| { | |
| "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": 6.8759, | |
| "eval_samples_per_second": 46.248, | |
| "eval_steps_per_second": 0.436, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008827293993026437, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 6.1897, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0017654587986052875, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 6.113, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0026481881979079315, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 5.7287, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003530917597210575, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 5.1703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0044136469965132185, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00015, | |
| "loss": 4.7298, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005296376395815863, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 4.4011, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0061791057951185065, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 4.1527, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.00706183519442115, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 3.8858, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.007944564593723794, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00027, | |
| "loss": 3.6934, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.008827293993026437, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0003, | |
| "loss": 3.5197, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.009710023392329082, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00029999941284073584, | |
| "loss": 3.3917, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.010592752791631726, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0002999976513675402, | |
| "loss": 3.2917, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01147548219093437, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00029999471559420324, | |
| "loss": 3.2102, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.012358211590237013, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00029999060554370847, | |
| "loss": 3.1378, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.013240940989539656, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00029998532124823267, | |
| "loss": 3.0807, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0141236703888423, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00029997886274914557, | |
| "loss": 3.0547, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.015006399788144944, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002999712300970092, | |
| "loss": 3.0153, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.015889129187447587, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0002999624233515781, | |
| "loss": 2.9845, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.016771858586750232, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00029995244258179844, | |
| "loss": 2.9619, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.017654587986052874, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 0.00029994128786580765, | |
| "loss": 2.9415, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01853731738535552, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0002999289592909335, | |
| "loss": 2.9384, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.019420046784658165, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0002999154569536939, | |
| "loss": 2.8946, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.020302776183960806, | |
| "grad_norm": 1.625, | |
| "learning_rate": 0.00029990078095979577, | |
| "loss": 2.8861, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.02118550558326345, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.0002998849314241344, | |
| "loss": 2.8779, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.022068234982566094, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.0002998679084707926, | |
| "loss": 2.8641, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02295096438186874, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029984971223303935, | |
| "loss": 2.8526, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02383369378117138, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00029983034285332923, | |
| "loss": 2.8404, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.024716423180474026, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.000299809800483301, | |
| "loss": 2.8211, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02559915257977667, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00029978808528377673, | |
| "loss": 2.8257, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.026481881979079313, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00029976519742476, | |
| "loss": 2.8099, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.027364611378381958, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00029974113708543503, | |
| "loss": 2.8003, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.0282473407776846, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00029971590445416525, | |
| "loss": 2.8083, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.029130070176987245, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0002996894997284917, | |
| "loss": 2.7965, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.030012799576289887, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.0002996619231151313, | |
| "loss": 2.8107, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.030895528975592532, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0002996331748299757, | |
| "loss": 2.8088, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.031778258374895174, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00029960325509808904, | |
| "loss": 2.7825, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.03266098777419782, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00029957216415370675, | |
| "loss": 2.7949, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.033543717173500465, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.0002995399022402333, | |
| "loss": 2.7735, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.03442644657280311, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00029950646961024034, | |
| "loss": 2.7799, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.03530917597210575, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00029947186652546504, | |
| "loss": 2.7506, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.036191905371408394, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00029943609325680764, | |
| "loss": 2.7706, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.03707463477071104, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00029939915008432946, | |
| "loss": 2.7553, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.037957364170013684, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00029936103729725106, | |
| "loss": 2.7546, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.03884009356931633, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0002993217551939492, | |
| "loss": 2.7679, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.03972282296861897, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00029928130408195545, | |
| "loss": 2.7471, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.04060555236792161, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.000299239684277953, | |
| "loss": 2.7374, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.04148828176722426, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.0002991968961077745, | |
| "loss": 2.7535, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.0423710111665269, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0002991529399063997, | |
| "loss": 2.7307, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.04325374056582954, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.0002991078160179523, | |
| "loss": 2.7239, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.04413646996513219, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00029906152479569777, | |
| "loss": 2.718, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04501919936443483, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0002990140666020405, | |
| "loss": 2.7328, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.04590192876373748, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0002989654418085207, | |
| "loss": 2.7221, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.04678465816304012, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.0002989156507958117, | |
| "loss": 2.7349, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.04766738756234276, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.00029886469395371673, | |
| "loss": 2.7242, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.048550116961645406, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.0002988125716811664, | |
| "loss": 2.7256, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.04943284636094805, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00029875928438621494, | |
| "loss": 2.7113, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.0503155757602507, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00029870483248603747, | |
| "loss": 2.7249, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.05119830515955334, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002986492164069264, | |
| "loss": 2.7274, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.05208103455885598, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0002985924365842885, | |
| "loss": 2.7165, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.052963763958158626, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00029853449346264096, | |
| "loss": 2.724, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.05384649335746127, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 0.00029847538749560833, | |
| "loss": 2.7059, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.054729222756763916, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00029841511914591883, | |
| "loss": 2.7108, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.055611952156066555, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00029835368888540077, | |
| "loss": 2.7205, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.0564946815553692, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00029829109719497874, | |
| "loss": 2.7139, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.057377410954671845, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00029822734456466996, | |
| "loss": 2.7014, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.05826014035397449, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0002981624314935804, | |
| "loss": 2.7063, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.059142869753277136, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0002980963584899009, | |
| "loss": 2.7028, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.060025599152579774, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.000298029126070903, | |
| "loss": 2.704, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.06090832855188242, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00029796073476293534, | |
| "loss": 2.6997, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.061791057951185065, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00029789118510141916, | |
| "loss": 2.6918, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0626737873504877, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002978204776308441, | |
| "loss": 2.705, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.06355651674979035, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.0002977486129047641, | |
| "loss": 2.6877, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.064439246149093, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.000297675591485793, | |
| "loss": 2.6792, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.06532197554839564, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00029760141394560017, | |
| "loss": 2.6927, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.06620470494769828, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00029752608086490613, | |
| "loss": 2.6883, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.06708743434700093, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.0002974495928334776, | |
| "loss": 2.7073, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.06797016374630357, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.00029737195045012344, | |
| "loss": 2.6876, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.06885289314560622, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00029729315432268957, | |
| "loss": 2.6725, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.06973562254490885, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002972132050680543, | |
| "loss": 2.6828, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.0706183519442115, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0002971321033121236, | |
| "loss": 2.6855, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.07150108134351414, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.000297049849689826, | |
| "loss": 2.69, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.07238381074281679, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 0.000296966444845108, | |
| "loss": 2.69, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.07326654014211943, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.0002968818894309285, | |
| "loss": 2.6839, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.07414926954142208, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00029679618410925426, | |
| "loss": 2.6771, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.07503199894072472, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.0002967093295510541, | |
| "loss": 2.6793, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.07591472834002737, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.00029662132643629423, | |
| "loss": 2.6858, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.07679745773933001, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00029653217545393254, | |
| "loss": 2.6784, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.07768018713863266, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00029644187730191334, | |
| "loss": 2.6795, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.07856291653793529, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00029635043268716194, | |
| "loss": 2.6725, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.07944564593723794, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00029625784232557904, | |
| "loss": 2.6658, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.08032837533654058, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00029616410694203516, | |
| "loss": 2.6789, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.08121110473584323, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.000296069227270365, | |
| "loss": 2.6772, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.08209383413514587, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002959732040533615, | |
| "loss": 2.6559, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.08297656353444852, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00029587603804277036, | |
| "loss": 2.6646, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.08385929293375116, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00029577772999928384, | |
| "loss": 2.6766, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.0847420223330538, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00029567828069253504, | |
| "loss": 2.674, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.08562475173235645, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00029557769090109165, | |
| "loss": 2.6703, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.08650748113165908, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0002954759614124501, | |
| "loss": 2.6708, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.08739021053096173, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0002953730930230292, | |
| "loss": 2.6539, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.08827293993026437, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0002952690865381639, | |
| "loss": 2.6705, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.08915566932956702, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 0.0002951639427720993, | |
| "loss": 2.6647, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.09003839872886966, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00029505766254798375, | |
| "loss": 2.66, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.09092112812817231, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002949502466978629, | |
| "loss": 2.6782, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.09180385752747496, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.000294841696062673, | |
| "loss": 2.6625, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.0926865869267776, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.00029473201149223395, | |
| "loss": 2.6778, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.09356931632608025, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0002946211938452434, | |
| "loss": 2.6638, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.09445204572538288, | |
| "grad_norm": 14.375, | |
| "learning_rate": 0.00029450924398926947, | |
| "loss": 2.669, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.09533477512468552, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00029439616280074407, | |
| "loss": 2.664, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.09621750452398817, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00029428195116495607, | |
| "loss": 2.6569, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.09710023392329081, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002941666099760444, | |
| "loss": 2.6622, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.09798296332259346, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00029405014013699106, | |
| "loss": 2.6524, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.0988656927218961, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00029393254255961394, | |
| "loss": 2.6551, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.09974842212119875, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.0002938138181645599, | |
| "loss": 2.6588, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.1006311515205014, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.00029369396788129717, | |
| "loss": 2.6663, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.10151388091980404, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0002935729926481086, | |
| "loss": 2.6635, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.10239661031910668, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0002934508934120839, | |
| "loss": 2.6799, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.10327933971840932, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00029332767112911226, | |
| "loss": 2.6464, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.10416206911771196, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00029320332676387515, | |
| "loss": 2.6563, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.1050447985170146, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.0002930778612898386, | |
| "loss": 2.6616, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.10592752791631725, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0002929512756892453, | |
| "loss": 2.6704, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.1068102573156199, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002928235709531075, | |
| "loss": 2.6715, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.10769298671492254, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00029269474808119874, | |
| "loss": 2.6581, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.10857571611422519, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00029256480808204625, | |
| "loss": 2.6485, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.10945844551352783, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.00029243375197292304, | |
| "loss": 2.6608, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.11034117491283048, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00029230158077983993, | |
| "loss": 2.652, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.11122390431213311, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00029216829553753735, | |
| "loss": 2.6633, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.11210663371143575, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00029203389728947756, | |
| "loss": 2.6567, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.1129893631107384, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00029189838708783625, | |
| "loss": 2.6433, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.11387209251004105, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00029176176599349436, | |
| "loss": 2.6554, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.11475482190934369, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00029162403507602974, | |
| "loss": 2.6621, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.11563755130864634, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0002914851954137089, | |
| "loss": 2.6466, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.11652028070794898, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.0002913452480934784, | |
| "loss": 2.6575, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.11740301010725163, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00029120419421095644, | |
| "loss": 2.6493, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.11828573950655427, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00029106203487042423, | |
| "loss": 2.6555, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.1191684689058569, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0002909187711848175, | |
| "loss": 2.6565, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.12005119830515955, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0002907744042757175, | |
| "loss": 2.6348, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.1209339277044622, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002906289352733426, | |
| "loss": 2.6356, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.12181665710376484, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.000290482365316539, | |
| "loss": 2.649, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.12269938650306748, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0002903346955527721, | |
| "loss": 2.6384, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.12358211590237013, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.0002901859271381176, | |
| "loss": 2.6474, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.12446484530167277, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00029003606123725217, | |
| "loss": 2.6363, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.1253475747009754, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0002898850990234445, | |
| "loss": 2.6352, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.12623030410027805, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00028973304167854606, | |
| "loss": 2.6277, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.1271130334995807, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.000289579890392982, | |
| "loss": 2.6459, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.12799576289888334, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.0002894256463657414, | |
| "loss": 2.6389, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.128878492298186, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002892703108043686, | |
| "loss": 2.6463, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.12976122169748863, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00028911388492495305, | |
| "loss": 2.6326, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.13064395109679128, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00028895636995212003, | |
| "loss": 2.6475, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.13152668049609392, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.0002887977671190214, | |
| "loss": 2.6531, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.13240940989539657, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00028863807766732534, | |
| "loss": 2.6387, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1332921392946992, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.0002884773028472071, | |
| "loss": 2.6455, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.13417486869400186, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00028831544391733896, | |
| "loss": 2.6414, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.1350575980933045, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.00028815250214488065, | |
| "loss": 2.641, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.13594032749260715, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00028798847880546894, | |
| "loss": 2.6405, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.1368230568919098, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002878233751832083, | |
| "loss": 2.6463, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.13770578629121244, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.0002876571925706603, | |
| "loss": 2.638, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.13858851569051509, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0002874899322688337, | |
| "loss": 2.6433, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.1394712450898177, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00028732159558717436, | |
| "loss": 2.6306, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.14035397448912035, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0002871521838435548, | |
| "loss": 2.6424, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.141236703888423, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002869816983642641, | |
| "loss": 2.6376, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.14211943328772564, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 0.0002868101404839972, | |
| "loss": 2.6285, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.14300216268702828, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00028663751154584476, | |
| "loss": 2.6356, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.14388489208633093, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.0002864638129012826, | |
| "loss": 2.6422, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.14476762148563357, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00028628904591016094, | |
| "loss": 2.6375, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.14565035088493622, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0002861132119406939, | |
| "loss": 2.6285, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.14653308028423886, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00028593631236944865, | |
| "loss": 2.6344, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.1474158096835415, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00028575834858133486, | |
| "loss": 2.6449, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.14829853908284416, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.0002855793219695937, | |
| "loss": 2.6217, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.1491812684821468, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.000285399233935787, | |
| "loss": 2.6386, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.15006399788144945, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00028521808588978607, | |
| "loss": 2.6391, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1509467272807521, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00028503587924976105, | |
| "loss": 2.6406, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.15182945668005474, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.0002848526154421695, | |
| "loss": 2.6373, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.15271218607935738, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00028466829590174524, | |
| "loss": 2.634, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.15359491547866003, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002844829220714873, | |
| "loss": 2.6369, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.15447764487796267, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00028429649540264866, | |
| "loss": 2.6264, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.15536037427726532, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0002841090173547244, | |
| "loss": 2.6287, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.15624310367656793, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00028392048939544084, | |
| "loss": 2.6445, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.15712583307587058, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.0002837309130007439, | |
| "loss": 2.6431, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.15800856247517323, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0002835402896547874, | |
| "loss": 2.631, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.15889129187447587, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0002833486208499215, | |
| "loss": 2.6255, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.15977402127377852, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.000283155908086681, | |
| "loss": 2.6332, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.16065675067308116, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0002829621528737738, | |
| "loss": 2.6187, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.1615394800723838, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.00028276735672806865, | |
| "loss": 2.6424, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.16242220947168645, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00028257152117458387, | |
| "loss": 2.6333, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.1633049388709891, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.0002823746477464748, | |
| "loss": 2.6279, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.16418766827029174, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0002821767379850223, | |
| "loss": 2.6298, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.1650703976695944, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002819777934396203, | |
| "loss": 2.631, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.16595312706889703, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00028177781566776395, | |
| "loss": 2.6396, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.16683585646819968, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00028157680623503724, | |
| "loss": 2.6374, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.16771858586750232, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00028137476671510084, | |
| "loss": 2.6368, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.16860131526680497, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.00028117169868967977, | |
| "loss": 2.6431, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.1694840446661076, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.000280967603748551, | |
| "loss": 2.6362, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.17036677406541026, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.000280762483489531, | |
| "loss": 2.6294, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.1712495034647129, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00028055633951846326, | |
| "loss": 2.6319, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.17213223286401555, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.0002803491734492056, | |
| "loss": 2.6329, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.17301496226331817, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0002801409869036178, | |
| "loss": 2.6246, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.1738976916626208, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00027993178151154865, | |
| "loss": 2.6275, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.17478042106192346, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.0002797215589108232, | |
| "loss": 2.6286, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.1756631504612261, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0002795103207472301, | |
| "loss": 2.6312, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.17654587986052875, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.00027929806867450866, | |
| "loss": 2.6497, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.17654587986052875, | |
| "eval_accuracy": 0.4939203354297694, | |
| "eval_loss": 2.517970323562622, | |
| "eval_runtime": 6.889, | |
| "eval_samples_per_second": 46.16, | |
| "eval_steps_per_second": 0.435, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1774286092598314, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.00027908480435433577, | |
| "loss": 2.6208, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.17831133865913404, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.0002788705294563131, | |
| "loss": 2.6338, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.17919406805843668, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.0002786552456579539, | |
| "loss": 2.6267, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.18007679745773933, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00027843895464466973, | |
| "loss": 2.6156, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.18095952685704197, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0002782216581097576, | |
| "loss": 2.6295, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.18184225625634462, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00027800335775438653, | |
| "loss": 2.63, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.18272498565564727, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00027778405528758424, | |
| "loss": 2.6245, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.1836077150549499, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.0002775637524262236, | |
| "loss": 2.6392, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.18449044445425256, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.0002773424508950096, | |
| "loss": 2.6146, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.1853731738535552, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0002771201524264655, | |
| "loss": 2.6128, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.18625590325285785, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.00027689685876091925, | |
| "loss": 2.6248, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.1871386326521605, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00027667257164649023, | |
| "loss": 2.6279, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.18802136205146314, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0002764472928390751, | |
| "loss": 2.6274, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.18890409145076575, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00027622102410233455, | |
| "loss": 2.6296, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.1897868208500684, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.0002759937672076789, | |
| "loss": 2.6125, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.19066955024937104, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.0002757655239342547, | |
| "loss": 2.6157, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.1915522796486737, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0002755362960689306, | |
| "loss": 2.6042, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.19243500904797634, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00027530608540628367, | |
| "loss": 2.6159, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.19331773844727898, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.0002750748937485847, | |
| "loss": 2.6407, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.19420046784658163, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00027484272290578474, | |
| "loss": 2.6275, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.19508319724588427, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002746095746955006, | |
| "loss": 2.6112, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.19596592664518692, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002743754509430007, | |
| "loss": 2.6232, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.19684865604448956, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.00027414035348119074, | |
| "loss": 2.6257, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.1977313854437922, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0002739042841505995, | |
| "loss": 2.6136, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.19861411484309485, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00027366724479936416, | |
| "loss": 2.6288, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.1994968442423975, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00027342923728321613, | |
| "loss": 2.6155, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.20037957364170014, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00027319026346546627, | |
| "loss": 2.6276, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.2012623030410028, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0002729503252169905, | |
| "loss": 2.6128, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.20214503244030543, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.000272709424416215, | |
| "loss": 2.6296, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.20302776183960808, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.0002724675629491016, | |
| "loss": 2.6403, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.20391049123891072, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.00027222474270913297, | |
| "loss": 2.6113, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.20479322063821337, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.0002719809655972979, | |
| "loss": 2.6046, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.205675950037516, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00027173623352207604, | |
| "loss": 2.6251, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.20655867943681863, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00027149054839942374, | |
| "loss": 2.6284, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.20744140883612128, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.000271243912152758, | |
| "loss": 2.6155, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.20832413823542392, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0002709963267129425, | |
| "loss": 2.6159, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.20920686763472657, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002707477940182716, | |
| "loss": 2.613, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.2100895970340292, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.0002704983160144556, | |
| "loss": 2.6217, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.21097232643333186, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00027024789465460543, | |
| "loss": 2.6239, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.2118550558326345, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0002699965318992174, | |
| "loss": 2.6215, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.21273778523193715, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0002697442297161578, | |
| "loss": 2.6324, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.2136205146312398, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0002694909900806475, | |
| "loss": 2.6237, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.21450324403054244, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0002692368149752464, | |
| "loss": 2.6295, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.21538597342984508, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00026898170638983823, | |
| "loss": 2.6259, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.21626870282914773, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00026872566632161435, | |
| "loss": 2.6244, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.21715143222845038, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00026846869677505893, | |
| "loss": 2.6245, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.21803416162775302, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00026821079976193247, | |
| "loss": 2.622, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.21891689102705567, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.0002679519773012568, | |
| "loss": 2.6229, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.2197996204263583, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.0002676922314192984, | |
| "loss": 2.6047, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.22068234982566096, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00026743156414955346, | |
| "loss": 2.6234, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.22156507922496357, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0002671699775327313, | |
| "loss": 2.6256, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.22244780862426622, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00026690747361673864, | |
| "loss": 2.6223, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.22333053802356886, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.0002666440544566636, | |
| "loss": 2.6169, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.2242132674228715, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00026637972211475945, | |
| "loss": 2.6141, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.22509599682217415, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.0002661144786604287, | |
| "loss": 2.6062, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.2259787262214768, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.00026584832617020663, | |
| "loss": 2.6223, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.22686145562077945, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00026558126672774534, | |
| "loss": 2.6163, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.2277441850200821, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00026531330242379714, | |
| "loss": 2.6112, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.22862691441938474, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.0002650444353561983, | |
| "loss": 2.6084, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.22950964381868738, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00026477466762985267, | |
| "loss": 2.6233, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.23039237321799003, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00026450400135671524, | |
| "loss": 2.606, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.23127510261729267, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0002642324386557755, | |
| "loss": 2.6044, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.23215783201659532, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002639599816530407, | |
| "loss": 2.6028, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.23304056141589796, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.0002636866324815198, | |
| "loss": 2.6186, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.2339232908152006, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.00026341239328120586, | |
| "loss": 2.6127, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.23480602021450325, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00026313726619906023, | |
| "loss": 2.607, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.2356887496138059, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.000262861253388995, | |
| "loss": 2.6007, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.23657147901310854, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00026258435701185655, | |
| "loss": 2.622, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.2374542084124112, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00026230657923540857, | |
| "loss": 2.6156, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.2383369378117138, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.000262027922234315, | |
| "loss": 2.6091, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.23921966721101645, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.000261748388190123, | |
| "loss": 2.6141, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.2401023966103191, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00026146797929124595, | |
| "loss": 2.6105, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.24098512600962174, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.0002611866977329463, | |
| "loss": 2.6185, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.2418678554089244, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.0002609045457173184, | |
| "loss": 2.6201, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.24275058480822703, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00026062152545327106, | |
| "loss": 2.5987, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.24363331420752968, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0002603376391565106, | |
| "loss": 2.6092, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.24451604360683232, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.0002600528890495234, | |
| "loss": 2.5916, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.24539877300613497, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.00025976727736155814, | |
| "loss": 2.6037, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.2462815024054376, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00025948080632860884, | |
| "loss": 2.6151, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.24716423180474026, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00025919347819339716, | |
| "loss": 2.6018, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.2480469612040429, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00025890529520535477, | |
| "loss": 2.6175, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.24892969060334555, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0002586162596206058, | |
| "loss": 2.6189, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.2498124200026482, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.0002583263737019492, | |
| "loss": 2.6085, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.2506951494019508, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.000258035639718841, | |
| "loss": 2.6078, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.2515778788012535, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0002577440599473766, | |
| "loss": 2.5885, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2524606082005561, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.0002574516366702728, | |
| "loss": 2.6067, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.2533433375998588, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0002571583721768501, | |
| "loss": 2.6125, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.2542260669991614, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00025686426876301465, | |
| "loss": 2.6029, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.25510879639846407, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0002565693287312405, | |
| "loss": 2.6327, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.2559915257977667, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.0002562735543905511, | |
| "loss": 2.6102, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.25687425519706936, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.0002559769480565019, | |
| "loss": 2.6051, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.257756984596372, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0002556795120511615, | |
| "loss": 2.6248, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.25863971399567465, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.000255381248703094, | |
| "loss": 2.5961, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.25952244339497726, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0002550821603473406, | |
| "loss": 2.6216, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.26040517279427994, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002547822493254013, | |
| "loss": 2.6258, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.26128790219358256, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00025448151798521665, | |
| "loss": 2.6331, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.26217063159288523, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00025417996868114907, | |
| "loss": 2.6173, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.26305336099218785, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.0002538776037739648, | |
| "loss": 2.6219, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.26393609039149046, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0002535744256308153, | |
| "loss": 2.6108, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.26481881979079314, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0002532704366252187, | |
| "loss": 2.6182, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.26570154919009575, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002529656391370411, | |
| "loss": 2.6089, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.2665842785893984, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00025266003555247805, | |
| "loss": 2.6119, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.26746700798870104, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.00025235362826403593, | |
| "loss": 2.6085, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.2683497373880037, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.0002520464196705131, | |
| "loss": 2.6002, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.26923246678730633, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0002517384121769812, | |
| "loss": 2.5967, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.270115196186609, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00025142960819476626, | |
| "loss": 2.6124, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.2709979255859116, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0002511200101414298, | |
| "loss": 2.5974, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.2718806549852143, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00025080962044075016, | |
| "loss": 2.6028, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.2727633843845169, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.0002504984415227031, | |
| "loss": 2.6069, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.2736461137838196, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.0002501864758234431, | |
| "loss": 2.6175, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.2745288431831222, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00024987372578528415, | |
| "loss": 2.6103, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.2754115725824249, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.0002495601938566807, | |
| "loss": 2.6319, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.2762943019817275, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00024924588249220837, | |
| "loss": 2.6215, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.27717703138103017, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.000248930794152545, | |
| "loss": 2.6246, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.2780597607803328, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.000248614931304451, | |
| "loss": 2.6115, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.2789424901796354, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.0002482982964207504, | |
| "loss": 2.6173, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.2798252195789381, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00024798089198031125, | |
| "loss": 2.6188, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.2807079489782407, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0002476627204680263, | |
| "loss": 2.5963, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.28159067837754337, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00024734378437479354, | |
| "loss": 2.6178, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.282473407776846, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00024702408619749677, | |
| "loss": 2.6126, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.28335613717614866, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00024670362843898594, | |
| "loss": 2.5998, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.2842388665754513, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0002463824136080576, | |
| "loss": 2.607, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.28512159597475395, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00024606044421943526, | |
| "loss": 2.6125, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.28600432537405657, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0002457377227937497, | |
| "loss": 2.5971, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.28688705477335924, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.00024541425185751933, | |
| "loss": 2.6118, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.28776978417266186, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.00024509003394313015, | |
| "loss": 2.6107, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.28865251357196453, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.0002447650715888162, | |
| "loss": 2.6028, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.28953524297126715, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.0002444393673386396, | |
| "loss": 2.6225, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.2904179723705698, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00024411292374247056, | |
| "loss": 2.6139, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.29130070176987244, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00024378574335596752, | |
| "loss": 2.6059, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.2921834311691751, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002434578287405571, | |
| "loss": 2.6034, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.29306616056847773, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00024312918246341403, | |
| "loss": 2.6263, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.2939488899677804, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00024279980709744097, | |
| "loss": 2.6106, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.294831619367083, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.00024246970522124865, | |
| "loss": 2.5977, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.29571434876638564, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00024213887941913536, | |
| "loss": 2.6201, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.2965970781656883, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0002418073322810669, | |
| "loss": 2.6075, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.29747980756499093, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00024147506640265624, | |
| "loss": 2.6012, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.2983625369642936, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00024114208438514323, | |
| "loss": 2.6081, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.2992452663635962, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00024080838883537426, | |
| "loss": 2.6042, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.3001279957628989, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00024047398236578174, | |
| "loss": 2.6035, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3010107251622015, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.0002401388675943637, | |
| "loss": 2.5915, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.3018934545615042, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00023980304714466355, | |
| "loss": 2.623, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.3027761839608068, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00023946652364574894, | |
| "loss": 2.6025, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.3036589133601095, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00023912929973219187, | |
| "loss": 2.6203, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.3045416427594121, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00023879137804404753, | |
| "loss": 2.6022, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.30542437215871476, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00023845276122683395, | |
| "loss": 2.6081, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.3063071015580174, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0002381134519315111, | |
| "loss": 2.6173, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.30718983095732005, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.00023777345281446031, | |
| "loss": 2.6129, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.30807256035662267, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.0002374327665374634, | |
| "loss": 2.6077, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.30895528975592534, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00023709139576768155, | |
| "loss": 2.6067, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.30983801915522796, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0002367493431776351, | |
| "loss": 2.6163, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.31072074855453063, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0002364066114451819, | |
| "loss": 2.6097, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.31160347795383325, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0002360632032534968, | |
| "loss": 2.6074, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.31248620735313587, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00023571912129105048, | |
| "loss": 2.5909, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.31336893675243854, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00023537436825158843, | |
| "loss": 2.6162, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.31425166615174116, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00023502894683410992, | |
| "loss": 2.6105, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.31513439555104383, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00023468285974284663, | |
| "loss": 2.6269, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.31601712495034645, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 0.00023433610968724187, | |
| "loss": 2.5995, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.3168998543496491, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00023398869938192908, | |
| "loss": 2.6132, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.31778258374895174, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.00023364063154671062, | |
| "loss": 2.6204, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.3186653131482544, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00023329190890653657, | |
| "loss": 2.6149, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.31954804254755703, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00023294253419148324, | |
| "loss": 2.6035, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.3204307719468597, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.0002325925101367321, | |
| "loss": 2.6204, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.3213135013461623, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0002322418394825479, | |
| "loss": 2.6171, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.322196230745465, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00023189052497425767, | |
| "loss": 2.6158, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.3230789601447676, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00023153856936222907, | |
| "loss": 2.6138, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.3239616895440703, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0002311859754018486, | |
| "loss": 2.5937, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.3248444189433729, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.0002308327458535005, | |
| "loss": 2.605, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.3257271483426756, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00023047888348254474, | |
| "loss": 2.6208, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.3266098777419782, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00023012439105929563, | |
| "loss": 2.6216, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.32749260714128087, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.0002297692713589999, | |
| "loss": 2.6049, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.3283753365405835, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00022941352716181524, | |
| "loss": 2.6095, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.3292580659398861, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.0002290571612527883, | |
| "loss": 2.6131, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.3301407953391888, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.00022870017642183298, | |
| "loss": 2.6189, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.3310235247384914, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0002283425754637087, | |
| "loss": 2.6052, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.33190625413779407, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00022798436117799832, | |
| "loss": 2.6156, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.3327889835370967, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.00022762553636908626, | |
| "loss": 2.595, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.33367171293639936, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00022726610384613666, | |
| "loss": 2.6064, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.334554442335702, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.0002269060664230714, | |
| "loss": 2.6054, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.33543717173500465, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.0002265454269185479, | |
| "loss": 2.6095, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.33631990113430726, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00022618418815593722, | |
| "loss": 2.6084, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.33720263053360994, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00022582235296330177, | |
| "loss": 2.5958, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.33808535993291255, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00022545992417337328, | |
| "loss": 2.6117, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.3389680893322152, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00022509690462353088, | |
| "loss": 2.5986, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.33985081873151785, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00022473329715577838, | |
| "loss": 2.6048, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.3407335481308205, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00022436910461672236, | |
| "loss": 2.6027, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.34161627753012314, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00022400432985754978, | |
| "loss": 2.5819, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.3424990069294258, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0002236389757340057, | |
| "loss": 2.6089, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.3433817363287284, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00022327304510637094, | |
| "loss": 2.5947, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.3442644657280311, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00022290654083943951, | |
| "loss": 2.6035, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.3451471951273337, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00022253946580249654, | |
| "loss": 2.6094, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.34602992452663633, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.0002221718228692954, | |
| "loss": 2.6152, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.346912653925939, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.0002218036149180355, | |
| "loss": 2.6111, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.3477953833252416, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00022143484483133966, | |
| "loss": 2.6132, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.3486781127245443, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00022106551549623145, | |
| "loss": 2.5919, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.3495608421238469, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0002206956298041128, | |
| "loss": 2.6089, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.3504435715231496, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00022032519065074118, | |
| "loss": 2.6031, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.3513263009224522, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.000219954200936207, | |
| "loss": 2.6035, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.3522090303217549, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00021958266356491088, | |
| "loss": 2.6095, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.3530917597210575, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00021921058144554092, | |
| "loss": 2.6174, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3530917597210575, | |
| "eval_accuracy": 0.4966042343400834, | |
| "eval_loss": 2.4940457344055176, | |
| "eval_runtime": 6.9346, | |
| "eval_samples_per_second": 45.857, | |
| "eval_steps_per_second": 0.433, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.35397448912036017, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.0002188379574910501, | |
| "loss": 2.5982, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.3548572185196628, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.000218464794618633, | |
| "loss": 2.6218, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.35573994791896546, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00021809109574970356, | |
| "loss": 2.6048, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.3566226773182681, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00021771686380987176, | |
| "loss": 2.6039, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.35750540671757075, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00021734210172892106, | |
| "loss": 2.607, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.35838813611687337, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.000216966812440785, | |
| "loss": 2.5955, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.35927086551617604, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00021659099888352473, | |
| "loss": 2.5995, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.36015359491547866, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00021621466399930582, | |
| "loss": 2.5947, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.3610363243147813, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.000215837810734375, | |
| "loss": 2.5966, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.36191905371408395, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00021546044203903756, | |
| "loss": 2.5998, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.36280178311338657, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00021508256086763368, | |
| "loss": 2.5955, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.36368451251268924, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0002147041701785159, | |
| "loss": 2.5944, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.36456724191199186, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00021432527293402544, | |
| "loss": 2.5891, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.36544997131129453, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00021394587210046937, | |
| "loss": 2.6239, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.36633270071059715, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.00021356597064809725, | |
| "loss": 2.6111, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.3672154301098998, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00021318557155107772, | |
| "loss": 2.6024, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.36809815950920244, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00021280467778747562, | |
| "loss": 2.5955, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.3689808889085051, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00021242329233922823, | |
| "loss": 2.5902, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.36986361830780773, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00021204141819212231, | |
| "loss": 2.6004, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.3707463477071104, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00021165905833577035, | |
| "loss": 2.6114, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.371629077106413, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.00021127621576358743, | |
| "loss": 2.6061, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.3725118065057157, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0002108928934727678, | |
| "loss": 2.5979, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.3733945359050183, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00021050909446426117, | |
| "loss": 2.5945, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.374277265304321, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00021012482174274954, | |
| "loss": 2.6105, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.3751599947036236, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00020974007831662332, | |
| "loss": 2.6202, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.3760427241029263, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0002093548671979581, | |
| "loss": 2.5993, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.3769254535022289, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00020896919140249093, | |
| "loss": 2.5922, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.3778081829015315, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00020858305394959672, | |
| "loss": 2.6189, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.3786909123008342, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00020819645786226458, | |
| "loss": 2.6111, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.3795736417001368, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00020780940616707406, | |
| "loss": 2.6019, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.38045637109943947, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00020742190189417174, | |
| "loss": 2.5998, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.3813391004987421, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00020703394807724726, | |
| "loss": 2.6115, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.38222182989804476, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0002066455477535097, | |
| "loss": 2.6019, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.3831045592973474, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.00020625670396366366, | |
| "loss": 2.5943, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.38398728869665005, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00020586741975188558, | |
| "loss": 2.6045, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.38487001809595267, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00020547769816579988, | |
| "loss": 2.5918, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.38575274749525534, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00020508754225645513, | |
| "loss": 2.5942, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.38663547689455796, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00020469695507830004, | |
| "loss": 2.6249, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.38751820629386063, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.0002043059396891597, | |
| "loss": 2.6117, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.38840093569316325, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0002039144991502116, | |
| "loss": 2.5974, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.3892836650924659, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00020352263652596148, | |
| "loss": 2.6114, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.39016639449176854, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00020313035488421963, | |
| "loss": 2.6044, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.3910491238910712, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 0.00020273765729607684, | |
| "loss": 2.5984, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.39193185329037383, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00020234454683588002, | |
| "loss": 2.5931, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.3928145826896765, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0002019510265812085, | |
| "loss": 2.5915, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.3936973120889791, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00020155709961284978, | |
| "loss": 2.6093, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.39458004148828174, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00020116276901477548, | |
| "loss": 2.6098, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.3954627708875844, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.00020076803787411716, | |
| "loss": 2.6025, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.39634550028688703, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00020037290928114196, | |
| "loss": 2.598, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.3972282296861897, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00019997738632922886, | |
| "loss": 2.5909, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.3981109590854923, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00019958147211484402, | |
| "loss": 2.6027, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.398993688484795, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.00019918516973751674, | |
| "loss": 2.6077, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.3998764178840976, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.00019878848229981522, | |
| "loss": 2.6142, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.4007591472834003, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00019839141290732215, | |
| "loss": 2.624, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.4016418766827029, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.0001979939646686105, | |
| "loss": 2.6037, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.4025246060820056, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00019759614069521914, | |
| "loss": 2.5959, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.4034073354813082, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00019719794410162853, | |
| "loss": 2.5839, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.40429006488061087, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00019679937800523624, | |
| "loss": 2.6024, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.4051727942799135, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00019640044552633257, | |
| "loss": 2.6189, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.40605552367921616, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00019600114978807618, | |
| "loss": 2.6165, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.4069382530785188, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00019560149391646976, | |
| "loss": 2.6029, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.40782098247782145, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00019520148104033513, | |
| "loss": 2.5895, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.40870371187712407, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00019480111429128922, | |
| "loss": 2.6012, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.40958644127642674, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00019440039680371934, | |
| "loss": 2.6103, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.41046917067572936, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0001939993317147586, | |
| "loss": 2.6065, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.411351900075032, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.0001935979221642614, | |
| "loss": 2.6046, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.41223462947433465, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00019319617129477898, | |
| "loss": 2.608, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.41311735887363726, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00019279408225153453, | |
| "loss": 2.6015, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.41400008827293994, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0001923916581823988, | |
| "loss": 2.5999, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.41488281767224255, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00019198890223786542, | |
| "loss": 2.6072, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.4157655470715452, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019158581757102618, | |
| "loss": 2.5948, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.41664827647084784, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00019118240733754642, | |
| "loss": 2.5934, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.4175310058701505, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0001907786746956401, | |
| "loss": 2.5961, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.41841373526945314, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00019037462280604544, | |
| "loss": 2.5949, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.4192964646687558, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00018997025483199987, | |
| "loss": 2.6071, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.4201791940680584, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00018956557393921545, | |
| "loss": 2.5987, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.4210619234673611, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.000189160583295854, | |
| "loss": 2.5984, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.4219446528666637, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00018875528607250225, | |
| "loss": 2.5762, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.4228273822659664, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00018834968544214717, | |
| "loss": 2.5946, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.423710111665269, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00018794378458015094, | |
| "loss": 2.6053, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4245928410645717, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.0001875375866642263, | |
| "loss": 2.5866, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.4254755704638743, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.0001871310948744115, | |
| "loss": 2.6037, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.42635829986317697, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0001867243123930454, | |
| "loss": 2.5994, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.4272410292624796, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00018631724240474276, | |
| "loss": 2.5731, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.4281237586617822, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00018590988809636918, | |
| "loss": 2.6008, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.4290064880610849, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00018550225265701596, | |
| "loss": 2.6024, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.4298892174603875, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0001850943392779755, | |
| "loss": 2.59, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.43077194685969017, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00018468615115271598, | |
| "loss": 2.578, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.4316546762589928, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.0001842776914768567, | |
| "loss": 2.6038, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.43253740565829546, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00018386896344814265, | |
| "loss": 2.6159, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.4334201350575981, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.0001834599702664199, | |
| "loss": 2.5928, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.43430286445690075, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001830507151336102, | |
| "loss": 2.6126, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.43518559385620337, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00018264120125368615, | |
| "loss": 2.6007, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.43606832325550604, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.000182231431832646, | |
| "loss": 2.6017, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.43695105265480866, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.0001818214100784885, | |
| "loss": 2.591, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.43783378205411133, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.0001814111392011881, | |
| "loss": 2.5996, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.43871651145341395, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00018100062241266924, | |
| "loss": 2.6052, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.4395992408527166, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00018058986292678178, | |
| "loss": 2.5944, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.44048197025201924, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00018017886395927564, | |
| "loss": 2.5944, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.4413646996513219, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00017976762872777538, | |
| "loss": 2.6063, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.44224742905062453, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00017935616045175543, | |
| "loss": 2.5938, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.44313015844992715, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.00017894446235251456, | |
| "loss": 2.6043, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.4440128878492298, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.00017853253765315083, | |
| "loss": 2.5911, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.44489561724853244, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00017812038957853624, | |
| "loss": 2.6054, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.4457783466478351, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00017770802135529164, | |
| "loss": 2.6108, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.44666107604713773, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00017729543621176132, | |
| "loss": 2.5967, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.4475438054464404, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00017688263737798777, | |
| "loss": 2.6069, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.448426534845743, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00017646962808568652, | |
| "loss": 2.5937, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.4493092642450457, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00017605641156822057, | |
| "loss": 2.5942, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.4501919936443483, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001756429910605754, | |
| "loss": 2.5916, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.451074723043651, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00017522936979933346, | |
| "loss": 2.6076, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.4519574524429536, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00017481555102264872, | |
| "loss": 2.6115, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.4528401818422563, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00017440153797022165, | |
| "loss": 2.608, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.4537229112415589, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00017398733388327347, | |
| "loss": 2.5951, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.45460564064086156, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00017357294200452122, | |
| "loss": 2.5901, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.4554883700401642, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00017315836557815184, | |
| "loss": 2.604, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.45637109943946685, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00017274360784979727, | |
| "loss": 2.5979, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.45725382883876947, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.0001723286720665087, | |
| "loss": 2.5969, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.45813655823807214, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00017191356147673134, | |
| "loss": 2.6108, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.45901928763737476, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00017149827933027892, | |
| "loss": 2.5887, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.4599020170366774, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00017108282887830817, | |
| "loss": 2.5945, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.46078474643598005, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00017066721337329356, | |
| "loss": 2.5939, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.46166747583528267, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00017025143606900166, | |
| "loss": 2.5943, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.46255020523458534, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00016983550022046581, | |
| "loss": 2.6079, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.46343293463388796, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00016941940908396042, | |
| "loss": 2.5886, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.46431566403319063, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00016900316591697572, | |
| "loss": 2.6052, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.46519839343249325, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00016858677397819214, | |
| "loss": 2.6119, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.4660811228317959, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.0001681702365274548, | |
| "loss": 2.6098, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.46696385223109854, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00016775355682574803, | |
| "loss": 2.5927, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.4678465816304012, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0001673367381351698, | |
| "loss": 2.61, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.46872931102970383, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00016691978371890612, | |
| "loss": 2.5895, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.4696120404290065, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00016650269684120566, | |
| "loss": 2.6085, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.4704947698283091, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.0001660854807673541, | |
| "loss": 2.6033, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.4713774992276118, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00016566813876364858, | |
| "loss": 2.596, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.4722602286269144, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00016525067409737195, | |
| "loss": 2.5848, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.4731429580262171, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00016483309003676765, | |
| "loss": 2.6044, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.4740256874255197, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00016441538985101354, | |
| "loss": 2.5872, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.4749084168248224, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00016399757681019686, | |
| "loss": 2.5911, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.475791146224125, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0001635796541852882, | |
| "loss": 2.6005, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.4766738756234276, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00016316162524811605, | |
| "loss": 2.6079, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.4775566050227303, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00016274349327134128, | |
| "loss": 2.597, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.4784393344220329, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001623252615284314, | |
| "loss": 2.6009, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.4793220638213356, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00016190693329363505, | |
| "loss": 2.592, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.4802047932206382, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.00016148851184195616, | |
| "loss": 2.5821, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.48108752261994087, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.0001610700004491285, | |
| "loss": 2.6081, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.4819702520192435, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00016065140239158987, | |
| "loss": 2.601, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.48285298141854616, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00016023272094645673, | |
| "loss": 2.5955, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.4837357108178488, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00015981395939149823, | |
| "loss": 2.5929, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.48461844021715145, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00015939512100511077, | |
| "loss": 2.6101, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.48550116961645406, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00015897620906629218, | |
| "loss": 2.5993, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.48638389901575674, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00015855722685461607, | |
| "loss": 2.5878, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.48726662841505936, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00015813817765020636, | |
| "loss": 2.6051, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.48814935781436203, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0001577190647337113, | |
| "loss": 2.6019, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.48903208721366465, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.000157299891386278, | |
| "loss": 2.5948, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.4899148166129673, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.0001568806608895266, | |
| "loss": 2.6066, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.49079754601226994, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00015646137652552473, | |
| "loss": 2.6193, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.4916802754115726, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00015604204157676176, | |
| "loss": 2.5749, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.4925630048108752, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00015562265932612298, | |
| "loss": 2.6002, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.49344573421017784, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.000155203233056864, | |
| "loss": 2.5942, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.4943284636094805, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.0001547837660525851, | |
| "loss": 2.6, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.49521119300878313, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00015436426159720553, | |
| "loss": 2.6046, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.4960939224080858, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00015394472297493753, | |
| "loss": 2.5969, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.4969766518073884, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00015352515347026103, | |
| "loss": 2.5947, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.4978593812066911, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00015310555636789767, | |
| "loss": 2.5921, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.4987421106059937, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.000152685934952785, | |
| "loss": 2.6147, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.4996248400052964, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0001522662925100512, | |
| "loss": 2.5908, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.5005075694045991, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00015184663232498878, | |
| "loss": 2.608, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.5013902988039016, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001514269576830294, | |
| "loss": 2.5922, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.5022730282032043, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00015100727186971762, | |
| "loss": 2.6042, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.503155757602507, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00015058757817068577, | |
| "loss": 2.5941, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5040384870018096, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00015016787987162767, | |
| "loss": 2.5957, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.5049212164011122, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00014974818025827332, | |
| "loss": 2.5909, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.5058039458004149, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00014932848261636287, | |
| "loss": 2.6006, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.5066866751997176, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.0001489087902316211, | |
| "loss": 2.603, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.5075694045990202, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.0001484891063897317, | |
| "loss": 2.5961, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.5084521339983228, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00014806943437631134, | |
| "loss": 2.602, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.5093348633976255, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00014764977747688422, | |
| "loss": 2.6007, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.5102175927969281, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 0.00014723013897685613, | |
| "loss": 2.5926, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.5111003221962307, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00014681052216148886, | |
| "loss": 2.601, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.5119830515955334, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00014639093031587432, | |
| "loss": 2.5952, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.512865780994836, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00014597136672490915, | |
| "loss": 2.6029, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.5137485103941387, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.0001455518346732687, | |
| "loss": 2.6121, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.5146312397934413, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0001451323374453812, | |
| "loss": 2.5955, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.515513969192744, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00014471287832540264, | |
| "loss": 2.589, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.5163966985920466, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00014429346059719033, | |
| "loss": 2.6054, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.5172794279913493, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00014387408754427776, | |
| "loss": 2.5854, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.5181621573906519, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00014345476244984845, | |
| "loss": 2.6012, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.5190448867899545, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00014303548859671069, | |
| "loss": 2.595, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.5199276161892572, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00014261626926727146, | |
| "loss": 2.5809, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.5208103455885599, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00014219710774351094, | |
| "loss": 2.5891, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.5216930749878624, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.00014177800730695678, | |
| "loss": 2.599, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.5225758043871651, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00014135897123865833, | |
| "loss": 2.5937, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.5234585337864678, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.0001409400028191611, | |
| "loss": 2.5942, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.5243412631857705, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.0001405211053284808, | |
| "loss": 2.59, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.525223992585073, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001401022820460782, | |
| "loss": 2.6084, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.5261067219843757, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00013968353625083279, | |
| "loss": 2.6112, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.5269894513836784, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00013926487122101753, | |
| "loss": 2.5961, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.5278721807829809, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00013884629023427314, | |
| "loss": 2.5982, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.5287549101822836, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00013842779656758234, | |
| "loss": 2.5956, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.5296376395815863, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00013800939349724426, | |
| "loss": 2.5972, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5296376395815863, | |
| "eval_accuracy": 0.49770467128957696, | |
| "eval_loss": 2.4865996837615967, | |
| "eval_runtime": 6.9231, | |
| "eval_samples_per_second": 45.933, | |
| "eval_steps_per_second": 0.433, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.530520368980889, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00013759108429884867, | |
| "loss": 2.5862, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.5314030983801915, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0001371728722472506, | |
| "loss": 2.5956, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.5322858277794942, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.0001367547606165444, | |
| "loss": 2.5962, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.5331685571787969, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.0001363367526800383, | |
| "loss": 2.5976, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.5340512865780995, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00013591885171022886, | |
| "loss": 2.6077, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.5349340159774021, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00013550106097877496, | |
| "loss": 2.5817, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.5358167453767048, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.0001350833837564726, | |
| "loss": 2.6126, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.5366994747760074, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00013466582331322905, | |
| "loss": 2.6022, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.5375822041753101, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0001342483829180376, | |
| "loss": 2.5966, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.5384649335746127, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00013383106583895137, | |
| "loss": 2.609, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.5393476629739153, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00013341387534305827, | |
| "loss": 2.6015, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.540230392373218, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00013299681469645513, | |
| "loss": 2.5923, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.5411131217725207, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.0001325798871642223, | |
| "loss": 2.5937, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.5419958511718233, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 0.0001321630960103979, | |
| "loss": 2.5939, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.5428785805711259, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00013174644449795244, | |
| "loss": 2.5918, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.5437613099704286, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00013132993588876323, | |
| "loss": 2.6067, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.5446440393697312, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00013091357344358873, | |
| "loss": 2.6008, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.5455267687690338, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00013049736042204318, | |
| "loss": 2.5962, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.5464094981683365, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00013008130008257098, | |
| "loss": 2.5948, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.5472922275676392, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00012966539568242132, | |
| "loss": 2.6013, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.5481749569669417, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00012924965047762243, | |
| "loss": 2.586, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.5490576863662444, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00012883406772295618, | |
| "loss": 2.6065, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.5499404157655471, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.000128418650671933, | |
| "loss": 2.603, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.5508231451648498, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0001280034025767656, | |
| "loss": 2.5907, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.5517058745641523, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00012758832668834438, | |
| "loss": 2.5947, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.552588603963455, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.0001271734262562112, | |
| "loss": 2.6048, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.5534713333627577, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00012675870452853464, | |
| "loss": 2.5926, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.5543540627620603, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00012634416475208401, | |
| "loss": 2.5901, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.5552367921613629, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00012592981017220425, | |
| "loss": 2.5909, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.5561195215606656, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00012551564403279048, | |
| "loss": 2.5953, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.5570022509599682, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00012510166957626248, | |
| "loss": 2.6004, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.5578849803592708, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00012468789004353942, | |
| "loss": 2.5984, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.5587677097585735, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00012427430867401434, | |
| "loss": 2.5956, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.5596504391578762, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00012386092870552917, | |
| "loss": 2.5942, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.5605331685571788, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00012344775337434896, | |
| "loss": 2.5981, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.5614158979564814, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0001230347859151365, | |
| "loss": 2.5976, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.5622986273557841, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00012262202956092754, | |
| "loss": 2.6007, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.5631813567550867, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00012220948754310492, | |
| "loss": 2.5985, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.5640640861543894, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0001217971630913736, | |
| "loss": 2.5844, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.564946815553692, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0001213850594337351, | |
| "loss": 2.5874, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.5658295449529946, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.0001209731797964626, | |
| "loss": 2.6036, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.5667122743522973, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00012056152740407529, | |
| "loss": 2.5904, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.5675950037516, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00012015010547931336, | |
| "loss": 2.6085, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.5684777331509026, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00011973891724311284, | |
| "loss": 2.5961, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.5693604625502052, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00011932796591458007, | |
| "loss": 2.6053, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.5702431919495079, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00011891725471096684, | |
| "loss": 2.5982, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.5711259213488106, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00011850678684764488, | |
| "loss": 2.5927, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.5720086507481131, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 0.00011809656553808112, | |
| "loss": 2.5926, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.5728913801474158, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00011768659399381203, | |
| "loss": 2.5924, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.5737741095467185, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00011727687542441882, | |
| "loss": 2.5924, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.574656838946021, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00011686741303750225, | |
| "loss": 2.5915, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.5755395683453237, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00011645821003865741, | |
| "loss": 2.6102, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.5764222977446264, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00011604926963144873, | |
| "loss": 2.596, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.5773050271439291, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00011564059501738481, | |
| "loss": 2.6111, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.5781877565432316, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00011523218939589354, | |
| "loss": 2.5835, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.5790704859425343, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00011482405596429679, | |
| "loss": 2.5978, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.579953215341837, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00011441619791778552, | |
| "loss": 2.5975, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.5808359447411396, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00011400861844939495, | |
| "loss": 2.5914, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.5817186741404422, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00011360132074997912, | |
| "loss": 2.5929, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.5826014035397449, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00011319430800818624, | |
| "loss": 2.5858, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.5834841329390476, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00011278758341043367, | |
| "loss": 2.5938, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.5843668623383502, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00011238115014088304, | |
| "loss": 2.5893, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.5852495917376528, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0001119750113814151, | |
| "loss": 2.5772, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.5861323211369555, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00011156917031160491, | |
| "loss": 2.5838, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.5870150505362581, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00011116363010869717, | |
| "loss": 2.5947, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.5878977799355608, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00011075839394758099, | |
| "loss": 2.599, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.5887805093348634, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00011035346500076523, | |
| "loss": 2.6109, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.589663238734166, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00010994884643835366, | |
| "loss": 2.5911, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.5905459681334687, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.0001095445414280202, | |
| "loss": 2.6052, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.5914286975327713, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00010914055313498382, | |
| "loss": 2.6236, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.592311426932074, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00010873688472198412, | |
| "loss": 2.588, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.5931941563313766, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00010833353934925652, | |
| "loss": 2.5957, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.5940768857306793, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0001079305201745073, | |
| "loss": 2.5877, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.5949596151299819, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.0001075278303528889, | |
| "loss": 2.5965, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.5958423445292845, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00010712547303697548, | |
| "loss": 2.588, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.5967250739285872, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00010672345137673814, | |
| "loss": 2.6023, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.5976078033278899, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00010632176851952005, | |
| "loss": 2.6022, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.5984905327271924, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00010592042761001198, | |
| "loss": 2.5867, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.5993732621264951, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00010551943179022779, | |
| "loss": 2.6013, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.6002559915257978, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 0.00010511878419947958, | |
| "loss": 2.5891, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6011387209251005, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.00010471848797435328, | |
| "loss": 2.59, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.602021450324403, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.000104318546248684, | |
| "loss": 2.5856, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.6029041797237057, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00010391896215353167, | |
| "loss": 2.5669, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.6037869091230084, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.00010351973881715632, | |
| "loss": 2.5969, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.604669638522311, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 0.00010312087936499361, | |
| "loss": 2.5938, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.6055523679216136, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00010272238691963064, | |
| "loss": 2.5898, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.6064350973209163, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00010232426460078106, | |
| "loss": 2.5899, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.607317826720219, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00010192651552526104, | |
| "loss": 2.5952, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.6082005561195215, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00010152914280696453, | |
| "loss": 2.5992, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.6090832855188242, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00010113214955683929, | |
| "loss": 2.5931, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.6099660149181269, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00010073553888286212, | |
| "loss": 2.5838, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 0.6108487443174295, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00010033931389001476, | |
| "loss": 2.586, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.6117314737167321, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 9.99434776802596e-05, | |
| "loss": 2.6036, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 0.6126142031160348, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 9.954803335251524e-05, | |
| "loss": 2.5978, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.6134969325153374, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 9.915298400263235e-05, | |
| "loss": 2.5901, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.6143796619146401, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 9.87583327233694e-05, | |
| "loss": 2.5769, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.6152623913139427, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 9.836408260436849e-05, | |
| "loss": 2.6066, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 0.6161451207132453, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 9.797023673213106e-05, | |
| "loss": 2.5825, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.617027850112548, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 9.757679818999374e-05, | |
| "loss": 2.583, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 0.6179105795118507, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 9.718377005810445e-05, | |
| "loss": 2.5874, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6187933089111533, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 9.679115541339793e-05, | |
| "loss": 2.5848, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 0.6196760383104559, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 9.639895732957188e-05, | |
| "loss": 2.6047, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.6205587677097586, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 9.600717887706269e-05, | |
| "loss": 2.6044, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 0.6214414971090613, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 9.561582312302181e-05, | |
| "loss": 2.5997, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.6223242265083638, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 9.522489313129128e-05, | |
| "loss": 2.594, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.6232069559076665, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 9.483439196237993e-05, | |
| "loss": 2.5925, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.6240896853069692, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 9.444432267343956e-05, | |
| "loss": 2.5932, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 0.6249724147062717, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 9.405468831824076e-05, | |
| "loss": 2.5924, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.6258551441055744, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 9.366549194714915e-05, | |
| "loss": 2.6024, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 0.6267378735048771, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 9.327673660710138e-05, | |
| "loss": 2.599, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.6276206029041798, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 9.288842534158163e-05, | |
| "loss": 2.6037, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 0.6285033323034823, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 9.250056119059715e-05, | |
| "loss": 2.5929, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.629386061702785, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 9.211314719065503e-05, | |
| "loss": 2.5814, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 0.6302687911020877, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 9.172618637473827e-05, | |
| "loss": 2.5902, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.6311515205013903, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 9.133968177228186e-05, | |
| "loss": 2.5986, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.6320342499006929, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 9.095363640914925e-05, | |
| "loss": 2.6009, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.6329169792999956, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 9.056805330760856e-05, | |
| "loss": 2.5972, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 0.6337997086992982, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 9.018293548630903e-05, | |
| "loss": 2.5858, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.6346824380986009, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.979828596025729e-05, | |
| "loss": 2.5969, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 0.6355651674979035, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 8.941410774079368e-05, | |
| "loss": 2.5973, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.6364478968972062, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.903040383556901e-05, | |
| "loss": 2.6079, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 0.6373306262965088, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.86471772485206e-05, | |
| "loss": 2.5989, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.6382133556958115, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 8.826443097984894e-05, | |
| "loss": 2.5726, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 0.6390960850951141, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.788216802599426e-05, | |
| "loss": 2.577, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.6399788144944167, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 8.750039137961314e-05, | |
| "loss": 2.597, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.6408615438937194, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.711910402955479e-05, | |
| "loss": 2.6039, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.641744273293022, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.67383089608378e-05, | |
| "loss": 2.6105, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 0.6426270026923246, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.63580091546269e-05, | |
| "loss": 2.5943, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.6435097320916273, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.597820758820957e-05, | |
| "loss": 2.5949, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 0.64439246149093, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 8.559890723497248e-05, | |
| "loss": 2.5896, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.6452751908902326, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 8.522011106437853e-05, | |
| "loss": 2.5816, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 0.6461579202895352, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.484182204194349e-05, | |
| "loss": 2.5905, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.6470406496888379, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 8.446404312921283e-05, | |
| "loss": 2.5994, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 0.6479233790881406, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.408677728373824e-05, | |
| "loss": 2.585, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.6488061084874431, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.371002745905492e-05, | |
| "loss": 2.5888, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.6496888378867458, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 8.333379660465829e-05, | |
| "loss": 2.5958, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.6505715672860485, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 8.295808766598068e-05, | |
| "loss": 2.5744, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 0.6514542966853512, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.258290358436853e-05, | |
| "loss": 2.6047, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.6523370260846537, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.220824729705939e-05, | |
| "loss": 2.5985, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 0.6532197554839564, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 8.183412173715862e-05, | |
| "loss": 2.6073, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.6541024848832591, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.146052983361675e-05, | |
| "loss": 2.6032, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 0.6549852142825617, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.108747451120652e-05, | |
| "loss": 2.5904, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.6558679436818643, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 8.071495869049968e-05, | |
| "loss": 2.5938, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 0.656750673081167, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 8.034298528784446e-05, | |
| "loss": 2.5933, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.6576334024804696, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 7.997155721534247e-05, | |
| "loss": 2.5932, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.6585161318797722, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 7.960067738082638e-05, | |
| "loss": 2.5882, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.6593988612790749, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.923034868783643e-05, | |
| "loss": 2.5971, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 0.6602815906783776, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.886057403559831e-05, | |
| "loss": 2.5925, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.6611643200776802, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.849135631900026e-05, | |
| "loss": 2.579, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 0.6620470494769828, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.812269842857017e-05, | |
| "loss": 2.594, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.6629297788762855, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.775460325045347e-05, | |
| "loss": 2.5999, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 0.6638125082755881, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.738707366638993e-05, | |
| "loss": 2.598, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.6646952376748908, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 7.702011255369163e-05, | |
| "loss": 2.6078, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 0.6655779670741934, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 7.665372278522018e-05, | |
| "loss": 2.5921, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.666460696473496, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.628790722936404e-05, | |
| "loss": 2.585, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.6673434258727987, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.592266875001666e-05, | |
| "loss": 2.5924, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.6682261552721014, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.555801020655339e-05, | |
| "loss": 2.5913, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 0.669108884671404, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.519393445380932e-05, | |
| "loss": 2.5881, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.6699916140707066, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 7.483044434205726e-05, | |
| "loss": 2.5966, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 0.6708743434700093, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 7.4467542716985e-05, | |
| "loss": 2.592, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.671757072869312, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 7.410523241967328e-05, | |
| "loss": 2.5957, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 0.6726398022686145, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 7.374351628657328e-05, | |
| "loss": 2.5918, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.6735225316679172, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.33823971494848e-05, | |
| "loss": 2.5837, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 0.6744052610672199, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 7.302187783553383e-05, | |
| "loss": 2.5819, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.6752879904665224, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.266196116715033e-05, | |
| "loss": 2.5864, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.6761707198658251, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 7.230264996204644e-05, | |
| "loss": 2.5926, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.6770534492651278, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.194394703319423e-05, | |
| "loss": 2.5866, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 0.6779361786644305, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 7.158585518880362e-05, | |
| "loss": 2.5889, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.678818908063733, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 7.122837723230051e-05, | |
| "loss": 2.6028, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 0.6797016374630357, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.087151596230486e-05, | |
| "loss": 2.597, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.6805843668623384, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.051527417260873e-05, | |
| "loss": 2.6031, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 0.681467096261641, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 7.01596546521543e-05, | |
| "loss": 2.6113, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.6823498256609436, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 6.980466018501203e-05, | |
| "loss": 2.5859, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 0.6832325550602463, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.945029355035939e-05, | |
| "loss": 2.5877, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.684115284459549, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 6.909655752245823e-05, | |
| "loss": 2.598, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.6849980138588516, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.874345487063377e-05, | |
| "loss": 2.588, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.6858807432581542, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 6.83909883592526e-05, | |
| "loss": 2.593, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 0.6867634726574569, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 6.803916074770099e-05, | |
| "loss": 2.5724, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.6876462020567595, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.768797479036363e-05, | |
| "loss": 2.5905, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 0.6885289314560622, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 6.733743323660155e-05, | |
| "loss": 2.586, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.6894116608553648, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 6.69875388307311e-05, | |
| "loss": 2.5853, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 0.6902943902546674, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 6.663829431200224e-05, | |
| "loss": 2.574, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.6911771196539701, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 6.628970241457684e-05, | |
| "loss": 2.6049, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 0.6920598490532727, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 6.594176586750796e-05, | |
| "loss": 2.5952, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.6929425784525753, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 6.559448739471771e-05, | |
| "loss": 2.5938, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.693825307851878, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.524786971497635e-05, | |
| "loss": 2.6017, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.6947080372511807, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 6.490191554188102e-05, | |
| "loss": 2.5923, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 0.6955907666504832, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.455662758383433e-05, | |
| "loss": 2.5872, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.6964734960497859, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 6.421200854402337e-05, | |
| "loss": 2.5829, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 0.6973562254490886, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 6.386806112039812e-05, | |
| "loss": 2.5778, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.6982389548483913, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 6.35247880056509e-05, | |
| "loss": 2.604, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 0.6991216842476938, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.318219188719493e-05, | |
| "loss": 2.6085, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.7000044136469965, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 6.284027544714325e-05, | |
| "loss": 2.5987, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 0.7008871430462992, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 6.249904136228796e-05, | |
| "loss": 2.5919, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.7017698724456018, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 6.21584923040792e-05, | |
| "loss": 2.5853, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.7026526018449044, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.181863093860399e-05, | |
| "loss": 2.6032, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.7035353312442071, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 6.147945992656566e-05, | |
| "loss": 2.5987, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 0.7044180606435098, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.114098192326296e-05, | |
| "loss": 2.5868, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.7053007900428124, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 6.0803199578569166e-05, | |
| "loss": 2.595, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 0.706183519442115, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 6.046611553691136e-05, | |
| "loss": 2.6022, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.706183519442115, | |
| "eval_accuracy": 0.49817080194438684, | |
| "eval_loss": 2.483642101287842, | |
| "eval_runtime": 6.8683, | |
| "eval_samples_per_second": 46.3, | |
| "eval_steps_per_second": 0.437, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7070662488414177, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 6.012973243724968e-05, | |
| "loss": 2.5992, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 0.7079489782407203, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.979405291305702e-05, | |
| "loss": 2.5919, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.7088317076400229, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 5.94590795922978e-05, | |
| "loss": 2.5916, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 0.7097144370393256, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 5.912481509740792e-05, | |
| "loss": 2.5939, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.7105971664386282, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.8791262045274036e-05, | |
| "loss": 2.5925, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.7114798958379309, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.84584230472129e-05, | |
| "loss": 2.5965, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.7123626252372335, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 5.81263007089513e-05, | |
| "loss": 2.5973, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 0.7132453546365362, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.7794897630605183e-05, | |
| "loss": 2.5888, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.7141280840358388, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 5.7464216406659916e-05, | |
| "loss": 2.5954, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 0.7150108134351415, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.713425962594933e-05, | |
| "loss": 2.5989, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.7158935428344441, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 5.680502987163573e-05, | |
| "loss": 2.5982, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 0.7167762722337467, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 5.6476529721189974e-05, | |
| "loss": 2.5899, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.7176590016330494, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 5.614876174637069e-05, | |
| "loss": 2.5835, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 0.7185417310323521, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 5.5821728513204663e-05, | |
| "loss": 2.5927, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.7194244604316546, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.549543258196636e-05, | |
| "loss": 2.6001, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.7203071898309573, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 5.516987650715823e-05, | |
| "loss": 2.5821, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.72118991923026, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.484506283749051e-05, | |
| "loss": 2.5814, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 0.7220726486295626, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 5.45209941158612e-05, | |
| "loss": 2.5928, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.7229553780288652, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 5.419767287933633e-05, | |
| "loss": 2.5924, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 0.7238381074281679, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 5.387510165913014e-05, | |
| "loss": 2.6029, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.7247208368274706, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 5.355328298058486e-05, | |
| "loss": 2.5877, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 0.7256035662267731, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 5.323221936315152e-05, | |
| "loss": 2.5766, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.7264862956260758, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.291191332036982e-05, | |
| "loss": 2.577, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 0.7273690250253785, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 5.259236735984849e-05, | |
| "loss": 2.5952, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.7282517544246812, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 5.2273583983245834e-05, | |
| "loss": 2.6003, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.7291344838239837, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 5.195556568625006e-05, | |
| "loss": 2.6014, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.7300172132232864, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.1638314958559724e-05, | |
| "loss": 2.596, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 0.7308999426225891, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 5.1321834283864105e-05, | |
| "loss": 2.6006, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.7317826720218917, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 5.100612613982405e-05, | |
| "loss": 2.6007, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 0.7326654014211943, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 5.0691192998052455e-05, | |
| "loss": 2.5807, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.733548130820497, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 5.037703732409469e-05, | |
| "loss": 2.5927, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 0.7344308602197996, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.0063661577409654e-05, | |
| "loss": 2.5976, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.7353135896191023, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 4.9751068211350385e-05, | |
| "loss": 2.587, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 0.7361963190184049, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.943925967314466e-05, | |
| "loss": 2.5885, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.7370790484177075, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.912823840387619e-05, | |
| "loss": 2.6006, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.7379617778170102, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 4.881800683846511e-05, | |
| "loss": 2.5985, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.7388445072163128, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.850856740564943e-05, | |
| "loss": 2.5753, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 0.7397272366156155, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 4.819992252796546e-05, | |
| "loss": 2.5878, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.7406099660149181, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.789207462172909e-05, | |
| "loss": 2.5761, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 0.7414926954142208, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 4.7585026097017165e-05, | |
| "loss": 2.5949, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.7423754248135234, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 4.7278779357648056e-05, | |
| "loss": 2.596, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 0.743258154212826, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 4.69733368011633e-05, | |
| "loss": 2.5932, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 0.7441408836121287, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 4.666870081880849e-05, | |
| "loss": 2.602, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 0.7450236130114314, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.636487379551486e-05, | |
| "loss": 2.5954, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 0.745906342410734, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 4.6061858109880494e-05, | |
| "loss": 2.5888, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.7467890718100366, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.575965613415152e-05, | |
| "loss": 2.5912, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 0.7476718012093393, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 4.5458270234203833e-05, | |
| "loss": 2.5882, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 0.748554530608642, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.515770276952449e-05, | |
| "loss": 2.5882, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.7494372600079445, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 4.4857956093192966e-05, | |
| "loss": 2.5898, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 0.7503199894072472, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 4.455903255186318e-05, | |
| "loss": 2.6044, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.7512027188065499, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 4.426093448574482e-05, | |
| "loss": 2.5954, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 0.7520854482058525, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 4.396366422858514e-05, | |
| "loss": 2.5937, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 0.7529681776051551, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.3667224107650515e-05, | |
| "loss": 2.5993, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 0.7538509070044578, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.33716164437085e-05, | |
| "loss": 2.5979, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 0.7547336364037605, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.307684355100953e-05, | |
| "loss": 2.5936, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.755616365803063, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 4.2782907737268644e-05, | |
| "loss": 2.6058, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 0.7564990952023657, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.24898113036477e-05, | |
| "loss": 2.5962, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 0.7573818246016684, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 4.219755654473722e-05, | |
| "loss": 2.5915, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 0.758264554000971, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 4.190614574853832e-05, | |
| "loss": 2.5913, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 0.7591472834002736, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.1615581196444996e-05, | |
| "loss": 2.5946, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.7600300127995763, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.132586516322625e-05, | |
| "loss": 2.5908, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 0.7609127421988789, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.103699991700799e-05, | |
| "loss": 2.6099, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 0.7617954715981816, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.074898771925573e-05, | |
| "loss": 2.5835, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 0.7626782009974842, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.04618308247564e-05, | |
| "loss": 2.5815, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.7635609303967869, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.017553148160126e-05, | |
| "loss": 2.5932, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.7644436597960895, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.989009193116768e-05, | |
| "loss": 2.5983, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 0.7653263891953922, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 3.960551440810203e-05, | |
| "loss": 2.5841, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 0.7662091185946948, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.9321801140302086e-05, | |
| "loss": 2.5952, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 0.7670918479939974, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 3.903895434889939e-05, | |
| "loss": 2.5844, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 0.7679745773933001, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 3.8756976248242246e-05, | |
| "loss": 2.5877, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.7688573067926028, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.847586904587793e-05, | |
| "loss": 2.5984, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 0.7697400361919053, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.819563494253582e-05, | |
| "loss": 2.5884, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 0.770622765591208, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.7916276132109915e-05, | |
| "loss": 2.5753, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 0.7715054949905107, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 3.763779480164165e-05, | |
| "loss": 2.584, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 0.7723882243898132, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.736019313130306e-05, | |
| "loss": 2.5967, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.7732709537891159, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 3.708347329437933e-05, | |
| "loss": 2.5982, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 0.7741536831884186, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.680763745725194e-05, | |
| "loss": 2.586, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 0.7750364125877213, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.6532687779381816e-05, | |
| "loss": 2.5978, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 0.7759191419870238, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 3.625862641329232e-05, | |
| "loss": 2.6061, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 0.7768018713863265, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 3.598545550455244e-05, | |
| "loss": 2.5981, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.7776846007856292, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.5713177191759795e-05, | |
| "loss": 2.6118, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 0.7785673301849318, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.544179360652421e-05, | |
| "loss": 2.6004, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 0.7794500595842344, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.517130687345089e-05, | |
| "loss": 2.5886, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 0.7803327889835371, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 3.490171911012364e-05, | |
| "loss": 2.5937, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 0.7812155183828398, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.463303242708851e-05, | |
| "loss": 2.5927, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.7820982477821424, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.436524892783725e-05, | |
| "loss": 2.5842, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 0.782980977181445, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.4098370708790574e-05, | |
| "loss": 2.5858, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 0.7838637065807477, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.3832399859282116e-05, | |
| "loss": 2.5962, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.7847464359800503, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 3.356733846154185e-05, | |
| "loss": 2.5992, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 0.785629165379353, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 3.33031885906799e-05, | |
| "loss": 2.5835, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.7865118947786556, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 3.3039952314670135e-05, | |
| "loss": 2.5954, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 0.7873946241779582, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.2777631694334056e-05, | |
| "loss": 2.5791, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 0.7882773535772609, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.2516228783324904e-05, | |
| "loss": 2.6001, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 0.7891600829765635, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.2255745628111116e-05, | |
| "loss": 2.5724, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 0.7900428123758662, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.19961842679607e-05, | |
| "loss": 2.5989, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.7909255417751688, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.1737546734925126e-05, | |
| "loss": 2.5959, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.7918082711744715, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 3.147983505382329e-05, | |
| "loss": 2.6015, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 0.7926910005737741, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 3.122305124222597e-05, | |
| "loss": 2.5994, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 0.7935737299730767, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.096719731043965e-05, | |
| "loss": 2.5719, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 0.7944564593723794, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.0712275261491145e-05, | |
| "loss": 2.6032, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.7953391887716821, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.0458287091111692e-05, | |
| "loss": 2.594, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 0.7962219181709846, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.02052347877213e-05, | |
| "loss": 2.6004, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 0.7971046475702873, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 2.9953120332413517e-05, | |
| "loss": 2.5977, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 0.79798737696959, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.970194569893937e-05, | |
| "loss": 2.5989, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 0.7988701063688927, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 2.9451712853692484e-05, | |
| "loss": 2.5855, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.7997528357681952, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.9202423755693193e-05, | |
| "loss": 2.6097, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 0.8006355651674979, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 2.8954080356573596e-05, | |
| "loss": 2.5833, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 0.8015182945668006, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 2.8706684600562098e-05, | |
| "loss": 2.5918, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 0.8024010239661032, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 2.846023842446808e-05, | |
| "loss": 2.5957, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 0.8032837533654058, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 2.8214743757666984e-05, | |
| "loss": 2.5923, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.8041664827647085, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.79702025220851e-05, | |
| "loss": 2.5864, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 0.8050492121640112, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 2.772661663218438e-05, | |
| "loss": 2.5944, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.8059319415633137, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.748398799494766e-05, | |
| "loss": 2.6034, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 0.8068146709626164, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.7242318509863708e-05, | |
| "loss": 2.6014, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 0.8076974003619191, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 2.7001610068912117e-05, | |
| "loss": 2.6088, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.8085801297612217, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.6761864556548805e-05, | |
| "loss": 2.591, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 0.8094628591605243, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 2.65230838496911e-05, | |
| "loss": 2.5863, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 0.810345588559827, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.62852698177031e-05, | |
| "loss": 2.5847, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 0.8112283179591296, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 2.6048424322380877e-05, | |
| "loss": 2.5979, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 0.8121110473584323, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.581254921793816e-05, | |
| "loss": 2.5939, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.8129937767577349, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.5577646350991686e-05, | |
| "loss": 2.5843, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 0.8138765061570375, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.534371756054664e-05, | |
| "loss": 2.5922, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 0.8147592355563402, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.5110764677982496e-05, | |
| "loss": 2.5972, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 0.8156419649556429, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 2.4878789527038523e-05, | |
| "loss": 2.5881, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 0.8165246943549455, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 2.4647793923799413e-05, | |
| "loss": 2.5987, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.8174074237542481, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.441777967668139e-05, | |
| "loss": 2.5866, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 0.8182901531535508, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.4188748586417588e-05, | |
| "loss": 2.5926, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 0.8191728825528535, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.396070244604454e-05, | |
| "loss": 2.591, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 0.820055611952156, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 2.3733643040887524e-05, | |
| "loss": 2.5941, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 0.8209383413514587, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.3507572148546928e-05, | |
| "loss": 2.5885, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.8218210707507614, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.328249153888449e-05, | |
| "loss": 2.6027, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 0.822703800150064, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.3058402974008954e-05, | |
| "loss": 2.5711, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 0.8235865295493666, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.283530820826276e-05, | |
| "loss": 2.5866, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 0.8244692589486693, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.261320898820795e-05, | |
| "loss": 2.5912, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 0.825351988347972, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 2.239210705261275e-05, | |
| "loss": 2.6169, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.8262347177472745, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 2.2172004132437858e-05, | |
| "loss": 2.591, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 0.8271174471465772, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 2.195290195082279e-05, | |
| "loss": 2.5965, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 0.8280001765458799, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.1734802223072584e-05, | |
| "loss": 2.5916, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 0.8288829059451825, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.151770665664429e-05, | |
| "loss": 2.5762, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 0.8297656353444851, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 2.1301616951133473e-05, | |
| "loss": 2.5888, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.8306483647437878, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.108653479826114e-05, | |
| "loss": 2.5969, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 0.8315310941430905, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.0872461881860354e-05, | |
| "loss": 2.5977, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 0.8324138235423931, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.0659399877863054e-05, | |
| "loss": 2.5946, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 0.8332965529416957, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 2.044735045428693e-05, | |
| "loss": 2.5796, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 0.8341792823409984, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.023631527122247e-05, | |
| "loss": 2.5779, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.835062011740301, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.002629598081985e-05, | |
| "loss": 2.6046, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 0.8359447411396037, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 1.9817294227275987e-05, | |
| "loss": 2.5939, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 0.8368274705389063, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.960931164682178e-05, | |
| "loss": 2.6001, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 0.8377101999382089, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.940234986770926e-05, | |
| "loss": 2.5967, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 0.8385929293375116, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.919641051019869e-05, | |
| "loss": 2.588, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.8394756587368142, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.899149518654613e-05, | |
| "loss": 2.5935, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 0.8403583881361169, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 1.8787605500990688e-05, | |
| "loss": 2.5986, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 0.8412411175354195, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.8584743049741864e-05, | |
| "loss": 2.6029, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 0.8421238469347222, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.8382909420967302e-05, | |
| "loss": 2.5967, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 0.8430065763340248, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.8182106194780044e-05, | |
| "loss": 2.6032, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.8438893057333274, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.7982334943226517e-05, | |
| "loss": 2.5997, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 0.8447720351326301, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.778359723027386e-05, | |
| "loss": 2.5832, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 0.8456547645319328, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.7585894611797967e-05, | |
| "loss": 2.591, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 0.8465374939312353, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.7389228635571168e-05, | |
| "loss": 2.5917, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 0.847420223330538, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 1.7193600841250095e-05, | |
| "loss": 2.5745, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.8483029527298407, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 1.6999012760363734e-05, | |
| "loss": 2.5985, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 0.8491856821291434, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.6805465916301304e-05, | |
| "loss": 2.5911, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 0.8500684115284459, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 1.661296182430047e-05, | |
| "loss": 2.5969, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 0.8509511409277486, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 1.64215019914354e-05, | |
| "loss": 2.6001, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 0.8518338703270513, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 1.623108791660486e-05, | |
| "loss": 2.6006, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.8527165997263539, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.6041721090520815e-05, | |
| "loss": 2.5933, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 0.8535993291256565, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 1.585340299569637e-05, | |
| "loss": 2.5961, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 0.8544820585249592, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.566613510643432e-05, | |
| "loss": 2.5971, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 0.8553647879242618, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 1.547991888881574e-05, | |
| "loss": 2.5917, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 0.8562475173235644, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.5294755800688307e-05, | |
| "loss": 2.5875, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.8571302467228671, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.5110647291655025e-05, | |
| "loss": 2.5895, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 0.8580129761221698, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.492759480306272e-05, | |
| "loss": 2.5911, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 0.8588957055214724, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.4745599767990945e-05, | |
| "loss": 2.6006, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 0.859778434920775, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 1.4564663611240696e-05, | |
| "loss": 2.592, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 0.8606611643200777, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.4384787749323095e-05, | |
| "loss": 2.5933, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.8615438937193803, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.420597359044856e-05, | |
| "loss": 2.5868, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 0.862426623118683, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.402822253451567e-05, | |
| "loss": 2.5935, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 0.8633093525179856, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 1.3851535973100081e-05, | |
| "loss": 2.5906, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 0.8641920819172882, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.3675915289443828e-05, | |
| "loss": 2.5728, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 0.8650748113165909, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.350136185844441e-05, | |
| "loss": 2.5725, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.8659575407158936, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 1.3327877046644036e-05, | |
| "loss": 2.6052, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 0.8668402701151962, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.3155462212218881e-05, | |
| "loss": 2.6, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 0.8677229995144988, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.298411870496845e-05, | |
| "loss": 2.593, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 0.8686057289138015, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.2813847866305204e-05, | |
| "loss": 2.5864, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 0.8694884583131042, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.2644651029243747e-05, | |
| "loss": 2.5912, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.8703711877124067, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.247652951839065e-05, | |
| "loss": 2.5825, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 0.8712539171117094, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 1.230948464993397e-05, | |
| "loss": 2.5887, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 0.8721366465110121, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 1.2143517731632878e-05, | |
| "loss": 2.6033, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 0.8730193759103146, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 1.197863006280761e-05, | |
| "loss": 2.6062, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 0.8739021053096173, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 1.1814822934329055e-05, | |
| "loss": 2.5908, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.87478483470892, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.1652097628608925e-05, | |
| "loss": 2.5967, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 0.8756675641082227, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.149045541958949e-05, | |
| "loss": 2.5904, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 0.8765502935075252, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.1329897572733614e-05, | |
| "loss": 2.6088, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 0.8774330229068279, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.117042534501511e-05, | |
| "loss": 2.6058, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 0.8783157523061306, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.1012039984908516e-05, | |
| "loss": 2.6058, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.8791984817054332, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.0854742732379612e-05, | |
| "loss": 2.5935, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 0.8800812111047358, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 1.0698534818875543e-05, | |
| "loss": 2.5994, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 0.8809639405040385, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 1.054341746731534e-05, | |
| "loss": 2.6002, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 0.8818466699033412, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.0389391892080185e-05, | |
| "loss": 2.5918, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 0.8827293993026438, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.0236459299003974e-05, | |
| "loss": 2.5999, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.8827293993026438, | |
| "eval_accuracy": 0.49828983036530206, | |
| "eval_loss": 2.4831273555755615, | |
| "eval_runtime": 6.931, | |
| "eval_samples_per_second": 45.881, | |
| "eval_steps_per_second": 0.433, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.8836121287019464, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.008462088536392e-05, | |
| "loss": 2.6038, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 0.8844948581012491, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 9.933877839871101e-06, | |
| "loss": 2.5905, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 0.8853775875005517, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 9.784231342661158e-06, | |
| "loss": 2.6041, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 0.8862603168998543, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 9.635682565285146e-06, | |
| "loss": 2.5878, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 0.887143046299157, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 9.488232670700296e-06, | |
| "loss": 2.603, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.8880257756984596, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 9.341882813260854e-06, | |
| "loss": 2.578, | |
| "step": 10060 | |
| }, | |
| { | |
| "epoch": 0.8889085050977623, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 9.196634138709186e-06, | |
| "loss": 2.5883, | |
| "step": 10070 | |
| }, | |
| { | |
| "epoch": 0.8897912344970649, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 9.052487784166656e-06, | |
| "loss": 2.5927, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 0.8906739638963675, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 8.909444878124866e-06, | |
| "loss": 2.6058, | |
| "step": 10090 | |
| }, | |
| { | |
| "epoch": 0.8915566932956702, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 8.767506540436675e-06, | |
| "loss": 2.5899, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.8924394226949729, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 8.626673882307578e-06, | |
| "loss": 2.5955, | |
| "step": 10110 | |
| }, | |
| { | |
| "epoch": 0.8933221520942755, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 8.486948006286926e-06, | |
| "loss": 2.5877, | |
| "step": 10120 | |
| }, | |
| { | |
| "epoch": 0.8942048814935781, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 8.348330006259234e-06, | |
| "loss": 2.5907, | |
| "step": 10130 | |
| }, | |
| { | |
| "epoch": 0.8950876108928808, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.210820967435778e-06, | |
| "loss": 2.5912, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 0.8959703402921835, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 8.074421966345952e-06, | |
| "loss": 2.5819, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.896853069691486, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 7.939134070828907e-06, | |
| "loss": 2.5954, | |
| "step": 10160 | |
| }, | |
| { | |
| "epoch": 0.8977357990907887, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 7.804958340025252e-06, | |
| "loss": 2.6031, | |
| "step": 10170 | |
| }, | |
| { | |
| "epoch": 0.8986185284900914, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 7.671895824368546e-06, | |
| "loss": 2.6098, | |
| "step": 10180 | |
| }, | |
| { | |
| "epoch": 0.8995012578893941, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 7.539947565577381e-06, | |
| "loss": 2.5844, | |
| "step": 10190 | |
| }, | |
| { | |
| "epoch": 0.9003839872886966, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 7.4091145966469645e-06, | |
| "loss": 2.5834, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.9012667166879993, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 7.279397941841136e-06, | |
| "loss": 2.5912, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 0.902149446087302, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 7.1507986166843944e-06, | |
| "loss": 2.5953, | |
| "step": 10220 | |
| }, | |
| { | |
| "epoch": 0.9030321754866045, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.02331762795385e-06, | |
| "loss": 2.5803, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 0.9039149048859072, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 6.896955973671487e-06, | |
| "loss": 2.5899, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 0.9047976342852099, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 6.771714643096093e-06, | |
| "loss": 2.585, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.9056803636845125, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 6.647594616715812e-06, | |
| "loss": 2.6141, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 0.9065630930838151, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.524596866240306e-06, | |
| "loss": 2.5963, | |
| "step": 10270 | |
| }, | |
| { | |
| "epoch": 0.9074458224831178, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 6.4027223545931315e-06, | |
| "loss": 2.598, | |
| "step": 10280 | |
| }, | |
| { | |
| "epoch": 0.9083285518824205, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 6.281972035904265e-06, | |
| "loss": 2.5859, | |
| "step": 10290 | |
| }, | |
| { | |
| "epoch": 0.9092112812817231, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 6.162346855502637e-06, | |
| "loss": 2.589, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.9100940106810257, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 6.043847749908676e-06, | |
| "loss": 2.5919, | |
| "step": 10310 | |
| }, | |
| { | |
| "epoch": 0.9109767400803284, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 5.9264756468269946e-06, | |
| "loss": 2.5981, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 0.911859469479631, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.810231465139198e-06, | |
| "loss": 2.5933, | |
| "step": 10330 | |
| }, | |
| { | |
| "epoch": 0.9127421988789337, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 5.695116114896603e-06, | |
| "loss": 2.6009, | |
| "step": 10340 | |
| }, | |
| { | |
| "epoch": 0.9136249282782363, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.581130497313096e-06, | |
| "loss": 2.5923, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.9145076576775389, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 5.468275504758157e-06, | |
| "loss": 2.5938, | |
| "step": 10360 | |
| }, | |
| { | |
| "epoch": 0.9153903870768416, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.35655202074991e-06, | |
| "loss": 2.5905, | |
| "step": 10370 | |
| }, | |
| { | |
| "epoch": 0.9162731164761443, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.245960919947984e-06, | |
| "loss": 2.5801, | |
| "step": 10380 | |
| }, | |
| { | |
| "epoch": 0.9171558458754469, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.136503068146963e-06, | |
| "loss": 2.5854, | |
| "step": 10390 | |
| }, | |
| { | |
| "epoch": 0.9180385752747495, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 5.0281793222693986e-06, | |
| "loss": 2.5779, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.9189213046740522, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.920990530359142e-06, | |
| "loss": 2.5931, | |
| "step": 10410 | |
| }, | |
| { | |
| "epoch": 0.9198040340733548, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.8149375315747865e-06, | |
| "loss": 2.5798, | |
| "step": 10420 | |
| }, | |
| { | |
| "epoch": 0.9206867634726574, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.710021156182969e-06, | |
| "loss": 2.5887, | |
| "step": 10430 | |
| }, | |
| { | |
| "epoch": 0.9215694928719601, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.606242225552015e-06, | |
| "loss": 2.5883, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 0.9224522222712628, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.503601552145403e-06, | |
| "loss": 2.6045, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.9233349516705653, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 4.402099939515374e-06, | |
| "loss": 2.5885, | |
| "step": 10460 | |
| }, | |
| { | |
| "epoch": 0.924217681069868, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 4.301738182296838e-06, | |
| "loss": 2.597, | |
| "step": 10470 | |
| }, | |
| { | |
| "epoch": 0.9251004104691707, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.202517066200872e-06, | |
| "loss": 2.5859, | |
| "step": 10480 | |
| }, | |
| { | |
| "epoch": 0.9259831398684734, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.104437368008817e-06, | |
| "loss": 2.6073, | |
| "step": 10490 | |
| }, | |
| { | |
| "epoch": 0.9268658692677759, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.00749985556601e-06, | |
| "loss": 2.5869, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9277485986670786, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.911705287775923e-06, | |
| "loss": 2.5927, | |
| "step": 10510 | |
| }, | |
| { | |
| "epoch": 0.9286313280663813, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 3.817054414594156e-06, | |
| "loss": 2.5923, | |
| "step": 10520 | |
| }, | |
| { | |
| "epoch": 0.9295140574656839, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.723547977022484e-06, | |
| "loss": 2.5858, | |
| "step": 10530 | |
| }, | |
| { | |
| "epoch": 0.9303967868649865, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.6311867071032006e-06, | |
| "loss": 2.6051, | |
| "step": 10540 | |
| }, | |
| { | |
| "epoch": 0.9312795162642892, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 3.539971327913338e-06, | |
| "loss": 2.5804, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.9321622456635918, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.4499025535589363e-06, | |
| "loss": 2.5944, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 0.9330449750628945, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 3.3609810891695345e-06, | |
| "loss": 2.5844, | |
| "step": 10570 | |
| }, | |
| { | |
| "epoch": 0.9339277044621971, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.273207630892638e-06, | |
| "loss": 2.605, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 0.9348104338614998, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.18658286588821e-06, | |
| "loss": 2.6071, | |
| "step": 10590 | |
| }, | |
| { | |
| "epoch": 0.9356931632608024, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 3.101107472323372e-06, | |
| "loss": 2.5912, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.936575892660105, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.016782119367045e-06, | |
| "loss": 2.5925, | |
| "step": 10610 | |
| }, | |
| { | |
| "epoch": 0.9374586220594077, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.933607467184751e-06, | |
| "loss": 2.5881, | |
| "step": 10620 | |
| }, | |
| { | |
| "epoch": 0.9383413514587103, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.851584166933352e-06, | |
| "loss": 2.5991, | |
| "step": 10630 | |
| }, | |
| { | |
| "epoch": 0.939224080858013, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 2.7707128607560693e-06, | |
| "loss": 2.5856, | |
| "step": 10640 | |
| }, | |
| { | |
| "epoch": 0.9401068102573156, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.690994181777406e-06, | |
| "loss": 2.6109, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.9409895396566182, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.6124287540981505e-06, | |
| "loss": 2.5893, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 0.9418722690559209, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 2.5350171927905617e-06, | |
| "loss": 2.5834, | |
| "step": 10670 | |
| }, | |
| { | |
| "epoch": 0.9427549984552236, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.4587601038935256e-06, | |
| "loss": 2.589, | |
| "step": 10680 | |
| }, | |
| { | |
| "epoch": 0.9436377278545262, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 2.3836580844077745e-06, | |
| "loss": 2.5996, | |
| "step": 10690 | |
| }, | |
| { | |
| "epoch": 0.9445204572538288, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 2.3097117222912744e-06, | |
| "loss": 2.6005, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.9454031866531315, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 2.2369215964545608e-06, | |
| "loss": 2.5932, | |
| "step": 10710 | |
| }, | |
| { | |
| "epoch": 0.9462859160524342, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.1652882767562773e-06, | |
| "loss": 2.6008, | |
| "step": 10720 | |
| }, | |
| { | |
| "epoch": 0.9471686454517367, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.0948123239986457e-06, | |
| "loss": 2.5907, | |
| "step": 10730 | |
| }, | |
| { | |
| "epoch": 0.9480513748510394, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.0254942899230853e-06, | |
| "loss": 2.5874, | |
| "step": 10740 | |
| }, | |
| { | |
| "epoch": 0.9489341042503421, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.9573347172059996e-06, | |
| "loss": 2.5755, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.9498168336496448, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.890334139454347e-06, | |
| "loss": 2.6, | |
| "step": 10760 | |
| }, | |
| { | |
| "epoch": 0.9506995630489473, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.8244930812015945e-06, | |
| "loss": 2.5805, | |
| "step": 10770 | |
| }, | |
| { | |
| "epoch": 0.95158229244825, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.7598120579035702e-06, | |
| "loss": 2.581, | |
| "step": 10780 | |
| }, | |
| { | |
| "epoch": 0.9524650218475527, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.6962915759344332e-06, | |
| "loss": 2.6009, | |
| "step": 10790 | |
| }, | |
| { | |
| "epoch": 0.9533477512468552, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.6339321325827437e-06, | |
| "loss": 2.5827, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.9542304806461579, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.57273421604745e-06, | |
| "loss": 2.5949, | |
| "step": 10810 | |
| }, | |
| { | |
| "epoch": 0.9551132100454606, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.5126983054342234e-06, | |
| "loss": 2.5984, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 0.9559959394447632, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.45382487075163e-06, | |
| "loss": 2.5951, | |
| "step": 10830 | |
| }, | |
| { | |
| "epoch": 0.9568786688440658, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 1.3961143729073975e-06, | |
| "loss": 2.5902, | |
| "step": 10840 | |
| }, | |
| { | |
| "epoch": 0.9577613982433685, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 1.339567263704938e-06, | |
| "loss": 2.5927, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.9586441276426712, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.2841839858396984e-06, | |
| "loss": 2.5969, | |
| "step": 10860 | |
| }, | |
| { | |
| "epoch": 0.9595268570419738, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.2299649728957306e-06, | |
| "loss": 2.6002, | |
| "step": 10870 | |
| }, | |
| { | |
| "epoch": 0.9604095864412764, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.1769106493423274e-06, | |
| "loss": 2.5813, | |
| "step": 10880 | |
| }, | |
| { | |
| "epoch": 0.9612923158405791, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.125021430530626e-06, | |
| "loss": 2.5912, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 0.9621750452398817, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.0742977226904759e-06, | |
| "loss": 2.6056, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.9630577746391844, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.0247399229271257e-06, | |
| "loss": 2.5852, | |
| "step": 10910 | |
| }, | |
| { | |
| "epoch": 0.963940504038487, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 9.763484192181758e-07, | |
| "loss": 2.5917, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 0.9648232334377896, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 9.291235904105632e-07, | |
| "loss": 2.5826, | |
| "step": 10930 | |
| }, | |
| { | |
| "epoch": 0.9657059628370923, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 8.830658062175811e-07, | |
| "loss": 2.5965, | |
| "step": 10940 | |
| }, | |
| { | |
| "epoch": 0.966588692236395, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 8.381754272159313e-07, | |
| "loss": 2.5802, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.9674714216356975, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 7.944528048430098e-07, | |
| "loss": 2.5859, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 0.9683541510350002, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 7.518982813940255e-07, | |
| "loss": 2.5864, | |
| "step": 10970 | |
| }, | |
| { | |
| "epoch": 0.9692368804343029, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 7.105121900194188e-07, | |
| "loss": 2.6153, | |
| "step": 10980 | |
| }, | |
| { | |
| "epoch": 0.9701196098336055, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 6.702948547221976e-07, | |
| "loss": 2.5774, | |
| "step": 10990 | |
| }, | |
| { | |
| "epoch": 0.9710023392329081, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 6.312465903554886e-07, | |
| "loss": 2.5937, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.9718850686322108, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 5.933677026199235e-07, | |
| "loss": 2.5982, | |
| "step": 11010 | |
| }, | |
| { | |
| "epoch": 0.9727677980315135, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 5.566584880613567e-07, | |
| "loss": 2.5894, | |
| "step": 11020 | |
| }, | |
| { | |
| "epoch": 0.973650527430816, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 5.211192340685843e-07, | |
| "loss": 2.5913, | |
| "step": 11030 | |
| }, | |
| { | |
| "epoch": 0.9745332568301187, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.867502188709127e-07, | |
| "loss": 2.5733, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 0.9754159862294214, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.535517115361764e-07, | |
| "loss": 2.5926, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.9762987156287241, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.215239719685071e-07, | |
| "loss": 2.5952, | |
| "step": 11060 | |
| }, | |
| { | |
| "epoch": 0.9771814450280266, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.9066725090636794e-07, | |
| "loss": 2.6035, | |
| "step": 11070 | |
| }, | |
| { | |
| "epoch": 0.9780641744273293, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.609817899205725e-07, | |
| "loss": 2.585, | |
| "step": 11080 | |
| }, | |
| { | |
| "epoch": 0.978946903826632, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.3246782141233553e-07, | |
| "loss": 2.5884, | |
| "step": 11090 | |
| }, | |
| { | |
| "epoch": 0.9798296332259346, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.051255686115417e-07, | |
| "loss": 2.5789, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.9807123626252372, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.7895524557497995e-07, | |
| "loss": 2.5947, | |
| "step": 11110 | |
| }, | |
| { | |
| "epoch": 0.9815950920245399, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 2.539570571845784e-07, | |
| "loss": 2.5986, | |
| "step": 11120 | |
| }, | |
| { | |
| "epoch": 0.9824778214238425, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.3013119914592204e-07, | |
| "loss": 2.5778, | |
| "step": 11130 | |
| }, | |
| { | |
| "epoch": 0.9833605508231452, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 2.0747785798667095e-07, | |
| "loss": 2.5914, | |
| "step": 11140 | |
| }, | |
| { | |
| "epoch": 0.9842432802224478, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 1.8599721105506116e-07, | |
| "loss": 2.587, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.9851260096217505, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 1.6568942651857263e-07, | |
| "loss": 2.5922, | |
| "step": 11160 | |
| }, | |
| { | |
| "epoch": 0.9860087390210531, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.4655466336261356e-07, | |
| "loss": 2.594, | |
| "step": 11170 | |
| }, | |
| { | |
| "epoch": 0.9868914684203557, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 1.2859307138920472e-07, | |
| "loss": 2.5975, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 0.9877741978196584, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.1180479121588037e-07, | |
| "loss": 2.588, | |
| "step": 11190 | |
| }, | |
| { | |
| "epoch": 0.988656927218961, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 9.618995427455589e-08, | |
| "loss": 2.584, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.9895396566182637, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 8.174868281054513e-08, | |
| "loss": 2.5937, | |
| "step": 11210 | |
| }, | |
| { | |
| "epoch": 0.9904223860175663, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 6.8481089881528e-08, | |
| "loss": 2.5943, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 0.9913051154168689, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 5.638727935666776e-08, | |
| "loss": 2.5936, | |
| "step": 11230 | |
| }, | |
| { | |
| "epoch": 0.9921878448161716, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.5467345915911615e-08, | |
| "loss": 2.5902, | |
| "step": 11240 | |
| }, | |
| { | |
| "epoch": 0.9930705742154743, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.57213750491081e-08, | |
| "loss": 2.5943, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.9939533036147769, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.7149443055424214e-08, | |
| "loss": 2.589, | |
| "step": 11260 | |
| }, | |
| { | |
| "epoch": 0.9948360330140795, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.9751617042712554e-08, | |
| "loss": 2.6015, | |
| "step": 11270 | |
| }, | |
| { | |
| "epoch": 0.9957187624133822, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.35279549269951e-08, | |
| "loss": 2.5815, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 0.9966014918126849, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 8.478505432030214e-09, | |
| "loss": 2.5944, | |
| "step": 11290 | |
| }, | |
| { | |
| "epoch": 0.9974842212119874, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.603308088879654e-09, | |
| "loss": 2.5838, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.9983669506112901, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.9023932356754258e-09, | |
| "loss": 2.5887, | |
| "step": 11310 | |
| }, | |
| { | |
| "epoch": 0.9992496800105928, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 3.757820173033721e-10, | |
| "loss": 2.5906, | |
| "step": 11320 | |
| }, | |
| { | |
| "epoch": 0.9999558635300348, | |
| "step": 11328, | |
| "total_flos": 1.9775705361382638e+20, | |
| "train_loss": 2.636019475715982, | |
| "train_runtime": 23008.8845, | |
| "train_samples_per_second": 126.042, | |
| "train_steps_per_second": 0.492 | |
| } | |
| ], | |
| "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": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.9775705361382638e+20, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |