| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9999558635300348, | |
| "eval_steps": 2000, | |
| "global_step": 11328, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.827293993026438e-05, | |
| "eval_accuracy": 0.31068875219818615, | |
| "eval_loss": 5.8817362785339355, | |
| "eval_runtime": 6.9767, | |
| "eval_samples_per_second": 45.58, | |
| "eval_steps_per_second": 0.43, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008827293993026437, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 5e-05, | |
| "loss": 6.0559, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0017654587986052875, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.0001, | |
| "loss": 5.1955, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0026481881979079315, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.00015, | |
| "loss": 4.4828, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003530917597210575, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.0002, | |
| "loss": 3.9175, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0044136469965132185, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00025, | |
| "loss": 3.5422, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005296376395815863, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0003, | |
| "loss": 3.2998, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0061791057951185065, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00035, | |
| "loss": 3.1526, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.00706183519442115, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0004, | |
| "loss": 3.0186, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.007944564593723794, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00045000000000000004, | |
| "loss": 2.9449, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.008827293993026437, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8837, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.009710023392329082, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0004999990214012265, | |
| "loss": 2.8424, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.010592752791631726, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.000499996085612567, | |
| "loss": 2.8241, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01147548219093437, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0004999911926570055, | |
| "loss": 2.8067, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.012358211590237013, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0004999843425728476, | |
| "loss": 2.7826, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.013240940989539656, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0004999755354137212, | |
| "loss": 2.7627, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0141236703888423, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.000499964771248576, | |
| "loss": 2.7586, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.015006399788144944, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.000499952050161682, | |
| "loss": 2.7426, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.015889129187447587, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0004999373722526303, | |
| "loss": 2.7326, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.016771858586750232, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.0004999207376363309, | |
| "loss": 2.7255, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.017654587986052874, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0004999021464430128, | |
| "loss": 2.7191, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01853731738535552, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0004998815988182225, | |
| "loss": 2.7228, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.019420046784658165, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0004998590949228232, | |
| "loss": 2.6945, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.020302776183960806, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.000499834634932993, | |
| "loss": 2.6945, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.02118550558326345, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0004998082190402241, | |
| "loss": 2.6918, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.022068234982566094, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0004997798474513211, | |
| "loss": 2.6832, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02295096438186874, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.000499749520388399, | |
| "loss": 2.6809, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02383369378117138, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0004997172380888822, | |
| "loss": 2.6734, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.024716423180474026, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0004996830008055017, | |
| "loss": 2.6599, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02559915257977667, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0004996468088062946, | |
| "loss": 2.6679, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.026481881979079313, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0004996086623746, | |
| "loss": 2.6578, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.027364611378381958, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.0004995685618090584, | |
| "loss": 2.651, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.0282473407776846, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0004995265074236088, | |
| "loss": 2.661, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.029130070176987245, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0004994824995474863, | |
| "loss": 2.6544, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.030012799576289887, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.0004994365385252189, | |
| "loss": 2.6704, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.030895528975592532, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.0004993886247166261, | |
| "loss": 2.6713, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.031778258374895174, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.000499338758496815, | |
| "loss": 2.6464, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.03266098777419782, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.000499286940256178, | |
| "loss": 2.6611, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.033543717173500465, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.0004992331704003889, | |
| "loss": 2.6427, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.03442644657280311, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0004991774493504007, | |
| "loss": 2.6505, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.03530917597210575, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0004991197775424418, | |
| "loss": 2.6229, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.036191905371408394, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0004990601554280128, | |
| "loss": 2.643, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.03707463477071104, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0004989985834738824, | |
| "loss": 2.6313, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.037957364170013684, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.0004989350621620851, | |
| "loss": 2.6317, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.03884009356931633, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.0004988695919899154, | |
| "loss": 2.6456, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.03972282296861897, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0004988021734699258, | |
| "loss": 2.6272, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.04060555236792161, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0004987328071299217, | |
| "loss": 2.6182, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.04148828176722426, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0004986614935129576, | |
| "loss": 2.6343, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.0423710111665269, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0004985882331773328, | |
| "loss": 2.614, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.04325374056582954, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0004985130266965871, | |
| "loss": 2.6091, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.04413646996513219, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0004984358746594964, | |
| "loss": 2.6036, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04501919936443483, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.0004983567776700676, | |
| "loss": 2.6175, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.04590192876373748, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.0004982757363475346, | |
| "loss": 2.6096, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.04678465816304012, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0004981927513263529, | |
| "loss": 2.6205, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.04766738756234276, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.0004981078232561947, | |
| "loss": 2.6116, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.048550116961645406, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0004980209528019441, | |
| "loss": 2.6113, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.04943284636094805, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.0004979321406436917, | |
| "loss": 2.6001, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.0503155757602507, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0004978413874767291, | |
| "loss": 2.6143, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.05119830515955334, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.0004977486940115441, | |
| "loss": 2.6178, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.05208103455885598, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0004976540609738143, | |
| "loss": 2.6071, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.052963763958158626, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0004975574891044017, | |
| "loss": 2.6138, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.05384649335746127, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0004974589791593472, | |
| "loss": 2.5974, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.054729222756763916, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0004973585319098648, | |
| "loss": 2.6022, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.055611952156066555, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0004972561481423346, | |
| "loss": 2.6128, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.0564946815553692, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.0004971518286582979, | |
| "loss": 2.6068, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.057377410954671845, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.0004970455742744499, | |
| "loss": 2.5949, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.05826014035397449, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0004969373858226341, | |
| "loss": 2.5997, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.059142869753277136, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.0004968272641498349, | |
| "loss": 2.5979, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.060025599152579774, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0004967152101181717, | |
| "loss": 2.5975, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.06090832855188242, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0004966012246048924, | |
| "loss": 2.5956, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.061791057951185065, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0004964853085023653, | |
| "loss": 2.5859, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0626737873504877, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0004963674627180735, | |
| "loss": 2.6004, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.06355651674979035, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0004962476881746068, | |
| "loss": 2.5842, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.064439246149093, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.000496125985809655, | |
| "loss": 2.576, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.06532197554839564, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.0004960023565760003, | |
| "loss": 2.5892, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.06620470494769828, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0004958768014415103, | |
| "loss": 2.585, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.06708743434700093, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0004957493213891295, | |
| "loss": 2.6027, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.06797016374630357, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.0004956199174168725, | |
| "loss": 2.5856, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.06885289314560622, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.000495488590537816, | |
| "loss": 2.5701, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.06973562254490885, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0004953553417800905, | |
| "loss": 2.5805, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.0706183519442115, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.0004952201721868726, | |
| "loss": 2.5825, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.07150108134351414, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0004950830828163767, | |
| "loss": 2.588, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.07238381074281679, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0004949440747418467, | |
| "loss": 2.589, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.07326654014211943, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0004948031490515476, | |
| "loss": 2.5865, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.07414926954142208, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.0004946603068487572, | |
| "loss": 2.5766, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.07503199894072472, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0004945155492517569, | |
| "loss": 2.578, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.07591472834002737, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0004943688773938237, | |
| "loss": 2.5858, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.07679745773933001, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.000494220292423221, | |
| "loss": 2.5783, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.07768018713863266, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.000494069795503189, | |
| "loss": 2.581, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.07856291653793529, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0004939173878119366, | |
| "loss": 2.5741, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.07944564593723794, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.0004937630705426318, | |
| "loss": 2.567, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.08032837533654058, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.000493606844903392, | |
| "loss": 2.5788, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.08121110473584323, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.000493448712117275, | |
| "loss": 2.5789, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.08209383413514587, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.0004932886734222693, | |
| "loss": 2.558, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.08297656353444852, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.000493126730071284, | |
| "loss": 2.5654, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.08385929293375116, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.0004929628833321397, | |
| "loss": 2.5772, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.0847420223330538, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.0004927971344875585, | |
| "loss": 2.5739, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.08562475173235645, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.0004926294848351528, | |
| "loss": 2.5725, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.08650748113165908, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.0004924599356874169, | |
| "loss": 2.5709, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.08739021053096173, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.0004922884883717154, | |
| "loss": 2.557, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.08827293993026437, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004921151442302732, | |
| "loss": 2.5722, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.08915566932956702, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.0004919399046201656, | |
| "loss": 2.567, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.09003839872886966, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.0004917627709133064, | |
| "loss": 2.5629, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.09092112812817231, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0004915837444964383, | |
| "loss": 2.5806, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.09180385752747496, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.0004914028267711217, | |
| "loss": 2.5642, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.0926865869267776, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0004912200191537233, | |
| "loss": 2.5789, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.09356931632608025, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0004910353230754057, | |
| "loss": 2.5657, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.09445204572538288, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0004908487399821158, | |
| "loss": 2.5721, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.09533477512468552, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.0004906602713345735, | |
| "loss": 2.5663, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.09621750452398817, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.0004904699186082602, | |
| "loss": 2.5603, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.09710023392329081, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.0004902776832934074, | |
| "loss": 2.5647, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.09798296332259346, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.0004900835668949852, | |
| "loss": 2.5559, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.0988656927218961, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00048988757093269, | |
| "loss": 2.5599, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.09974842212119875, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0004896896969409332, | |
| "loss": 2.5621, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.1006311515205014, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0004894899464688287, | |
| "loss": 2.5699, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.10151388091980404, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.000489288321080181, | |
| "loss": 2.566, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.10239661031910668, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.0004890848223534732, | |
| "loss": 2.5834, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.10327933971840932, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0004888794518818538, | |
| "loss": 2.5496, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.10416206911771196, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.0004886722112731253, | |
| "loss": 2.5583, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.1050447985170146, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.000488463102149731, | |
| "loss": 2.564, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.10592752791631725, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.0004882521261487422, | |
| "loss": 2.5742, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.1068102573156199, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0004880392849218459, | |
| "loss": 2.5743, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.10769298671492254, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.00048782458013533125, | |
| "loss": 2.5613, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.10857571611422519, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00048760801347007716, | |
| "loss": 2.5522, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.10945844551352783, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0004873895866215385, | |
| "loss": 2.5649, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.11034117491283048, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00048716930129973323, | |
| "loss": 2.5571, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.11122390431213311, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.0004869471592292289, | |
| "loss": 2.5664, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.11210663371143575, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0004867231621491293, | |
| "loss": 2.5603, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.1129893631107384, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00048649731181306047, | |
| "loss": 2.5477, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.11387209251004105, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00048626960998915733, | |
| "loss": 2.5609, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.11475482190934369, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0004860400584600496, | |
| "loss": 2.5668, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.11563755130864634, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0004858086590228482, | |
| "loss": 2.5504, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.11652028070794898, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.0004855754134891307, | |
| "loss": 2.5613, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.11740301010725163, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0004853403236849274, | |
| "loss": 2.5539, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.11828573950655427, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0004851033914507071, | |
| "loss": 2.5613, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.1191684689058569, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00048486461864136253, | |
| "loss": 2.5599, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.12005119830515955, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.0004846240071261959, | |
| "loss": 2.5392, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.1209339277044622, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00048438155878890434, | |
| "loss": 2.5409, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.12181665710376484, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00048413727552756505, | |
| "loss": 2.5521, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.12269938650306748, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00048389115925462025, | |
| "loss": 2.5431, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.12358211590237013, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00048364321189686276, | |
| "loss": 2.5519, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.12446484530167277, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00048339343539542033, | |
| "loss": 2.5424, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.1253475747009754, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0004831418317057409, | |
| "loss": 2.5411, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.12623030410027805, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.0004828884027975768, | |
| "loss": 2.5336, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.1271130334995807, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00048263315065497, | |
| "loss": 2.5506, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.12799576289888334, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0004823760772762358, | |
| "loss": 2.5433, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.128878492298186, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00048211718467394774, | |
| "loss": 2.5515, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.12976122169748863, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.0004818564748749218, | |
| "loss": 2.5384, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.13064395109679128, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0004815939499202001, | |
| "loss": 2.553, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.13152668049609392, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0004813296118650357, | |
| "loss": 2.5587, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.13240940989539657, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004810634627788756, | |
| "loss": 2.543, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1332921392946992, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0004807955047453452, | |
| "loss": 2.5504, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.13417486869400186, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0004805257398622317, | |
| "loss": 2.5478, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.1350575980933045, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0004802541702414678, | |
| "loss": 2.5459, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.13594032749260715, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.000479980798009115, | |
| "loss": 2.545, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.1368230568919098, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00047970562530534724, | |
| "loss": 2.5507, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.13770578629121244, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0004794286542844338, | |
| "loss": 2.5439, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.13858851569051509, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00047914988711472283, | |
| "loss": 2.5482, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.1394712450898177, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00047886932597862396, | |
| "loss": 2.5358, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.14035397448912035, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.0004785869730725914, | |
| "loss": 2.5476, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.141236703888423, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.0004783028306071069, | |
| "loss": 2.5427, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.14211943328772564, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00047801690080666206, | |
| "loss": 2.5343, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.14300216268702828, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00047772918590974136, | |
| "loss": 2.5404, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.14388489208633093, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00047743968816880446, | |
| "loss": 2.5495, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.14476762148563357, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0004771484098502683, | |
| "loss": 2.5438, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.14565035088493622, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0004768553532344899, | |
| "loss": 2.5342, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.14653308028423886, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0004765605206157478, | |
| "loss": 2.541, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.1474158096835415, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0004762639143022248, | |
| "loss": 2.5504, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.14829853908284416, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00047596553661598956, | |
| "loss": 2.5269, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.1491812684821468, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00047566538989297837, | |
| "loss": 2.5442, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.15006399788144945, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00047536347648297685, | |
| "loss": 2.5452, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1509467272807521, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.0004750597987496018, | |
| "loss": 2.5453, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.15182945668005474, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00047475435907028254, | |
| "loss": 2.5416, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.15271218607935738, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0004744471598362421, | |
| "loss": 2.5395, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.15359491547866003, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004741382034524789, | |
| "loss": 2.5423, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.15447764487796267, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.0004738274923377478, | |
| "loss": 2.5322, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.15536037427726532, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.0004735150289245407, | |
| "loss": 2.5331, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.15624310367656793, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00047320081565906813, | |
| "loss": 2.5491, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.15712583307587058, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.0004728848550012399, | |
| "loss": 2.5477, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.15800856247517323, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00047256714942464574, | |
| "loss": 2.5364, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.15889129187447587, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.0004722477014165358, | |
| "loss": 2.5308, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.15977402127377852, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.0004719265134778017, | |
| "loss": 2.5375, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.16065675067308116, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00047160358812295633, | |
| "loss": 2.5245, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.1615394800723838, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0004712789278801145, | |
| "loss": 2.5464, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.16242220947168645, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00047095253529097313, | |
| "loss": 2.5394, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.1633049388709891, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0004706244129107914, | |
| "loss": 2.5325, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.16418766827029174, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00047029456330837055, | |
| "loss": 2.5355, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.1650703976695944, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.0004699629890660339, | |
| "loss": 2.5369, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.16595312706889703, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00046962969277960663, | |
| "loss": 2.546, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.16683585646819968, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00046929467705839544, | |
| "loss": 2.5437, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.16771858586750232, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.0004689579445251681, | |
| "loss": 2.5414, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.16860131526680497, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.000468619497816133, | |
| "loss": 2.548, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.1694840446661076, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0004682793395809184, | |
| "loss": 2.5408, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.17036677406541026, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.0004679374724825517, | |
| "loss": 2.5349, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.1712495034647129, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00046759389919743876, | |
| "loss": 2.5381, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.17213223286401555, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.0004672486224153427, | |
| "loss": 2.5384, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.17301496226331817, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0004669016448393631, | |
| "loss": 2.5296, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.1738976916626208, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0004665529691859144, | |
| "loss": 2.5334, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.17478042106192346, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.00046620259818470536, | |
| "loss": 2.5341, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.1756631504612261, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0004658505345787169, | |
| "loss": 2.5369, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.17654587986052875, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.00046549678112418116, | |
| "loss": 2.556, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.17654587986052875, | |
| "eval_accuracy": 0.5045637800354782, | |
| "eval_loss": 2.4191172122955322, | |
| "eval_runtime": 6.9583, | |
| "eval_samples_per_second": 45.701, | |
| "eval_steps_per_second": 0.431, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1774286092598314, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.0004651413405905597, | |
| "loss": 2.5262, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.17831133865913404, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00046478421576052196, | |
| "loss": 2.5394, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.17919406805843668, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00046442540942992315, | |
| "loss": 2.5327, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.18007679745773933, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00046406492440778294, | |
| "loss": 2.5218, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.18095952685704197, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004637027635162627, | |
| "loss": 2.5349, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.18184225625634462, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00046333892959064425, | |
| "loss": 2.5356, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.18272498565564727, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.0004629734254793071, | |
| "loss": 2.5306, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.1836077150549499, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00046260625404370606, | |
| "loss": 2.5442, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.18449044445425256, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.0004622374181583494, | |
| "loss": 2.52, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.1853731738535552, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00046186692071077586, | |
| "loss": 2.5189, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.18625590325285785, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00046149476460153216, | |
| "loss": 2.5297, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.1871386326521605, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0004611209527441504, | |
| "loss": 2.5336, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.18802136205146314, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.0004607454880651253, | |
| "loss": 2.5322, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.18890409145076575, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0004603683735038909, | |
| "loss": 2.5361, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.1897868208500684, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00045998961201279814, | |
| "loss": 2.5197, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.19066955024937104, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00045960920655709113, | |
| "loss": 2.5213, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.1915522796486737, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004592271601148844, | |
| "loss": 2.5096, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.19243500904797634, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00045884347567713945, | |
| "loss": 2.5222, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.19331773844727898, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.0004584581562476412, | |
| "loss": 2.5463, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.19420046784658163, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0004580712048429746, | |
| "loss": 2.5331, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.19508319724588427, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.000457682624492501, | |
| "loss": 2.5172, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.19596592664518692, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004572924182383346, | |
| "loss": 2.5282, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.19684865604448956, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00045690058913531794, | |
| "loss": 2.5318, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.1977313854437922, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.0004565071402509992, | |
| "loss": 2.5202, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.19861411484309485, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.000456112074665607, | |
| "loss": 2.5342, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.1994968442423975, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0004557153954720269, | |
| "loss": 2.5226, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.20037957364170014, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0004553171057757772, | |
| "loss": 2.535, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.2012623030410028, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004549172086949842, | |
| "loss": 2.5182, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.20214503244030543, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0004545157073603584, | |
| "loss": 2.5335, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.20302776183960808, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.0004541126049151694, | |
| "loss": 2.545, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.20391049123891072, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00045370790451522165, | |
| "loss": 2.5158, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.20479322063821337, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004533016093288298, | |
| "loss": 2.5105, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.205675950037516, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.0004528937225367935, | |
| "loss": 2.5305, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.20655867943681863, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.0004524842473323729, | |
| "loss": 2.5335, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.20744140883612128, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0004520731869212634, | |
| "loss": 2.5196, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.20832413823542392, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.0004516605445215709, | |
| "loss": 2.5201, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.20920686763472657, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00045124632336378603, | |
| "loss": 2.5195, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.2100895970340292, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00045083052669075936, | |
| "loss": 2.5264, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.21097232643333186, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0004504131577576758, | |
| "loss": 2.5283, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.2118550558326345, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00044999421983202905, | |
| "loss": 2.5266, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.21273778523193715, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.00044957371619359644, | |
| "loss": 2.5356, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.2136205146312398, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00044915165013441257, | |
| "loss": 2.5281, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.21450324403054244, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0004487280249587441, | |
| "loss": 2.5335, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.21538597342984508, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00044830284398306375, | |
| "loss": 2.5303, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.21626870282914773, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.000447876110536024, | |
| "loss": 2.5297, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.21715143222845038, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.0004474478279584316, | |
| "loss": 2.5287, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.21803416162775302, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00044701799960322085, | |
| "loss": 2.525, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.21891689102705567, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.000446586628835428, | |
| "loss": 2.5278, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.2197996204263583, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00044615371903216407, | |
| "loss": 2.5087, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.22068234982566096, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00044571927358258917, | |
| "loss": 2.5294, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.22156507922496357, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0004452832958878856, | |
| "loss": 2.5296, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.22244780862426622, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0004448457893612311, | |
| "loss": 2.5267, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.22333053802356886, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.0004444067574277727, | |
| "loss": 2.5221, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.2242132674228715, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00044396620352459915, | |
| "loss": 2.5178, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.22509599682217415, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00044352413110071453, | |
| "loss": 2.5104, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.2259787262214768, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0004430805436170111, | |
| "loss": 2.5262, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.22686145562077945, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00044263544454624224, | |
| "loss": 2.5204, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.2277441850200821, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.00044218883737299526, | |
| "loss": 2.5159, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.22862691441938474, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00044174072559366386, | |
| "loss": 2.5133, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.22950964381868738, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00044129111271642117, | |
| "loss": 2.5293, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.23039237321799003, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0004408400022611921, | |
| "loss": 2.511, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.23127510261729267, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00044038739775962584, | |
| "loss": 2.5088, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.23215783201659532, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0004399333027550679, | |
| "loss": 2.5066, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.23304056141589796, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.000439477720802533, | |
| "loss": 2.5231, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.2339232908152006, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00043902065546867655, | |
| "loss": 2.5164, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.23480602021450325, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.0004385621103317671, | |
| "loss": 2.511, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.2356887496138059, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00043810208898165836, | |
| "loss": 2.5049, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.23657147901310854, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.000437640595019761, | |
| "loss": 2.5262, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.2374542084124112, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00043717763205901436, | |
| "loss": 2.5205, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.2383369378117138, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00043671320372385834, | |
| "loss": 2.5136, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.23921966721101645, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00043624731365020505, | |
| "loss": 2.5185, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.2401023966103191, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00043577996548541, | |
| "loss": 2.5148, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.24098512600962174, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00043531116288824393, | |
| "loss": 2.5225, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.2418678554089244, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00043484090952886404, | |
| "loss": 2.524, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.24275058480822703, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0004343692090887852, | |
| "loss": 2.5035, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.24363331420752968, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0004338960652608511, | |
| "loss": 2.5136, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.24451604360683232, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0004334214817492057, | |
| "loss": 2.4963, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.24539877300613497, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0004329454622692636, | |
| "loss": 2.5087, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.2462815024054376, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00043246801054768147, | |
| "loss": 2.5188, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.24716423180474026, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0004319891303223287, | |
| "loss": 2.5059, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.2480469612040429, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.000431508825342258, | |
| "loss": 2.5227, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.24892969060334555, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0004310270993676764, | |
| "loss": 2.5215, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.2498124200026482, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00043054395616991535, | |
| "loss": 2.513, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.2506951494019508, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004300593995314017, | |
| "loss": 2.5106, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.2515778788012535, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0004295734332456277, | |
| "loss": 2.4932, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2524606082005561, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00042908606111712136, | |
| "loss": 2.5118, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.2533433375998588, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0004285972869614169, | |
| "loss": 2.5157, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.2542260669991614, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00042810711460502447, | |
| "loss": 2.5073, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.25510879639846407, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00042761554788540084, | |
| "loss": 2.5368, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.2559915257977667, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0004271225906509186, | |
| "loss": 2.5143, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.25687425519706936, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0004266282467608365, | |
| "loss": 2.5078, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.257756984596372, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00042613252008526914, | |
| "loss": 2.5283, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.25863971399567465, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0004256354145051567, | |
| "loss": 2.5008, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.25952244339497726, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.0004251369339122344, | |
| "loss": 2.5261, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.26040517279427994, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00042463708220900225, | |
| "loss": 2.5284, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.26128790219358256, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00042413586330869446, | |
| "loss": 2.5355, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.26217063159288523, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00042363328113524846, | |
| "loss": 2.5201, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.26305336099218785, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004231293396232747, | |
| "loss": 2.5249, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.26393609039149046, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00042262404271802565, | |
| "loss": 2.5161, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.26481881979079314, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00042211739437536457, | |
| "loss": 2.522, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.26570154919009575, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0004216093985617352, | |
| "loss": 2.5132, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.2665842785893984, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.0004211000592541301, | |
| "loss": 2.5156, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.26746700798870104, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.0004205893804400599, | |
| "loss": 2.5121, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.2683497373880037, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004200773661175219, | |
| "loss": 2.5054, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.26923246678730633, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0004195640202949687, | |
| "loss": 2.5012, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.270115196186609, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00041904934699127713, | |
| "loss": 2.5139, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.2709979255859116, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.0004185333502357164, | |
| "loss": 2.501, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.2718806549852143, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.000418016034067917, | |
| "loss": 2.5065, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.2727633843845169, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00041749740253783853, | |
| "loss": 2.5107, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.2736461137838196, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00041697745970573855, | |
| "loss": 2.5222, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.2745288431831222, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00041645620964214023, | |
| "loss": 2.5137, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.2754115725824249, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0004159336564278012, | |
| "loss": 2.5347, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.2762943019817275, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0004154098041536807, | |
| "loss": 2.5243, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.27717703138103017, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00041488465692090837, | |
| "loss": 2.5265, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.2780597607803328, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00041435821884075176, | |
| "loss": 2.5145, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.2789424901796354, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00041383049403458403, | |
| "loss": 2.5187, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.2798252195789381, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0004133014866338521, | |
| "loss": 2.5217, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.2807079489782407, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00041277120078004383, | |
| "loss": 2.4996, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.28159067837754337, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004122396406246559, | |
| "loss": 2.5199, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.282473407776846, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.0004117068103291614, | |
| "loss": 2.516, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.28335613717614866, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00041117271406497665, | |
| "loss": 2.5019, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.2842388665754513, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00041063735601342934, | |
| "loss": 2.5114, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.28512159597475395, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0004101007403657255, | |
| "loss": 2.5148, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.28600432537405657, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00040956287132291625, | |
| "loss": 2.5008, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.28688705477335924, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00040902375309586557, | |
| "loss": 2.5153, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.28776978417266186, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00040848338990521696, | |
| "loss": 2.5147, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.28865251357196453, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00040794178598136033, | |
| "loss": 2.5062, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.28953524297126715, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.0004073989455643994, | |
| "loss": 2.526, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.2904179723705698, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00040685487290411765, | |
| "loss": 2.5167, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.29130070176987244, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0004063095722599459, | |
| "loss": 2.5083, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.2921834311691751, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00040576304790092857, | |
| "loss": 2.5069, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.29306616056847773, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00040521530410569007, | |
| "loss": 2.5289, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.2939488899677804, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0004046663451624016, | |
| "loss": 2.5135, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.294831619367083, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.0004041161753687478, | |
| "loss": 2.4992, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.29571434876638564, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00040356479903189233, | |
| "loss": 2.5231, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.2965970781656883, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0004030122204684449, | |
| "loss": 2.509, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.29747980756499093, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.0004024584440044271, | |
| "loss": 2.5044, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.2983625369642936, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00040190347397523873, | |
| "loss": 2.5097, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.2992452663635962, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0004013473147256238, | |
| "loss": 2.5068, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.3001279957628989, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0004007899706096363, | |
| "loss": 2.5048, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3010107251622015, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00040023144599060623, | |
| "loss": 2.4943, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.3018934545615042, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00039967174524110596, | |
| "loss": 2.5269, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.3027761839608068, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.000399110872742915, | |
| "loss": 2.5052, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.3036589133601095, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.0003985488328869865, | |
| "loss": 2.5229, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.3045416427594121, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0003979856300734126, | |
| "loss": 2.5035, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.30542437215871476, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00039742126871138996, | |
| "loss": 2.5109, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.3063071015580174, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0003968557532191852, | |
| "loss": 2.5181, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.30718983095732005, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00039628908802410057, | |
| "loss": 2.5159, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.30807256035662267, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00039572127756243904, | |
| "loss": 2.508, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.30895528975592534, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0003951523262794693, | |
| "loss": 2.5088, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.30983801915522796, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.00039458223862939184, | |
| "loss": 2.5191, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.31072074855453063, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00039401101907530323, | |
| "loss": 2.5113, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.31160347795383325, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0003934386720891614, | |
| "loss": 2.5097, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.31248620735313587, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.00039286520215175085, | |
| "loss": 2.4938, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.31336893675243854, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0003922906137526474, | |
| "loss": 2.5175, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.31425166615174116, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00039171491139018325, | |
| "loss": 2.5129, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.31513439555104383, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0003911380995714111, | |
| "loss": 2.5287, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.31601712495034645, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0003905601828120698, | |
| "loss": 2.5026, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.3168998543496491, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.0003899811656365485, | |
| "loss": 2.5138, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.31778258374895174, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0003894010525778511, | |
| "loss": 2.5213, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.3186653131482544, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.000388819848177561, | |
| "loss": 2.515, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.31954804254755703, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00038823755698580545, | |
| "loss": 2.5041, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.3204307719468597, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.0003876541835612202, | |
| "loss": 2.5209, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.3213135013461623, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0003870697324709132, | |
| "loss": 2.5182, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.322196230745465, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00038648420829042954, | |
| "loss": 2.5174, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.3230789601447676, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00038589761560371515, | |
| "loss": 2.5152, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.3239616895440703, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.00038530995900308107, | |
| "loss": 2.4956, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.3248444189433729, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00038472124308916753, | |
| "loss": 2.5054, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.3257271483426756, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00038413147247090795, | |
| "loss": 2.5215, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.3266098777419782, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00038354065176549274, | |
| "loss": 2.5235, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.32749260714128087, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00038294878559833317, | |
| "loss": 2.5044, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.3283753365405835, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.0003823558786030255, | |
| "loss": 2.5099, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.3292580659398861, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00038176193542131386, | |
| "loss": 2.5144, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.3301407953391888, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00038116696070305503, | |
| "loss": 2.5195, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.3310235247384914, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00038057095910618125, | |
| "loss": 2.5066, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.33190625413779407, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00037997393529666393, | |
| "loss": 2.5161, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.3327889835370967, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00037937589394847714, | |
| "loss": 2.4976, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.33367171293639936, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00037877683974356114, | |
| "loss": 2.5082, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.334554442335702, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003781767773717857, | |
| "loss": 2.5061, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.33543717173500465, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00037757571153091324, | |
| "loss": 2.5106, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.33631990113430726, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.000376973646926562, | |
| "loss": 2.509, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.33720263053360994, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00037637058827216964, | |
| "loss": 2.4963, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.33808535993291255, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00037576654028895554, | |
| "loss": 2.5123, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.3389680893322152, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00037516150770588487, | |
| "loss": 2.4981, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.33985081873151785, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00037455549525963066, | |
| "loss": 2.5043, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.3407335481308205, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.0003739485076945373, | |
| "loss": 2.5044, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.34161627753012314, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.000373340549762583, | |
| "loss": 2.483, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.3424990069294258, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0003727316262233429, | |
| "loss": 2.5101, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.3433817363287284, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.0003721217418439516, | |
| "loss": 2.4966, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.3442644657280311, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00037151090139906593, | |
| "loss": 2.505, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.3451471951273337, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00037089910967082765, | |
| "loss": 2.5099, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.34602992452663633, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0003702863714488257, | |
| "loss": 2.5158, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.346912653925939, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.0003696726915300592, | |
| "loss": 2.5127, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.3477953833252416, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0003690580747188995, | |
| "loss": 2.5141, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.3486781127245443, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00036844252582705244, | |
| "loss": 2.4919, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.3495608421238469, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0003678260496735214, | |
| "loss": 2.5091, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.3504435715231496, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0003672086510845687, | |
| "loss": 2.504, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.3513263009224522, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.00036659033489367835, | |
| "loss": 2.5042, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.3522090303217549, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0003659711059415182, | |
| "loss": 2.5083, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.3530917597210575, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0003653509690759016, | |
| "loss": 2.5196, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3530917597210575, | |
| "eval_accuracy": 0.5083680819529877, | |
| "eval_loss": 2.3899683952331543, | |
| "eval_runtime": 6.95, | |
| "eval_samples_per_second": 45.756, | |
| "eval_steps_per_second": 0.432, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.35397448912036017, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00036472992915175017, | |
| "loss": 2.4973, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.3548572185196628, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00036410799103105503, | |
| "loss": 2.5225, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.35573994791896546, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.0003634851595828393, | |
| "loss": 2.5052, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.3566226773182681, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00036286143968311963, | |
| "loss": 2.5045, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.35750540671757075, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00036223683621486845, | |
| "loss": 2.5086, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.35838813611687337, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00036161135406797504, | |
| "loss": 2.4969, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.35927086551617604, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.0003609849981392079, | |
| "loss": 2.4986, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.36015359491547866, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0003603577733321764, | |
| "loss": 2.4938, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.3610363243147813, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0003597296845572917, | |
| "loss": 2.4959, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.36191905371408395, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00035910073673172933, | |
| "loss": 2.5011, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.36280178311338657, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00035847093477938953, | |
| "loss": 2.4938, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.36368451251268924, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00035784028363085985, | |
| "loss": 2.4949, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.36456724191199186, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00035720878822337576, | |
| "loss": 2.4885, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.36544997131129453, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00035657645350078233, | |
| "loss": 2.5217, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.36633270071059715, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0003559432844134954, | |
| "loss": 2.5107, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.3672154301098998, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0003553092859184629, | |
| "loss": 2.5016, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.36809815950920244, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.0003546744629791261, | |
| "loss": 2.4954, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.3689808889085051, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.00035403882056538044, | |
| "loss": 2.4903, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.36986361830780773, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00035340236365353724, | |
| "loss": 2.5008, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.3707463477071104, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.000352765097226284, | |
| "loss": 2.5107, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.371629077106413, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0003521270262726458, | |
| "loss": 2.5058, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.3725118065057157, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00035148815578794635, | |
| "loss": 2.4966, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.3733945359050183, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0003508484907737687, | |
| "loss": 2.4942, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.374277265304321, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0003502080362379159, | |
| "loss": 2.5097, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.3751599947036236, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00034956679719437225, | |
| "loss": 2.5197, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.3760427241029263, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00034892477866326356, | |
| "loss": 2.4971, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.3769254535022289, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0003482819856708183, | |
| "loss": 2.4922, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.3778081829015315, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00034763842324932794, | |
| "loss": 2.5172, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.3786909123008342, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00034699409643710764, | |
| "loss": 2.51, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.3795736417001368, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00034634901027845677, | |
| "loss": 2.5022, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.38045637109943947, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.0003457031698236196, | |
| "loss": 2.4982, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.3813391004987421, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00034505658012874544, | |
| "loss": 2.5115, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.38222182989804476, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00034440924625584954, | |
| "loss": 2.4998, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.3831045592973474, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003437611732727728, | |
| "loss": 2.4927, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.38398728869665005, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0003431123662531427, | |
| "loss": 2.505, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.38487001809595267, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0003424628302763332, | |
| "loss": 2.4918, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.38575274749525534, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0003418125704274252, | |
| "loss": 2.4941, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.38663547689455796, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00034116159179716675, | |
| "loss": 2.5237, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.38751820629386063, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.0003405098994819329, | |
| "loss": 2.5107, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.38840093569316325, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00033985749858368605, | |
| "loss": 2.4944, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.3892836650924659, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003392043942099358, | |
| "loss": 2.5099, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.39016639449176854, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0003385505914736994, | |
| "loss": 2.5048, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.3910491238910712, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00033789609549346146, | |
| "loss": 2.4965, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.39193185329037383, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.0003372409113931334, | |
| "loss": 2.4925, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.3928145826896765, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0003365850443020142, | |
| "loss": 2.4915, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.3936973120889791, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00033592849935474965, | |
| "loss": 2.5076, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.39458004148828174, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.0003352712816912925, | |
| "loss": 2.5088, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.3954627708875844, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00033461339645686196, | |
| "loss": 2.5017, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.39634550028688703, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.0003339548488019033, | |
| "loss": 2.496, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.3972282296861897, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00033329564388204816, | |
| "loss": 2.4897, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.3981109590854923, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0003326357868580734, | |
| "loss": 2.4998, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.398993688484795, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0003319752828958613, | |
| "loss": 2.5061, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.3998764178840976, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0003313141371663587, | |
| "loss": 2.514, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.4007591472834003, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.000330652354845537, | |
| "loss": 2.5237, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.4016418766827029, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.0003299899411143509, | |
| "loss": 2.5028, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.4025246060820056, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.0003293269011586986, | |
| "loss": 2.4957, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.4034073354813082, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00032866324016938095, | |
| "loss": 2.4829, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.40429006488061087, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00032799896334206045, | |
| "loss": 2.5011, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.4051727942799135, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.000327334075877221, | |
| "loss": 2.5175, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.40605552367921616, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.000326668582980127, | |
| "loss": 2.5149, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.4069382530785188, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00032600248986078295, | |
| "loss": 2.5004, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.40782098247782145, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00032533580173389195, | |
| "loss": 2.488, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.40870371187712407, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0003246685238188154, | |
| "loss": 2.4982, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.40958644127642674, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00032400066133953225, | |
| "loss": 2.5085, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.41046917067572936, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.0003233322195245977, | |
| "loss": 2.506, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.411351900075032, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00032266320360710237, | |
| "loss": 2.5029, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.41223462947433465, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003219936188246317, | |
| "loss": 2.5053, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.41311735887363726, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.0003213234704192243, | |
| "loss": 2.5006, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.41400008827293994, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00032065276363733137, | |
| "loss": 2.4967, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.41488281767224255, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00031998150372977577, | |
| "loss": 2.5042, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.4157655470715452, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0003193096959517103, | |
| "loss": 2.4926, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.41664827647084784, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003186373455625774, | |
| "loss": 2.4898, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.4175310058701505, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.0003179644578260669, | |
| "loss": 2.4932, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.41841373526945314, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00031729103801007575, | |
| "loss": 2.493, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.4192964646687558, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.0003166170913866665, | |
| "loss": 2.5047, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.4201791940680584, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00031594262323202577, | |
| "loss": 2.4976, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.4210619234673611, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0003152676388264234, | |
| "loss": 2.4955, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.4219446528666637, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00031459214345417046, | |
| "loss": 2.4742, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.4228273822659664, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.00031391614240357864, | |
| "loss": 2.4919, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.423710111665269, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00031323964096691825, | |
| "loss": 2.5034, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4245928410645717, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.0003125626444403772, | |
| "loss": 2.4846, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.4254755704638743, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00031188515812401917, | |
| "loss": 2.5016, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.42635829986317697, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00031120718732174235, | |
| "loss": 2.4959, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.4272410292624796, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.000310528737341238, | |
| "loss": 2.4716, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.4281237586617822, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00030984981349394864, | |
| "loss": 2.4994, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.4290064880610849, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00030917042109502663, | |
| "loss": 2.499, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.4298892174603875, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00030849056546329253, | |
| "loss": 2.4879, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.43077194685969017, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.0003078102519211933, | |
| "loss": 2.4751, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.4316546762589928, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.0003071294857947612, | |
| "loss": 2.5005, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.43253740565829546, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0003064482724135711, | |
| "loss": 2.5128, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.4334201350575981, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00030576661711069985, | |
| "loss": 2.4908, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.43430286445690075, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0003050845252226837, | |
| "loss": 2.5086, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.43518559385620337, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0003044020020894769, | |
| "loss": 2.498, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.43606832325550604, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00030371905305441, | |
| "loss": 2.4986, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.43695105265480866, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0003030356834641476, | |
| "loss": 2.4886, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.43783378205411133, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0003023518986686469, | |
| "loss": 2.4948, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.43871651145341395, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0003016677040211154, | |
| "loss": 2.5024, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.4395992408527166, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00030098310487796965, | |
| "loss": 2.4908, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.44048197025201924, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00030029810659879273, | |
| "loss": 2.4913, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.4413646996513219, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00029961271454629235, | |
| "loss": 2.502, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.44224742905062453, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.0002989269340862591, | |
| "loss": 2.4902, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.44313015844992715, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0002982407705875243, | |
| "loss": 2.5009, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.4440128878492298, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00029755422942191805, | |
| "loss": 2.489, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.44489561724853244, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.0002968673159642271, | |
| "loss": 2.5023, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.4457783466478351, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00029618003559215276, | |
| "loss": 2.5067, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.44666107604713773, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.0002954923936862689, | |
| "loss": 2.4927, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.4475438054464404, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00029480439562997964, | |
| "loss": 2.5039, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.448426534845743, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00029411604680947755, | |
| "loss": 2.49, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.4493092642450457, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00029342735261370095, | |
| "loss": 2.4923, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.4501919936443483, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0002927383184342924, | |
| "loss": 2.487, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.451074723043651, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00029204894966555577, | |
| "loss": 2.5052, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.4519574524429536, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00029135925170441457, | |
| "loss": 2.5055, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.4528401818422563, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0002906692299503694, | |
| "loss": 2.5038, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.4537229112415589, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.00028997888980545586, | |
| "loss": 2.4906, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.45460564064086156, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00028928823667420206, | |
| "loss": 2.4871, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.4554883700401642, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00028859727596358643, | |
| "loss": 2.4991, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.45637109943946685, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00028790601308299545, | |
| "loss": 2.4943, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.45725382883876947, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.0002872144534441812, | |
| "loss": 2.4936, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.45813655823807214, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0002865226024612189, | |
| "loss": 2.5053, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.45901928763737476, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.00028583046555046487, | |
| "loss": 2.4854, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.4599020170366774, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0002851380481305136, | |
| "loss": 2.4905, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.46078474643598005, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00028444535562215594, | |
| "loss": 2.4897, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.46166747583528267, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00028375239344833616, | |
| "loss": 2.49, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.46255020523458534, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00028305916703410974, | |
| "loss": 2.5025, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.46343293463388796, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00028236568180660073, | |
| "loss": 2.4854, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.46431566403319063, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0002816719431949596, | |
| "loss": 2.4993, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.46519839343249325, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.0002809779566303203, | |
| "loss": 2.507, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.4660811228317959, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00028028372754575805, | |
| "loss": 2.5041, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.46696385223109854, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0002795892613762467, | |
| "loss": 2.4892, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.4678465816304012, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.00027889456355861635, | |
| "loss": 2.5044, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.46872931102970383, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00027819963953151024, | |
| "loss": 2.4867, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.4696120404290065, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.0002775044947353428, | |
| "loss": 2.5044, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.4704947698283091, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0002768091346122569, | |
| "loss": 2.4989, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.4713774992276118, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.000276113564606081, | |
| "loss": 2.4902, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.4722602286269144, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00027541779016228664, | |
| "loss": 2.4804, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.4731429580262171, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0002747218167279461, | |
| "loss": 2.5, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.4740256874255197, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.00027402564975168925, | |
| "loss": 2.484, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.4749084168248224, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.0002733292946836615, | |
| "loss": 2.4866, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.475791146224125, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0002726327569754803, | |
| "loss": 2.4957, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.4766738756234276, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00027193604208019346, | |
| "loss": 2.5038, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.4775566050227303, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0002712391554522355, | |
| "loss": 2.4937, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.4784393344220329, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0002705421025473857, | |
| "loss": 2.4954, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.4793220638213356, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0002698448888227251, | |
| "loss": 2.4861, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.4802047932206382, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.0002691475197365936, | |
| "loss": 2.4765, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.48108752261994087, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00026845000074854754, | |
| "loss": 2.5035, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.4819702520192435, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0002677523373193165, | |
| "loss": 2.4953, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.48285298141854616, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00026705453491076127, | |
| "loss": 2.4888, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.4837357108178488, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00026635659898583043, | |
| "loss": 2.4894, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.48461844021715145, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.000265658535008518, | |
| "loss": 2.5047, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.48550116961645406, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00026496034844382036, | |
| "loss": 2.4942, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.48638389901575674, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.0002642620447576935, | |
| "loss": 2.4822, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.48726662841505936, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0002635636294170106, | |
| "loss": 2.4992, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.48814935781436203, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00026286510788951886, | |
| "loss": 2.4969, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.48903208721366465, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0002621664856437967, | |
| "loss": 2.4898, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.4899148166129673, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00026146776814921105, | |
| "loss": 2.5002, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.49079754601226994, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0002607689608758746, | |
| "loss": 2.5127, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.4916802754115726, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.000260070069294603, | |
| "loss": 2.4692, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.4925630048108752, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00025937109887687164, | |
| "loss": 2.495, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.49344573421017784, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00025867205509477335, | |
| "loss": 2.4888, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.4943284636094805, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0002579729434209752, | |
| "loss": 2.4946, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.49521119300878313, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00025727376932867593, | |
| "loss": 2.499, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.4960939224080858, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00025657453829156256, | |
| "loss": 2.4931, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.4969766518073884, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00025587525578376843, | |
| "loss": 2.4886, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.4978593812066911, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.0002551759272798295, | |
| "loss": 2.4854, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.4987421106059937, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00025447655825464174, | |
| "loss": 2.5096, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.4996248400052964, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.0002537771541834187, | |
| "loss": 2.4855, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.5005075694045991, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00025307772054164804, | |
| "loss": 2.5022, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.5013902988039016, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.000252378262805049, | |
| "loss": 2.4872, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.5022730282032043, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0002516787864495294, | |
| "loss": 2.4977, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.503155757602507, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00025097929695114295, | |
| "loss": 2.4897, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5040384870018096, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00025027979978604615, | |
| "loss": 2.4892, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.5049212164011122, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0002495803004304556, | |
| "loss": 2.4854, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.5058039458004149, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0002488808043606048, | |
| "loss": 2.4952, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.5066866751997176, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0002481813170527019, | |
| "loss": 2.4973, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.5075694045990202, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.0002474818439828862, | |
| "loss": 2.4893, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.5084521339983228, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.0002467823906271856, | |
| "loss": 2.4936, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.5093348633976255, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00024608296246147375, | |
| "loss": 2.4944, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.5102175927969281, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00024538356496142693, | |
| "loss": 2.4862, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.5111003221962307, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00024468420360248145, | |
| "loss": 2.4958, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.5119830515955334, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00024398488385979055, | |
| "loss": 2.4893, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.512865780994836, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00024328561120818195, | |
| "loss": 2.4969, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.5137485103941387, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00024258639112211453, | |
| "loss": 2.5057, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.5146312397934413, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00024188722907563537, | |
| "loss": 2.4883, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.515513969192744, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00024118813054233774, | |
| "loss": 2.4837, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.5163966985920466, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00024048910099531726, | |
| "loss": 2.4996, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.5172794279913493, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00023979014590712962, | |
| "loss": 2.4796, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.5181621573906519, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00023909127074974744, | |
| "loss": 2.4937, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.5190448867899545, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00023839248099451782, | |
| "loss": 2.4882, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.5199276161892572, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00023769378211211916, | |
| "loss": 2.4752, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.5208103455885599, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00023699517957251825, | |
| "loss": 2.4822, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.5216930749878624, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00023629667884492799, | |
| "loss": 2.4913, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.5225758043871651, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00023559828539776394, | |
| "loss": 2.4878, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.5234585337864678, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00023490000469860185, | |
| "loss": 2.487, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.5243412631857705, | |
| "grad_norm": 0.125, | |
| "learning_rate": 0.0002342018422141347, | |
| "loss": 2.4839, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.525223992585073, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00023350380341013034, | |
| "loss": 2.5015, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.5261067219843757, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.000232805893751388, | |
| "loss": 2.5031, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.5269894513836784, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.0002321081187016959, | |
| "loss": 2.4887, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.5278721807829809, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00023141048372378863, | |
| "loss": 2.4911, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.5287549101822836, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00023071299427930396, | |
| "loss": 2.4885, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.5296376395815863, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00023001565582874046, | |
| "loss": 2.4919, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5296376395815863, | |
| "eval_accuracy": 0.5102518027046329, | |
| "eval_loss": 2.375392198562622, | |
| "eval_runtime": 6.985, | |
| "eval_samples_per_second": 45.526, | |
| "eval_steps_per_second": 0.429, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.530520368980889, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.00022931847383141446, | |
| "loss": 2.4796, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.5314030983801915, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00022862145374541768, | |
| "loss": 2.4886, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.5322858277794942, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00022792460102757407, | |
| "loss": 2.4902, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.5331685571787969, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00022722792113339722, | |
| "loss": 2.4898, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.5340512865780995, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0002265314195170481, | |
| "loss": 2.5015, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.5349340159774021, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00022583510163129162, | |
| "loss": 2.4765, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.5358167453767048, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.00022513897292745434, | |
| "loss": 2.5056, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.5366994747760074, | |
| "grad_norm": 0.125, | |
| "learning_rate": 0.00022444303885538178, | |
| "loss": 2.4953, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.5375822041753101, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.000223747304863396, | |
| "loss": 2.4893, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.5384649335746127, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0002230517763982523, | |
| "loss": 2.5002, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.5393476629739153, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.0002223564589050971, | |
| "loss": 2.4939, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.540230392373218, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00022166135782742525, | |
| "loss": 2.4857, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.5411131217725207, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0002209664786070372, | |
| "loss": 2.4857, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.5419958511718233, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00022027182668399653, | |
| "loss": 2.4873, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.5428785805711259, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.0002195774074965874, | |
| "loss": 2.486, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.5437613099704286, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00021888322648127206, | |
| "loss": 2.4984, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.5446440393697312, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.0002181892890726479, | |
| "loss": 2.4945, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.5455267687690338, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00021749560070340534, | |
| "loss": 2.4869, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.5464094981683365, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.000216802166804285, | |
| "loss": 2.4863, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.5472922275676392, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00021610899280403555, | |
| "loss": 2.495, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.5481749569669417, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00021541608412937075, | |
| "loss": 2.4794, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.5490576863662444, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.000214723446204927, | |
| "loss": 2.4992, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.5499404157655471, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00021403108445322168, | |
| "loss": 2.4956, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.5508231451648498, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.0002133390042946094, | |
| "loss": 2.4832, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.5517058745641523, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00021264721114724064, | |
| "loss": 2.4871, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.552588603963455, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0002119557104270187, | |
| "loss": 2.4968, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.5534713333627577, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00021126450754755774, | |
| "loss": 2.484, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.5543540627620603, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00021057360792014004, | |
| "loss": 2.4831, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.5552367921613629, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0002098830169536738, | |
| "loss": 2.4831, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.5561195215606656, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00020919274005465083, | |
| "loss": 2.4874, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.5570022509599682, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00020850278262710416, | |
| "loss": 2.492, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.5578849803592708, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.0002078131500725657, | |
| "loss": 2.4912, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.5587677097585735, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00020712384779002392, | |
| "loss": 2.4868, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.5596504391578762, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00020643488117588199, | |
| "loss": 2.486, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.5605331685571788, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00020574625562391494, | |
| "loss": 2.4893, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.5614158979564814, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00020505797652522751, | |
| "loss": 2.4902, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.5622986273557841, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00020437004926821255, | |
| "loss": 2.4928, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.5631813567550867, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00020368247923850826, | |
| "loss": 2.4896, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.5640640861543894, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00020299527181895602, | |
| "loss": 2.4764, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.564946815553692, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00020230843238955854, | |
| "loss": 2.4788, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.5658295449529946, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.0002016219663274377, | |
| "loss": 2.4956, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.5667122743522973, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00020093587900679217, | |
| "loss": 2.4832, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.5675950037516, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00020025017579885563, | |
| "loss": 2.499, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.5684777331509026, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00019956486207185477, | |
| "loss": 2.488, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.5693604625502052, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.0001988799431909668, | |
| "loss": 2.4964, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.5702431919495079, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00019819542451827808, | |
| "loss": 2.4901, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.5711259213488106, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00019751131141274147, | |
| "loss": 2.4834, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.5720086507481131, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0001968276092301352, | |
| "loss": 2.4859, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.5728913801474158, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00019614432332302006, | |
| "loss": 2.4835, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.5737741095467185, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00019546145904069808, | |
| "loss": 2.4865, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.574656838946021, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.00019477902172917045, | |
| "loss": 2.4848, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.5755395683453237, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0001940970167310957, | |
| "loss": 2.502, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.5764222977446264, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.0001934154493857479, | |
| "loss": 2.4871, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.5773050271439291, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0001927343250289747, | |
| "loss": 2.5027, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.5781877565432316, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00019205364899315593, | |
| "loss": 2.4757, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.5790704859425343, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.00019137342660716133, | |
| "loss": 2.4887, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.579953215341837, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00019069366319630923, | |
| "loss": 2.488, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.5808359447411396, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00019001436408232496, | |
| "loss": 2.4843, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.5817186741404422, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00018933553458329856, | |
| "loss": 2.4844, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.5826014035397449, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00018865718001364375, | |
| "loss": 2.477, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.5834841329390476, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00018797930568405612, | |
| "loss": 2.4859, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.5843668623383502, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00018730191690147176, | |
| "loss": 2.4824, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.5852495917376528, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00018662501896902519, | |
| "loss": 2.4692, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.5861323211369555, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.0001859486171860082, | |
| "loss": 2.4747, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.5870150505362581, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00018527271684782865, | |
| "loss": 2.4858, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.5878977799355608, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00018459732324596834, | |
| "loss": 2.4894, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.5887805093348634, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.0001839224416679421, | |
| "loss": 2.5035, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.589663238734166, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00018324807739725614, | |
| "loss": 2.4819, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.5905459681334687, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.000182574235713367, | |
| "loss": 2.4958, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.5914286975327713, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00018190092189163974, | |
| "loss": 2.5124, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.592311426932074, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00018122814120330688, | |
| "loss": 2.4775, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.5931941563313766, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00018055589891542758, | |
| "loss": 2.4865, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.5940768857306793, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00017988420029084551, | |
| "loss": 2.4782, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.5949596151299819, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00017921305058814818, | |
| "loss": 2.4903, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.5958423445292845, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00017854245506162582, | |
| "loss": 2.4786, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.5967250739285872, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00017787241896123024, | |
| "loss": 2.4919, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.5976078033278899, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00017720294753253345, | |
| "loss": 2.4921, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.5984905327271924, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00017653404601668666, | |
| "loss": 2.4775, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.5993732621264951, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00017586571965037966, | |
| "loss": 2.491, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.6002559915257978, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 0.0001751979736657993, | |
| "loss": 2.4803, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6011387209251005, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00017453081329058882, | |
| "loss": 2.4797, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.602021450324403, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.0001738642437478067, | |
| "loss": 2.4761, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.6029041797237057, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00017319827025588614, | |
| "loss": 2.4584, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.6037869091230084, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.0001725328980285939, | |
| "loss": 2.4875, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.604669638522311, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00017186813227498937, | |
| "loss": 2.4854, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.6055523679216136, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.0001712039781993844, | |
| "loss": 2.4811, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.6064350973209163, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00017054044100130178, | |
| "loss": 2.4804, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.607317826720219, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.0001698775258754351, | |
| "loss": 2.4854, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.6082005561195215, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00016921523801160756, | |
| "loss": 2.4895, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.6090832855188242, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00016855358259473217, | |
| "loss": 2.4823, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.6099660149181269, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00016789256480477023, | |
| "loss": 2.4752, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 0.6108487443174295, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00016723218981669127, | |
| "loss": 2.4758, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.6117314737167321, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 0.00016657246280043266, | |
| "loss": 2.4934, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 0.6126142031160348, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00016591338892085874, | |
| "loss": 2.488, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.6134969325153374, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0001652549733377206, | |
| "loss": 2.4797, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.6143796619146401, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00016459722120561567, | |
| "loss": 2.467, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.6152623913139427, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.0001639401376739475, | |
| "loss": 2.4972, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 0.6161451207132453, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.0001632837278868851, | |
| "loss": 2.4728, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.617027850112548, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00016262799698332292, | |
| "loss": 2.4723, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 0.6179105795118507, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00016197295009684077, | |
| "loss": 2.4765, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6187933089111533, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00016131859235566325, | |
| "loss": 2.476, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 0.6196760383104559, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00016066492888261983, | |
| "loss": 2.4959, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.6205587677097586, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00016001196479510448, | |
| "loss": 2.4946, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 0.6214414971090613, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00015935970520503638, | |
| "loss": 2.4895, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.6223242265083638, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 0.0001587081552188188, | |
| "loss": 2.4848, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.6232069559076665, | |
| "grad_norm": 0.125, | |
| "learning_rate": 0.0001580573199372999, | |
| "loss": 2.4822, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.6240896853069692, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00015740720445573262, | |
| "loss": 2.4835, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 0.6249724147062717, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00015675781386373462, | |
| "loss": 2.4818, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.6258551441055744, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0001561091532452486, | |
| "loss": 2.4918, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 0.6267378735048771, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00015546122767850232, | |
| "loss": 2.4882, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.6276206029041798, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00015481404223596939, | |
| "loss": 2.4933, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 0.6285033323034823, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.0001541676019843286, | |
| "loss": 2.4844, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.629386061702785, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00015352191198442507, | |
| "loss": 2.4719, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 0.6302687911020877, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00015287697729123045, | |
| "loss": 2.4803, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.6311515205013903, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.0001522328029538031, | |
| "loss": 2.4894, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.6320342499006929, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00015158939401524877, | |
| "loss": 2.491, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.6329169792999956, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00015094675551268096, | |
| "loss": 2.487, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 0.6337997086992982, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00015030489247718173, | |
| "loss": 2.4762, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.6346824380986009, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00014966380993376217, | |
| "loss": 2.4862, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 0.6355651674979035, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.0001490235129013228, | |
| "loss": 2.4853, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.6364478968972062, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.00014838400639261503, | |
| "loss": 2.4968, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 0.6373306262965088, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.000147745295414201, | |
| "loss": 2.4892, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.6382133556958115, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00014710738496641492, | |
| "loss": 2.4631, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 0.6390960850951141, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.0001464702800433238, | |
| "loss": 2.467, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.6399788144944167, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00014583398563268858, | |
| "loss": 2.4861, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.6408615438937194, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00014519850671592467, | |
| "loss": 2.493, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.641744273293022, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.000144563848268063, | |
| "loss": 2.4982, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 0.6426270026923246, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00014393001525771153, | |
| "loss": 2.4848, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.6435097320916273, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00014329701264701597, | |
| "loss": 2.4833, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 0.64439246149093, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.0001426648453916208, | |
| "loss": 2.4801, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.6452751908902326, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00014203351844063088, | |
| "loss": 2.4711, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 0.6461579202895352, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0001414030367365725, | |
| "loss": 2.4779, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.6470406496888379, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00014077340521535472, | |
| "loss": 2.4893, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 0.6479233790881406, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00014014462880623042, | |
| "loss": 2.4746, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.6488061084874431, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00013951671243175824, | |
| "loss": 2.4794, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.6496888378867458, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.00013888966100776386, | |
| "loss": 2.4845, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.6505715672860485, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00013826347944330116, | |
| "loss": 2.4637, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 0.6514542966853512, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00013763817264061425, | |
| "loss": 2.493, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.6523370260846537, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.00013701374549509899, | |
| "loss": 2.4888, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 0.6532197554839564, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00013639020289526438, | |
| "loss": 2.4966, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.6541024848832591, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00013576754972269463, | |
| "loss": 2.4908, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 0.6549852142825617, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.0001351457908520109, | |
| "loss": 2.4792, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.6558679436818643, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.0001345249311508328, | |
| "loss": 2.4821, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 0.656750673081167, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00013390497547974078, | |
| "loss": 2.4825, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.6576334024804696, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00013328592869223747, | |
| "loss": 2.4828, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.6585161318797722, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00013266779563471064, | |
| "loss": 2.4791, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.6593988612790749, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00013205058114639407, | |
| "loss": 2.486, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 0.6602815906783776, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.00013143429005933052, | |
| "loss": 2.484, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.6611643200776802, | |
| "grad_norm": 0.125, | |
| "learning_rate": 0.00013081892719833378, | |
| "loss": 2.4683, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 0.6620470494769828, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 0.0001302044973809503, | |
| "loss": 2.484, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.6629297788762855, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00012959100541742248, | |
| "loss": 2.4898, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 0.6638125082755881, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.0001289784561106499, | |
| "loss": 2.4868, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.6646952376748908, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00012836685425615275, | |
| "loss": 2.4984, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 0.6655779670741934, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 0.00012775620464203365, | |
| "loss": 2.4814, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.666460696473496, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.0001271465120489401, | |
| "loss": 2.4742, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.6673434258727987, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.0001265377812500278, | |
| "loss": 2.4828, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.6682261552721014, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00012593001701092233, | |
| "loss": 2.482, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 0.669108884671404, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.00012532322408968221, | |
| "loss": 2.4777, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.6699916140707066, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00012471740723676213, | |
| "loss": 2.4852, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 0.6708743434700093, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 0.000124112571194975, | |
| "loss": 2.4819, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.671757072869312, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00012350872069945547, | |
| "loss": 2.4826, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 0.6726398022686145, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00012290586047762216, | |
| "loss": 2.4811, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.6735225316679172, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00012230399524914136, | |
| "loss": 2.472, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 0.6744052610672199, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00012170312972588974, | |
| "loss": 2.469, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.6752879904665224, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 0.00012110326861191722, | |
| "loss": 2.4757, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.6761707198658251, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00012050441660341074, | |
| "loss": 2.4811, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.6770534492651278, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00011990657838865706, | |
| "loss": 2.4748, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 0.6779361786644305, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00011930975864800603, | |
| "loss": 2.478, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.678818908063733, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.0001187139620538342, | |
| "loss": 2.4918, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 0.6797016374630357, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.0001181191932705081, | |
| "loss": 2.4842, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.6805843668623384, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00011752545695434788, | |
| "loss": 2.491, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 0.681467096261641, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00011693275775359049, | |
| "loss": 2.4993, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.6823498256609436, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00011634110030835341, | |
| "loss": 2.4743, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 0.6832325550602463, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.000115750489250599, | |
| "loss": 2.4776, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.684115284459549, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00011516092920409706, | |
| "loss": 2.4867, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.6849980138588516, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 0.00011457242478438962, | |
| "loss": 2.4774, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.6858807432581542, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00011398498059875434, | |
| "loss": 2.4809, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 0.6867634726574569, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00011339860124616833, | |
| "loss": 2.4625, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.6876462020567595, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00011281329131727272, | |
| "loss": 2.4775, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 0.6885289314560622, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00011222905539433593, | |
| "loss": 2.4737, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.6894116608553648, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.00011164589805121852, | |
| "loss": 2.4744, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 0.6902943902546674, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00011106382385333708, | |
| "loss": 2.4642, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.6911771196539701, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 0.00011048283735762806, | |
| "loss": 2.4921, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 0.6920598490532727, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 0.00010990294311251328, | |
| "loss": 2.4849, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.6929425784525753, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 0.00010932414565786286, | |
| "loss": 2.4835, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.693825307851878, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 0.0001087464495249606, | |
| "loss": 2.49, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.6947080372511807, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 0.00010816985923646838, | |
| "loss": 2.48, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 0.6955907666504832, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00010759437930639058, | |
| "loss": 2.4782, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.6964734960497859, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 0.00010702001424003896, | |
| "loss": 2.4723, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 0.6973562254490886, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 0.00010644676853399688, | |
| "loss": 2.4662, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.6982389548483913, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 0.00010587464667608484, | |
| "loss": 2.493, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 0.6991216842476938, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00010530365314532488, | |
| "loss": 2.4957, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.7000044136469965, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00010473379241190542, | |
| "loss": 2.4864, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 0.7008871430462992, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00010416506893714662, | |
| "loss": 2.48, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.7017698724456018, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00010359748717346534, | |
| "loss": 2.4745, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.7026526018449044, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 0.00010303105156433998, | |
| "loss": 2.4918, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.7035353312442071, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 0.00010246576654427611, | |
| "loss": 2.4878, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 0.7044180606435098, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 0.0001019016365387716, | |
| "loss": 2.4767, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.7053007900428124, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 0.00010133866596428196, | |
| "loss": 2.4814, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 0.706183519442115, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 0.0001007768592281856, | |
| "loss": 2.4884, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.706183519442115, | |
| "eval_accuracy": 0.5112470338885433, | |
| "eval_loss": 2.367577075958252, | |
| "eval_runtime": 6.9898, | |
| "eval_samples_per_second": 45.495, | |
| "eval_steps_per_second": 0.429, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7070662488414177, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00010021622072874948, | |
| "loss": 2.4866, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 0.7079489782407203, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 9.965675485509504e-05, | |
| "loss": 2.4817, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.7088317076400229, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 9.909846598716302e-05, | |
| "loss": 2.4782, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 0.7097144370393256, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 9.854135849567988e-05, | |
| "loss": 2.4829, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.7105971664386282, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 9.79854367421234e-05, | |
| "loss": 2.48, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.7114798958379309, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 9.743070507868818e-05, | |
| "loss": 2.4849, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.7123626252372335, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 9.687716784825218e-05, | |
| "loss": 2.485, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 0.7132453546365362, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 9.632482938434197e-05, | |
| "loss": 2.477, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.7141280840358388, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 9.577369401109987e-05, | |
| "loss": 2.4836, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 0.7150108134351415, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 9.522376604324889e-05, | |
| "loss": 2.4868, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.7158935428344441, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 9.467504978605956e-05, | |
| "loss": 2.4857, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 0.7167762722337467, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 9.412754953531663e-05, | |
| "loss": 2.4781, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.7176590016330494, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 9.35812695772845e-05, | |
| "loss": 2.4731, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 0.7185417310323521, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 9.303621418867444e-05, | |
| "loss": 2.4815, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.7194244604316546, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 9.24923876366106e-05, | |
| "loss": 2.4876, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.7203071898309573, | |
| "grad_norm": 0.125, | |
| "learning_rate": 9.194979417859705e-05, | |
| "loss": 2.4689, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.72118991923026, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 9.14084380624842e-05, | |
| "loss": 2.4709, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 0.7220726486295626, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 9.086832352643535e-05, | |
| "loss": 2.4813, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.7229553780288652, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 9.032945479889391e-05, | |
| "loss": 2.4797, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 0.7238381074281679, | |
| "grad_norm": 0.125, | |
| "learning_rate": 8.979183609855024e-05, | |
| "loss": 2.4909, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.7247208368274706, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 8.925547163430812e-05, | |
| "loss": 2.4756, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 0.7256035662267731, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 8.872036560525254e-05, | |
| "loss": 2.466, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.7264862956260758, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 8.818652220061638e-05, | |
| "loss": 2.4651, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 0.7273690250253785, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 8.76539455997475e-05, | |
| "loss": 2.482, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.7282517544246812, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 8.71226399720764e-05, | |
| "loss": 2.4888, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.7291344838239837, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 8.659260947708344e-05, | |
| "loss": 2.4901, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.7300172132232864, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 8.606385826426621e-05, | |
| "loss": 2.4829, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 0.7308999426225891, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 8.553639047310685e-05, | |
| "loss": 2.4875, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.7317826720218917, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 8.50102102330401e-05, | |
| "loss": 2.4883, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 0.7326654014211943, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 8.448532166342077e-05, | |
| "loss": 2.4676, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.733548130820497, | |
| "grad_norm": 0.125, | |
| "learning_rate": 8.396172887349115e-05, | |
| "loss": 2.4799, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 0.7344308602197996, | |
| "grad_norm": 0.125, | |
| "learning_rate": 8.343943596234943e-05, | |
| "loss": 2.4862, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.7353135896191023, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 8.291844701891732e-05, | |
| "loss": 2.4746, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 0.7361963190184049, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 8.239876612190778e-05, | |
| "loss": 2.4751, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.7370790484177075, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 8.188039733979366e-05, | |
| "loss": 2.4868, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.7379617778170102, | |
| "grad_norm": 0.125, | |
| "learning_rate": 8.136334473077519e-05, | |
| "loss": 2.4861, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.7388445072163128, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 8.084761234274906e-05, | |
| "loss": 2.4648, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 0.7397272366156155, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 8.033320421327578e-05, | |
| "loss": 2.4733, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.7406099660149181, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.982012436954849e-05, | |
| "loss": 2.4632, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 0.7414926954142208, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 7.930837682836195e-05, | |
| "loss": 2.4826, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.7423754248135234, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 7.87979655960801e-05, | |
| "loss": 2.4837, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 0.743258154212826, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 7.828889466860551e-05, | |
| "loss": 2.4816, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 0.7441408836121287, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 7.77811680313475e-05, | |
| "loss": 2.4895, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 0.7450236130114314, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 7.727478965919144e-05, | |
| "loss": 2.4837, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 0.745906342410734, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 7.67697635164675e-05, | |
| "loss": 2.4751, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.7467890718100366, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 7.626609355691922e-05, | |
| "loss": 2.4792, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 0.7476718012093393, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 7.576378372367306e-05, | |
| "loss": 2.4755, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 0.748554530608642, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 7.52628379492075e-05, | |
| "loss": 2.4761, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.7494372600079445, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 7.476326015532162e-05, | |
| "loss": 2.4775, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 0.7503199894072472, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 7.426505425310531e-05, | |
| "loss": 2.4919, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.7512027188065499, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 7.376822414290804e-05, | |
| "loss": 2.483, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 0.7520854482058525, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 7.327277371430858e-05, | |
| "loss": 2.4805, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 0.7529681776051551, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.27787068460842e-05, | |
| "loss": 2.4868, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 0.7538509070044578, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.228602740618085e-05, | |
| "loss": 2.4842, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 0.7547336364037605, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 7.179473925168256e-05, | |
| "loss": 2.4829, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.755616365803063, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 7.130484622878108e-05, | |
| "loss": 2.4933, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 0.7564990952023657, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.081635217274617e-05, | |
| "loss": 2.4848, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 0.7573818246016684, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.032926090789537e-05, | |
| "loss": 2.4788, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 0.758264554000971, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 6.984357624756388e-05, | |
| "loss": 2.4789, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 0.7591472834002736, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 6.935930199407501e-05, | |
| "loss": 2.4821, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.7600300127995763, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 6.887644193871042e-05, | |
| "loss": 2.4781, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 0.7609127421988789, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 6.839499986167999e-05, | |
| "loss": 2.4975, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 0.7617954715981816, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 6.791497953209289e-05, | |
| "loss": 2.4711, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 0.7626782009974842, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 6.743638470792735e-05, | |
| "loss": 2.4693, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.7635609303967869, | |
| "grad_norm": 0.125, | |
| "learning_rate": 6.695921913600212e-05, | |
| "loss": 2.4798, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.7644436597960895, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 6.648348655194613e-05, | |
| "loss": 2.4843, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 0.7653263891953922, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 6.600919068017006e-05, | |
| "loss": 2.4709, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 0.7662091185946948, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 6.553633523383682e-05, | |
| "loss": 2.4822, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 0.7670918479939974, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 6.506492391483232e-05, | |
| "loss": 2.4729, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 0.7679745773933001, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 6.459496041373708e-05, | |
| "loss": 2.4769, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.7688573067926028, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 6.412644840979656e-05, | |
| "loss": 2.4866, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 0.7697400361919053, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 6.365939157089304e-05, | |
| "loss": 2.4759, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 0.770622765591208, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 6.319379355351653e-05, | |
| "loss": 2.4632, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 0.7715054949905107, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 6.272965800273608e-05, | |
| "loss": 2.4707, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 0.7723882243898132, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 6.226698855217178e-05, | |
| "loss": 2.4834, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.7732709537891159, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 6.180578882396556e-05, | |
| "loss": 2.485, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 0.7741536831884186, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 6.134606242875324e-05, | |
| "loss": 2.4726, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 0.7750364125877213, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 6.088781296563636e-05, | |
| "loss": 2.4861, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 0.7759191419870238, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 6.043104402215388e-05, | |
| "loss": 2.4927, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 0.7768018713863265, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 5.9975759174254075e-05, | |
| "loss": 2.4853, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.7776846007856292, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 5.952196198626633e-05, | |
| "loss": 2.4986, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 0.7785673301849318, | |
| "grad_norm": 0.125, | |
| "learning_rate": 5.906965601087369e-05, | |
| "loss": 2.4872, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 0.7794500595842344, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 5.861884478908483e-05, | |
| "loss": 2.475, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 0.7803327889835371, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 5.816953185020607e-05, | |
| "loss": 2.482, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 0.7812155183828398, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 5.7721720711814195e-05, | |
| "loss": 2.4814, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.7820982477821424, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 5.727541487972876e-05, | |
| "loss": 2.4722, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 0.782980977181445, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 5.68306178479843e-05, | |
| "loss": 2.4738, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 0.7838637065807477, | |
| "grad_norm": 0.125, | |
| "learning_rate": 5.638733309880353e-05, | |
| "loss": 2.4844, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.7847464359800503, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 5.5945564102569764e-05, | |
| "loss": 2.4867, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 0.785629165379353, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 5.550531431779984e-05, | |
| "loss": 2.4711, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.7865118947786556, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 5.50665871911169e-05, | |
| "loss": 2.4816, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 0.7873946241779582, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 5.4629386157223434e-05, | |
| "loss": 2.4665, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 0.7882773535772609, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 5.4193714638874845e-05, | |
| "loss": 2.4881, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 0.7891600829765635, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 5.375957604685186e-05, | |
| "loss": 2.4593, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 0.7900428123758662, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 5.3326973779934506e-05, | |
| "loss": 2.4863, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.7909255417751688, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 5.289591122487522e-05, | |
| "loss": 2.4839, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.7918082711744715, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 5.246639175637216e-05, | |
| "loss": 2.4887, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 0.7926910005737741, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 5.203841873704329e-05, | |
| "loss": 2.4873, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 0.7935737299730767, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 5.161199551739942e-05, | |
| "loss": 2.4577, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 0.7944564593723794, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 5.1187125435818575e-05, | |
| "loss": 2.4905, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.7953391887716821, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 5.0763811818519494e-05, | |
| "loss": 2.4824, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 0.7962219181709846, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 5.0342057979535507e-05, | |
| "loss": 2.487, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 0.7971046475702873, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 4.99218672206892e-05, | |
| "loss": 2.4845, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 0.79798737696959, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 4.950324283156562e-05, | |
| "loss": 2.4858, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 0.7988701063688927, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 4.908618808948748e-05, | |
| "loss": 2.4718, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.7997528357681952, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 4.867070625948866e-05, | |
| "loss": 2.4967, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 0.8006355651674979, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 4.825680059428933e-05, | |
| "loss": 2.4717, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 0.8015182945668006, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 4.784447433427016e-05, | |
| "loss": 2.4791, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 0.8024010239661032, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 4.7433730707446805e-05, | |
| "loss": 2.483, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 0.8032837533654058, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 4.702457292944498e-05, | |
| "loss": 2.4791, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.8041664827647085, | |
| "grad_norm": 0.125, | |
| "learning_rate": 4.661700420347517e-05, | |
| "loss": 2.4739, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 0.8050492121640112, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 4.62110277203073e-05, | |
| "loss": 2.4814, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.8059319415633137, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 4.5806646658246104e-05, | |
| "loss": 2.4909, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 0.8068146709626164, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 4.5403864183106184e-05, | |
| "loss": 2.4878, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 0.8076974003619191, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 4.5002683448186866e-05, | |
| "loss": 2.4955, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.8085801297612217, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 4.460310759424802e-05, | |
| "loss": 2.4799, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 0.8094628591605243, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 4.420513974948517e-05, | |
| "loss": 2.4743, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 0.810345588559827, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 4.3808783029505166e-05, | |
| "loss": 2.4726, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 0.8112283179591296, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 4.341404053730147e-05, | |
| "loss": 2.485, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 0.8121110473584323, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 4.3020915363230274e-05, | |
| "loss": 2.4825, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.8129937767577349, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 4.262941058498615e-05, | |
| "loss": 2.4725, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 0.8138765061570375, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 4.2239529267577736e-05, | |
| "loss": 2.4801, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 0.8147592355563402, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 4.1851274463304165e-05, | |
| "loss": 2.4842, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 0.8156419649556429, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 4.146464921173088e-05, | |
| "loss": 2.4747, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 0.8165246943549455, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 4.1079656539665696e-05, | |
| "loss": 2.4856, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.8174074237542481, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 4.069629946113565e-05, | |
| "loss": 2.4736, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 0.8182901531535508, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 4.0314580977362655e-05, | |
| "loss": 2.4811, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 0.8191728825528535, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 3.99345040767409e-05, | |
| "loss": 2.4788, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 0.820055611952156, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 3.955607173481254e-05, | |
| "loss": 2.4807, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 0.8209383413514587, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 3.9179286914244884e-05, | |
| "loss": 2.4758, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.8218210707507614, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 3.880415256480749e-05, | |
| "loss": 2.4897, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 0.822703800150064, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.843067162334826e-05, | |
| "loss": 2.4591, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 0.8235865295493666, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 3.805884701377127e-05, | |
| "loss": 2.4757, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 0.8244692589486693, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.768868164701325e-05, | |
| "loss": 2.4784, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 0.825351988347972, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.732017842102126e-05, | |
| "loss": 2.5042, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.8262347177472745, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 3.695334022072977e-05, | |
| "loss": 2.4785, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 0.8271174471465772, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 3.658816991803798e-05, | |
| "loss": 2.4855, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 0.8280001765458799, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 3.622467037178765e-05, | |
| "loss": 2.4773, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 0.8288829059451825, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 3.586284442774049e-05, | |
| "loss": 2.4639, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 0.8297656353444851, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.550269491855579e-05, | |
| "loss": 2.4763, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.8306483647437878, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 3.514422466376857e-05, | |
| "loss": 2.4841, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 0.8315310941430905, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 3.478743646976726e-05, | |
| "loss": 2.4838, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 0.8324138235423931, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 3.443233312977176e-05, | |
| "loss": 2.4814, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 0.8332965529416957, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.4078917423811556e-05, | |
| "loss": 2.4671, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 0.8341792823409984, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.372719211870412e-05, | |
| "loss": 2.4649, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.835062011740301, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 3.3377159968033085e-05, | |
| "loss": 2.4915, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 0.8359447411396037, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 3.302882371212665e-05, | |
| "loss": 2.4789, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 0.8368274705389063, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 3.2682186078036304e-05, | |
| "loss": 2.4878, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 0.8377101999382089, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.2337249779515436e-05, | |
| "loss": 2.4843, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 0.8385929293375116, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 3.199401751699782e-05, | |
| "loss": 2.4744, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.8394756587368142, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 3.1652491977576883e-05, | |
| "loss": 2.481, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 0.8403583881361169, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 3.131267583498448e-05, | |
| "loss": 2.4857, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 0.8412411175354195, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 3.097457174956977e-05, | |
| "loss": 2.489, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 0.8421238469347222, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 3.063818236827884e-05, | |
| "loss": 2.4832, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 0.8430065763340248, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 3.030351032463341e-05, | |
| "loss": 2.491, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.8438893057333274, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 2.9970558238710865e-05, | |
| "loss": 2.4858, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 0.8447720351326301, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 2.9639328717123104e-05, | |
| "loss": 2.4705, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 0.8456547645319328, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 2.9309824352996618e-05, | |
| "loss": 2.4779, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 0.8465374939312353, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 2.898204772595195e-05, | |
| "loss": 2.4793, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 0.847420223330538, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 2.865600140208349e-05, | |
| "loss": 2.4618, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.8483029527298407, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 2.833168793393956e-05, | |
| "loss": 2.4848, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 0.8491856821291434, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.8009109860502174e-05, | |
| "loss": 2.4771, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 0.8500684115284459, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 2.768826970716745e-05, | |
| "loss": 2.4828, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 0.8509511409277486, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 2.736916998572567e-05, | |
| "loss": 2.487, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 0.8518338703270513, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 2.705181319434144e-05, | |
| "loss": 2.4868, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.8527165997263539, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.6736201817534696e-05, | |
| "loss": 2.4807, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 0.8535993291256565, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.6422338326160618e-05, | |
| "loss": 2.4829, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 0.8544820585249592, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.6110225177390534e-05, | |
| "loss": 2.4849, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 0.8553647879242618, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 2.5799864814692902e-05, | |
| "loss": 2.4787, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 0.8562475173235644, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 2.549125966781385e-05, | |
| "loss": 2.4753, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.8571302467228671, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.518441215275838e-05, | |
| "loss": 2.4752, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 0.8580129761221698, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.48793246717712e-05, | |
| "loss": 2.4789, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 0.8588957055214724, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 2.4575999613318245e-05, | |
| "loss": 2.4878, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 0.859778434920775, | |
| "grad_norm": 0.125, | |
| "learning_rate": 2.4274439352067828e-05, | |
| "loss": 2.4793, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 0.8606611643200777, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 2.3974646248871827e-05, | |
| "loss": 2.4803, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.8615438937193803, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.3676622650747603e-05, | |
| "loss": 2.4748, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 0.862426623118683, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.3380370890859454e-05, | |
| "loss": 2.4787, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 0.8633093525179856, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.3085893288500136e-05, | |
| "loss": 2.4785, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 0.8641920819172882, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 2.279319214907305e-05, | |
| "loss": 2.4603, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 0.8650748113165909, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 2.2502269764074017e-05, | |
| "loss": 2.4593, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.8659575407158936, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 2.2213128411073396e-05, | |
| "loss": 2.4895, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 0.8668402701151962, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.1925770353698137e-05, | |
| "loss": 2.4858, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 0.8677229995144988, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 2.1640197841614083e-05, | |
| "loss": 2.4799, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 0.8686057289138015, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 2.1356413110508675e-05, | |
| "loss": 2.4731, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 0.8694884583131042, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.1074418382072912e-05, | |
| "loss": 2.4795, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.8703711877124067, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.0794215863984417e-05, | |
| "loss": 2.4697, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 0.8712539171117094, | |
| "grad_norm": 0.125, | |
| "learning_rate": 2.0515807749889954e-05, | |
| "loss": 2.4764, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 0.8721366465110121, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 2.0239196219388133e-05, | |
| "loss": 2.4901, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 0.8730193759103146, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.9964383438012685e-05, | |
| "loss": 2.4936, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 0.8739021053096173, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 1.969137155721509e-05, | |
| "loss": 2.4794, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.87478483470892, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.942016271434821e-05, | |
| "loss": 2.4851, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 0.8756675641082227, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 1.915075903264915e-05, | |
| "loss": 2.4784, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 0.8765502935075252, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.8883162621222693e-05, | |
| "loss": 2.4947, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 0.8774330229068279, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.8617375575025186e-05, | |
| "loss": 2.4925, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 0.8783157523061306, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.835339997484753e-05, | |
| "loss": 2.4929, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.8791984817054332, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 1.8091237887299357e-05, | |
| "loss": 2.4804, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 0.8800812111047358, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 1.783089136479257e-05, | |
| "loss": 2.4877, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 0.8809639405040385, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.757236244552557e-05, | |
| "loss": 2.4866, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 0.8818466699033412, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.7315653153466977e-05, | |
| "loss": 2.4786, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 0.8827293993026438, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.7060765498339958e-05, | |
| "loss": 2.4865, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.8827293993026438, | |
| "eval_accuracy": 0.5115012171615945, | |
| "eval_loss": 2.3658125400543213, | |
| "eval_runtime": 6.9822, | |
| "eval_samples_per_second": 45.544, | |
| "eval_steps_per_second": 0.43, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.8836121287019464, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 1.6807701475606534e-05, | |
| "loss": 2.4905, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 0.8844948581012491, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.6556463066451837e-05, | |
| "loss": 2.4761, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 0.8853775875005517, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 1.63070522377686e-05, | |
| "loss": 2.4894, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 0.8862603168998543, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.6059470942141912e-05, | |
| "loss": 2.4742, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 0.887143046299157, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.5813721117833828e-05, | |
| "loss": 2.4906, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.8880257756984596, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 1.5569804688768092e-05, | |
| "loss": 2.4644, | |
| "step": 10060 | |
| }, | |
| { | |
| "epoch": 0.8889085050977623, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 1.532772356451531e-05, | |
| "loss": 2.4756, | |
| "step": 10070 | |
| }, | |
| { | |
| "epoch": 0.8897912344970649, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 1.5087479640277763e-05, | |
| "loss": 2.4801, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 0.8906739638963675, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 1.4849074796874779e-05, | |
| "loss": 2.493, | |
| "step": 10090 | |
| }, | |
| { | |
| "epoch": 0.8915566932956702, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.4612510900727794e-05, | |
| "loss": 2.478, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.8924394226949729, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 1.4377789803845964e-05, | |
| "loss": 2.4824, | |
| "step": 10110 | |
| }, | |
| { | |
| "epoch": 0.8933221520942755, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 1.4144913343811544e-05, | |
| "loss": 2.4745, | |
| "step": 10120 | |
| }, | |
| { | |
| "epoch": 0.8942048814935781, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 1.3913883343765394e-05, | |
| "loss": 2.4786, | |
| "step": 10130 | |
| }, | |
| { | |
| "epoch": 0.8950876108928808, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.3684701612392963e-05, | |
| "loss": 2.4768, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 0.8959703402921835, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.345736994390992e-05, | |
| "loss": 2.4689, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.896853069691486, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.3231890118048179e-05, | |
| "loss": 2.4817, | |
| "step": 10160 | |
| }, | |
| { | |
| "epoch": 0.8977357990907887, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 1.300826390004209e-05, | |
| "loss": 2.49, | |
| "step": 10170 | |
| }, | |
| { | |
| "epoch": 0.8986185284900914, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.2786493040614245e-05, | |
| "loss": 2.4965, | |
| "step": 10180 | |
| }, | |
| { | |
| "epoch": 0.8995012578893941, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.2566579275962303e-05, | |
| "loss": 2.4725, | |
| "step": 10190 | |
| }, | |
| { | |
| "epoch": 0.9003839872886966, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.2348524327744943e-05, | |
| "loss": 2.4706, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.9012667166879993, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 1.2132329903068563e-05, | |
| "loss": 2.4772, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 0.902149446087302, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.1917997694473992e-05, | |
| "loss": 2.4823, | |
| "step": 10220 | |
| }, | |
| { | |
| "epoch": 0.9030321754866045, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 1.1705529379923085e-05, | |
| "loss": 2.4674, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 0.9039149048859072, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 1.1494926622785811e-05, | |
| "loss": 2.4772, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 0.9047976342852099, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.1286191071826823e-05, | |
| "loss": 2.472, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.9056803636845125, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 1.1079324361193022e-05, | |
| "loss": 2.5008, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 0.9065630930838151, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.0874328110400511e-05, | |
| "loss": 2.4843, | |
| "step": 10270 | |
| }, | |
| { | |
| "epoch": 0.9074458224831178, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.0671203924321887e-05, | |
| "loss": 2.4851, | |
| "step": 10280 | |
| }, | |
| { | |
| "epoch": 0.9083285518824205, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 1.0469953393173776e-05, | |
| "loss": 2.4739, | |
| "step": 10290 | |
| }, | |
| { | |
| "epoch": 0.9092112812817231, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 1.0270578092504396e-05, | |
| "loss": 2.4763, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.9100940106810257, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.0073079583181126e-05, | |
| "loss": 2.4805, | |
| "step": 10310 | |
| }, | |
| { | |
| "epoch": 0.9109767400803284, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 9.877459411378325e-06, | |
| "loss": 2.4858, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 0.911859469479631, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 9.683719108565331e-06, | |
| "loss": 2.4798, | |
| "step": 10330 | |
| }, | |
| { | |
| "epoch": 0.9127421988789337, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 9.49186019149434e-06, | |
| "loss": 2.4887, | |
| "step": 10340 | |
| }, | |
| { | |
| "epoch": 0.9136249282782363, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 9.301884162188496e-06, | |
| "loss": 2.4803, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.9145076576775389, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 9.113792507930263e-06, | |
| "loss": 2.4804, | |
| "step": 10360 | |
| }, | |
| { | |
| "epoch": 0.9153903870768416, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 8.927586701249852e-06, | |
| "loss": 2.4766, | |
| "step": 10370 | |
| }, | |
| { | |
| "epoch": 0.9162731164761443, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 8.743268199913307e-06, | |
| "loss": 2.4674, | |
| "step": 10380 | |
| }, | |
| { | |
| "epoch": 0.9171558458754469, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 8.560838446911607e-06, | |
| "loss": 2.473, | |
| "step": 10390 | |
| }, | |
| { | |
| "epoch": 0.9180385752747495, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 8.380298870449e-06, | |
| "loss": 2.4637, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.9189213046740522, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 8.201650883931904e-06, | |
| "loss": 2.4808, | |
| "step": 10410 | |
| }, | |
| { | |
| "epoch": 0.9198040340733548, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 8.024895885957978e-06, | |
| "loss": 2.4655, | |
| "step": 10420 | |
| }, | |
| { | |
| "epoch": 0.9206867634726574, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 7.85003526030495e-06, | |
| "loss": 2.4755, | |
| "step": 10430 | |
| }, | |
| { | |
| "epoch": 0.9215694928719601, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 7.677070375920026e-06, | |
| "loss": 2.4738, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 0.9224522222712628, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 7.506002586909006e-06, | |
| "loss": 2.4909, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.9233349516705653, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 7.336833232525625e-06, | |
| "loss": 2.4758, | |
| "step": 10460 | |
| }, | |
| { | |
| "epoch": 0.924217681069868, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 7.169563637161397e-06, | |
| "loss": 2.4819, | |
| "step": 10470 | |
| }, | |
| { | |
| "epoch": 0.9251004104691707, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 7.004195110334788e-06, | |
| "loss": 2.4731, | |
| "step": 10480 | |
| }, | |
| { | |
| "epoch": 0.9259831398684734, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 6.840728946681363e-06, | |
| "loss": 2.4937, | |
| "step": 10490 | |
| }, | |
| { | |
| "epoch": 0.9268658692677759, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 6.679166425943351e-06, | |
| "loss": 2.473, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9277485986670786, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 6.519508812959873e-06, | |
| "loss": 2.4801, | |
| "step": 10510 | |
| }, | |
| { | |
| "epoch": 0.9286313280663813, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 6.3617573576569274e-06, | |
| "loss": 2.479, | |
| "step": 10520 | |
| }, | |
| { | |
| "epoch": 0.9295140574656839, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 6.205913295037474e-06, | |
| "loss": 2.4722, | |
| "step": 10530 | |
| }, | |
| { | |
| "epoch": 0.9303967868649865, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 6.051977845172002e-06, | |
| "loss": 2.4912, | |
| "step": 10540 | |
| }, | |
| { | |
| "epoch": 0.9312795162642892, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 5.899952213188897e-06, | |
| "loss": 2.4677, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.9321622456635918, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 5.749837589264895e-06, | |
| "loss": 2.4811, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 0.9330449750628945, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 5.601635148615891e-06, | |
| "loss": 2.4728, | |
| "step": 10570 | |
| }, | |
| { | |
| "epoch": 0.9339277044621971, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 5.4553460514877304e-06, | |
| "loss": 2.49, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 0.9348104338614998, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 5.3109714431470165e-06, | |
| "loss": 2.4939, | |
| "step": 10590 | |
| }, | |
| { | |
| "epoch": 0.9356931632608024, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 5.168512453872287e-06, | |
| "loss": 2.4792, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.936575892660105, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 5.027970198945076e-06, | |
| "loss": 2.4792, | |
| "step": 10610 | |
| }, | |
| { | |
| "epoch": 0.9374586220594077, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 4.889345778641252e-06, | |
| "loss": 2.4766, | |
| "step": 10620 | |
| }, | |
| { | |
| "epoch": 0.9383413514587103, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 4.752640278222254e-06, | |
| "loss": 2.4851, | |
| "step": 10630 | |
| }, | |
| { | |
| "epoch": 0.939224080858013, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 4.617854767926782e-06, | |
| "loss": 2.4703, | |
| "step": 10640 | |
| }, | |
| { | |
| "epoch": 0.9401068102573156, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 4.484990302962344e-06, | |
| "loss": 2.4966, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.9409895396566182, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 4.354047923496917e-06, | |
| "loss": 2.4759, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 0.9418722690559209, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 4.2250286546509365e-06, | |
| "loss": 2.4697, | |
| "step": 10670 | |
| }, | |
| { | |
| "epoch": 0.9427549984552236, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 4.09793350648921e-06, | |
| "loss": 2.4771, | |
| "step": 10680 | |
| }, | |
| { | |
| "epoch": 0.9436377278545262, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.9727634740129585e-06, | |
| "loss": 2.486, | |
| "step": 10690 | |
| }, | |
| { | |
| "epoch": 0.9445204572538288, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 3.849519537152124e-06, | |
| "loss": 2.4853, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.9454031866531315, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.7282026607576016e-06, | |
| "loss": 2.48, | |
| "step": 10710 | |
| }, | |
| { | |
| "epoch": 0.9462859160524342, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 3.608813794593796e-06, | |
| "loss": 2.4858, | |
| "step": 10720 | |
| }, | |
| { | |
| "epoch": 0.9471686454517367, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 3.491353873331077e-06, | |
| "loss": 2.4776, | |
| "step": 10730 | |
| }, | |
| { | |
| "epoch": 0.9480513748510394, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 3.3758238165384757e-06, | |
| "loss": 2.474, | |
| "step": 10740 | |
| }, | |
| { | |
| "epoch": 0.9489341042503421, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 3.262224528676666e-06, | |
| "loss": 2.4638, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.9498168336496448, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 3.1505568990905787e-06, | |
| "loss": 2.4872, | |
| "step": 10760 | |
| }, | |
| { | |
| "epoch": 0.9506995630489473, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 3.040821802002658e-06, | |
| "loss": 2.4675, | |
| "step": 10770 | |
| }, | |
| { | |
| "epoch": 0.95158229244825, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 2.9330200965059507e-06, | |
| "loss": 2.4689, | |
| "step": 10780 | |
| }, | |
| { | |
| "epoch": 0.9524650218475527, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.827152626557389e-06, | |
| "loss": 2.4866, | |
| "step": 10790 | |
| }, | |
| { | |
| "epoch": 0.9533477512468552, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 2.72322022097124e-06, | |
| "loss": 2.4686, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.9542304806461579, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.621223693412417e-06, | |
| "loss": 2.482, | |
| "step": 10810 | |
| }, | |
| { | |
| "epoch": 0.9551132100454606, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 2.5211638423903725e-06, | |
| "loss": 2.4855, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 0.9559959394447632, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.4230414512527166e-06, | |
| "loss": 2.4815, | |
| "step": 10830 | |
| }, | |
| { | |
| "epoch": 0.9568786688440658, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.326857288178996e-06, | |
| "loss": 2.477, | |
| "step": 10840 | |
| }, | |
| { | |
| "epoch": 0.9577613982433685, | |
| "grad_norm": 0.125, | |
| "learning_rate": 2.232612106174897e-06, | |
| "loss": 2.4788, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.9586441276426712, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 2.1403066430661644e-06, | |
| "loss": 2.4843, | |
| "step": 10860 | |
| }, | |
| { | |
| "epoch": 0.9595268570419738, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 2.0499416214928844e-06, | |
| "loss": 2.4884, | |
| "step": 10870 | |
| }, | |
| { | |
| "epoch": 0.9604095864412764, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.9615177489038792e-06, | |
| "loss": 2.4683, | |
| "step": 10880 | |
| }, | |
| { | |
| "epoch": 0.9612923158405791, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 1.8750357175510435e-06, | |
| "loss": 2.478, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 0.9621750452398817, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 1.7904962044841266e-06, | |
| "loss": 2.493, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.9630577746391844, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 1.70789987154521e-06, | |
| "loss": 2.4737, | |
| "step": 10910 | |
| }, | |
| { | |
| "epoch": 0.963940504038487, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 1.6272473653636266e-06, | |
| "loss": 2.4757, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 0.9648232334377896, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.5485393173509388e-06, | |
| "loss": 2.4702, | |
| "step": 10930 | |
| }, | |
| { | |
| "epoch": 0.9657059628370923, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.4717763436959685e-06, | |
| "loss": 2.4828, | |
| "step": 10940 | |
| }, | |
| { | |
| "epoch": 0.966588692236395, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 1.3969590453598858e-06, | |
| "loss": 2.467, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.9674714216356975, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 1.3240880080716832e-06, | |
| "loss": 2.4732, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 0.9683541510350002, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.2531638023233761e-06, | |
| "loss": 2.4722, | |
| "step": 10970 | |
| }, | |
| { | |
| "epoch": 0.9692368804343029, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.1841869833656981e-06, | |
| "loss": 2.5025, | |
| "step": 10980 | |
| }, | |
| { | |
| "epoch": 0.9701196098336055, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.1171580912036627e-06, | |
| "loss": 2.463, | |
| "step": 10990 | |
| }, | |
| { | |
| "epoch": 0.9710023392329081, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.0520776505924812e-06, | |
| "loss": 2.4808, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.9718850686322108, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 9.889461710332059e-07, | |
| "loss": 2.4864, | |
| "step": 11010 | |
| }, | |
| { | |
| "epoch": 0.9727677980315135, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 9.277641467689279e-07, | |
| "loss": 2.477, | |
| "step": 11020 | |
| }, | |
| { | |
| "epoch": 0.973650527430816, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 8.685320567809741e-07, | |
| "loss": 2.4772, | |
| "step": 11030 | |
| }, | |
| { | |
| "epoch": 0.9745332568301187, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 8.112503647848546e-07, | |
| "loss": 2.4618, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 0.9754159862294214, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 7.559195192269608e-07, | |
| "loss": 2.4775, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.9762987156287241, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 7.025399532808452e-07, | |
| "loss": 2.4819, | |
| "step": 11060 | |
| }, | |
| { | |
| "epoch": 0.9771814450280266, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 6.511120848439467e-07, | |
| "loss": 2.4891, | |
| "step": 11070 | |
| }, | |
| { | |
| "epoch": 0.9780641744273293, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 6.016363165342875e-07, | |
| "loss": 2.4727, | |
| "step": 11080 | |
| }, | |
| { | |
| "epoch": 0.978946903826632, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 5.54113035687226e-07, | |
| "loss": 2.4755, | |
| "step": 11090 | |
| }, | |
| { | |
| "epoch": 0.9798296332259346, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 5.085426143525695e-07, | |
| "loss": 2.4663, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.9807123626252372, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 4.649254092916333e-07, | |
| "loss": 2.4814, | |
| "step": 11110 | |
| }, | |
| { | |
| "epoch": 0.9815950920245399, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 4.2326176197429735e-07, | |
| "loss": 2.4864, | |
| "step": 11120 | |
| }, | |
| { | |
| "epoch": 0.9824778214238425, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 3.835519985765368e-07, | |
| "loss": 2.4657, | |
| "step": 11130 | |
| }, | |
| { | |
| "epoch": 0.9833605508231452, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 3.457964299777849e-07, | |
| "loss": 2.4785, | |
| "step": 11140 | |
| }, | |
| { | |
| "epoch": 0.9842432802224478, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 3.099953517584353e-07, | |
| "loss": 2.4743, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.9851260096217505, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 2.761490441976211e-07, | |
| "loss": 2.4793, | |
| "step": 11160 | |
| }, | |
| { | |
| "epoch": 0.9860087390210531, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 2.4425777227102265e-07, | |
| "loss": 2.4827, | |
| "step": 11170 | |
| }, | |
| { | |
| "epoch": 0.9868914684203557, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 2.1432178564867455e-07, | |
| "loss": 2.4843, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 0.9877741978196584, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.8634131869313397e-07, | |
| "loss": 2.4731, | |
| "step": 11190 | |
| }, | |
| { | |
| "epoch": 0.988656927218961, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 1.6031659045759318e-07, | |
| "loss": 2.4696, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.9895396566182637, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 1.3624780468424192e-07, | |
| "loss": 2.4817, | |
| "step": 11210 | |
| }, | |
| { | |
| "epoch": 0.9904223860175663, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 1.1413514980254669e-07, | |
| "loss": 2.4791, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 0.9913051154168689, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 9.397879892777961e-08, | |
| "loss": 2.4808, | |
| "step": 11230 | |
| }, | |
| { | |
| "epoch": 0.9921878448161716, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 7.577890985985269e-08, | |
| "loss": 2.4776, | |
| "step": 11240 | |
| }, | |
| { | |
| "epoch": 0.9930705742154743, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 5.953562508184684e-08, | |
| "loss": 2.4801, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.9939533036147769, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 4.524907175904036e-08, | |
| "loss": 2.4766, | |
| "step": 11260 | |
| }, | |
| { | |
| "epoch": 0.9948360330140795, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 3.2919361737854256e-08, | |
| "loss": 2.4889, | |
| "step": 11270 | |
| }, | |
| { | |
| "epoch": 0.9957187624133822, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 2.2546591544991833e-08, | |
| "loss": 2.4668, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 0.9966014918126849, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 1.4130842386717025e-08, | |
| "loss": 2.4816, | |
| "step": 11290 | |
| }, | |
| { | |
| "epoch": 0.9974842212119874, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 7.672180148132757e-09, | |
| "loss": 2.4718, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.9983669506112901, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 3.1706553927923763e-09, | |
| "loss": 2.4761, | |
| "step": 11310 | |
| }, | |
| { | |
| "epoch": 0.9992496800105928, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 6.263033621722869e-10, | |
| "loss": 2.4773, | |
| "step": 11320 | |
| }, | |
| { | |
| "epoch": 0.9999558635300348, | |
| "step": 11328, | |
| "total_flos": 2.202782600225295e+20, | |
| "train_loss": 2.520355888510828, | |
| "train_runtime": 24592.232, | |
| "train_samples_per_second": 117.927, | |
| "train_steps_per_second": 0.461 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 11328, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.202782600225295e+20, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |