step": 77220 + }, + { + "epoch": 0.38615, + "grad_norm": 0.3165797293186188, + "learning_rate": 0.0003420717059217151, + "loss": 2.754459190368652, + "step": 77230 + }, + { + "epoch": 0.3862, + "grad_norm": 0.31249934434890747, + "learning_rate": 0.0003420348262643546, + "loss": 2.7247575759887694, + "step": 77240 + }, + { + "epoch": 0.38625, + "grad_norm": 0.3243383765220642, + "learning_rate": 0.0003419979442900179, + "loss": 2.718629837036133, + "step": 77250 + }, + { + "epoch": 0.3863, + "grad_norm": 0.34437477588653564, + "learning_rate": 0.0003419610599996336, + "loss": 2.766041564941406, + "step": 77260 + }, + { + "epoch": 0.38635, + "grad_norm": 0.39469996094703674, + "learning_rate": 0.0003419241733941303, + "loss": 2.708729553222656, + "step": 77270 + }, + { + "epoch": 0.3864, + "grad_norm": 0.31425002217292786, + "learning_rate": 0.00034188728447443644, + "loss": 2.7232826232910154, + "step": 77280 + }, + { + "epoch": 0.38645, + "grad_norm": 0.37066712975502014, + "learning_rate": 0.0003418503932414809, + "loss": 2.732994270324707, + "step": 77290 + }, + { + "epoch": 0.3865, + "grad_norm": 0.35038265585899353, + "learning_rate": 0.0003418134996961923, + "loss": 2.809754180908203, + "step": 77300 + }, + { + "epoch": 0.38655, + "grad_norm": 0.3626604676246643, + "learning_rate": 0.0003417766038394994, + "loss": 2.735930633544922, + "step": 77310 + }, + { + "epoch": 0.3866, + "grad_norm": 0.35391271114349365, + "learning_rate": 0.00034173970567233115, + "loss": 2.741852569580078, + "step": 77320 + }, + { + "epoch": 0.38665, + "grad_norm": 0.3544723391532898, + "learning_rate": 0.0003417028051956165, + "loss": 2.7224170684814455, + "step": 77330 + }, + { + "epoch": 0.3867, + "grad_norm": 0.28926360607147217, + "learning_rate": 0.00034166590241028425, + "loss": 2.7044118881225585, + "step": 77340 + }, + { + "epoch": 0.38675, + "grad_norm": 0.33009597659111023, + "learning_rate": 0.00034162899731726357, + "loss": 2.7683694839477537, + "step": 77350 + }, + { + "epoch": 0.3868, + "grad_norm": 0.29237139225006104, + "learning_rate": 0.00034159208991748334, + "loss": 2.75208740234375, + "step": 77360 + }, + { + "epoch": 0.38685, + "grad_norm": 0.3236350417137146, + "learning_rate": 0.00034155518021187304, + "loss": 2.75323371887207, + "step": 77370 + }, + { + "epoch": 0.3869, + "grad_norm": 0.3165416717529297, + "learning_rate": 0.0003415182682013617, + "loss": 2.7282066345214844, + "step": 77380 + }, + { + "epoch": 0.38695, + "grad_norm": 0.34823518991470337, + "learning_rate": 0.00034148135388687844, + "loss": 2.787122917175293, + "step": 77390 + }, + { + "epoch": 0.387, + "grad_norm": 0.32457801699638367, + "learning_rate": 0.00034144443726935277, + "loss": 2.7806982040405273, + "step": 77400 + }, + { + "epoch": 0.38705, + "grad_norm": 0.34088289737701416, + "learning_rate": 0.000341407518349714, + "loss": 2.6984819412231444, + "step": 77410 + }, + { + "epoch": 0.3871, + "grad_norm": 0.3108751177787781, + "learning_rate": 0.00034137059712889155, + "loss": 2.7482385635375977, + "step": 77420 + }, + { + "epoch": 0.38715, + "grad_norm": 0.34820491075515747, + "learning_rate": 0.00034133367360781496, + "loss": 2.7079505920410156, + "step": 77430 + }, + { + "epoch": 0.3872, + "grad_norm": 0.3299744725227356, + "learning_rate": 0.0003412967477874137, + "loss": 2.769305419921875, + "step": 77440 + }, + { + "epoch": 0.38725, + "grad_norm": 0.33684754371643066, + "learning_rate": 0.00034125981966861755, + "loss": 2.7585433959960937, + "step": 77450 + }, + { + "epoch": 0.3873, + "grad_norm": 0.3511207401752472, + "learning_rate": 0.0003412228892523559, + "loss": 2.7866960525512696, + "step": 77460 + }, + { + "epoch": 0.38735, + "grad_norm": 0.36231017112731934, + "learning_rate": 0.0003411859565395587, + "loss": 2.7109018325805665, + "step": 77470 + }, + { + "epoch": 0.3874, + "grad_norm": 0.38500455021858215, + "learning_rate": 0.0003411490215311556, + "loss": 2.7394287109375, + "step": 77480 + }, + { + "epoch": 0.38745, + "grad_norm": 0.3244083523750305, + "learning_rate": 0.00034111208422807647, + "loss": 2.6768394470214845, + "step": 77490 + }, + { + "epoch": 0.3875, + "grad_norm": 0.3911060392856598, + "learning_rate": 0.00034107514463125135, + "loss": 2.7369735717773436, + "step": 77500 + }, + { + "epoch": 0.38755, + "grad_norm": 0.31788793206214905, + "learning_rate": 0.00034103820274161, + "loss": 2.7428802490234374, + "step": 77510 + }, + { + "epoch": 0.3876, + "grad_norm": 0.3561636805534363, + "learning_rate": 0.0003410012585600825, + "loss": 2.7351207733154297, + "step": 77520 + }, + { + "epoch": 0.38765, + "grad_norm": 0.3609500527381897, + "learning_rate": 0.00034096431208759895, + "loss": 2.7238927841186524, + "step": 77530 + }, + { + "epoch": 0.3877, + "grad_norm": 0.38404494524002075, + "learning_rate": 0.00034092736332508945, + "loss": 2.7841865539550783, + "step": 77540 + }, + { + "epoch": 0.38775, + "grad_norm": 0.3517138957977295, + "learning_rate": 0.0003408904122734842, + "loss": 2.720384979248047, + "step": 77550 + }, + { + "epoch": 0.3878, + "grad_norm": 0.33729252219200134, + "learning_rate": 0.0003408534589337133, + "loss": 2.727949523925781, + "step": 77560 + }, + { + "epoch": 0.38785, + "grad_norm": 0.3366270959377289, + "learning_rate": 0.00034081650330670737, + "loss": 2.708819580078125, + "step": 77570 + }, + { + "epoch": 0.3879, + "grad_norm": 0.3117811381816864, + "learning_rate": 0.0003407795453933964, + "loss": 2.750870704650879, + "step": 77580 + }, + { + "epoch": 0.38795, + "grad_norm": 0.3143388628959656, + "learning_rate": 0.00034074258519471115, + "loss": 2.7146942138671877, + "step": 77590 + }, + { + "epoch": 0.388, + "grad_norm": 0.40443482995033264, + "learning_rate": 0.0003407056227115818, + "loss": 2.761505126953125, + "step": 77600 + }, + { + "epoch": 0.38805, + "grad_norm": 0.34380802512168884, + "learning_rate": 0.00034066865794493906, + "loss": 2.705966567993164, + "step": 77610 + }, + { + "epoch": 0.3881, + "grad_norm": 0.34850114583969116, + "learning_rate": 0.00034063169089571345, + "loss": 2.6913660049438475, + "step": 77620 + }, + { + "epoch": 0.38815, + "grad_norm": 0.3629045784473419, + "learning_rate": 0.0003405947215648356, + "loss": 2.7541213989257813, + "step": 77630 + }, + { + "epoch": 0.3882, + "grad_norm": 0.36075496673583984, + "learning_rate": 0.00034055774995323624, + "loss": 2.773616981506348, + "step": 77640 + }, + { + "epoch": 0.38825, + "grad_norm": 0.3470989763736725, + "learning_rate": 0.00034052077606184616, + "loss": 2.735066223144531, + "step": 77650 + }, + { + "epoch": 0.3883, + "grad_norm": 0.33111587166786194, + "learning_rate": 0.0003404837998915961, + "loss": 2.774809455871582, + "step": 77660 + }, + { + "epoch": 0.38835, + "grad_norm": 0.3435176908969879, + "learning_rate": 0.0003404468214434171, + "loss": 2.7402950286865235, + "step": 77670 + }, + { + "epoch": 0.3884, + "grad_norm": 0.32562315464019775, + "learning_rate": 0.0003404098407182399, + "loss": 2.7711177825927735, + "step": 77680 + }, + { + "epoch": 0.38845, + "grad_norm": 0.33111822605133057, + "learning_rate": 0.00034037285771699555, + "loss": 2.689879608154297, + "step": 77690 + }, + { + "epoch": 0.3885, + "grad_norm": 0.33542925119400024, + "learning_rate": 0.00034033587244061507, + "loss": 2.6959407806396483, + "step": 77700 + }, + { + "epoch": 0.38855, + "grad_norm": 0.32149428129196167, + "learning_rate": 0.0003402988848900297, + "loss": 2.7121536254882814, + "step": 77710 + }, + { + "epoch": 0.3886, + "grad_norm": 0.37184372544288635, + "learning_rate": 0.00034026189506617045, + "loss": 2.7342496871948243, + "step": 77720 + }, + { + "epoch": 0.38865, + "grad_norm": 0.3581389784812927, + "learning_rate": 0.0003402249029699687, + "loss": 2.7395883560180665, + "step": 77730 + }, + { + "epoch": 0.3887, + "grad_norm": 0.32202616333961487, + "learning_rate": 0.0003401879086023554, + "loss": 2.782870864868164, + "step": 77740 + }, + { + "epoch": 0.38875, + "grad_norm": 0.33824393153190613, + "learning_rate": 0.0003401509119642623, + "loss": 2.788253974914551, + "step": 77750 + }, + { + "epoch": 0.3888, + "grad_norm": 0.48616138100624084, + "learning_rate": 0.0003401139130566205, + "loss": 2.782031440734863, + "step": 77760 + }, + { + "epoch": 0.38885, + "grad_norm": 0.3612343370914459, + "learning_rate": 0.00034007691188036156, + "loss": 2.741411590576172, + "step": 77770 + }, + { + "epoch": 0.3889, + "grad_norm": 0.3562510907649994, + "learning_rate": 0.00034003990843641695, + "loss": 2.7574848175048827, + "step": 77780 + }, + { + "epoch": 0.38895, + "grad_norm": 0.3382144868373871, + "learning_rate": 0.0003400029027257183, + "loss": 2.717171478271484, + "step": 77790 + }, + { + "epoch": 0.389, + "grad_norm": 0.3614339232444763, + "learning_rate": 0.00033996589474919717, + "loss": 2.7131675720214843, + "step": 77800 + }, + { + "epoch": 0.38905, + "grad_norm": 0.3204841613769531, + "learning_rate": 0.0003399288845077852, + "loss": 2.754654312133789, + "step": 77810 + }, + { + "epoch": 0.3891, + "grad_norm": 0.3277742564678192, + "learning_rate": 0.0003398918720024141, + "loss": 2.746202087402344, + "step": 77820 + }, + { + "epoch": 0.38915, + "grad_norm": 0.35719722509384155, + "learning_rate": 0.00033985485723401587, + "loss": 2.779203414916992, + "step": 77830 + }, + { + "epoch": 0.3892, + "grad_norm": 0.32802972197532654, + "learning_rate": 0.0003398178402035221, + "loss": 2.762490653991699, + "step": 77840 + }, + { + "epoch": 0.38925, + "grad_norm": 0.3283425569534302, + "learning_rate": 0.0003397808209118649, + "loss": 2.7611656188964844, + "step": 77850 + }, + { + "epoch": 0.3893, + "grad_norm": 0.3914891481399536, + "learning_rate": 0.000339743799359976, + "loss": 2.7324623107910155, + "step": 77860 + }, + { + "epoch": 0.38935, + "grad_norm": 0.3602728545665741, + "learning_rate": 0.0003397067755487877, + "loss": 2.731901741027832, + "step": 77870 + }, + { + "epoch": 0.3894, + "grad_norm": 0.33397719264030457, + "learning_rate": 0.0003396697494792318, + "loss": 2.7389415740966796, + "step": 77880 + }, + { + "epoch": 0.38945, + "grad_norm": 0.3334265947341919, + "learning_rate": 0.00033963272115224066, + "loss": 2.720175933837891, + "step": 77890 + }, + { + "epoch": 0.3895, + "grad_norm": 0.3189128339290619, + "learning_rate": 0.0003395956905687463, + "loss": 2.7484264373779297, + "step": 77900 + }, + { + "epoch": 0.38955, + "grad_norm": 0.3301943838596344, + "learning_rate": 0.00033955865772968105, + "loss": 2.7585472106933593, + "step": 77910 + }, + { + "epoch": 0.3896, + "grad_norm": 0.32978811860084534, + "learning_rate": 0.0003395216226359772, + "loss": 2.7411048889160154, + "step": 77920 + }, + { + "epoch": 0.38965, + "grad_norm": 0.33206459879875183, + "learning_rate": 0.00033948458528856705, + "loss": 2.7623329162597656, + "step": 77930 + }, + { + "epoch": 0.3897, + "grad_norm": 0.3438114523887634, + "learning_rate": 0.0003394475456883831, + "loss": 2.7076995849609373, + "step": 77940 + }, + { + "epoch": 0.38975, + "grad_norm": 0.35692691802978516, + "learning_rate": 0.0003394105038363578, + "loss": 2.780703544616699, + "step": 77950 + }, + { + "epoch": 0.3898, + "grad_norm": 0.3410336971282959, + "learning_rate": 0.0003393734597334236, + "loss": 2.7109086990356444, + "step": 77960 + }, + { + "epoch": 0.38985, + "grad_norm": 0.379363089799881, + "learning_rate": 0.00033933641338051323, + "loss": 2.756540298461914, + "step": 77970 + }, + { + "epoch": 0.3899, + "grad_norm": 0.3429560959339142, + "learning_rate": 0.0003392993647785592, + "loss": 2.6977874755859377, + "step": 77980 + }, + { + "epoch": 0.38995, + "grad_norm": 0.3538571894168854, + "learning_rate": 0.00033926231392849426, + "loss": 2.718812370300293, + "step": 77990 + }, + { + "epoch": 0.39, + "grad_norm": 0.3579083979129791, + "learning_rate": 0.00033922526083125115, + "loss": 2.764999771118164, + "step": 78000 + }, + { + "epoch": 0.39005, + "grad_norm": 0.3405689001083374, + "learning_rate": 0.00033918820548776274, + "loss": 2.743444633483887, + "step": 78010 + }, + { + "epoch": 0.3901, + "grad_norm": 0.3428488075733185, + "learning_rate": 0.0003391511478989618, + "loss": 2.7537668228149412, + "step": 78020 + }, + { + "epoch": 0.39015, + "grad_norm": 0.3700137436389923, + "learning_rate": 0.00033911408806578134, + "loss": 2.747916412353516, + "step": 78030 + }, + { + "epoch": 0.3902, + "grad_norm": 0.30836230516433716, + "learning_rate": 0.0003390770259891543, + "loss": 2.7168390274047853, + "step": 78040 + }, + { + "epoch": 0.39025, + "grad_norm": 0.29916414618492126, + "learning_rate": 0.00033903996167001377, + "loss": 2.7402217864990233, + "step": 78050 + }, + { + "epoch": 0.3903, + "grad_norm": 0.34360671043395996, + "learning_rate": 0.0003390028951092927, + "loss": 2.727975082397461, + "step": 78060 + }, + { + "epoch": 0.39035, + "grad_norm": 0.3709241449832916, + "learning_rate": 0.0003389658263079244, + "loss": 2.789530563354492, + "step": 78070 + }, + { + "epoch": 0.3904, + "grad_norm": 0.34582656621932983, + "learning_rate": 0.0003389287552668421, + "loss": 2.760502243041992, + "step": 78080 + }, + { + "epoch": 0.39045, + "grad_norm": 0.3317185938358307, + "learning_rate": 0.00033889168198697884, + "loss": 2.7171730041503905, + "step": 78090 + }, + { + "epoch": 0.3905, + "grad_norm": 0.32787877321243286, + "learning_rate": 0.0003388546064692682, + "loss": 2.700154495239258, + "step": 78100 + }, + { + "epoch": 0.39055, + "grad_norm": 0.3724697530269623, + "learning_rate": 0.00033881752871464335, + "loss": 2.768468475341797, + "step": 78110 + }, + { + "epoch": 0.3906, + "grad_norm": 0.3350907862186432, + "learning_rate": 0.00033878044872403786, + "loss": 2.6814130783081054, + "step": 78120 + }, + { + "epoch": 0.39065, + "grad_norm": 0.36388564109802246, + "learning_rate": 0.0003387433664983851, + "loss": 2.7164752960205076, + "step": 78130 + }, + { + "epoch": 0.3907, + "grad_norm": 0.35382241010665894, + "learning_rate": 0.00033870628203861873, + "loss": 2.738734817504883, + "step": 78140 + }, + { + "epoch": 0.39075, + "grad_norm": 0.3483697474002838, + "learning_rate": 0.0003386691953456723, + "loss": 2.7143056869506834, + "step": 78150 + }, + { + "epoch": 0.3908, + "grad_norm": 0.324856162071228, + "learning_rate": 0.0003386321064204795, + "loss": 2.7025146484375, + "step": 78160 + }, + { + "epoch": 0.39085, + "grad_norm": 0.3295777440071106, + "learning_rate": 0.00033859501526397397, + "loss": 2.646902847290039, + "step": 78170 + }, + { + "epoch": 0.3909, + "grad_norm": 0.3352917432785034, + "learning_rate": 0.0003385579218770895, + "loss": 2.7619279861450194, + "step": 78180 + }, + { + "epoch": 0.39095, + "grad_norm": 0.36215442419052124, + "learning_rate": 0.00033852082626076, + "loss": 2.7416887283325195, + "step": 78190 + }, + { + "epoch": 0.391, + "grad_norm": 0.3319534957408905, + "learning_rate": 0.0003384837284159192, + "loss": 2.7075042724609375, + "step": 78200 + }, + { + "epoch": 0.39105, + "grad_norm": 0.3543872833251953, + "learning_rate": 0.0003384466283435012, + "loss": 2.8108493804931642, + "step": 78210 + }, + { + "epoch": 0.3911, + "grad_norm": 0.33563095331192017, + "learning_rate": 0.00033840952604443996, + "loss": 2.7358827590942383, + "step": 78220 + }, + { + "epoch": 0.39115, + "grad_norm": 0.36528974771499634, + "learning_rate": 0.0003383724215196694, + "loss": 2.7248329162597655, + "step": 78230 + }, + { + "epoch": 0.3912, + "grad_norm": 0.30759575963020325, + "learning_rate": 0.0003383353147701238, + "loss": 2.702420234680176, + "step": 78240 + }, + { + "epoch": 0.39125, + "grad_norm": 0.35378286242485046, + "learning_rate": 0.0003382982057967372, + "loss": 2.733050727844238, + "step": 78250 + }, + { + "epoch": 0.3913, + "grad_norm": 0.3360952138900757, + "learning_rate": 0.00033826109460044387, + "loss": 2.7511301040649414, + "step": 78260 + }, + { + "epoch": 0.39135, + "grad_norm": 0.3211061656475067, + "learning_rate": 0.000338223981182178, + "loss": 2.6717191696166993, + "step": 78270 + }, + { + "epoch": 0.3914, + "grad_norm": 0.3161756694316864, + "learning_rate": 0.00033818686554287406, + "loss": 2.7521448135375977, + "step": 78280 + }, + { + "epoch": 0.39145, + "grad_norm": 0.3400798439979553, + "learning_rate": 0.00033814974768346635, + "loss": 2.7446014404296877, + "step": 78290 + }, + { + "epoch": 0.3915, + "grad_norm": 0.3682464063167572, + "learning_rate": 0.0003381126276048894, + "loss": 2.753184700012207, + "step": 78300 + }, + { + "epoch": 0.39155, + "grad_norm": 0.3382704257965088, + "learning_rate": 0.0003380755053080775, + "loss": 2.778424072265625, + "step": 78310 + }, + { + "epoch": 0.3916, + "grad_norm": 0.32964566349983215, + "learning_rate": 0.00033803838079396546, + "loss": 2.695981979370117, + "step": 78320 + }, + { + "epoch": 0.39165, + "grad_norm": 0.32127586007118225, + "learning_rate": 0.0003380012540634878, + "loss": 2.721969985961914, + "step": 78330 + }, + { + "epoch": 0.3917, + "grad_norm": 0.3166256546974182, + "learning_rate": 0.00033796412511757906, + "loss": 2.7388708114624025, + "step": 78340 + }, + { + "epoch": 0.39175, + "grad_norm": 0.30842354893684387, + "learning_rate": 0.00033792699395717406, + "loss": 2.6933021545410156, + "step": 78350 + }, + { + "epoch": 0.3918, + "grad_norm": 0.34266379475593567, + "learning_rate": 0.00033788986058320763, + "loss": 2.717488098144531, + "step": 78360 + }, + { + "epoch": 0.39185, + "grad_norm": 0.39386314153671265, + "learning_rate": 0.00033785272499661453, + "loss": 2.7423629760742188, + "step": 78370 + }, + { + "epoch": 0.3919, + "grad_norm": 0.32296955585479736, + "learning_rate": 0.0003378155871983297, + "loss": 2.6992256164550783, + "step": 78380 + }, + { + "epoch": 0.39195, + "grad_norm": 0.3410117030143738, + "learning_rate": 0.00033777844718928794, + "loss": 2.8038143157958983, + "step": 78390 + }, + { + "epoch": 0.392, + "grad_norm": 0.335631400346756, + "learning_rate": 0.0003377413049704245, + "loss": 2.6869701385498046, + "step": 78400 + }, + { + "epoch": 0.39205, + "grad_norm": 0.3137277066707611, + "learning_rate": 0.00033770416054267424, + "loss": 2.759465217590332, + "step": 78410 + }, + { + "epoch": 0.3921, + "grad_norm": 0.3468489348888397, + "learning_rate": 0.00033766701390697224, + "loss": 2.719141960144043, + "step": 78420 + }, + { + "epoch": 0.39215, + "grad_norm": 0.3454684019088745, + "learning_rate": 0.00033762986506425383, + "loss": 2.813299560546875, + "step": 78430 + }, + { + "epoch": 0.3922, + "grad_norm": 0.3192676603794098, + "learning_rate": 0.00033759271401545413, + "loss": 2.7759477615356447, + "step": 78440 + }, + { + "epoch": 0.39225, + "grad_norm": 0.3453798294067383, + "learning_rate": 0.0003375555607615084, + "loss": 2.7504884719848635, + "step": 78450 + }, + { + "epoch": 0.3923, + "grad_norm": 0.3337611258029938, + "learning_rate": 0.00033751840530335216, + "loss": 2.751864433288574, + "step": 78460 + }, + { + "epoch": 0.39235, + "grad_norm": 0.37088027596473694, + "learning_rate": 0.0003374812476419205, + "loss": 2.791861343383789, + "step": 78470 + }, + { + "epoch": 0.3924, + "grad_norm": 0.36009761691093445, + "learning_rate": 0.0003374440877781491, + "loss": 2.777669143676758, + "step": 78480 + }, + { + "epoch": 0.39245, + "grad_norm": 0.33465316891670227, + "learning_rate": 0.0003374069257129734, + "loss": 2.737460136413574, + "step": 78490 + }, + { + "epoch": 0.3925, + "grad_norm": 0.3305393159389496, + "learning_rate": 0.00033736976144732876, + "loss": 2.7242992401123045, + "step": 78500 + }, + { + "epoch": 0.39255, + "grad_norm": 0.3299550414085388, + "learning_rate": 0.00033733259498215114, + "loss": 2.6843116760253904, + "step": 78510 + }, + { + "epoch": 0.3926, + "grad_norm": 0.3345004916191101, + "learning_rate": 0.00033729542631837593, + "loss": 2.79488525390625, + "step": 78520 + }, + { + "epoch": 0.39265, + "grad_norm": 0.32980337738990784, + "learning_rate": 0.00033725825545693897, + "loss": 2.731135940551758, + "step": 78530 + }, + { + "epoch": 0.3927, + "grad_norm": 0.3275350034236908, + "learning_rate": 0.000337221082398776, + "loss": 2.728012466430664, + "step": 78540 + }, + { + "epoch": 0.39275, + "grad_norm": 0.3857799768447876, + "learning_rate": 0.0003371839071448229, + "loss": 2.776804733276367, + "step": 78550 + }, + { + "epoch": 0.3928, + "grad_norm": 0.3601057827472687, + "learning_rate": 0.0003371467296960155, + "loss": 2.7497493743896486, + "step": 78560 + }, + { + "epoch": 0.39285, + "grad_norm": 0.3278190791606903, + "learning_rate": 0.0003371095500532898, + "loss": 2.721324920654297, + "step": 78570 + }, + { + "epoch": 0.3929, + "grad_norm": 0.32511481642723083, + "learning_rate": 0.0003370723682175817, + "loss": 2.781374931335449, + "step": 78580 + }, + { + "epoch": 0.39295, + "grad_norm": 0.3344132900238037, + "learning_rate": 0.0003370351841898273, + "loss": 2.7711320877075196, + "step": 78590 + }, + { + "epoch": 0.393, + "grad_norm": 0.32795292139053345, + "learning_rate": 0.0003369979979709628, + "loss": 2.7513864517211912, + "step": 78600 + }, + { + "epoch": 0.39305, + "grad_norm": 0.5028729438781738, + "learning_rate": 0.0003369608095619242, + "loss": 2.697469139099121, + "step": 78610 + }, + { + "epoch": 0.3931, + "grad_norm": 0.33486801385879517, + "learning_rate": 0.00033692361896364785, + "loss": 2.702000045776367, + "step": 78620 + }, + { + "epoch": 0.39315, + "grad_norm": 0.4207281172275543, + "learning_rate": 0.00033688642617707, + "loss": 2.739105224609375, + "step": 78630 + }, + { + "epoch": 0.3932, + "grad_norm": 0.36517027020454407, + "learning_rate": 0.00033684923120312686, + "loss": 2.7531700134277344, + "step": 78640 + }, + { + "epoch": 0.39325, + "grad_norm": 0.39459729194641113, + "learning_rate": 0.00033681203404275497, + "loss": 2.762449836730957, + "step": 78650 + }, + { + "epoch": 0.3933, + "grad_norm": 0.32563066482543945, + "learning_rate": 0.0003367748346968907, + "loss": 2.701444625854492, + "step": 78660 + }, + { + "epoch": 0.39335, + "grad_norm": 0.39625000953674316, + "learning_rate": 0.00033673763316647046, + "loss": 2.789453887939453, + "step": 78670 + }, + { + "epoch": 0.3934, + "grad_norm": 0.31863877177238464, + "learning_rate": 0.0003367004294524311, + "loss": 2.7260457992553713, + "step": 78680 + }, + { + "epoch": 0.39345, + "grad_norm": 0.35417845845222473, + "learning_rate": 0.0003366632235557088, + "loss": 2.7140512466430664, + "step": 78690 + }, + { + "epoch": 0.3935, + "grad_norm": 0.36556947231292725, + "learning_rate": 0.0003366260154772405, + "loss": 2.736545181274414, + "step": 78700 + }, + { + "epoch": 0.39355, + "grad_norm": 0.34387293457984924, + "learning_rate": 0.0003365888052179628, + "loss": 2.7531620025634767, + "step": 78710 + }, + { + "epoch": 0.3936, + "grad_norm": 0.3327597677707672, + "learning_rate": 0.0003365515927788125, + "loss": 2.7830888748168947, + "step": 78720 + }, + { + "epoch": 0.39365, + "grad_norm": 0.3589378595352173, + "learning_rate": 0.00033651437816072647, + "loss": 2.7627735137939453, + "step": 78730 + }, + { + "epoch": 0.3937, + "grad_norm": 0.37446457147598267, + "learning_rate": 0.00033647716136464155, + "loss": 2.7500946044921877, + "step": 78740 + }, + { + "epoch": 0.39375, + "grad_norm": 0.3628113567829132, + "learning_rate": 0.00033643994239149466, + "loss": 2.756759452819824, + "step": 78750 + }, + { + "epoch": 0.3938, + "grad_norm": 0.3749728202819824, + "learning_rate": 0.00033640272124222287, + "loss": 2.7477014541625975, + "step": 78760 + }, + { + "epoch": 0.39385, + "grad_norm": 0.3298308551311493, + "learning_rate": 0.00033636549791776305, + "loss": 2.770255279541016, + "step": 78770 + }, + { + "epoch": 0.3939, + "grad_norm": 0.3207687437534332, + "learning_rate": 0.0003363282724190524, + "loss": 2.722528648376465, + "step": 78780 + }, + { + "epoch": 0.39395, + "grad_norm": 0.37033987045288086, + "learning_rate": 0.00033629104474702814, + "loss": 2.7694997787475586, + "step": 78790 + }, + { + "epoch": 0.394, + "grad_norm": 0.34310182929039, + "learning_rate": 0.00033625381490262737, + "loss": 2.765543556213379, + "step": 78800 + }, + { + "epoch": 0.39405, + "grad_norm": 0.32884183526039124, + "learning_rate": 0.00033621658288678736, + "loss": 2.7402166366577148, + "step": 78810 + }, + { + "epoch": 0.3941, + "grad_norm": 0.3429965376853943, + "learning_rate": 0.0003361793487004455, + "loss": 2.7171239852905273, + "step": 78820 + }, + { + "epoch": 0.39415, + "grad_norm": 0.3499976396560669, + "learning_rate": 0.00033614211234453905, + "loss": 2.778386688232422, + "step": 78830 + }, + { + "epoch": 0.3942, + "grad_norm": 0.3602846562862396, + "learning_rate": 0.0003361048738200056, + "loss": 2.7248788833618165, + "step": 78840 + }, + { + "epoch": 0.39425, + "grad_norm": 0.34764567017555237, + "learning_rate": 0.0003360676331277824, + "loss": 2.7534454345703123, + "step": 78850 + }, + { + "epoch": 0.3943, + "grad_norm": 0.342232882976532, + "learning_rate": 0.0003360303902688072, + "loss": 2.678942108154297, + "step": 78860 + }, + { + "epoch": 0.39435, + "grad_norm": 0.3643326759338379, + "learning_rate": 0.0003359931452440175, + "loss": 2.73077278137207, + "step": 78870 + }, + { + "epoch": 0.3944, + "grad_norm": 0.3262583315372467, + "learning_rate": 0.0003359558980543509, + "loss": 2.692401123046875, + "step": 78880 + }, + { + "epoch": 0.39445, + "grad_norm": 0.334274560213089, + "learning_rate": 0.00033591864870074513, + "loss": 2.7148229598999025, + "step": 78890 + }, + { + "epoch": 0.3945, + "grad_norm": 0.3393418490886688, + "learning_rate": 0.000335881397184138, + "loss": 2.7474035263061523, + "step": 78900 + }, + { + "epoch": 0.39455, + "grad_norm": 0.32113027572631836, + "learning_rate": 0.0003358441435054673, + "loss": 2.7325998306274415, + "step": 78910 + }, + { + "epoch": 0.3946, + "grad_norm": 0.3045094907283783, + "learning_rate": 0.0003358068876656708, + "loss": 2.7080392837524414, + "step": 78920 + }, + { + "epoch": 0.39465, + "grad_norm": 0.352377325296402, + "learning_rate": 0.0003357696296656865, + "loss": 2.7441564559936524, + "step": 78930 + }, + { + "epoch": 0.3947, + "grad_norm": 0.3227420151233673, + "learning_rate": 0.0003357323695064522, + "loss": 2.693105125427246, + "step": 78940 + }, + { + "epoch": 0.39475, + "grad_norm": 0.36694908142089844, + "learning_rate": 0.00033569510718890635, + "loss": 2.6935188293457033, + "step": 78950 + }, + { + "epoch": 0.3948, + "grad_norm": 0.3505591154098511, + "learning_rate": 0.00033565784271398654, + "loss": 2.771173858642578, + "step": 78960 + }, + { + "epoch": 0.39485, + "grad_norm": 0.3621535301208496, + "learning_rate": 0.0003356205760826312, + "loss": 2.722396469116211, + "step": 78970 + }, + { + "epoch": 0.3949, + "grad_norm": 0.3931739330291748, + "learning_rate": 0.0003355833072957784, + "loss": 2.751246452331543, + "step": 78980 + }, + { + "epoch": 0.39495, + "grad_norm": 0.35196489095687866, + "learning_rate": 0.00033554603635436645, + "loss": 2.758851432800293, + "step": 78990 + }, + { + "epoch": 0.395, + "grad_norm": 0.36760368943214417, + "learning_rate": 0.0003355087632593335, + "loss": 2.754213333129883, + "step": 79000 + }, + { + "epoch": 0.39505, + "grad_norm": 0.34103894233703613, + "learning_rate": 0.0003354714880116181, + "loss": 2.6748588562011717, + "step": 79010 + }, + { + "epoch": 0.3951, + "grad_norm": 0.3093791604042053, + "learning_rate": 0.00033543421061215853, + "loss": 2.7336231231689454, + "step": 79020 + }, + { + "epoch": 0.39515, + "grad_norm": 0.32208508253097534, + "learning_rate": 0.0003353969310618933, + "loss": 2.720635414123535, + "step": 79030 + }, + { + "epoch": 0.3952, + "grad_norm": 0.34164658188819885, + "learning_rate": 0.00033535964936176084, + "loss": 2.7377178192138674, + "step": 79040 + }, + { + "epoch": 0.39525, + "grad_norm": 0.4157669246196747, + "learning_rate": 0.0003353223655126999, + "loss": 2.727446746826172, + "step": 79050 + }, + { + "epoch": 0.3953, + "grad_norm": 0.3673497438430786, + "learning_rate": 0.00033528507951564884, + "loss": 2.6821561813354493, + "step": 79060 + }, + { + "epoch": 0.39535, + "grad_norm": 0.3452690541744232, + "learning_rate": 0.0003352477913715465, + "loss": 2.7596450805664063, + "step": 79070 + }, + { + "epoch": 0.3954, + "grad_norm": 0.3470686078071594, + "learning_rate": 0.00033521050108133166, + "loss": 2.7228876113891602, + "step": 79080 + }, + { + "epoch": 0.39545, + "grad_norm": 0.3333761394023895, + "learning_rate": 0.00033517320864594305, + "loss": 2.743459701538086, + "step": 79090 + }, + { + "epoch": 0.3955, + "grad_norm": 0.37148818373680115, + "learning_rate": 0.00033513591406631935, + "loss": 2.7396724700927733, + "step": 79100 + }, + { + "epoch": 0.39555, + "grad_norm": 0.33320683240890503, + "learning_rate": 0.0003350986173433997, + "loss": 2.6966819763183594, + "step": 79110 + }, + { + "epoch": 0.3956, + "grad_norm": 0.3404179811477661, + "learning_rate": 0.00033506131847812283, + "loss": 2.729949188232422, + "step": 79120 + }, + { + "epoch": 0.39565, + "grad_norm": 0.32534778118133545, + "learning_rate": 0.0003350240174714279, + "loss": 2.6932220458984375, + "step": 79130 + }, + { + "epoch": 0.3957, + "grad_norm": 0.3262384831905365, + "learning_rate": 0.0003349867143242539, + "loss": 2.7072372436523438, + "step": 79140 + }, + { + "epoch": 0.39575, + "grad_norm": 0.42557230591773987, + "learning_rate": 0.0003349494090375399, + "loss": 2.7571359634399415, + "step": 79150 + }, + { + "epoch": 0.3958, + "grad_norm": 0.3843659460544586, + "learning_rate": 0.00033491210161222507, + "loss": 2.7474733352661134, + "step": 79160 + }, + { + "epoch": 0.39585, + "grad_norm": 0.3395986258983612, + "learning_rate": 0.00033487479204924876, + "loss": 2.76611213684082, + "step": 79170 + }, + { + "epoch": 0.3959, + "grad_norm": 0.3662030100822449, + "learning_rate": 0.00033483748034954995, + "loss": 2.697552490234375, + "step": 79180 + }, + { + "epoch": 0.39595, + "grad_norm": 0.3279010057449341, + "learning_rate": 0.00033480016651406833, + "loss": 2.6956201553344727, + "step": 79190 + }, + { + "epoch": 0.396, + "grad_norm": 0.32425200939178467, + "learning_rate": 0.00033476285054374297, + "loss": 2.7508012771606447, + "step": 79200 + }, + { + "epoch": 0.39605, + "grad_norm": 0.3898378312587738, + "learning_rate": 0.00033472553243951345, + "loss": 2.7426631927490233, + "step": 79210 + }, + { + "epoch": 0.3961, + "grad_norm": 0.3368774950504303, + "learning_rate": 0.00033468821220231917, + "loss": 2.6982404708862306, + "step": 79220 + }, + { + "epoch": 0.39615, + "grad_norm": 0.3713764250278473, + "learning_rate": 0.00033465088983309976, + "loss": 2.713998222351074, + "step": 79230 + }, + { + "epoch": 0.3962, + "grad_norm": 0.313637912273407, + "learning_rate": 0.0003346135653327947, + "loss": 2.727707862854004, + "step": 79240 + }, + { + "epoch": 0.39625, + "grad_norm": 0.3629339337348938, + "learning_rate": 0.0003345762387023438, + "loss": 2.7119794845581056, + "step": 79250 + }, + { + "epoch": 0.3963, + "grad_norm": 0.31440940499305725, + "learning_rate": 0.0003345389099426866, + "loss": 2.7279052734375, + "step": 79260 + }, + { + "epoch": 0.39635, + "grad_norm": 0.3552584946155548, + "learning_rate": 0.000334501579054763, + "loss": 2.774293327331543, + "step": 79270 + }, + { + "epoch": 0.3964, + "grad_norm": 0.3542127013206482, + "learning_rate": 0.0003344642460395126, + "loss": 2.728692626953125, + "step": 79280 + }, + { + "epoch": 0.39645, + "grad_norm": 0.3316887319087982, + "learning_rate": 0.0003344269108978754, + "loss": 2.7257963180541993, + "step": 79290 + }, + { + "epoch": 0.3965, + "grad_norm": 0.3868963122367859, + "learning_rate": 0.00033438957363079124, + "loss": 2.695109748840332, + "step": 79300 + }, + { + "epoch": 0.39655, + "grad_norm": 0.3624626398086548, + "learning_rate": 0.00033435223423920014, + "loss": 2.7436939239501954, + "step": 79310 + }, + { + "epoch": 0.3966, + "grad_norm": 0.37883704900741577, + "learning_rate": 0.00033431489272404215, + "loss": 2.76599063873291, + "step": 79320 + }, + { + "epoch": 0.39665, + "grad_norm": 0.3033877909183502, + "learning_rate": 0.0003342775490862573, + "loss": 2.7308723449707033, + "step": 79330 + }, + { + "epoch": 0.3967, + "grad_norm": 0.356518030166626, + "learning_rate": 0.00033424020332678566, + "loss": 2.792422294616699, + "step": 79340 + }, + { + "epoch": 0.39675, + "grad_norm": 0.35419702529907227, + "learning_rate": 0.0003342028554465676, + "loss": 2.744423675537109, + "step": 79350 + }, + { + "epoch": 0.3968, + "grad_norm": 0.34673264622688293, + "learning_rate": 0.00033416550544654306, + "loss": 2.744533920288086, + "step": 79360 + }, + { + "epoch": 0.39685, + "grad_norm": 0.39543431997299194, + "learning_rate": 0.00033412815332765255, + "loss": 2.753851318359375, + "step": 79370 + }, + { + "epoch": 0.3969, + "grad_norm": 0.3983745276927948, + "learning_rate": 0.00033409079909083636, + "loss": 2.7134611129760744, + "step": 79380 + }, + { + "epoch": 0.39695, + "grad_norm": 0.38377779722213745, + "learning_rate": 0.0003340534427370348, + "loss": 2.7525379180908205, + "step": 79390 + }, + { + "epoch": 0.397, + "grad_norm": 0.3263362944126129, + "learning_rate": 0.0003340160842671884, + "loss": 2.758576202392578, + "step": 79400 + }, + { + "epoch": 0.39705, + "grad_norm": 0.3324289917945862, + "learning_rate": 0.0003339787236822377, + "loss": 2.755950164794922, + "step": 79410 + }, + { + "epoch": 0.3971, + "grad_norm": 0.3557564616203308, + "learning_rate": 0.0003339413609831231, + "loss": 2.7538055419921874, + "step": 79420 + }, + { + "epoch": 0.39715, + "grad_norm": 0.3998018503189087, + "learning_rate": 0.00033390399617078543, + "loss": 2.751588249206543, + "step": 79430 + }, + { + "epoch": 0.3972, + "grad_norm": 0.37593087553977966, + "learning_rate": 0.00033386662924616515, + "loss": 2.6828075408935548, + "step": 79440 + }, + { + "epoch": 0.39725, + "grad_norm": 0.3414742648601532, + "learning_rate": 0.000333829260210203, + "loss": 2.7886810302734375, + "step": 79450 + }, + { + "epoch": 0.3973, + "grad_norm": 0.3363836705684662, + "learning_rate": 0.0003337918890638398, + "loss": 2.796514701843262, + "step": 79460 + }, + { + "epoch": 0.39735, + "grad_norm": 0.32729631662368774, + "learning_rate": 0.00033375451580801643, + "loss": 2.7787363052368166, + "step": 79470 + }, + { + "epoch": 0.3974, + "grad_norm": 0.3441981375217438, + "learning_rate": 0.00033371714044367354, + "loss": 2.7997684478759766, + "step": 79480 + }, + { + "epoch": 0.39745, + "grad_norm": 0.350349098443985, + "learning_rate": 0.0003336797629717523, + "loss": 2.7667055130004883, + "step": 79490 + }, + { + "epoch": 0.3975, + "grad_norm": 0.33774885535240173, + "learning_rate": 0.0003336423833931936, + "loss": 2.6895734786987306, + "step": 79500 + }, + { + "epoch": 0.39755, + "grad_norm": 0.33865681290626526, + "learning_rate": 0.0003336050017089384, + "loss": 2.7073850631713867, + "step": 79510 + }, + { + "epoch": 0.3976, + "grad_norm": 0.325399786233902, + "learning_rate": 0.00033356761791992784, + "loss": 2.7255828857421873, + "step": 79520 + }, + { + "epoch": 0.39765, + "grad_norm": 0.3173513412475586, + "learning_rate": 0.0003335302320271031, + "loss": 2.7023229598999023, + "step": 79530 + }, + { + "epoch": 0.3977, + "grad_norm": 0.360154390335083, + "learning_rate": 0.00033349284403140527, + "loss": 2.80300350189209, + "step": 79540 + }, + { + "epoch": 0.39775, + "grad_norm": 0.33755072951316833, + "learning_rate": 0.00033345545393377574, + "loss": 2.691950225830078, + "step": 79550 + }, + { + "epoch": 0.3978, + "grad_norm": 0.35533618927001953, + "learning_rate": 0.0003334180617351556, + "loss": 2.724526786804199, + "step": 79560 + }, + { + "epoch": 0.39785, + "grad_norm": 0.38837650418281555, + "learning_rate": 0.00033338066743648643, + "loss": 2.738611030578613, + "step": 79570 + }, + { + "epoch": 0.3979, + "grad_norm": 0.3529309332370758, + "learning_rate": 0.00033334327103870943, + "loss": 2.6906206130981447, + "step": 79580 + }, + { + "epoch": 0.39795, + "grad_norm": 0.34783247113227844, + "learning_rate": 0.00033330587254276617, + "loss": 2.696307373046875, + "step": 79590 + }, + { + "epoch": 0.398, + "grad_norm": 0.3418363630771637, + "learning_rate": 0.0003332684719495981, + "loss": 2.8037681579589844, + "step": 79600 + }, + { + "epoch": 0.39805, + "grad_norm": 0.34326857328414917, + "learning_rate": 0.00033323106926014685, + "loss": 2.6790802001953127, + "step": 79610 + }, + { + "epoch": 0.3981, + "grad_norm": 0.3473598062992096, + "learning_rate": 0.00033319366447535394, + "loss": 2.805570602416992, + "step": 79620 + }, + { + "epoch": 0.39815, + "grad_norm": 0.3263952136039734, + "learning_rate": 0.0003331562575961612, + "loss": 2.748282241821289, + "step": 79630 + }, + { + "epoch": 0.3982, + "grad_norm": 0.3524402976036072, + "learning_rate": 0.00033311884862351004, + "loss": 2.79309139251709, + "step": 79640 + }, + { + "epoch": 0.39825, + "grad_norm": 0.3586866855621338, + "learning_rate": 0.0003330814375583426, + "loss": 2.7610570907592775, + "step": 79650 + }, + { + "epoch": 0.3983, + "grad_norm": 0.35161128640174866, + "learning_rate": 0.00033304402440160043, + "loss": 2.7249261856079103, + "step": 79660 + }, + { + "epoch": 0.39835, + "grad_norm": 0.36174502968788147, + "learning_rate": 0.0003330066091542255, + "loss": 2.6964391708374023, + "step": 79670 + }, + { + "epoch": 0.3984, + "grad_norm": 0.31281620264053345, + "learning_rate": 0.0003329691918171597, + "loss": 2.741942024230957, + "step": 79680 + }, + { + "epoch": 0.39845, + "grad_norm": 0.33472612500190735, + "learning_rate": 0.00033293177239134507, + "loss": 2.73425350189209, + "step": 79690 + }, + { + "epoch": 0.3985, + "grad_norm": 0.312970906496048, + "learning_rate": 0.00033289435087772366, + "loss": 2.7738637924194336, + "step": 79700 + }, + { + "epoch": 0.39855, + "grad_norm": 0.3812415897846222, + "learning_rate": 0.0003328569272772376, + "loss": 2.780020332336426, + "step": 79710 + }, + { + "epoch": 0.3986, + "grad_norm": 0.33721181750297546, + "learning_rate": 0.0003328195015908288, + "loss": 2.762278366088867, + "step": 79720 + }, + { + "epoch": 0.39865, + "grad_norm": 0.37641507387161255, + "learning_rate": 0.0003327820738194397, + "loss": 2.7614463806152343, + "step": 79730 + }, + { + "epoch": 0.3987, + "grad_norm": 0.355243057012558, + "learning_rate": 0.00033274464396401243, + "loss": 2.759571838378906, + "step": 79740 + }, + { + "epoch": 0.39875, + "grad_norm": 0.40870916843414307, + "learning_rate": 0.00033270721202548925, + "loss": 2.67846622467041, + "step": 79750 + }, + { + "epoch": 0.3988, + "grad_norm": 0.3453720510005951, + "learning_rate": 0.0003326697780048127, + "loss": 2.7703550338745115, + "step": 79760 + }, + { + "epoch": 0.39885, + "grad_norm": 0.353984534740448, + "learning_rate": 0.0003326323419029249, + "loss": 2.747265625, + "step": 79770 + }, + { + "epoch": 0.3989, + "grad_norm": 0.3533434271812439, + "learning_rate": 0.0003325949037207685, + "loss": 2.7655271530151366, + "step": 79780 + }, + { + "epoch": 0.39895, + "grad_norm": 0.321186900138855, + "learning_rate": 0.0003325574634592859, + "loss": 2.778809928894043, + "step": 79790 + }, + { + "epoch": 0.399, + "grad_norm": 0.3413957357406616, + "learning_rate": 0.0003325200211194197, + "loss": 2.783908462524414, + "step": 79800 + }, + { + "epoch": 0.39905, + "grad_norm": 0.3302249610424042, + "learning_rate": 0.0003324825767021126, + "loss": 2.8062822341918947, + "step": 79810 + }, + { + "epoch": 0.3991, + "grad_norm": 0.41056329011917114, + "learning_rate": 0.0003324451302083072, + "loss": 2.724323272705078, + "step": 79820 + }, + { + "epoch": 0.39915, + "grad_norm": 0.34984004497528076, + "learning_rate": 0.0003324076816389461, + "loss": 2.7021787643432615, + "step": 79830 + }, + { + "epoch": 0.3992, + "grad_norm": 0.323360413312912, + "learning_rate": 0.0003323702309949722, + "loss": 2.7864025115966795, + "step": 79840 + }, + { + "epoch": 0.39925, + "grad_norm": 0.341980904340744, + "learning_rate": 0.00033233277827732834, + "loss": 2.7185779571533204, + "step": 79850 + }, + { + "epoch": 0.3993, + "grad_norm": 0.3540818691253662, + "learning_rate": 0.00033229532348695725, + "loss": 2.740917205810547, + "step": 79860 + }, + { + "epoch": 0.39935, + "grad_norm": 0.38069668412208557, + "learning_rate": 0.00033225786662480195, + "loss": 2.778292655944824, + "step": 79870 + }, + { + "epoch": 0.3994, + "grad_norm": 0.3594958782196045, + "learning_rate": 0.00033222040769180543, + "loss": 2.689220428466797, + "step": 79880 + }, + { + "epoch": 0.39945, + "grad_norm": 0.35864919424057007, + "learning_rate": 0.0003321829466889107, + "loss": 2.7481849670410154, + "step": 79890 + }, + { + "epoch": 0.3995, + "grad_norm": 0.3349226713180542, + "learning_rate": 0.0003321454836170609, + "loss": 2.749246597290039, + "step": 79900 + }, + { + "epoch": 0.39955, + "grad_norm": 0.3418237268924713, + "learning_rate": 0.00033210801847719895, + "loss": 2.705695152282715, + "step": 79910 + }, + { + "epoch": 0.3996, + "grad_norm": 0.3573083281517029, + "learning_rate": 0.0003320705512702683, + "loss": 2.7194705963134767, + "step": 79920 + }, + { + "epoch": 0.39965, + "grad_norm": 0.32021090388298035, + "learning_rate": 0.00033203308199721205, + "loss": 2.7355894088745116, + "step": 79930 + }, + { + "epoch": 0.3997, + "grad_norm": 0.3958079516887665, + "learning_rate": 0.00033199561065897355, + "loss": 2.753602409362793, + "step": 79940 + }, + { + "epoch": 0.39975, + "grad_norm": 0.31060492992401123, + "learning_rate": 0.000331958137256496, + "loss": 2.719281768798828, + "step": 79950 + }, + { + "epoch": 0.3998, + "grad_norm": 0.3302314579486847, + "learning_rate": 0.00033192066179072295, + "loss": 2.7713344573974608, + "step": 79960 + }, + { + "epoch": 0.39985, + "grad_norm": 0.3275400698184967, + "learning_rate": 0.0003318831842625978, + "loss": 2.747918701171875, + "step": 79970 + }, + { + "epoch": 0.3999, + "grad_norm": 0.31618717312812805, + "learning_rate": 0.00033184570467306403, + "loss": 2.7607038497924803, + "step": 79980 + }, + { + "epoch": 0.39995, + "grad_norm": 0.3143027424812317, + "learning_rate": 0.00033180822302306514, + "loss": 2.7596664428710938, + "step": 79990 + }, + { + "epoch": 0.4, + "grad_norm": 0.3453507125377655, + "learning_rate": 0.00033177073931354485, + "loss": 2.753369140625, + "step": 80000 + } + ], + "logging_steps": 10, + "max_steps": 200000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 10000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.8156221022208e+17, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}