| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 4347, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00644122383252818, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 2.727272727272727e-05, | |
| "loss": 1.3628, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01288244766505636, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 5.454545454545454e-05, | |
| "loss": 1.3272, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01932367149758454, | |
| "grad_norm": 0.375, | |
| "learning_rate": 8.18181818181818e-05, | |
| "loss": 1.2626, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02576489533011272, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00010909090909090908, | |
| "loss": 1.2028, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00013636363636363634, | |
| "loss": 1.1822, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03864734299516908, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001636363636363636, | |
| "loss": 1.2029, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04508856682769726, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.0001909090909090909, | |
| "loss": 1.1609, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05152979066022544, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00021818181818181816, | |
| "loss": 1.1137, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.057971014492753624, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00024545454545454545, | |
| "loss": 1.085, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.0002727272727272727, | |
| "loss": 1.1052, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07085346215780998, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0003, | |
| "loss": 1.0712, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.07729468599033816, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00029999936035650057, | |
| "loss": 1.0588, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08373590982286634, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.0002999974414314574, | |
| "loss": 1.0531, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09017713365539452, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00029999424324123633, | |
| "loss": 0.9953, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0002999897658131134, | |
| "loss": 0.9887, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10305958132045089, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.0002999840091852746, | |
| "loss": 0.9945, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.10950080515297907, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00029997697340681585, | |
| "loss": 0.9306, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.11594202898550725, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00029996865853774236, | |
| "loss": 0.9458, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.12238325281803543, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00029995906464896807, | |
| "loss": 0.9487, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002999481918223153, | |
| "loss": 0.9144, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13526570048309178, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.0002999360401505139, | |
| "loss": 0.9289, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.14170692431561996, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00029992260973720023, | |
| "loss": 0.882, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00029990790069691665, | |
| "loss": 0.9031, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.15458937198067632, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00029989191315511055, | |
| "loss": 0.9127, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0002998746472481328, | |
| "loss": 0.8803, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16747181964573268, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002998561031232371, | |
| "loss": 0.8764, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00029983628093857855, | |
| "loss": 0.9189, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.18035426731078905, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00029981518086321225, | |
| "loss": 0.8169, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.18679549114331723, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00029979280307709176, | |
| "loss": 0.8672, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002997691477710679, | |
| "loss": 0.8387, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1996779388083736, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002997442151468869, | |
| "loss": 0.8039, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.20611916264090177, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029971800541718854, | |
| "loss": 0.8294, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.21256038647342995, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002996905188055046, | |
| "loss": 0.8228, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.21900161030595813, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029966175554625696, | |
| "loss": 0.8341, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00029963171588475525, | |
| "loss": 0.8095, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2318840579710145, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002996004000771952, | |
| "loss": 0.8285, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.23832528180354268, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00029956780839065616, | |
| "loss": 0.8123, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.24476650563607086, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00029953394110309887, | |
| "loss": 0.7612, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.25120772946859904, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0002994987985033633, | |
| "loss": 0.7723, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002994623808911659, | |
| "loss": 0.8202, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2640901771336554, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00029942468857709715, | |
| "loss": 0.7324, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.27053140096618356, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.000299385721882619, | |
| "loss": 0.7818, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.27697262479871176, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.000299345481140062, | |
| "loss": 0.7693, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2834138486312399, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00029930396669262255, | |
| "loss": 0.7481, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00029926117889435993, | |
| "loss": 0.7478, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00029921711811019334, | |
| "loss": 0.7581, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3027375201288245, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00029917178471589864, | |
| "loss": 0.7131, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.30917874396135264, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002991251790981053, | |
| "loss": 0.7121, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.31561996779388085, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002990773016542932, | |
| "loss": 0.7385, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00029902815279278874, | |
| "loss": 0.743, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3285024154589372, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00029897773293276214, | |
| "loss": 0.6984, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.33494363929146537, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.000298926042504223, | |
| "loss": 0.7278, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3413848631239936, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00029887308194801745, | |
| "loss": 0.7043, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00029881885171582364, | |
| "loss": 0.7455, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002987633522701486, | |
| "loss": 0.7314, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3607085346215781, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00029870658408432375, | |
| "loss": 0.7344, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3671497584541063, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0002986485476425011, | |
| "loss": 0.7324, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.37359098228663445, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002985892434396491, | |
| "loss": 0.7197, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.38003220611916266, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00029852867198154837, | |
| "loss": 0.6616, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002984668337847874, | |
| "loss": 0.6325, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.392914653784219, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002984037293767583, | |
| "loss": 0.6445, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3993558776167472, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00029833935929565194, | |
| "loss": 0.6846, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4057971014492754, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00029827372409045377, | |
| "loss": 0.6976, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.41223832528180354, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0002982068243209389, | |
| "loss": 0.7165, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00029813866055766736, | |
| "loss": 0.6647, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4251207729468599, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00029806923338197925, | |
| "loss": 0.6809, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.43156199677938806, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029799854338598974, | |
| "loss": 0.7285, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.43800322061191627, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0002979265911725842, | |
| "loss": 0.6978, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029785337735541276, | |
| "loss": 0.6598, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002977789025588854, | |
| "loss": 0.6534, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4573268921095008, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0002977031674181663, | |
| "loss": 0.7261, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.463768115942029, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00029762617257916873, | |
| "loss": 0.6762, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.47020933977455714, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0002975479186985493, | |
| "loss": 0.6625, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.47665056360708535, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002974684064437025, | |
| "loss": 0.6617, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00029738763649275496, | |
| "loss": 0.6886, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4895330112721417, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002973056095345596, | |
| "loss": 0.6623, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.49597423510466987, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00029722232626869, | |
| "loss": 0.6568, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5024154589371981, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002971377874054341, | |
| "loss": 0.6281, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5088566827697263, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0002970519936657884, | |
| "loss": 0.6618, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00029696494578145157, | |
| "loss": 0.6797, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0002968766444948185, | |
| "loss": 0.6756, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5281803542673108, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002967870905589739, | |
| "loss": 0.698, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.534621578099839, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002966962847376855, | |
| "loss": 0.6431, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5410628019323671, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00029660422780539814, | |
| "loss": 0.6713, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00029651092054722665, | |
| "loss": 0.615, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5539452495974235, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002964163637589495, | |
| "loss": 0.7173, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5603864734299517, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0002963205582470017, | |
| "loss": 0.6808, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5668276972624798, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00029622350482846844, | |
| "loss": 0.6684, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.573268921095008, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00029612520433107734, | |
| "loss": 0.644, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002960256575931922, | |
| "loss": 0.6599, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5861513687600645, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0002959248654638053, | |
| "loss": 0.7006, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00029582282880253035, | |
| "loss": 0.625, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5990338164251208, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002957195484795952, | |
| "loss": 0.7234, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.605475040257649, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002956150253758344, | |
| "loss": 0.6556, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00029550926038268146, | |
| "loss": 0.6402, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6183574879227053, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002954022544021617, | |
| "loss": 0.6446, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.6247987117552335, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00029529400834688415, | |
| "loss": 0.6379, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.6312399355877617, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00029518452314003394, | |
| "loss": 0.644, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.6376811594202898, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0002950737997153645, | |
| "loss": 0.6413, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029496183901718927, | |
| "loss": 0.6249, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6505636070853462, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00029484864200037415, | |
| "loss": 0.5929, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6570048309178744, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002947342096303289, | |
| "loss": 0.6447, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6634460547504025, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002946185428829991, | |
| "loss": 0.641, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6698872785829307, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0002945016427448579, | |
| "loss": 0.6878, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002943835102128975, | |
| "loss": 0.6646, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6827697262479872, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002942641462946206, | |
| "loss": 0.613, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6892109500805152, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00029414355200803197, | |
| "loss": 0.6135, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002940217283816296, | |
| "loss": 0.6145, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.7020933977455717, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002938986764543961, | |
| "loss": 0.6199, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002937743972757895, | |
| "loss": 0.6566, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.714975845410628, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002936488919057349, | |
| "loss": 0.6536, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.7214170692431562, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002935221614146148, | |
| "loss": 0.6586, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.7278582930756844, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0002933942068832604, | |
| "loss": 0.6234, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.7342995169082126, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00029326502940294207, | |
| "loss": 0.6115, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00029313463007536034, | |
| "loss": 0.6205, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7471819645732689, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002930030100126363, | |
| "loss": 0.6185, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.7536231884057971, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002928701703373021, | |
| "loss": 0.6305, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.7600644122383253, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029273611218229165, | |
| "loss": 0.6464, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.7665056360708534, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002926008366909307, | |
| "loss": 0.6488, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00029246434501692685, | |
| "loss": 0.6148, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7793880837359098, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00029232663832436047, | |
| "loss": 0.5946, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.785829307568438, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002921877177876741, | |
| "loss": 0.5898, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7922705314009661, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002920475845916626, | |
| "loss": 0.6435, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7987117552334944, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00029190623993146313, | |
| "loss": 0.6605, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002917636850125449, | |
| "loss": 0.6297, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8115942028985508, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00029161992105069905, | |
| "loss": 0.6313, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.8180354267310789, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002914749492720279, | |
| "loss": 0.5953, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.8244766505636071, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00029132877091293493, | |
| "loss": 0.6615, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.8309178743961353, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.000291181387220114, | |
| "loss": 0.6771, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002910327994505387, | |
| "loss": 0.5889, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8438003220611916, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002908830088714516, | |
| "loss": 0.5781, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.8502415458937198, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00029073201676035383, | |
| "loss": 0.6182, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.856682769726248, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00029057982440499356, | |
| "loss": 0.6226, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.8631239935587761, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00029042643310335547, | |
| "loss": 0.6547, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00029027184416364956, | |
| "loss": 0.6114, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8760064412238325, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002901160589043, | |
| "loss": 0.6491, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8824476650563607, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00028995907865393385, | |
| "loss": 0.6375, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00028980090475136963, | |
| "loss": 0.6083, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.895330112721417, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002896415385456062, | |
| "loss": 0.5879, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.000289480981395811, | |
| "loss": 0.6596, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9082125603864735, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00028931923467130855, | |
| "loss": 0.5774, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.9146537842190016, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00028915629975156867, | |
| "loss": 0.6118, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.9210950080515298, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0002889921780261949, | |
| "loss": 0.615, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.927536231884058, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00028882687089491234, | |
| "loss": 0.6225, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002886603797675563, | |
| "loss": 0.5626, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9404186795491143, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002884927060640596, | |
| "loss": 0.5886, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.9468599033816425, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0002883238512144409, | |
| "loss": 0.6251, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.9533011272141707, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002881538166587921, | |
| "loss": 0.6326, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.9597423510466989, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002879826038472667, | |
| "loss": 0.5666, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00028781021424006677, | |
| "loss": 0.5282, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9726247987117552, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00028763664930743087, | |
| "loss": 0.6628, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.9790660225442834, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00028746191052962146, | |
| "loss": 0.5669, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9855072463768116, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00028728599939691215, | |
| "loss": 0.5955, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9919484702093397, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00028710891740957507, | |
| "loss": 0.5995, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00028693066607786823, | |
| "loss": 0.5813, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0048309178743962, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002867512469220222, | |
| "loss": 0.5306, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.0112721417069244, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00028657066147222773, | |
| "loss": 0.4918, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.0177133655394526, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00028638891126862224, | |
| "loss": 0.5198, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.0241545893719808, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0002862059978612769, | |
| "loss": 0.5673, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.0305958132045088, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00028602192281018327, | |
| "loss": 0.5127, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002858366876852403, | |
| "loss": 0.5517, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002856502940662403, | |
| "loss": 0.5209, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.0499194847020934, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00028546274354285646, | |
| "loss": 0.5362, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.0563607085346216, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00028527403771462826, | |
| "loss": 0.5256, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.0628019323671498, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00028508417819094844, | |
| "loss": 0.5257, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.069243156199678, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0002848931665910492, | |
| "loss": 0.4971, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.075684380032206, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002847010045439882, | |
| "loss": 0.5214, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.0821256038647342, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0002845076936886349, | |
| "loss": 0.5283, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0885668276972624, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002843132356736563, | |
| "loss": 0.5024, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.0950080515297906, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002841176321575032, | |
| "loss": 0.5515, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.1014492753623188, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002839208848083958, | |
| "loss": 0.5493, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.107890499194847, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0002837229953043096, | |
| "loss": 0.4908, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.1143317230273753, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002835239653329611, | |
| "loss": 0.5136, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.1207729468599035, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002833237965917934, | |
| "loss": 0.5379, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.1272141706924317, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002831224907879614, | |
| "loss": 0.5059, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1336553945249597, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00028292004963831796, | |
| "loss": 0.5231, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.1400966183574879, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00028271647486939855, | |
| "loss": 0.5223, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.146537842190016, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002825117682174069, | |
| "loss": 0.4907, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.1529790660225443, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002823059314282, | |
| "loss": 0.4996, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.1594202898550725, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0002820989662572734, | |
| "loss": 0.5084, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1658615136876007, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002818908744697461, | |
| "loss": 0.4909, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.1723027375201287, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00028168165784034566, | |
| "loss": 0.5245, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.178743961352657, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00028147131815339267, | |
| "loss": 0.5307, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00028125985720278614, | |
| "loss": 0.5213, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.1916264090177133, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0002810472767919876, | |
| "loss": 0.5257, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1980676328502415, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0002808335787340061, | |
| "loss": 0.4913, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.2045088566827697, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00028061876485138264, | |
| "loss": 0.5331, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.210950080515298, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00028040283697617464, | |
| "loss": 0.5055, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002801857969499402, | |
| "loss": 0.5318, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.2238325281803544, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0002799676466237225, | |
| "loss": 0.4991, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.2302737520128824, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002797483878580342, | |
| "loss": 0.5059, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.2367149758454106, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00027952802252284104, | |
| "loss": 0.5043, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.2431561996779388, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0002793065524975465, | |
| "loss": 0.5747, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.249597423510467, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0002790839796709755, | |
| "loss": 0.5082, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.2560386473429952, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00027886030594135805, | |
| "loss": 0.5369, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2624798711755234, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002786355332163135, | |
| "loss": 0.5423, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.2689210950080514, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.000278409663412834, | |
| "loss": 0.4882, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.2753623188405796, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002781826984572683, | |
| "loss": 0.504, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.2818035426731078, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0002779546402853051, | |
| "loss": 0.4872, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.288244766505636, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00027772549084195675, | |
| "loss": 0.5348, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2946859903381642, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00027749525208154265, | |
| "loss": 0.5718, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.3011272141706924, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002772639259676726, | |
| "loss": 0.5393, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.3075684380032206, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00027703151447322965, | |
| "loss": 0.5421, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.3140096618357489, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002767980195803539, | |
| "loss": 0.5555, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.320450885668277, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0002765634432804253, | |
| "loss": 0.553, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.3268921095008053, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00027632778757404655, | |
| "loss": 0.5075, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002760910544710261, | |
| "loss": 0.4933, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.3397745571658615, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00027585324599036133, | |
| "loss": 0.5039, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.3462157809983897, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00027561436416022073, | |
| "loss": 0.5175, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.3526570048309179, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00027537441101792715, | |
| "loss": 0.5375, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.359098228663446, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002751333886099402, | |
| "loss": 0.5235, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.3655394524959743, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002748912989918387, | |
| "loss": 0.4882, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.3719806763285023, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002746481442283034, | |
| "loss": 0.5032, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.3784219001610305, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002744039263930991, | |
| "loss": 0.5052, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.3848631239935587, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002741586475690571, | |
| "loss": 0.5538, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002739123098480576, | |
| "loss": 0.5457, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.3977455716586151, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00027366491533101147, | |
| "loss": 0.5111, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.4041867954911433, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0002734164661278426, | |
| "loss": 0.4902, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.4106280193236715, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00027316696435747, | |
| "loss": 0.5504, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.4170692431561998, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00027291641214778937, | |
| "loss": 0.5234, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.423510466988728, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002726648116356554, | |
| "loss": 0.5052, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.4299516908212562, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.000272412164966863, | |
| "loss": 0.5189, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.4363929146537842, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00027215847429612965, | |
| "loss": 0.4982, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.4428341384863124, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002719037417870765, | |
| "loss": 0.4916, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.4492753623188406, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00027164796961221015, | |
| "loss": 0.5149, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4557165861513688, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002713911599529039, | |
| "loss": 0.5636, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.462157809983897, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00027113331499937967, | |
| "loss": 0.5191, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.4685990338164252, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00027087443695068873, | |
| "loss": 0.4786, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.4750402576489532, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0002706145280146931, | |
| "loss": 0.5033, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00027035359040804703, | |
| "loss": 0.4753, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.4879227053140096, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002700916263561778, | |
| "loss": 0.5255, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.4943639291465378, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0002698286380932667, | |
| "loss": 0.5472, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.500805152979066, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0002695646278622302, | |
| "loss": 0.4944, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.5072463768115942, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002692995979147007, | |
| "loss": 0.4677, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.5136876006441224, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00026903355051100734, | |
| "loss": 0.5152, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.5201288244766507, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002687664879201565, | |
| "loss": 0.5287, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.5265700483091789, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00026849841241981313, | |
| "loss": 0.5185, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.533011272141707, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00026822932629628034, | |
| "loss": 0.4925, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.539452495974235, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002679592318444808, | |
| "loss": 0.4938, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.5458937198067633, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0002676881313679366, | |
| "loss": 0.4962, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5523349436392915, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002674160271787498, | |
| "loss": 0.4962, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.5587761674718197, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002671429215975828, | |
| "loss": 0.5142, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00026686881695363833, | |
| "loss": 0.5361, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.5716586151368759, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002665937155846399, | |
| "loss": 0.519, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.578099838969404, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0002663176198368114, | |
| "loss": 0.5055, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5845410628019323, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0002660405320648576, | |
| "loss": 0.5256, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.5909822866344605, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002657624546319437, | |
| "loss": 0.5103, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.5974235104669887, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0002654833899096753, | |
| "loss": 0.5249, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.603864734299517, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00026520334027807827, | |
| "loss": 0.4895, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.6103059581320451, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0002649223081255782, | |
| "loss": 0.5061, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6167471819645733, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00026464029584898036, | |
| "loss": 0.4781, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.6231884057971016, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00026435730585344896, | |
| "loss": 0.4885, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.6296296296296298, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002640733405524869, | |
| "loss": 0.5188, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.636070853462158, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00026378840236791485, | |
| "loss": 0.5386, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.642512077294686, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.000263502493729851, | |
| "loss": 0.5438, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6489533011272142, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00026321561707668995, | |
| "loss": 0.5121, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.6553945249597424, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002629277748550823, | |
| "loss": 0.4868, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.6618357487922706, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002626389695199134, | |
| "loss": 0.5199, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.6682769726247986, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002623492035342826, | |
| "loss": 0.5424, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.6747181964573268, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00026205847936948244, | |
| "loss": 0.4983, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.681159420289855, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00026176679950497706, | |
| "loss": 0.5323, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.6876006441223832, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002614741664283816, | |
| "loss": 0.5964, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.6940418679549114, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00026118058263544056, | |
| "loss": 0.5227, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.7004830917874396, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00026088605063000696, | |
| "loss": 0.464, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.7069243156199678, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0002605905729240205, | |
| "loss": 0.4978, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.713365539452496, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00026029415203748633, | |
| "loss": 0.4983, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.7198067632850242, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0002599967904984539, | |
| "loss": 0.5166, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.7262479871175525, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00025969849084299466, | |
| "loss": 0.5683, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.7326892109500807, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00025939925561518126, | |
| "loss": 0.486, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002590990873670652, | |
| "loss": 0.4655, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.7455716586151369, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00025879798865865533, | |
| "loss": 0.4689, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.752012882447665, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002584959620578962, | |
| "loss": 0.424, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.7584541062801933, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00025819301014064574, | |
| "loss": 0.5134, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.7648953301127213, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002578891354906537, | |
| "loss": 0.4893, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.7713365539452495, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00025758434069953927, | |
| "loss": 0.4887, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0002572786283667692, | |
| "loss": 0.5153, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.7842190016103059, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00025697200109963563, | |
| "loss": 0.5056, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.790660225442834, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002566644615132337, | |
| "loss": 0.5319, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.7971014492753623, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00025635601223043933, | |
| "loss": 0.5182, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.8035426731078905, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.000256046655881887, | |
| "loss": 0.5028, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.8099838969404187, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.000255736395105947, | |
| "loss": 0.5006, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 1.816425120772947, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0002554252325487032, | |
| "loss": 0.5234, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 1.8228663446054751, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002551131708639303, | |
| "loss": 0.5544, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 1.8293075684380034, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00025480021271307156, | |
| "loss": 0.4766, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 1.8357487922705316, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00025448636076521534, | |
| "loss": 0.4615, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8421900161030595, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002541716176970732, | |
| "loss": 0.504, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 1.8486312399355878, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002538559861929566, | |
| "loss": 0.5873, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.855072463768116, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.000253539468944754, | |
| "loss": 0.5917, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 1.8615136876006442, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002532220686519081, | |
| "loss": 0.4924, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 1.8679549114331722, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00025290378802139273, | |
| "loss": 0.4582, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8743961352657004, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0002525846297676896, | |
| "loss": 0.5639, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.8808373590982286, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0002522645966127655, | |
| "loss": 0.5198, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 1.8872785829307568, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002519436912860488, | |
| "loss": 0.4766, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 1.893719806763285, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002516219165244062, | |
| "loss": 0.4583, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.9001610305958132, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0002512992750721195, | |
| "loss": 0.549, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.9066022544283414, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002509757696808622, | |
| "loss": 0.4792, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0002506514031096758, | |
| "loss": 0.4834, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 1.9194847020933978, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00025032617812494664, | |
| "loss": 0.4969, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 1.925925925925926, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00025000009750038196, | |
| "loss": 0.5553, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 1.9323671497584543, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00024967316401698647, | |
| "loss": 0.536, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.9388083735909822, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00024934538046303856, | |
| "loss": 0.4848, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.9452495974235104, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002490167496340664, | |
| "loss": 0.4984, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 1.9516908212560387, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0002486872743328244, | |
| "loss": 0.4993, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 1.9581320450885669, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.000248356957369269, | |
| "loss": 0.5265, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 1.9645732689210949, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0002480258015605349, | |
| "loss": 0.5287, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.971014492753623, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002476938097309108, | |
| "loss": 0.5616, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 1.9774557165861513, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002473609847118156, | |
| "loss": 0.4542, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 1.9838969404186795, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0002470273293417741, | |
| "loss": 0.4813, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.9903381642512077, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00024669284646639287, | |
| "loss": 0.5336, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 1.996779388083736, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00024635753893833585, | |
| "loss": 0.5528, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.003220611916264, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00024602140961730006, | |
| "loss": 0.4706, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 2.0096618357487923, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00024568446136999134, | |
| "loss": 0.4093, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 2.0161030595813205, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00024534669707009974, | |
| "loss": 0.3899, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 2.0225442834138487, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0002450081195982752, | |
| "loss": 0.361, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 2.028985507246377, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00024466873184210273, | |
| "loss": 0.3999, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.035426731078905, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00024432853669607786, | |
| "loss": 0.3753, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 2.0418679549114334, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00024398753706158225, | |
| "loss": 0.3951, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 2.0483091787439616, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00024364573584685848, | |
| "loss": 0.3791, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 2.0547504025764893, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00024330313596698553, | |
| "loss": 0.4148, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 2.0611916264090175, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00024295974034385396, | |
| "loss": 0.3767, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.0676328502415457, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00024261555190614072, | |
| "loss": 0.3743, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 2.074074074074074, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00024227057358928452, | |
| "loss": 0.3847, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 2.080515297906602, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00024192480833546044, | |
| "loss": 0.3627, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00024157825909355523, | |
| "loss": 0.4324, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 2.0933977455716586, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0002412309288191417, | |
| "loss": 0.4302, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.099838969404187, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00024088282047445396, | |
| "loss": 0.3788, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 2.106280193236715, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00024053393702836185, | |
| "loss": 0.399, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 2.112721417069243, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0002401842814563457, | |
| "loss": 0.387, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 2.1191626409017714, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00023983385674047113, | |
| "loss": 0.3905, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 2.1256038647342996, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00023948266586936324, | |
| "loss": 0.3715, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.132045088566828, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00023913071183818155, | |
| "loss": 0.4474, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 2.138486312399356, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00023877799764859416, | |
| "loss": 0.3759, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 2.1449275362318843, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00023842452630875216, | |
| "loss": 0.373, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 2.151368760064412, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0002380703008332643, | |
| "loss": 0.4218, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 2.1578099838969402, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0002377153242431708, | |
| "loss": 0.4234, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.1642512077294684, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00023735959956591786, | |
| "loss": 0.3971, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 2.1706924315619966, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.0002370031298353319, | |
| "loss": 0.4211, | |
| "step": 1348 | |
| }, | |
| { | |
| "epoch": 2.177133655394525, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00023664591809159353, | |
| "loss": 0.3972, | |
| "step": 1352 | |
| }, | |
| { | |
| "epoch": 2.183574879227053, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00023628796738121169, | |
| "loss": 0.4185, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 2.1900161030595813, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00023592928075699763, | |
| "loss": 0.402, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1964573268921095, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00023556986127803894, | |
| "loss": 0.4056, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 2.2028985507246377, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00023520971200967334, | |
| "loss": 0.4506, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 2.209339774557166, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00023484883602346274, | |
| "loss": 0.4093, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 2.215780998389694, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.0002344872363971668, | |
| "loss": 0.4717, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00023412491621471694, | |
| "loss": 0.3948, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.2286634460547505, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00023376187856618972, | |
| "loss": 0.3925, | |
| "step": 1384 | |
| }, | |
| { | |
| "epoch": 2.2351046698872787, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00023339812654778083, | |
| "loss": 0.4324, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 2.241545893719807, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0002330336632617784, | |
| "loss": 0.4557, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 2.247987117552335, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00023266849181653683, | |
| "loss": 0.4301, | |
| "step": 1396 | |
| }, | |
| { | |
| "epoch": 2.2544283413848634, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00023230261532644985, | |
| "loss": 0.3799, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.260869565217391, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002319360369119245, | |
| "loss": 0.3826, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 2.2673107890499193, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00023156875969935405, | |
| "loss": 0.3862, | |
| "step": 1408 | |
| }, | |
| { | |
| "epoch": 2.2737520128824475, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00023120078682109158, | |
| "loss": 0.4269, | |
| "step": 1412 | |
| }, | |
| { | |
| "epoch": 2.2801932367149758, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00023083212141542328, | |
| "loss": 0.4139, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 2.286634460547504, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00023046276662654143, | |
| "loss": 0.3579, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.293075684380032, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00023009272560451803, | |
| "loss": 0.4, | |
| "step": 1424 | |
| }, | |
| { | |
| "epoch": 2.2995169082125604, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00022972200150527745, | |
| "loss": 0.3937, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 2.3059581320450886, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00022935059749056992, | |
| "loss": 0.4553, | |
| "step": 1432 | |
| }, | |
| { | |
| "epoch": 2.312399355877617, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00022897851672794417, | |
| "loss": 0.396, | |
| "step": 1436 | |
| }, | |
| { | |
| "epoch": 2.318840579710145, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00022860576239072084, | |
| "loss": 0.5137, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.325281803542673, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00022823233765796502, | |
| "loss": 0.4085, | |
| "step": 1444 | |
| }, | |
| { | |
| "epoch": 2.3317230273752014, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.0002278582457144595, | |
| "loss": 0.3963, | |
| "step": 1448 | |
| }, | |
| { | |
| "epoch": 2.3381642512077296, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00022748348975067733, | |
| "loss": 0.4377, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 2.3446054750402574, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00022710807296275472, | |
| "loss": 0.4275, | |
| "step": 1456 | |
| }, | |
| { | |
| "epoch": 2.3510466988727856, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.0002267319985524637, | |
| "loss": 0.4089, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.357487922705314, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00022635526972718508, | |
| "loss": 0.4386, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 2.363929146537842, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002259778896998807, | |
| "loss": 0.4172, | |
| "step": 1468 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00022559986168906637, | |
| "loss": 0.4022, | |
| "step": 1472 | |
| }, | |
| { | |
| "epoch": 2.3768115942028984, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00022522118891878418, | |
| "loss": 0.4665, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 2.3832528180354267, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00022484187461857517, | |
| "loss": 0.3916, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.389694041867955, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00022446192202345156, | |
| "loss": 0.3918, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 2.396135265700483, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00022408133437386968, | |
| "loss": 0.4198, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 2.4025764895330113, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00022370011491570162, | |
| "loss": 0.3635, | |
| "step": 1492 | |
| }, | |
| { | |
| "epoch": 2.4090177133655395, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.000223318266900208, | |
| "loss": 0.4297, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 2.4154589371980677, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00022293579358401023, | |
| "loss": 0.3819, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.421900161030596, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002225526982290625, | |
| "loss": 0.4068, | |
| "step": 1504 | |
| }, | |
| { | |
| "epoch": 2.428341384863124, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00022216898410262428, | |
| "loss": 0.3808, | |
| "step": 1508 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00022178465447723214, | |
| "loss": 0.4037, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 2.4412238325281805, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.000221399712630672, | |
| "loss": 0.452, | |
| "step": 1516 | |
| }, | |
| { | |
| "epoch": 2.4476650563607087, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0002210141618459513, | |
| "loss": 0.4127, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.454106280193237, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00022062800541127064, | |
| "loss": 0.3894, | |
| "step": 1524 | |
| }, | |
| { | |
| "epoch": 2.4605475040257647, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00022024124661999613, | |
| "loss": 0.4256, | |
| "step": 1528 | |
| }, | |
| { | |
| "epoch": 2.466988727858293, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00021985388877063104, | |
| "loss": 0.4556, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 2.473429951690821, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00021946593516678777, | |
| "loss": 0.4504, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 2.4798711755233493, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00021907738911715964, | |
| "loss": 0.4062, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.4863123993558776, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00021868825393549275, | |
| "loss": 0.4386, | |
| "step": 1544 | |
| }, | |
| { | |
| "epoch": 2.4927536231884058, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0002182985329405576, | |
| "loss": 0.3559, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 2.499194847020934, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00021790822945612088, | |
| "loss": 0.4244, | |
| "step": 1552 | |
| }, | |
| { | |
| "epoch": 2.505636070853462, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0002175173468109171, | |
| "loss": 0.4028, | |
| "step": 1556 | |
| }, | |
| { | |
| "epoch": 2.5120772946859904, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00021712588833862014, | |
| "loss": 0.3726, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.5185185185185186, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00021673385737781492, | |
| "loss": 0.481, | |
| "step": 1564 | |
| }, | |
| { | |
| "epoch": 2.524959742351047, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00021634125727196883, | |
| "loss": 0.3778, | |
| "step": 1568 | |
| }, | |
| { | |
| "epoch": 2.531400966183575, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00021594809136940327, | |
| "loss": 0.4438, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 2.537842190016103, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00021555436302326514, | |
| "loss": 0.4399, | |
| "step": 1576 | |
| }, | |
| { | |
| "epoch": 2.544283413848631, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00021516007559149803, | |
| "loss": 0.3979, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.550724637681159, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00021476523243681397, | |
| "loss": 0.4085, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 2.5571658615136874, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0002143698369266643, | |
| "loss": 0.3875, | |
| "step": 1588 | |
| }, | |
| { | |
| "epoch": 2.5636070853462156, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0002139738924332113, | |
| "loss": 0.4288, | |
| "step": 1592 | |
| }, | |
| { | |
| "epoch": 2.570048309178744, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0002135774023332992, | |
| "loss": 0.4155, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 2.576489533011272, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00021318037000842558, | |
| "loss": 0.377, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.5829307568438002, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00021278279884471242, | |
| "loss": 0.4134, | |
| "step": 1604 | |
| }, | |
| { | |
| "epoch": 2.5893719806763285, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0002123846922328771, | |
| "loss": 0.3668, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 2.5958132045088567, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00021198605356820377, | |
| "loss": 0.4207, | |
| "step": 1612 | |
| }, | |
| { | |
| "epoch": 2.602254428341385, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00021158688625051416, | |
| "loss": 0.434, | |
| "step": 1616 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00021118719368413866, | |
| "loss": 0.3963, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.6151368760064413, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0002107869792778873, | |
| "loss": 0.4366, | |
| "step": 1624 | |
| }, | |
| { | |
| "epoch": 2.6215780998389695, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00021038624644502063, | |
| "loss": 0.3604, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 2.6280193236714977, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00020998499860322073, | |
| "loss": 0.4029, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 2.634460547504026, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00020958323917456186, | |
| "loss": 0.429, | |
| "step": 1636 | |
| }, | |
| { | |
| "epoch": 2.640901771336554, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00020918097158548145, | |
| "loss": 0.426, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.6473429951690823, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0002087781992667509, | |
| "loss": 0.4129, | |
| "step": 1644 | |
| }, | |
| { | |
| "epoch": 2.6537842190016105, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.000208374925653446, | |
| "loss": 0.3946, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 2.6602254428341388, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00020797115418491816, | |
| "loss": 0.3564, | |
| "step": 1652 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00020756688830476453, | |
| "loss": 0.4553, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 2.6731078904991947, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0002071621314607991, | |
| "loss": 0.4497, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.679549114331723, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00020675688710502293, | |
| "loss": 0.3987, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 2.685990338164251, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00020635115869359498, | |
| "loss": 0.4695, | |
| "step": 1668 | |
| }, | |
| { | |
| "epoch": 2.6924315619967794, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0002059449496868024, | |
| "loss": 0.4566, | |
| "step": 1672 | |
| }, | |
| { | |
| "epoch": 2.6988727858293076, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00020553826354903121, | |
| "loss": 0.4199, | |
| "step": 1676 | |
| }, | |
| { | |
| "epoch": 2.7053140096618358, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00020513110374873676, | |
| "loss": 0.3612, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.711755233494364, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00020472347375841384, | |
| "loss": 0.383, | |
| "step": 1684 | |
| }, | |
| { | |
| "epoch": 2.718196457326892, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002043153770545675, | |
| "loss": 0.4051, | |
| "step": 1688 | |
| }, | |
| { | |
| "epoch": 2.7246376811594204, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00020390681711768312, | |
| "loss": 0.4408, | |
| "step": 1692 | |
| }, | |
| { | |
| "epoch": 2.7310789049919486, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00020349779743219682, | |
| "loss": 0.4155, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 2.7375201288244764, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0002030883214864657, | |
| "loss": 0.4164, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7439613526570046, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0002026783927727381, | |
| "loss": 0.4013, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 2.750402576489533, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00020226801478712383, | |
| "loss": 0.3839, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 2.756843800322061, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00020185719102956438, | |
| "loss": 0.4691, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 2.763285024154589, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0002014459250038031, | |
| "loss": 0.3949, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 2.7697262479871174, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00020103422021735507, | |
| "loss": 0.3918, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.7761674718196456, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00020062208018147755, | |
| "loss": 0.4027, | |
| "step": 1724 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00020020950841113984, | |
| "loss": 0.4319, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 2.789049919484702, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00019979650842499324, | |
| "loss": 0.4255, | |
| "step": 1732 | |
| }, | |
| { | |
| "epoch": 2.7954911433172303, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00019938308374534115, | |
| "loss": 0.4403, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 2.8019323671497585, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00019896923789810905, | |
| "loss": 0.4311, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.8083735909822867, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019855497441281436, | |
| "loss": 0.407, | |
| "step": 1744 | |
| }, | |
| { | |
| "epoch": 2.814814814814815, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00019814029682253644, | |
| "loss": 0.4184, | |
| "step": 1748 | |
| }, | |
| { | |
| "epoch": 2.821256038647343, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00019772520866388605, | |
| "loss": 0.3812, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 2.8276972624798713, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019730971347697602, | |
| "loss": 0.4228, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 2.8341384863123995, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00019689381480539014, | |
| "loss": 0.4321, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.8405797101449277, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00019647751619615353, | |
| "loss": 0.4321, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 2.847020933977456, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00019606082119970214, | |
| "loss": 0.4502, | |
| "step": 1768 | |
| }, | |
| { | |
| "epoch": 2.853462157809984, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00019564373336985268, | |
| "loss": 0.4298, | |
| "step": 1772 | |
| }, | |
| { | |
| "epoch": 2.8599033816425123, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00019522625626377198, | |
| "loss": 0.4469, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 2.86634460547504, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019480839344194695, | |
| "loss": 0.4033, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.8727858293075683, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00019439014846815413, | |
| "loss": 0.4381, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.8792270531400965, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00019397152490942919, | |
| "loss": 0.4205, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 2.8856682769726247, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019355252633603668, | |
| "loss": 0.4187, | |
| "step": 1792 | |
| }, | |
| { | |
| "epoch": 2.892109500805153, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019313315632143944, | |
| "loss": 0.3912, | |
| "step": 1796 | |
| }, | |
| { | |
| "epoch": 2.898550724637681, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019271341844226812, | |
| "loss": 0.4236, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.9049919484702094, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001922933162782909, | |
| "loss": 0.3677, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 2.9114331723027376, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00019187285341238261, | |
| "loss": 0.3979, | |
| "step": 1808 | |
| }, | |
| { | |
| "epoch": 2.917874396135266, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00019145203343049453, | |
| "loss": 0.3967, | |
| "step": 1812 | |
| }, | |
| { | |
| "epoch": 2.924315619967794, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019103085992162343, | |
| "loss": 0.4128, | |
| "step": 1816 | |
| }, | |
| { | |
| "epoch": 2.930756843800322, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00019060933647778135, | |
| "loss": 0.3968, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.9371980676328504, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019018746669396464, | |
| "loss": 0.4208, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 2.943639291465378, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018976525416812358, | |
| "loss": 0.413, | |
| "step": 1828 | |
| }, | |
| { | |
| "epoch": 2.9500805152979064, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00018934270250113135, | |
| "loss": 0.4122, | |
| "step": 1832 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00018891981529675376, | |
| "loss": 0.3961, | |
| "step": 1836 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018849659616161808, | |
| "loss": 0.4498, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.969404186795491, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018807304870518263, | |
| "loss": 0.3935, | |
| "step": 1844 | |
| }, | |
| { | |
| "epoch": 2.975845410628019, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00018764917653970567, | |
| "loss": 0.4183, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 2.9822866344605474, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.000187224983280215, | |
| "loss": 0.4101, | |
| "step": 1852 | |
| }, | |
| { | |
| "epoch": 2.9887278582930756, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00018680047254447665, | |
| "loss": 0.4587, | |
| "step": 1856 | |
| }, | |
| { | |
| "epoch": 2.995169082125604, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001863756479529644, | |
| "loss": 0.4216, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.001610305958132, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00018595051312882892, | |
| "loss": 0.3842, | |
| "step": 1864 | |
| }, | |
| { | |
| "epoch": 3.0080515297906603, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018552507169786634, | |
| "loss": 0.3189, | |
| "step": 1868 | |
| }, | |
| { | |
| "epoch": 3.0144927536231885, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018509932728848804, | |
| "loss": 0.3061, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 3.0209339774557167, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018467328353168934, | |
| "loss": 0.3166, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 3.027375201288245, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00018424694406101838, | |
| "loss": 0.3081, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.033816425120773, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001838203125125455, | |
| "loss": 0.2944, | |
| "step": 1884 | |
| }, | |
| { | |
| "epoch": 3.0402576489533013, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00018339339252483196, | |
| "loss": 0.285, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 3.0466988727858295, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00018296618773889912, | |
| "loss": 0.2926, | |
| "step": 1892 | |
| }, | |
| { | |
| "epoch": 3.0531400966183573, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.000182538701798197, | |
| "loss": 0.3019, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 3.0595813204508855, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00018211093834857379, | |
| "loss": 0.2984, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.0660225442834137, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018168290103824422, | |
| "loss": 0.3185, | |
| "step": 1904 | |
| }, | |
| { | |
| "epoch": 3.072463768115942, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018125459351775873, | |
| "loss": 0.3192, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 3.07890499194847, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00018082601943997232, | |
| "loss": 0.3459, | |
| "step": 1912 | |
| }, | |
| { | |
| "epoch": 3.0853462157809983, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018039718246001325, | |
| "loss": 0.2837, | |
| "step": 1916 | |
| }, | |
| { | |
| "epoch": 3.0917874396135265, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.000179968086235252, | |
| "loss": 0.3134, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.0982286634460547, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00017953873442527008, | |
| "loss": 0.2907, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 3.104669887278583, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00017910913069182872, | |
| "loss": 0.3076, | |
| "step": 1928 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017867927869883775, | |
| "loss": 0.293, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 3.1175523349436394, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017824918211232422, | |
| "loss": 0.2775, | |
| "step": 1936 | |
| }, | |
| { | |
| "epoch": 3.1239935587761676, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017781884460040136, | |
| "loss": 0.3037, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00017738826983323703, | |
| "loss": 0.3139, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 3.136876006441224, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00017695746148302252, | |
| "loss": 0.3081, | |
| "step": 1948 | |
| }, | |
| { | |
| "epoch": 3.143317230273752, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017652642322394142, | |
| "loss": 0.344, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 3.14975845410628, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00017609515873213787, | |
| "loss": 0.3006, | |
| "step": 1956 | |
| }, | |
| { | |
| "epoch": 3.156199677938808, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017566367168568572, | |
| "loss": 0.2933, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.1626409017713364, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017523196576455663, | |
| "loss": 0.2869, | |
| "step": 1964 | |
| }, | |
| { | |
| "epoch": 3.1690821256038646, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017480004465058918, | |
| "loss": 0.2935, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 3.175523349436393, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00017436791202745706, | |
| "loss": 0.3451, | |
| "step": 1972 | |
| }, | |
| { | |
| "epoch": 3.181964573268921, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00017393557158063803, | |
| "loss": 0.3047, | |
| "step": 1976 | |
| }, | |
| { | |
| "epoch": 3.1884057971014492, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00017350302699738204, | |
| "loss": 0.327, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.1948470209339774, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00017307028196668028, | |
| "loss": 0.3238, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 3.2012882447665056, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001726373401792333, | |
| "loss": 0.2957, | |
| "step": 1988 | |
| }, | |
| { | |
| "epoch": 3.207729468599034, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00017220420532741977, | |
| "loss": 0.3124, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 3.214170692431562, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00017177088110526486, | |
| "loss": 0.2852, | |
| "step": 1996 | |
| }, | |
| { | |
| "epoch": 3.2206119162640903, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00017133737120840907, | |
| "loss": 0.3084, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2270531400966185, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.000170903679334076, | |
| "loss": 0.2671, | |
| "step": 2004 | |
| }, | |
| { | |
| "epoch": 3.2334943639291467, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00017046980918104164, | |
| "loss": 0.2851, | |
| "step": 2008 | |
| }, | |
| { | |
| "epoch": 3.239935587761675, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001700357644496022, | |
| "loss": 0.2921, | |
| "step": 2012 | |
| }, | |
| { | |
| "epoch": 3.246376811594203, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00016960154884154298, | |
| "loss": 0.2898, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 3.2528180354267313, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00016916716606010646, | |
| "loss": 0.3277, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.259259259259259, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00016873261980996095, | |
| "loss": 0.3301, | |
| "step": 2024 | |
| }, | |
| { | |
| "epoch": 3.2657004830917873, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00016829791379716896, | |
| "loss": 0.3639, | |
| "step": 2028 | |
| }, | |
| { | |
| "epoch": 3.2721417069243155, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00016786305172915544, | |
| "loss": 0.3492, | |
| "step": 2032 | |
| }, | |
| { | |
| "epoch": 3.2785829307568437, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001674280373146763, | |
| "loss": 0.3233, | |
| "step": 2036 | |
| }, | |
| { | |
| "epoch": 3.285024154589372, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00016699287426378683, | |
| "loss": 0.3232, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.2914653784219, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001665575662878099, | |
| "loss": 0.326, | |
| "step": 2044 | |
| }, | |
| { | |
| "epoch": 3.2979066022544283, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00016612211709930442, | |
| "loss": 0.3169, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 3.3043478260869565, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00016568653041203356, | |
| "loss": 0.3036, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 3.3107890499194848, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00016525080994093328, | |
| "loss": 0.2987, | |
| "step": 2056 | |
| }, | |
| { | |
| "epoch": 3.317230273752013, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00016481495940208046, | |
| "loss": 0.3108, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.323671497584541, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0001643789825126613, | |
| "loss": 0.3119, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 3.3301127214170694, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0001639428829909396, | |
| "loss": 0.342, | |
| "step": 2068 | |
| }, | |
| { | |
| "epoch": 3.3365539452495976, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00016350666455622497, | |
| "loss": 0.3025, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 3.342995169082126, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001630703309288412, | |
| "loss": 0.3136, | |
| "step": 2076 | |
| }, | |
| { | |
| "epoch": 3.3494363929146536, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00016263388583009463, | |
| "loss": 0.2957, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.3558776167471818, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001621973329822421, | |
| "loss": 0.2948, | |
| "step": 2084 | |
| }, | |
| { | |
| "epoch": 3.36231884057971, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00016176067610845958, | |
| "loss": 0.3298, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 3.368760064412238, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00016132391893281003, | |
| "loss": 0.327, | |
| "step": 2092 | |
| }, | |
| { | |
| "epoch": 3.3752012882447664, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001608870651802121, | |
| "loss": 0.3009, | |
| "step": 2096 | |
| }, | |
| { | |
| "epoch": 3.3816425120772946, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00016045011857640783, | |
| "loss": 0.3148, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.388083735909823, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001600130828479314, | |
| "loss": 0.3282, | |
| "step": 2104 | |
| }, | |
| { | |
| "epoch": 3.394524959742351, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001595759617220769, | |
| "loss": 0.3203, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 3.4009661835748792, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00015913875892686685, | |
| "loss": 0.2977, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 3.4074074074074074, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00015870147819102025, | |
| "loss": 0.2806, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 3.4138486312399356, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00015826412324392085, | |
| "loss": 0.3096, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.420289855072464, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00015782669781558528, | |
| "loss": 0.301, | |
| "step": 2124 | |
| }, | |
| { | |
| "epoch": 3.426731078904992, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00015738920563663136, | |
| "loss": 0.3055, | |
| "step": 2128 | |
| }, | |
| { | |
| "epoch": 3.4331723027375203, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00015695165043824605, | |
| "loss": 0.3187, | |
| "step": 2132 | |
| }, | |
| { | |
| "epoch": 3.4396135265700485, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00015651403595215392, | |
| "loss": 0.308, | |
| "step": 2136 | |
| }, | |
| { | |
| "epoch": 3.4460547504025767, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00015607636591058506, | |
| "loss": 0.3033, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.452495974235105, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001556386440462435, | |
| "loss": 0.3313, | |
| "step": 2144 | |
| }, | |
| { | |
| "epoch": 3.4589371980676327, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001552008740922751, | |
| "loss": 0.2891, | |
| "step": 2148 | |
| }, | |
| { | |
| "epoch": 3.465378421900161, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00015476305978223606, | |
| "loss": 0.3416, | |
| "step": 2152 | |
| }, | |
| { | |
| "epoch": 3.471819645732689, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00015432520485006055, | |
| "loss": 0.2768, | |
| "step": 2156 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00015388731303002954, | |
| "loss": 0.3216, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.4847020933977455, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001534493880567384, | |
| "loss": 0.3112, | |
| "step": 2164 | |
| }, | |
| { | |
| "epoch": 3.4911433172302737, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00015301143366506527, | |
| "loss": 0.323, | |
| "step": 2168 | |
| }, | |
| { | |
| "epoch": 3.497584541062802, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00015257345359013928, | |
| "loss": 0.3406, | |
| "step": 2172 | |
| }, | |
| { | |
| "epoch": 3.50402576489533, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00015213545156730847, | |
| "loss": 0.2904, | |
| "step": 2176 | |
| }, | |
| { | |
| "epoch": 3.5104669887278583, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00015169743133210814, | |
| "loss": 0.3107, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.5169082125603865, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001512593966202289, | |
| "loss": 0.3377, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 3.5233494363929148, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00015082135116748483, | |
| "loss": 0.3491, | |
| "step": 2188 | |
| }, | |
| { | |
| "epoch": 3.529790660225443, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00015038329870978168, | |
| "loss": 0.2865, | |
| "step": 2192 | |
| }, | |
| { | |
| "epoch": 3.536231884057971, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00014994524298308479, | |
| "loss": 0.2913, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 3.542673107890499, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001495071877233875, | |
| "loss": 0.3163, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.549114331723027, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00014906913666667913, | |
| "loss": 0.2722, | |
| "step": 2204 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00014863109354891317, | |
| "loss": 0.3163, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 3.5619967793880836, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00014819306210597536, | |
| "loss": 0.3735, | |
| "step": 2212 | |
| }, | |
| { | |
| "epoch": 3.5684380032206118, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00014775504607365196, | |
| "loss": 0.3303, | |
| "step": 2216 | |
| }, | |
| { | |
| "epoch": 3.57487922705314, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00014731704918759765, | |
| "loss": 0.2946, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.581320450885668, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.000146879075183304, | |
| "loss": 0.3434, | |
| "step": 2224 | |
| }, | |
| { | |
| "epoch": 3.5877616747181964, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00014644112779606727, | |
| "loss": 0.3063, | |
| "step": 2228 | |
| }, | |
| { | |
| "epoch": 3.5942028985507246, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00014600321076095683, | |
| "loss": 0.2962, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 3.600644122383253, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00014556532781278316, | |
| "loss": 0.3006, | |
| "step": 2236 | |
| }, | |
| { | |
| "epoch": 3.607085346215781, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00014512748268606592, | |
| "loss": 0.3688, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.6135265700483092, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00014468967911500242, | |
| "loss": 0.3348, | |
| "step": 2244 | |
| }, | |
| { | |
| "epoch": 3.6199677938808374, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0001442519208334353, | |
| "loss": 0.3128, | |
| "step": 2248 | |
| }, | |
| { | |
| "epoch": 3.6264090177133657, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00014381421157482125, | |
| "loss": 0.3488, | |
| "step": 2252 | |
| }, | |
| { | |
| "epoch": 3.632850241545894, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001433765550721985, | |
| "loss": 0.2614, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 3.639291465378422, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00014293895505815575, | |
| "loss": 0.2984, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.6457326892109503, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00014250141526479953, | |
| "loss": 0.3257, | |
| "step": 2264 | |
| }, | |
| { | |
| "epoch": 3.6521739130434785, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00014206393942372314, | |
| "loss": 0.3235, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 3.6586151368760067, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001416265312659741, | |
| "loss": 0.3435, | |
| "step": 2272 | |
| }, | |
| { | |
| "epoch": 3.6650563607085345, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00014118919452202306, | |
| "loss": 0.3191, | |
| "step": 2276 | |
| }, | |
| { | |
| "epoch": 3.6714975845410627, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00014075193292173126, | |
| "loss": 0.2869, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.677938808373591, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00014031475019431934, | |
| "loss": 0.3089, | |
| "step": 2284 | |
| }, | |
| { | |
| "epoch": 3.684380032206119, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00013987765006833518, | |
| "loss": 0.3332, | |
| "step": 2288 | |
| }, | |
| { | |
| "epoch": 3.6908212560386473, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001394406362716221, | |
| "loss": 0.3127, | |
| "step": 2292 | |
| }, | |
| { | |
| "epoch": 3.6972624798711755, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00013900371253128727, | |
| "loss": 0.3177, | |
| "step": 2296 | |
| }, | |
| { | |
| "epoch": 3.7037037037037037, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001385668825736697, | |
| "loss": 0.3324, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.710144927536232, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001381301501243087, | |
| "loss": 0.2785, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 3.71658615136876, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00013769351890791185, | |
| "loss": 0.3274, | |
| "step": 2308 | |
| }, | |
| { | |
| "epoch": 3.7230273752012883, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00013725699264832344, | |
| "loss": 0.3041, | |
| "step": 2312 | |
| }, | |
| { | |
| "epoch": 3.7294685990338166, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00013682057506849256, | |
| "loss": 0.3343, | |
| "step": 2316 | |
| }, | |
| { | |
| "epoch": 3.7359098228663448, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00013638426989044148, | |
| "loss": 0.2785, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.7423510466988725, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00013594808083523376, | |
| "loss": 0.3454, | |
| "step": 2324 | |
| }, | |
| { | |
| "epoch": 3.7487922705314007, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00013551201162294275, | |
| "loss": 0.312, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 3.755233494363929, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00013507606597261946, | |
| "loss": 0.2885, | |
| "step": 2332 | |
| }, | |
| { | |
| "epoch": 3.761674718196457, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00013464024760226142, | |
| "loss": 0.3328, | |
| "step": 2336 | |
| }, | |
| { | |
| "epoch": 3.7681159420289854, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.0001342045602287803, | |
| "loss": 0.3078, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.7745571658615136, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00013376900756797085, | |
| "loss": 0.3126, | |
| "step": 2344 | |
| }, | |
| { | |
| "epoch": 3.780998389694042, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00013333359333447865, | |
| "loss": 0.2941, | |
| "step": 2348 | |
| }, | |
| { | |
| "epoch": 3.78743961352657, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001328983212417689, | |
| "loss": 0.3251, | |
| "step": 2352 | |
| }, | |
| { | |
| "epoch": 3.793880837359098, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001324631950020945, | |
| "loss": 0.3367, | |
| "step": 2356 | |
| }, | |
| { | |
| "epoch": 3.8003220611916264, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0001320282183264643, | |
| "loss": 0.3164, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.8067632850241546, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00013159339492461176, | |
| "loss": 0.3584, | |
| "step": 2364 | |
| }, | |
| { | |
| "epoch": 3.813204508856683, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00013115872850496293, | |
| "loss": 0.3307, | |
| "step": 2368 | |
| }, | |
| { | |
| "epoch": 3.819645732689211, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001307242227746053, | |
| "loss": 0.3475, | |
| "step": 2372 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00013028988143925553, | |
| "loss": 0.3058, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 3.8325281803542675, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00012985570820322868, | |
| "loss": 0.2718, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.8389694041867957, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00012942170676940576, | |
| "loss": 0.3074, | |
| "step": 2384 | |
| }, | |
| { | |
| "epoch": 3.845410628019324, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00012898788083920282, | |
| "loss": 0.3177, | |
| "step": 2388 | |
| }, | |
| { | |
| "epoch": 3.851851851851852, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001285542341125389, | |
| "loss": 0.3012, | |
| "step": 2392 | |
| }, | |
| { | |
| "epoch": 3.8582930756843803, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001281207702878049, | |
| "loss": 0.3024, | |
| "step": 2396 | |
| }, | |
| { | |
| "epoch": 3.864734299516908, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00012768749306183165, | |
| "loss": 0.3092, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.8711755233494363, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00012725440612985868, | |
| "loss": 0.2978, | |
| "step": 2404 | |
| }, | |
| { | |
| "epoch": 3.8776167471819645, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001268215131855025, | |
| "loss": 0.3337, | |
| "step": 2408 | |
| }, | |
| { | |
| "epoch": 3.8840579710144927, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00012638881792072522, | |
| "loss": 0.3278, | |
| "step": 2412 | |
| }, | |
| { | |
| "epoch": 3.890499194847021, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00012595632402580305, | |
| "loss": 0.3051, | |
| "step": 2416 | |
| }, | |
| { | |
| "epoch": 3.896940418679549, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00012552403518929472, | |
| "loss": 0.2764, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.9033816425120773, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.0001250919550980102, | |
| "loss": 0.3124, | |
| "step": 2424 | |
| }, | |
| { | |
| "epoch": 3.9098228663446055, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00012466008743697906, | |
| "loss": 0.3407, | |
| "step": 2428 | |
| }, | |
| { | |
| "epoch": 3.9162640901771337, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00012422843588941925, | |
| "loss": 0.3336, | |
| "step": 2432 | |
| }, | |
| { | |
| "epoch": 3.922705314009662, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00012379700413670547, | |
| "loss": 0.2992, | |
| "step": 2436 | |
| }, | |
| { | |
| "epoch": 3.92914653784219, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00012336579585833798, | |
| "loss": 0.3341, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.9355877616747184, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00012293481473191103, | |
| "loss": 0.3153, | |
| "step": 2444 | |
| }, | |
| { | |
| "epoch": 3.942028985507246, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00012250406443308168, | |
| "loss": 0.2993, | |
| "step": 2448 | |
| }, | |
| { | |
| "epoch": 3.9484702093397743, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00012207354863553825, | |
| "loss": 0.3144, | |
| "step": 2452 | |
| }, | |
| { | |
| "epoch": 3.9549114331723025, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00012164327101096923, | |
| "loss": 0.3251, | |
| "step": 2456 | |
| }, | |
| { | |
| "epoch": 3.9613526570048307, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00012121323522903167, | |
| "loss": 0.2799, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.967793880837359, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00012078344495732028, | |
| "loss": 0.3188, | |
| "step": 2464 | |
| }, | |
| { | |
| "epoch": 3.974235104669887, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00012035390386133558, | |
| "loss": 0.3052, | |
| "step": 2468 | |
| }, | |
| { | |
| "epoch": 3.9806763285024154, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00011992461560445337, | |
| "loss": 0.2771, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 3.9871175523349436, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00011949558384789271, | |
| "loss": 0.3164, | |
| "step": 2476 | |
| }, | |
| { | |
| "epoch": 3.993558776167472, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00011906681225068535, | |
| "loss": 0.2902, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00011863830446964417, | |
| "loss": 0.3142, | |
| "step": 2484 | |
| }, | |
| { | |
| "epoch": 4.006441223832528, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00011821006415933199, | |
| "loss": 0.2147, | |
| "step": 2488 | |
| }, | |
| { | |
| "epoch": 4.012882447665056, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00011778209497203062, | |
| "loss": 0.2092, | |
| "step": 2492 | |
| }, | |
| { | |
| "epoch": 4.019323671497585, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00011735440055770945, | |
| "loss": 0.2548, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 4.025764895330113, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00011692698456399458, | |
| "loss": 0.2183, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.032206119162641, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001164998506361374, | |
| "loss": 0.2009, | |
| "step": 2504 | |
| }, | |
| { | |
| "epoch": 4.038647342995169, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00011607300241698387, | |
| "loss": 0.218, | |
| "step": 2508 | |
| }, | |
| { | |
| "epoch": 4.0450885668276975, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00011564644354694312, | |
| "loss": 0.2201, | |
| "step": 2512 | |
| }, | |
| { | |
| "epoch": 4.051529790660226, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00011522017766395665, | |
| "loss": 0.2078, | |
| "step": 2516 | |
| }, | |
| { | |
| "epoch": 4.057971014492754, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00011479420840346706, | |
| "loss": 0.1932, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.064412238325282, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00011436853939838734, | |
| "loss": 0.2217, | |
| "step": 2524 | |
| }, | |
| { | |
| "epoch": 4.07085346215781, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001139431742790696, | |
| "loss": 0.2448, | |
| "step": 2528 | |
| }, | |
| { | |
| "epoch": 4.0772946859903385, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001135181166732743, | |
| "loss": 0.2254, | |
| "step": 2532 | |
| }, | |
| { | |
| "epoch": 4.083735909822867, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00011309337020613922, | |
| "loss": 0.2665, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 4.090177133655395, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001126689385001486, | |
| "loss": 0.2365, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.096618357487923, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00011224482517510224, | |
| "loss": 0.2341, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 4.1030595813204505, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00011182103384808444, | |
| "loss": 0.2015, | |
| "step": 2548 | |
| }, | |
| { | |
| "epoch": 4.109500805152979, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00011139756813343359, | |
| "loss": 0.2334, | |
| "step": 2552 | |
| }, | |
| { | |
| "epoch": 4.115942028985507, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00011097443164271075, | |
| "loss": 0.246, | |
| "step": 2556 | |
| }, | |
| { | |
| "epoch": 4.122383252818035, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00011055162798466948, | |
| "loss": 0.2322, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.128824476650563, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00011012916076522443, | |
| "loss": 0.2178, | |
| "step": 2564 | |
| }, | |
| { | |
| "epoch": 4.1352657004830915, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00010970703358742127, | |
| "loss": 0.2147, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 4.14170692431562, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00010928525005140521, | |
| "loss": 0.2315, | |
| "step": 2572 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00010886381375439105, | |
| "loss": 0.2284, | |
| "step": 2576 | |
| }, | |
| { | |
| "epoch": 4.154589371980676, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001084427282906318, | |
| "loss": 0.2568, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.161030595813204, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00010802199725138869, | |
| "loss": 0.2163, | |
| "step": 2584 | |
| }, | |
| { | |
| "epoch": 4.1674718196457325, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00010760162422489987, | |
| "loss": 0.2267, | |
| "step": 2588 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00010718161279635048, | |
| "loss": 0.2263, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 4.180354267310789, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00010676196654784144, | |
| "loss": 0.2395, | |
| "step": 2596 | |
| }, | |
| { | |
| "epoch": 4.186795491143317, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00010634268905835949, | |
| "loss": 0.2454, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.193236714975845, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00010592378390374612, | |
| "loss": 0.2186, | |
| "step": 2604 | |
| }, | |
| { | |
| "epoch": 4.199677938808374, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00010550525465666751, | |
| "loss": 0.2302, | |
| "step": 2608 | |
| }, | |
| { | |
| "epoch": 4.206119162640902, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00010508710488658385, | |
| "loss": 0.2475, | |
| "step": 2612 | |
| }, | |
| { | |
| "epoch": 4.21256038647343, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00010466933815971884, | |
| "loss": 0.1988, | |
| "step": 2616 | |
| }, | |
| { | |
| "epoch": 4.219001610305958, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00010425195803902948, | |
| "loss": 0.2137, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.225442834138486, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00010383496808417547, | |
| "loss": 0.2564, | |
| "step": 2624 | |
| }, | |
| { | |
| "epoch": 4.231884057971015, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00010341837185148903, | |
| "loss": 0.2361, | |
| "step": 2628 | |
| }, | |
| { | |
| "epoch": 4.238325281803543, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00010300217289394443, | |
| "loss": 0.2324, | |
| "step": 2632 | |
| }, | |
| { | |
| "epoch": 4.244766505636071, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00010258637476112782, | |
| "loss": 0.2175, | |
| "step": 2636 | |
| }, | |
| { | |
| "epoch": 4.251207729468599, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00010217098099920676, | |
| "loss": 0.2533, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.2576489533011275, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00010175599515090026, | |
| "loss": 0.2155, | |
| "step": 2644 | |
| }, | |
| { | |
| "epoch": 4.264090177133656, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00010134142075544824, | |
| "loss": 0.2299, | |
| "step": 2648 | |
| }, | |
| { | |
| "epoch": 4.270531400966184, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00010092726134858168, | |
| "loss": 0.2776, | |
| "step": 2652 | |
| }, | |
| { | |
| "epoch": 4.276972624798712, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00010051352046249213, | |
| "loss": 0.2079, | |
| "step": 2656 | |
| }, | |
| { | |
| "epoch": 4.28341384863124, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00010010020162580192, | |
| "loss": 0.198, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.2898550724637685, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.96873083635337e-05, | |
| "loss": 0.223, | |
| "step": 2664 | |
| }, | |
| { | |
| "epoch": 4.296296296296296, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.927484419708076e-05, | |
| "loss": 0.187, | |
| "step": 2668 | |
| }, | |
| { | |
| "epoch": 4.302737520128824, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.88628126441768e-05, | |
| "loss": 0.2339, | |
| "step": 2672 | |
| }, | |
| { | |
| "epoch": 4.309178743961352, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.84512172188657e-05, | |
| "loss": 0.2164, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.3156199677938805, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.804006143147212e-05, | |
| "loss": 0.2328, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.322061191626409, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.762934878857105e-05, | |
| "loss": 0.2577, | |
| "step": 2684 | |
| }, | |
| { | |
| "epoch": 4.328502415458937, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.721908279295812e-05, | |
| "loss": 0.2256, | |
| "step": 2688 | |
| }, | |
| { | |
| "epoch": 4.334943639291465, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.680926694361964e-05, | |
| "loss": 0.2344, | |
| "step": 2692 | |
| }, | |
| { | |
| "epoch": 4.341384863123993, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.639990473570294e-05, | |
| "loss": 0.2238, | |
| "step": 2696 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.599099966048627e-05, | |
| "loss": 0.1847, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.35426731078905, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.558255520534937e-05, | |
| "loss": 0.2451, | |
| "step": 2704 | |
| }, | |
| { | |
| "epoch": 4.360708534621578, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.517457485374336e-05, | |
| "loss": 0.2112, | |
| "step": 2708 | |
| }, | |
| { | |
| "epoch": 4.367149758454106, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.476706208516138e-05, | |
| "loss": 0.2048, | |
| "step": 2712 | |
| }, | |
| { | |
| "epoch": 4.373590982286634, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.43600203751086e-05, | |
| "loss": 0.2036, | |
| "step": 2716 | |
| }, | |
| { | |
| "epoch": 4.3800322061191626, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.395345319507287e-05, | |
| "loss": 0.2125, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.386473429951691, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.354736401249486e-05, | |
| "loss": 0.2199, | |
| "step": 2724 | |
| }, | |
| { | |
| "epoch": 4.392914653784219, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.31417562907387e-05, | |
| "loss": 0.2064, | |
| "step": 2728 | |
| }, | |
| { | |
| "epoch": 4.399355877616747, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.273663348906222e-05, | |
| "loss": 0.2183, | |
| "step": 2732 | |
| }, | |
| { | |
| "epoch": 4.405797101449275, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.233199906258766e-05, | |
| "loss": 0.2639, | |
| "step": 2736 | |
| }, | |
| { | |
| "epoch": 4.412238325281804, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.192785646227217e-05, | |
| "loss": 0.251, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.418679549114332, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.152420913487814e-05, | |
| "loss": 0.2386, | |
| "step": 2744 | |
| }, | |
| { | |
| "epoch": 4.42512077294686, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.112106052294418e-05, | |
| "loss": 0.217, | |
| "step": 2748 | |
| }, | |
| { | |
| "epoch": 4.431561996779388, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.071841406475539e-05, | |
| "loss": 0.2102, | |
| "step": 2752 | |
| }, | |
| { | |
| "epoch": 4.438003220611916, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.03162731943144e-05, | |
| "loss": 0.2282, | |
| "step": 2756 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.991464134131166e-05, | |
| "loss": 0.2395, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.450885668276973, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 8.951352193109673e-05, | |
| "loss": 0.2379, | |
| "step": 2764 | |
| }, | |
| { | |
| "epoch": 4.457326892109501, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 8.911291838464838e-05, | |
| "loss": 0.2319, | |
| "step": 2768 | |
| }, | |
| { | |
| "epoch": 4.463768115942029, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 8.871283411854619e-05, | |
| "loss": 0.2066, | |
| "step": 2772 | |
| }, | |
| { | |
| "epoch": 4.4702093397745575, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 8.831327254494066e-05, | |
| "loss": 0.2086, | |
| "step": 2776 | |
| }, | |
| { | |
| "epoch": 4.476650563607086, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 8.791423707152482e-05, | |
| "loss": 0.2454, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.483091787439614, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 8.751573110150443e-05, | |
| "loss": 0.2254, | |
| "step": 2784 | |
| }, | |
| { | |
| "epoch": 4.489533011272142, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 8.711775803356971e-05, | |
| "loss": 0.2446, | |
| "step": 2788 | |
| }, | |
| { | |
| "epoch": 4.49597423510467, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 8.672032126186566e-05, | |
| "loss": 0.2154, | |
| "step": 2792 | |
| }, | |
| { | |
| "epoch": 4.5024154589371985, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 8.632342417596365e-05, | |
| "loss": 0.2269, | |
| "step": 2796 | |
| }, | |
| { | |
| "epoch": 4.508856682769727, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 8.592707016083221e-05, | |
| "loss": 0.2134, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.515297906602254, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 8.553126259680828e-05, | |
| "loss": 0.25, | |
| "step": 2804 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 8.513600485956835e-05, | |
| "loss": 0.2262, | |
| "step": 2808 | |
| }, | |
| { | |
| "epoch": 4.5281803542673105, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 8.474130032009951e-05, | |
| "loss": 0.2062, | |
| "step": 2812 | |
| }, | |
| { | |
| "epoch": 4.534621578099839, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 8.434715234467123e-05, | |
| "loss": 0.2293, | |
| "step": 2816 | |
| }, | |
| { | |
| "epoch": 4.541062801932367, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.395356429480587e-05, | |
| "loss": 0.204, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.547504025764895, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 8.356053952725072e-05, | |
| "loss": 0.2207, | |
| "step": 2824 | |
| }, | |
| { | |
| "epoch": 4.553945249597423, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 8.316808139394876e-05, | |
| "loss": 0.2231, | |
| "step": 2828 | |
| }, | |
| { | |
| "epoch": 4.5603864734299515, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.277619324201081e-05, | |
| "loss": 0.2321, | |
| "step": 2832 | |
| }, | |
| { | |
| "epoch": 4.56682769726248, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 8.238487841368617e-05, | |
| "loss": 0.2298, | |
| "step": 2836 | |
| }, | |
| { | |
| "epoch": 4.573268921095008, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 8.199414024633473e-05, | |
| "loss": 0.1997, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.579710144927536, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.160398207239805e-05, | |
| "loss": 0.2359, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 4.586151368760064, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 8.121440721937157e-05, | |
| "loss": 0.216, | |
| "step": 2848 | |
| }, | |
| { | |
| "epoch": 4.592592592592593, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 8.082541900977542e-05, | |
| "loss": 0.2374, | |
| "step": 2852 | |
| }, | |
| { | |
| "epoch": 4.599033816425121, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 8.04370207611267e-05, | |
| "loss": 0.2363, | |
| "step": 2856 | |
| }, | |
| { | |
| "epoch": 4.605475040257649, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 8.004921578591091e-05, | |
| "loss": 0.214, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.611916264090177, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 7.966200739155389e-05, | |
| "loss": 0.2214, | |
| "step": 2864 | |
| }, | |
| { | |
| "epoch": 4.618357487922705, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.927539888039339e-05, | |
| "loss": 0.2431, | |
| "step": 2868 | |
| }, | |
| { | |
| "epoch": 4.624798711755234, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 7.888939354965093e-05, | |
| "loss": 0.2104, | |
| "step": 2872 | |
| }, | |
| { | |
| "epoch": 4.631239935587762, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 7.850399469140393e-05, | |
| "loss": 0.204, | |
| "step": 2876 | |
| }, | |
| { | |
| "epoch": 4.63768115942029, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.811920559255736e-05, | |
| "loss": 0.2263, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.644122383252818, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 7.773502953481585e-05, | |
| "loss": 0.2161, | |
| "step": 2884 | |
| }, | |
| { | |
| "epoch": 4.650563607085346, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 7.73514697946556e-05, | |
| "loss": 0.2279, | |
| "step": 2888 | |
| }, | |
| { | |
| "epoch": 4.657004830917875, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 7.696852964329655e-05, | |
| "loss": 0.2615, | |
| "step": 2892 | |
| }, | |
| { | |
| "epoch": 4.663446054750403, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 7.658621234667443e-05, | |
| "loss": 0.2407, | |
| "step": 2896 | |
| }, | |
| { | |
| "epoch": 4.669887278582931, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 7.620452116541291e-05, | |
| "loss": 0.2101, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.676328502415459, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 7.582345935479569e-05, | |
| "loss": 0.2191, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 4.6827697262479875, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 7.544303016473894e-05, | |
| "loss": 0.2159, | |
| "step": 2908 | |
| }, | |
| { | |
| "epoch": 4.689210950080515, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 7.506323683976344e-05, | |
| "loss": 0.2251, | |
| "step": 2912 | |
| }, | |
| { | |
| "epoch": 4.695652173913043, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 7.468408261896701e-05, | |
| "loss": 0.1935, | |
| "step": 2916 | |
| }, | |
| { | |
| "epoch": 4.702093397745571, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 7.430557073599662e-05, | |
| "loss": 0.2123, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.708534621578099, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 7.392770441902116e-05, | |
| "loss": 0.2466, | |
| "step": 2924 | |
| }, | |
| { | |
| "epoch": 4.714975845410628, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.355048689070389e-05, | |
| "loss": 0.2332, | |
| "step": 2928 | |
| }, | |
| { | |
| "epoch": 4.721417069243156, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.317392136817453e-05, | |
| "loss": 0.2364, | |
| "step": 2932 | |
| }, | |
| { | |
| "epoch": 4.727858293075684, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.279801106300231e-05, | |
| "loss": 0.2662, | |
| "step": 2936 | |
| }, | |
| { | |
| "epoch": 4.734299516908212, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 7.242275918116832e-05, | |
| "loss": 0.2174, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.204816892303833e-05, | |
| "loss": 0.2135, | |
| "step": 2944 | |
| }, | |
| { | |
| "epoch": 4.747181964573269, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 7.16742434833352e-05, | |
| "loss": 0.231, | |
| "step": 2948 | |
| }, | |
| { | |
| "epoch": 4.753623188405797, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.1300986051112e-05, | |
| "loss": 0.2569, | |
| "step": 2952 | |
| }, | |
| { | |
| "epoch": 4.760064412238325, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 7.09283998097246e-05, | |
| "loss": 0.2072, | |
| "step": 2956 | |
| }, | |
| { | |
| "epoch": 4.766505636070853, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 7.055648793680466e-05, | |
| "loss": 0.2059, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.7729468599033815, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.018525360423217e-05, | |
| "loss": 0.2429, | |
| "step": 2964 | |
| }, | |
| { | |
| "epoch": 4.77938808373591, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 6.981469997810892e-05, | |
| "loss": 0.2203, | |
| "step": 2968 | |
| }, | |
| { | |
| "epoch": 4.785829307568438, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 6.944483021873115e-05, | |
| "loss": 0.232, | |
| "step": 2972 | |
| }, | |
| { | |
| "epoch": 4.792270531400966, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 6.907564748056273e-05, | |
| "loss": 0.2124, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 4.798711755233494, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.870715491220808e-05, | |
| "loss": 0.2184, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.805152979066023, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 6.833935565638559e-05, | |
| "loss": 0.238, | |
| "step": 2984 | |
| }, | |
| { | |
| "epoch": 4.811594202898551, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 6.797225284990064e-05, | |
| "loss": 0.2283, | |
| "step": 2988 | |
| }, | |
| { | |
| "epoch": 4.818035426731079, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 6.760584962361888e-05, | |
| "loss": 0.2351, | |
| "step": 2992 | |
| }, | |
| { | |
| "epoch": 4.824476650563607, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 6.72401491024396e-05, | |
| "loss": 0.2019, | |
| "step": 2996 | |
| }, | |
| { | |
| "epoch": 4.830917874396135, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.687515440526882e-05, | |
| "loss": 0.242, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.837359098228664, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 6.651086864499305e-05, | |
| "loss": 0.2196, | |
| "step": 3004 | |
| }, | |
| { | |
| "epoch": 4.843800322061192, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 6.614729492845258e-05, | |
| "loss": 0.2146, | |
| "step": 3008 | |
| }, | |
| { | |
| "epoch": 4.85024154589372, | |
| "grad_norm": 0.375, | |
| "learning_rate": 6.578443635641497e-05, | |
| "loss": 0.2232, | |
| "step": 3012 | |
| }, | |
| { | |
| "epoch": 4.856682769726248, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 6.542229602354847e-05, | |
| "loss": 0.2319, | |
| "step": 3016 | |
| }, | |
| { | |
| "epoch": 4.8631239935587764, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.506087701839593e-05, | |
| "loss": 0.2156, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 6.470018242334825e-05, | |
| "loss": 0.2372, | |
| "step": 3024 | |
| }, | |
| { | |
| "epoch": 4.876006441223833, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 6.434021531461818e-05, | |
| "loss": 0.2077, | |
| "step": 3028 | |
| }, | |
| { | |
| "epoch": 4.882447665056361, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 6.398097876221385e-05, | |
| "loss": 0.2183, | |
| "step": 3032 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 6.362247582991317e-05, | |
| "loss": 0.2104, | |
| "step": 3036 | |
| }, | |
| { | |
| "epoch": 4.8953301127214175, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 6.326470957523686e-05, | |
| "loss": 0.2048, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.901771336553946, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.29076830494232e-05, | |
| "loss": 0.2346, | |
| "step": 3044 | |
| }, | |
| { | |
| "epoch": 4.908212560386474, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 6.255139929740129e-05, | |
| "loss": 0.2068, | |
| "step": 3048 | |
| }, | |
| { | |
| "epoch": 4.914653784219001, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.219586135776575e-05, | |
| "loss": 0.239, | |
| "step": 3052 | |
| }, | |
| { | |
| "epoch": 4.921095008051529, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 6.184107226275038e-05, | |
| "loss": 0.1814, | |
| "step": 3056 | |
| }, | |
| { | |
| "epoch": 4.927536231884058, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 6.148703503820224e-05, | |
| "loss": 0.2272, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.933977455716586, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.113375270355617e-05, | |
| "loss": 0.2418, | |
| "step": 3064 | |
| }, | |
| { | |
| "epoch": 4.940418679549114, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 6.078122827180879e-05, | |
| "loss": 0.2723, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 4.946859903381642, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 6.042946474949302e-05, | |
| "loss": 0.2407, | |
| "step": 3072 | |
| }, | |
| { | |
| "epoch": 4.9533011272141705, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 6.007846513665207e-05, | |
| "loss": 0.2153, | |
| "step": 3076 | |
| }, | |
| { | |
| "epoch": 4.959742351046699, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 5.972823242681426e-05, | |
| "loss": 0.2206, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.966183574879227, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 5.937876960696727e-05, | |
| "loss": 0.2105, | |
| "step": 3084 | |
| }, | |
| { | |
| "epoch": 4.972624798711755, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 5.903007965753279e-05, | |
| "loss": 0.2526, | |
| "step": 3088 | |
| }, | |
| { | |
| "epoch": 4.979066022544283, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.868216555234081e-05, | |
| "loss": 0.2168, | |
| "step": 3092 | |
| }, | |
| { | |
| "epoch": 4.9855072463768115, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.833503025860469e-05, | |
| "loss": 0.2174, | |
| "step": 3096 | |
| }, | |
| { | |
| "epoch": 4.99194847020934, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 5.798867673689553e-05, | |
| "loss": 0.2365, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.998389694041868, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 5.764310794111711e-05, | |
| "loss": 0.1766, | |
| "step": 3104 | |
| }, | |
| { | |
| "epoch": 5.004830917874396, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.7298326818480427e-05, | |
| "loss": 0.1998, | |
| "step": 3108 | |
| }, | |
| { | |
| "epoch": 5.011272141706924, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.695433630947894e-05, | |
| "loss": 0.1838, | |
| "step": 3112 | |
| }, | |
| { | |
| "epoch": 5.017713365539453, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 5.661113934786321e-05, | |
| "loss": 0.2045, | |
| "step": 3116 | |
| }, | |
| { | |
| "epoch": 5.024154589371981, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.626873886061597e-05, | |
| "loss": 0.1917, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 5.030595813204509, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.592713776792723e-05, | |
| "loss": 0.204, | |
| "step": 3124 | |
| }, | |
| { | |
| "epoch": 5.037037037037037, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 5.5586338983169076e-05, | |
| "loss": 0.1471, | |
| "step": 3128 | |
| }, | |
| { | |
| "epoch": 5.043478260869565, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 5.52463454128714e-05, | |
| "loss": 0.1966, | |
| "step": 3132 | |
| }, | |
| { | |
| "epoch": 5.049919484702094, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 5.490715995669641e-05, | |
| "loss": 0.1782, | |
| "step": 3136 | |
| }, | |
| { | |
| "epoch": 5.056360708534622, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 5.456878550741453e-05, | |
| "loss": 0.1877, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 5.06280193236715, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 5.423122495087915e-05, | |
| "loss": 0.1643, | |
| "step": 3144 | |
| }, | |
| { | |
| "epoch": 5.069243156199678, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 5.3894481166002674e-05, | |
| "loss": 0.1792, | |
| "step": 3148 | |
| }, | |
| { | |
| "epoch": 5.0756843800322065, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.355855702473125e-05, | |
| "loss": 0.1567, | |
| "step": 3152 | |
| }, | |
| { | |
| "epoch": 5.082125603864735, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 5.322345539202086e-05, | |
| "loss": 0.2051, | |
| "step": 3156 | |
| }, | |
| { | |
| "epoch": 5.088566827697263, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 5.288917912581257e-05, | |
| "loss": 0.1754, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 5.095008051529791, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 5.255573107700832e-05, | |
| "loss": 0.1824, | |
| "step": 3164 | |
| }, | |
| { | |
| "epoch": 5.101449275362318, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 5.222311408944635e-05, | |
| "loss": 0.2092, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 5.107890499194847, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 5.189133099987731e-05, | |
| "loss": 0.146, | |
| "step": 3172 | |
| }, | |
| { | |
| "epoch": 5.114331723027375, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 5.156038463793981e-05, | |
| "loss": 0.1692, | |
| "step": 3176 | |
| }, | |
| { | |
| "epoch": 5.120772946859903, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 5.123027782613636e-05, | |
| "loss": 0.1877, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 5.127214170692431, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 5.09010133798094e-05, | |
| "loss": 0.154, | |
| "step": 3184 | |
| }, | |
| { | |
| "epoch": 5.1336553945249594, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.0572594107116974e-05, | |
| "loss": 0.1559, | |
| "step": 3188 | |
| }, | |
| { | |
| "epoch": 5.140096618357488, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 5.0245022809009155e-05, | |
| "loss": 0.171, | |
| "step": 3192 | |
| }, | |
| { | |
| "epoch": 5.146537842190016, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 4.991830227920398e-05, | |
| "loss": 0.1774, | |
| "step": 3196 | |
| }, | |
| { | |
| "epoch": 5.152979066022544, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 4.9592435304163675e-05, | |
| "loss": 0.1813, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.159420289855072, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 4.926742466307069e-05, | |
| "loss": 0.1557, | |
| "step": 3204 | |
| }, | |
| { | |
| "epoch": 5.1658615136876005, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 4.8943273127804345e-05, | |
| "loss": 0.1574, | |
| "step": 3208 | |
| }, | |
| { | |
| "epoch": 5.172302737520129, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 4.8619983462916935e-05, | |
| "loss": 0.1548, | |
| "step": 3212 | |
| }, | |
| { | |
| "epoch": 5.178743961352657, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 4.829755842561025e-05, | |
| "loss": 0.1888, | |
| "step": 3216 | |
| }, | |
| { | |
| "epoch": 5.185185185185185, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 4.797600076571194e-05, | |
| "loss": 0.2004, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 5.191626409017713, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 4.7655313225652294e-05, | |
| "loss": 0.1587, | |
| "step": 3224 | |
| }, | |
| { | |
| "epoch": 5.1980676328502415, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 4.7335498540440606e-05, | |
| "loss": 0.1669, | |
| "step": 3228 | |
| }, | |
| { | |
| "epoch": 5.20450885668277, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 4.7016559437642084e-05, | |
| "loss": 0.171, | |
| "step": 3232 | |
| }, | |
| { | |
| "epoch": 5.210950080515298, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 4.6698498637354225e-05, | |
| "loss": 0.1566, | |
| "step": 3236 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 4.6381318852184194e-05, | |
| "loss": 0.1936, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 5.223832528180354, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 4.606502278722503e-05, | |
| "loss": 0.1897, | |
| "step": 3244 | |
| }, | |
| { | |
| "epoch": 5.230273752012883, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 4.574961314003304e-05, | |
| "loss": 0.1935, | |
| "step": 3248 | |
| }, | |
| { | |
| "epoch": 5.236714975845411, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 4.5435092600604676e-05, | |
| "loss": 0.159, | |
| "step": 3252 | |
| }, | |
| { | |
| "epoch": 5.243156199677939, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 4.5121463851353476e-05, | |
| "loss": 0.2065, | |
| "step": 3256 | |
| }, | |
| { | |
| "epoch": 5.249597423510467, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.48087295670874e-05, | |
| "loss": 0.1993, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 5.256038647342995, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 4.449689241498569e-05, | |
| "loss": 0.1717, | |
| "step": 3264 | |
| }, | |
| { | |
| "epoch": 5.262479871175524, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 4.41859550545765e-05, | |
| "loss": 0.1907, | |
| "step": 3268 | |
| }, | |
| { | |
| "epoch": 5.268921095008052, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 4.387592013771396e-05, | |
| "loss": 0.1691, | |
| "step": 3272 | |
| }, | |
| { | |
| "epoch": 5.27536231884058, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 4.356679030855573e-05, | |
| "loss": 0.198, | |
| "step": 3276 | |
| }, | |
| { | |
| "epoch": 5.281803542673108, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 4.32585682035402e-05, | |
| "loss": 0.2026, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 5.2882447665056365, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 4.2951256451364264e-05, | |
| "loss": 0.1975, | |
| "step": 3284 | |
| }, | |
| { | |
| "epoch": 5.294685990338165, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 4.264485767296081e-05, | |
| "loss": 0.1686, | |
| "step": 3288 | |
| }, | |
| { | |
| "epoch": 5.301127214170693, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 4.233937448147635e-05, | |
| "loss": 0.1583, | |
| "step": 3292 | |
| }, | |
| { | |
| "epoch": 5.30756843800322, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 4.203480948224866e-05, | |
| "loss": 0.1777, | |
| "step": 3296 | |
| }, | |
| { | |
| "epoch": 5.314009661835748, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.173116527278471e-05, | |
| "loss": 0.1616, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.320450885668277, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 4.142844444273845e-05, | |
| "loss": 0.1731, | |
| "step": 3304 | |
| }, | |
| { | |
| "epoch": 5.326892109500805, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.1126649573888696e-05, | |
| "loss": 0.2219, | |
| "step": 3308 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 4.082578324011716e-05, | |
| "loss": 0.1937, | |
| "step": 3312 | |
| }, | |
| { | |
| "epoch": 5.339774557165861, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 4.052584800738636e-05, | |
| "loss": 0.1891, | |
| "step": 3316 | |
| }, | |
| { | |
| "epoch": 5.3462157809983895, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 4.0226846433717954e-05, | |
| "loss": 0.1811, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 5.352657004830918, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 3.992878106917079e-05, | |
| "loss": 0.1768, | |
| "step": 3324 | |
| }, | |
| { | |
| "epoch": 5.359098228663446, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 3.963165445581922e-05, | |
| "loss": 0.1985, | |
| "step": 3328 | |
| }, | |
| { | |
| "epoch": 5.365539452495974, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 3.933546912773119e-05, | |
| "loss": 0.154, | |
| "step": 3332 | |
| }, | |
| { | |
| "epoch": 5.371980676328502, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.904022761094715e-05, | |
| "loss": 0.1788, | |
| "step": 3336 | |
| }, | |
| { | |
| "epoch": 5.3784219001610305, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.874593242345785e-05, | |
| "loss": 0.1964, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 5.384863123993559, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.845258607518344e-05, | |
| "loss": 0.1822, | |
| "step": 3344 | |
| }, | |
| { | |
| "epoch": 5.391304347826087, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.816019106795157e-05, | |
| "loss": 0.1711, | |
| "step": 3348 | |
| }, | |
| { | |
| "epoch": 5.397745571658615, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 3.7868749895476624e-05, | |
| "loss": 0.1785, | |
| "step": 3352 | |
| }, | |
| { | |
| "epoch": 5.404186795491143, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.7578265043337834e-05, | |
| "loss": 0.1891, | |
| "step": 3356 | |
| }, | |
| { | |
| "epoch": 5.4106280193236715, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 3.72887389889586e-05, | |
| "loss": 0.1766, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 5.4170692431562, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 3.700017420158486e-05, | |
| "loss": 0.1733, | |
| "step": 3364 | |
| }, | |
| { | |
| "epoch": 5.423510466988728, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 3.671257314226471e-05, | |
| "loss": 0.1895, | |
| "step": 3368 | |
| }, | |
| { | |
| "epoch": 5.429951690821256, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 3.642593826382663e-05, | |
| "loss": 0.1867, | |
| "step": 3372 | |
| }, | |
| { | |
| "epoch": 5.436392914653784, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 3.6140272010859166e-05, | |
| "loss": 0.1946, | |
| "step": 3376 | |
| }, | |
| { | |
| "epoch": 5.442834138486313, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.585557681968979e-05, | |
| "loss": 0.1684, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 5.449275362318841, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 3.5571855118364236e-05, | |
| "loss": 0.1886, | |
| "step": 3384 | |
| }, | |
| { | |
| "epoch": 5.455716586151369, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.528910932662577e-05, | |
| "loss": 0.199, | |
| "step": 3388 | |
| }, | |
| { | |
| "epoch": 5.462157809983897, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 3.5007341855894394e-05, | |
| "loss": 0.1877, | |
| "step": 3392 | |
| }, | |
| { | |
| "epoch": 5.468599033816425, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 3.472655510924656e-05, | |
| "loss": 0.1856, | |
| "step": 3396 | |
| }, | |
| { | |
| "epoch": 5.475040257648954, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 3.4446751481394516e-05, | |
| "loss": 0.1818, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.481481481481482, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.4167933358665936e-05, | |
| "loss": 0.2009, | |
| "step": 3404 | |
| }, | |
| { | |
| "epoch": 5.48792270531401, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.3890103118983366e-05, | |
| "loss": 0.1824, | |
| "step": 3408 | |
| }, | |
| { | |
| "epoch": 5.494363929146537, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.3613263131844294e-05, | |
| "loss": 0.1746, | |
| "step": 3412 | |
| }, | |
| { | |
| "epoch": 5.500805152979066, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 3.333741575830069e-05, | |
| "loss": 0.1769, | |
| "step": 3416 | |
| }, | |
| { | |
| "epoch": 5.507246376811594, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 3.306256335093898e-05, | |
| "loss": 0.1822, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 5.513687600644122, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 3.278870825385983e-05, | |
| "loss": 0.1925, | |
| "step": 3424 | |
| }, | |
| { | |
| "epoch": 5.52012882447665, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.251585280265839e-05, | |
| "loss": 0.1923, | |
| "step": 3428 | |
| }, | |
| { | |
| "epoch": 5.526570048309178, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.224399932440419e-05, | |
| "loss": 0.1815, | |
| "step": 3432 | |
| }, | |
| { | |
| "epoch": 5.533011272141707, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 3.1973150137621364e-05, | |
| "loss": 0.1738, | |
| "step": 3436 | |
| }, | |
| { | |
| "epoch": 5.539452495974235, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 3.170330755226893e-05, | |
| "loss": 0.191, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 5.545893719806763, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 3.1434473869720804e-05, | |
| "loss": 0.1538, | |
| "step": 3444 | |
| }, | |
| { | |
| "epoch": 5.552334943639291, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.116665138274676e-05, | |
| "loss": 0.1748, | |
| "step": 3448 | |
| }, | |
| { | |
| "epoch": 5.5587761674718195, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 3.0899842375492145e-05, | |
| "loss": 0.1893, | |
| "step": 3452 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.063404912345897e-05, | |
| "loss": 0.1727, | |
| "step": 3456 | |
| }, | |
| { | |
| "epoch": 5.571658615136876, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.036927389348625e-05, | |
| "loss": 0.1804, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 5.578099838969404, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 3.010551894373075e-05, | |
| "loss": 0.1778, | |
| "step": 3464 | |
| }, | |
| { | |
| "epoch": 5.584541062801932, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.9842786523647582e-05, | |
| "loss": 0.1679, | |
| "step": 3468 | |
| }, | |
| { | |
| "epoch": 5.5909822866344605, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.9581078873971248e-05, | |
| "loss": 0.1812, | |
| "step": 3472 | |
| }, | |
| { | |
| "epoch": 5.597423510466989, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 2.9320398226696367e-05, | |
| "loss": 0.188, | |
| "step": 3476 | |
| }, | |
| { | |
| "epoch": 5.603864734299517, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 2.9060746805058738e-05, | |
| "loss": 0.1541, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 5.610305958132045, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 2.8802126823516193e-05, | |
| "loss": 0.1671, | |
| "step": 3484 | |
| }, | |
| { | |
| "epoch": 5.616747181964573, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 2.8544540487729984e-05, | |
| "loss": 0.1609, | |
| "step": 3488 | |
| }, | |
| { | |
| "epoch": 5.6231884057971016, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 2.828798999454577e-05, | |
| "loss": 0.1488, | |
| "step": 3492 | |
| }, | |
| { | |
| "epoch": 5.62962962962963, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 2.8032477531974984e-05, | |
| "loss": 0.2012, | |
| "step": 3496 | |
| }, | |
| { | |
| "epoch": 5.636070853462158, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 2.7778005279176053e-05, | |
| "loss": 0.208, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.642512077294686, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 2.7524575406435955e-05, | |
| "loss": 0.192, | |
| "step": 3504 | |
| }, | |
| { | |
| "epoch": 5.648953301127214, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.7272190075151655e-05, | |
| "loss": 0.1582, | |
| "step": 3508 | |
| }, | |
| { | |
| "epoch": 5.655394524959743, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 2.7020851437811608e-05, | |
| "loss": 0.1762, | |
| "step": 3512 | |
| }, | |
| { | |
| "epoch": 5.661835748792271, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 2.6770561637977556e-05, | |
| "loss": 0.1678, | |
| "step": 3516 | |
| }, | |
| { | |
| "epoch": 5.668276972624799, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 2.652132281026598e-05, | |
| "loss": 0.1822, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 5.674718196457327, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 2.6273137080330225e-05, | |
| "loss": 0.183, | |
| "step": 3524 | |
| }, | |
| { | |
| "epoch": 5.681159420289855, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.6026006564842106e-05, | |
| "loss": 0.2009, | |
| "step": 3528 | |
| }, | |
| { | |
| "epoch": 5.687600644122384, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 2.577993337147406e-05, | |
| "loss": 0.1858, | |
| "step": 3532 | |
| }, | |
| { | |
| "epoch": 5.694041867954912, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 2.5534919598880887e-05, | |
| "loss": 0.203, | |
| "step": 3536 | |
| }, | |
| { | |
| "epoch": 5.70048309178744, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 2.5290967336682266e-05, | |
| "loss": 0.1588, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 5.706924315619968, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 2.5048078665444497e-05, | |
| "loss": 0.1622, | |
| "step": 3544 | |
| }, | |
| { | |
| "epoch": 5.713365539452496, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 2.4806255656663092e-05, | |
| "loss": 0.185, | |
| "step": 3548 | |
| }, | |
| { | |
| "epoch": 5.719806763285024, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 2.4565500372744845e-05, | |
| "loss": 0.1904, | |
| "step": 3552 | |
| }, | |
| { | |
| "epoch": 5.726247987117552, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 2.4325814866990583e-05, | |
| "loss": 0.175, | |
| "step": 3556 | |
| }, | |
| { | |
| "epoch": 5.73268921095008, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 2.4087201183577205e-05, | |
| "loss": 0.1699, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 5.739130434782608, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 2.384966135754063e-05, | |
| "loss": 0.1823, | |
| "step": 3564 | |
| }, | |
| { | |
| "epoch": 5.745571658615137, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 2.3613197414758273e-05, | |
| "loss": 0.1788, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.752012882447665, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.3377811371931793e-05, | |
| "loss": 0.1794, | |
| "step": 3572 | |
| }, | |
| { | |
| "epoch": 5.758454106280193, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.3143505236569915e-05, | |
| "loss": 0.1684, | |
| "step": 3576 | |
| }, | |
| { | |
| "epoch": 5.764895330112721, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 2.2910281006971164e-05, | |
| "loss": 0.157, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 5.7713365539452495, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 2.26781406722071e-05, | |
| "loss": 0.1833, | |
| "step": 3584 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 2.2447086212105143e-05, | |
| "loss": 0.1945, | |
| "step": 3588 | |
| }, | |
| { | |
| "epoch": 5.784219001610306, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 2.2217119597231747e-05, | |
| "loss": 0.1801, | |
| "step": 3592 | |
| }, | |
| { | |
| "epoch": 5.790660225442834, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 2.1988242788875532e-05, | |
| "loss": 0.1735, | |
| "step": 3596 | |
| }, | |
| { | |
| "epoch": 5.797101449275362, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.1760457739030695e-05, | |
| "loss": 0.1755, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.8035426731078905, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 2.1533766390380254e-05, | |
| "loss": 0.1674, | |
| "step": 3604 | |
| }, | |
| { | |
| "epoch": 5.809983896940419, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.1308170676279547e-05, | |
| "loss": 0.138, | |
| "step": 3608 | |
| }, | |
| { | |
| "epoch": 5.816425120772947, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.108367252073961e-05, | |
| "loss": 0.1521, | |
| "step": 3612 | |
| }, | |
| { | |
| "epoch": 5.822866344605475, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 2.0860273838410928e-05, | |
| "loss": 0.1771, | |
| "step": 3616 | |
| }, | |
| { | |
| "epoch": 5.829307568438003, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 2.0637976534567046e-05, | |
| "loss": 0.1628, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 5.835748792270532, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 2.0416782505088347e-05, | |
| "loss": 0.1631, | |
| "step": 3624 | |
| }, | |
| { | |
| "epoch": 5.84219001610306, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 2.0196693636445727e-05, | |
| "loss": 0.1731, | |
| "step": 3628 | |
| }, | |
| { | |
| "epoch": 5.848631239935588, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 1.9977711805684706e-05, | |
| "loss": 0.1748, | |
| "step": 3632 | |
| }, | |
| { | |
| "epoch": 5.855072463768116, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 1.975983888040945e-05, | |
| "loss": 0.1872, | |
| "step": 3636 | |
| }, | |
| { | |
| "epoch": 5.861513687600644, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 1.9543076718766538e-05, | |
| "loss": 0.1883, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 5.867954911433173, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.932742716942946e-05, | |
| "loss": 0.1543, | |
| "step": 3644 | |
| }, | |
| { | |
| "epoch": 5.874396135265701, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 1.911289207158254e-05, | |
| "loss": 0.1807, | |
| "step": 3648 | |
| }, | |
| { | |
| "epoch": 5.880837359098229, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.8899473254905672e-05, | |
| "loss": 0.1775, | |
| "step": 3652 | |
| }, | |
| { | |
| "epoch": 5.887278582930757, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 1.8687172539558208e-05, | |
| "loss": 0.1767, | |
| "step": 3656 | |
| }, | |
| { | |
| "epoch": 5.8937198067632846, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 1.8475991736163835e-05, | |
| "loss": 0.1662, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 5.900161030595813, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 1.8265932645794827e-05, | |
| "loss": 0.1575, | |
| "step": 3664 | |
| }, | |
| { | |
| "epoch": 5.906602254428341, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.805699705995708e-05, | |
| "loss": 0.1778, | |
| "step": 3668 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.7849186760574346e-05, | |
| "loss": 0.1661, | |
| "step": 3672 | |
| }, | |
| { | |
| "epoch": 5.919484702093397, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 1.7642503519973432e-05, | |
| "loss": 0.1603, | |
| "step": 3676 | |
| }, | |
| { | |
| "epoch": 5.925925925925926, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 1.7436949100868864e-05, | |
| "loss": 0.1603, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 5.932367149758454, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 1.7232525256348013e-05, | |
| "loss": 0.1907, | |
| "step": 3684 | |
| }, | |
| { | |
| "epoch": 5.938808373590982, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.7029233729855883e-05, | |
| "loss": 0.1848, | |
| "step": 3688 | |
| }, | |
| { | |
| "epoch": 5.94524959742351, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.6827076255180593e-05, | |
| "loss": 0.1719, | |
| "step": 3692 | |
| }, | |
| { | |
| "epoch": 5.951690821256038, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.6626054556438322e-05, | |
| "loss": 0.1819, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 5.958132045088567, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 1.6426170348058703e-05, | |
| "loss": 0.1669, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.964573268921095, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 1.6227425334770245e-05, | |
| "loss": 0.169, | |
| "step": 3704 | |
| }, | |
| { | |
| "epoch": 5.971014492753623, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.6029821211585592e-05, | |
| "loss": 0.1723, | |
| "step": 3708 | |
| }, | |
| { | |
| "epoch": 5.977455716586151, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.5833359663787392e-05, | |
| "loss": 0.2008, | |
| "step": 3712 | |
| }, | |
| { | |
| "epoch": 5.9838969404186795, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 1.563804236691364e-05, | |
| "loss": 0.1523, | |
| "step": 3716 | |
| }, | |
| { | |
| "epoch": 5.990338164251208, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.5443870986743562e-05, | |
| "loss": 0.1592, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 5.996779388083736, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.5250847179283243e-05, | |
| "loss": 0.2154, | |
| "step": 3724 | |
| }, | |
| { | |
| "epoch": 6.003220611916264, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.505897259075171e-05, | |
| "loss": 0.1917, | |
| "step": 3728 | |
| }, | |
| { | |
| "epoch": 6.009661835748792, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 1.4868248857566734e-05, | |
| "loss": 0.1512, | |
| "step": 3732 | |
| }, | |
| { | |
| "epoch": 6.0161030595813205, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 1.4678677606330964e-05, | |
| "loss": 0.1889, | |
| "step": 3736 | |
| }, | |
| { | |
| "epoch": 6.022544283413849, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 1.4490260453817898e-05, | |
| "loss": 0.1694, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 6.028985507246377, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 1.4302999006958342e-05, | |
| "loss": 0.1365, | |
| "step": 3744 | |
| }, | |
| { | |
| "epoch": 6.035426731078905, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 1.411689486282654e-05, | |
| "loss": 0.148, | |
| "step": 3748 | |
| }, | |
| { | |
| "epoch": 6.041867954911433, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.393194960862657e-05, | |
| "loss": 0.1744, | |
| "step": 3752 | |
| }, | |
| { | |
| "epoch": 6.048309178743962, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.3748164821678759e-05, | |
| "loss": 0.1642, | |
| "step": 3756 | |
| }, | |
| { | |
| "epoch": 6.05475040257649, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 1.3565542069406433e-05, | |
| "loss": 0.1826, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 6.061191626409018, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 1.3384082909322375e-05, | |
| "loss": 0.1911, | |
| "step": 3764 | |
| }, | |
| { | |
| "epoch": 6.067632850241546, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 1.320378888901546e-05, | |
| "loss": 0.134, | |
| "step": 3768 | |
| }, | |
| { | |
| "epoch": 6.074074074074074, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 1.3024661546137694e-05, | |
| "loss": 0.1778, | |
| "step": 3772 | |
| }, | |
| { | |
| "epoch": 6.080515297906603, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.2846702408390975e-05, | |
| "loss": 0.1542, | |
| "step": 3776 | |
| }, | |
| { | |
| "epoch": 6.086956521739131, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 1.2669912993514036e-05, | |
| "loss": 0.185, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 6.093397745571659, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 1.2494294809269512e-05, | |
| "loss": 0.1937, | |
| "step": 3784 | |
| }, | |
| { | |
| "epoch": 6.099838969404187, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.2319849353431154e-05, | |
| "loss": 0.1642, | |
| "step": 3788 | |
| }, | |
| { | |
| "epoch": 6.106280193236715, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.2146578113771005e-05, | |
| "loss": 0.156, | |
| "step": 3792 | |
| }, | |
| { | |
| "epoch": 6.112721417069243, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.1974482568046694e-05, | |
| "loss": 0.172, | |
| "step": 3796 | |
| }, | |
| { | |
| "epoch": 6.119162640901771, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 1.1803564183988812e-05, | |
| "loss": 0.1655, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.125603864734299, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 1.1633824419288474e-05, | |
| "loss": 0.1741, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 6.132045088566827, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 1.146526472158487e-05, | |
| "loss": 0.1805, | |
| "step": 3808 | |
| }, | |
| { | |
| "epoch": 6.138486312399356, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.1297886528452882e-05, | |
| "loss": 0.1617, | |
| "step": 3812 | |
| }, | |
| { | |
| "epoch": 6.144927536231884, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 1.1131691267390757e-05, | |
| "loss": 0.1863, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 6.151368760064412, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 1.0966680355808122e-05, | |
| "loss": 0.2013, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 6.15780998389694, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 1.080285520101371e-05, | |
| "loss": 0.1683, | |
| "step": 3824 | |
| }, | |
| { | |
| "epoch": 6.164251207729468, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 1.0640217200203466e-05, | |
| "loss": 0.1729, | |
| "step": 3828 | |
| }, | |
| { | |
| "epoch": 6.170692431561997, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.047876774044863e-05, | |
| "loss": 0.1736, | |
| "step": 3832 | |
| }, | |
| { | |
| "epoch": 6.177133655394525, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 1.0318508198683734e-05, | |
| "loss": 0.1757, | |
| "step": 3836 | |
| }, | |
| { | |
| "epoch": 6.183574879227053, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 1.015943994169523e-05, | |
| "loss": 0.1824, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 6.190016103059581, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 1.0001564326109363e-05, | |
| "loss": 0.1536, | |
| "step": 3844 | |
| }, | |
| { | |
| "epoch": 6.1964573268921095, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.844882698381013e-06, | |
| "loss": 0.2085, | |
| "step": 3848 | |
| }, | |
| { | |
| "epoch": 6.202898550724638, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.689396394781923e-06, | |
| "loss": 0.1665, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 6.209339774557166, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.535106741389542e-06, | |
| "loss": 0.1714, | |
| "step": 3856 | |
| }, | |
| { | |
| "epoch": 6.215780998389694, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.382015054075465e-06, | |
| "loss": 0.1639, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.230122638494408e-06, | |
| "loss": 0.1391, | |
| "step": 3864 | |
| }, | |
| { | |
| "epoch": 6.2286634460547505, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.079430790072972e-06, | |
| "loss": 0.1514, | |
| "step": 3868 | |
| }, | |
| { | |
| "epoch": 6.235104669887279, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.92994079399868e-06, | |
| "loss": 0.1324, | |
| "step": 3872 | |
| }, | |
| { | |
| "epoch": 6.241545893719807, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.781653925208887e-06, | |
| "loss": 0.1629, | |
| "step": 3876 | |
| }, | |
| { | |
| "epoch": 6.247987117552335, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 8.634571448380056e-06, | |
| "loss": 0.1899, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 6.254428341384863, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 8.488694617916785e-06, | |
| "loss": 0.1696, | |
| "step": 3884 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 8.344024677941346e-06, | |
| "loss": 0.1652, | |
| "step": 3888 | |
| }, | |
| { | |
| "epoch": 6.26731078904992, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 8.200562862282912e-06, | |
| "loss": 0.1642, | |
| "step": 3892 | |
| }, | |
| { | |
| "epoch": 6.273752012882448, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 8.058310394466994e-06, | |
| "loss": 0.1458, | |
| "step": 3896 | |
| }, | |
| { | |
| "epoch": 6.280193236714976, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 7.917268487705175e-06, | |
| "loss": 0.1519, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.286634460547504, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 7.777438344884645e-06, | |
| "loss": 0.1745, | |
| "step": 3904 | |
| }, | |
| { | |
| "epoch": 6.293075684380033, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.638821158557962e-06, | |
| "loss": 0.1696, | |
| "step": 3908 | |
| }, | |
| { | |
| "epoch": 6.29951690821256, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 7.501418110932872e-06, | |
| "loss": 0.1634, | |
| "step": 3912 | |
| }, | |
| { | |
| "epoch": 6.305958132045088, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 7.365230373862274e-06, | |
| "loss": 0.1589, | |
| "step": 3916 | |
| }, | |
| { | |
| "epoch": 6.312399355877616, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 7.2302591088341576e-06, | |
| "loss": 0.1675, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 6.318840579710145, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 7.096505466961794e-06, | |
| "loss": 0.1718, | |
| "step": 3924 | |
| }, | |
| { | |
| "epoch": 6.325281803542673, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.963970588973761e-06, | |
| "loss": 0.1412, | |
| "step": 3928 | |
| }, | |
| { | |
| "epoch": 6.331723027375201, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 6.832655605204401e-06, | |
| "loss": 0.165, | |
| "step": 3932 | |
| }, | |
| { | |
| "epoch": 6.338164251207729, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.702561635584047e-06, | |
| "loss": 0.1496, | |
| "step": 3936 | |
| }, | |
| { | |
| "epoch": 6.344605475040257, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.57368978962956e-06, | |
| "loss": 0.1829, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 6.351046698872786, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 6.44604116643474e-06, | |
| "loss": 0.182, | |
| "step": 3944 | |
| }, | |
| { | |
| "epoch": 6.357487922705314, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.3196168546610634e-06, | |
| "loss": 0.1714, | |
| "step": 3948 | |
| }, | |
| { | |
| "epoch": 6.363929146537842, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 6.194417932528478e-06, | |
| "loss": 0.2102, | |
| "step": 3952 | |
| }, | |
| { | |
| "epoch": 6.37037037037037, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 6.070445467805923e-06, | |
| "loss": 0.1761, | |
| "step": 3956 | |
| }, | |
| { | |
| "epoch": 6.3768115942028984, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 5.947700517802523e-06, | |
| "loss": 0.1816, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 6.383252818035427, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 5.826184129358358e-06, | |
| "loss": 0.2008, | |
| "step": 3964 | |
| }, | |
| { | |
| "epoch": 6.389694041867955, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 5.705897338835724e-06, | |
| "loss": 0.1857, | |
| "step": 3968 | |
| }, | |
| { | |
| "epoch": 6.396135265700483, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 5.58684117211009e-06, | |
| "loss": 0.1681, | |
| "step": 3972 | |
| }, | |
| { | |
| "epoch": 6.402576489533011, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.469016644561519e-06, | |
| "loss": 0.1409, | |
| "step": 3976 | |
| }, | |
| { | |
| "epoch": 6.4090177133655395, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 5.352424761065926e-06, | |
| "loss": 0.1647, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 6.415458937198068, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.2370665159865045e-06, | |
| "loss": 0.1556, | |
| "step": 3984 | |
| }, | |
| { | |
| "epoch": 6.421900161030596, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 5.1229428931652775e-06, | |
| "loss": 0.1912, | |
| "step": 3988 | |
| }, | |
| { | |
| "epoch": 6.428341384863124, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.010054865914676e-06, | |
| "loss": 0.1521, | |
| "step": 3992 | |
| }, | |
| { | |
| "epoch": 6.434782608695652, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 4.898403397009293e-06, | |
| "loss": 0.192, | |
| "step": 3996 | |
| }, | |
| { | |
| "epoch": 6.4412238325281805, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 4.787989438677625e-06, | |
| "loss": 0.1464, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.447665056360709, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.678813932593911e-06, | |
| "loss": 0.1608, | |
| "step": 4004 | |
| }, | |
| { | |
| "epoch": 6.454106280193237, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 4.570877809870188e-06, | |
| "loss": 0.1778, | |
| "step": 4008 | |
| }, | |
| { | |
| "epoch": 6.460547504025765, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 4.464181991048349e-06, | |
| "loss": 0.1657, | |
| "step": 4012 | |
| }, | |
| { | |
| "epoch": 6.466988727858293, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 4.358727386092198e-06, | |
| "loss": 0.1707, | |
| "step": 4016 | |
| }, | |
| { | |
| "epoch": 6.473429951690822, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 4.254514894379774e-06, | |
| "loss": 0.1641, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 6.47987117552335, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 4.1515454046956384e-06, | |
| "loss": 0.1767, | |
| "step": 4024 | |
| }, | |
| { | |
| "epoch": 6.486312399355878, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 4.049819795223336e-06, | |
| "loss": 0.1914, | |
| "step": 4028 | |
| }, | |
| { | |
| "epoch": 6.492753623188406, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 3.949338933537843e-06, | |
| "loss": 0.1588, | |
| "step": 4032 | |
| }, | |
| { | |
| "epoch": 6.499194847020934, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.850103676598265e-06, | |
| "loss": 0.1517, | |
| "step": 4036 | |
| }, | |
| { | |
| "epoch": 6.505636070853463, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 3.752114870740386e-06, | |
| "loss": 0.1713, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 6.512077294685991, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.6553733516695937e-06, | |
| "loss": 0.1704, | |
| "step": 4044 | |
| }, | |
| { | |
| "epoch": 6.518518518518518, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 3.5598799444536697e-06, | |
| "loss": 0.1545, | |
| "step": 4048 | |
| }, | |
| { | |
| "epoch": 6.524959742351046, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.465635463515792e-06, | |
| "loss": 0.1684, | |
| "step": 4052 | |
| }, | |
| { | |
| "epoch": 6.531400966183575, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 3.3726407126275112e-06, | |
| "loss": 0.1472, | |
| "step": 4056 | |
| }, | |
| { | |
| "epoch": 6.537842190016103, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 3.2808964849020513e-06, | |
| "loss": 0.1268, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 6.544283413848631, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 3.190403562787369e-06, | |
| "loss": 0.1694, | |
| "step": 4064 | |
| }, | |
| { | |
| "epoch": 6.550724637681159, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.1011627180596075e-06, | |
| "loss": 0.1688, | |
| "step": 4068 | |
| }, | |
| { | |
| "epoch": 6.557165861513687, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.0131747118164018e-06, | |
| "loss": 0.1946, | |
| "step": 4072 | |
| }, | |
| { | |
| "epoch": 6.563607085346216, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 2.9264402944705665e-06, | |
| "loss": 0.1714, | |
| "step": 4076 | |
| }, | |
| { | |
| "epoch": 6.570048309178744, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 2.8409602057434865e-06, | |
| "loss": 0.1467, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 6.576489533011272, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 2.7567351746589363e-06, | |
| "loss": 0.1615, | |
| "step": 4084 | |
| }, | |
| { | |
| "epoch": 6.5829307568438, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 2.6737659195368354e-06, | |
| "loss": 0.1674, | |
| "step": 4088 | |
| }, | |
| { | |
| "epoch": 6.5893719806763285, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 2.592053147987105e-06, | |
| "loss": 0.1651, | |
| "step": 4092 | |
| }, | |
| { | |
| "epoch": 6.595813204508857, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.5115975569036718e-06, | |
| "loss": 0.1263, | |
| "step": 4096 | |
| }, | |
| { | |
| "epoch": 6.602254428341385, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.4323998324584536e-06, | |
| "loss": 0.148, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 2.354460650095602e-06, | |
| "loss": 0.1537, | |
| "step": 4104 | |
| }, | |
| { | |
| "epoch": 6.615136876006441, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.2777806745256534e-06, | |
| "loss": 0.1553, | |
| "step": 4108 | |
| }, | |
| { | |
| "epoch": 6.6215780998389695, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 2.202360559719918e-06, | |
| "loss": 0.1763, | |
| "step": 4112 | |
| }, | |
| { | |
| "epoch": 6.628019323671498, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 2.1282009489048847e-06, | |
| "loss": 0.1717, | |
| "step": 4116 | |
| }, | |
| { | |
| "epoch": 6.634460547504026, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 2.055302474556708e-06, | |
| "loss": 0.1761, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 6.640901771336554, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.9836657583958806e-06, | |
| "loss": 0.1569, | |
| "step": 4124 | |
| }, | |
| { | |
| "epoch": 6.647342995169082, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.9132914113818677e-06, | |
| "loss": 0.18, | |
| "step": 4128 | |
| }, | |
| { | |
| "epoch": 6.6537842190016105, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.8441800337078982e-06, | |
| "loss": 0.1563, | |
| "step": 4132 | |
| }, | |
| { | |
| "epoch": 6.660225442834139, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.7763322147958836e-06, | |
| "loss": 0.1637, | |
| "step": 4136 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.7097485332913885e-06, | |
| "loss": 0.173, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 6.673107890499195, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.6444295570586518e-06, | |
| "loss": 0.1901, | |
| "step": 4144 | |
| }, | |
| { | |
| "epoch": 6.679549114331723, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.580375843175824e-06, | |
| "loss": 0.1777, | |
| "step": 4148 | |
| }, | |
| { | |
| "epoch": 6.685990338164252, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.5175879379300704e-06, | |
| "loss": 0.1655, | |
| "step": 4152 | |
| }, | |
| { | |
| "epoch": 6.692431561996779, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 1.4560663768131253e-06, | |
| "loss": 0.1555, | |
| "step": 4156 | |
| }, | |
| { | |
| "epoch": 6.698872785829307, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 1.3958116845164958e-06, | |
| "loss": 0.1506, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 6.705314009661835, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.3368243749271813e-06, | |
| "loss": 0.1987, | |
| "step": 4164 | |
| }, | |
| { | |
| "epoch": 6.7117552334943635, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.2791049511231277e-06, | |
| "loss": 0.1656, | |
| "step": 4168 | |
| }, | |
| { | |
| "epoch": 6.718196457326892, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 1.2226539053690975e-06, | |
| "loss": 0.193, | |
| "step": 4172 | |
| }, | |
| { | |
| "epoch": 6.72463768115942, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.1674717191123228e-06, | |
| "loss": 0.164, | |
| "step": 4176 | |
| }, | |
| { | |
| "epoch": 6.731078904991948, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.1135588629785252e-06, | |
| "loss": 0.1586, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 6.737520128824476, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.0609157967677695e-06, | |
| "loss": 0.1586, | |
| "step": 4184 | |
| }, | |
| { | |
| "epoch": 6.743961352657005, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.0095429694506829e-06, | |
| "loss": 0.1947, | |
| "step": 4188 | |
| }, | |
| { | |
| "epoch": 6.750402576489533, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.594408191645254e-07, | |
| "loss": 0.1751, | |
| "step": 4192 | |
| }, | |
| { | |
| "epoch": 6.756843800322061, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.106097732095085e-07, | |
| "loss": 0.1543, | |
| "step": 4196 | |
| }, | |
| { | |
| "epoch": 6.763285024154589, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 8.630502480450996e-07, | |
| "loss": 0.195, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.769726247987117, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 8.167626492865064e-07, | |
| "loss": 0.1806, | |
| "step": 4204 | |
| }, | |
| { | |
| "epoch": 6.776167471819646, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 7.717473717012312e-07, | |
| "loss": 0.1726, | |
| "step": 4208 | |
| }, | |
| { | |
| "epoch": 6.782608695652174, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 7.280047992056725e-07, | |
| "loss": 0.1747, | |
| "step": 4212 | |
| }, | |
| { | |
| "epoch": 6.789049919484702, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 6.855353048618284e-07, | |
| "loss": 0.1681, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 6.79549114331723, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 6.443392508742151e-07, | |
| "loss": 0.1712, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 6.8019323671497585, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 6.0441698858667e-07, | |
| "loss": 0.1498, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 6.808373590982287, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 5.657688584793874e-07, | |
| "loss": 0.1483, | |
| "step": 4228 | |
| }, | |
| { | |
| "epoch": 6.814814814814815, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 5.283951901660366e-07, | |
| "loss": 0.1557, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 6.821256038647343, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 4.922963023909321e-07, | |
| "loss": 0.1984, | |
| "step": 4236 | |
| }, | |
| { | |
| "epoch": 6.827697262479871, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 4.574725030263515e-07, | |
| "loss": 0.1322, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 6.8341384863123995, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 4.239240890698381e-07, | |
| "loss": 0.1781, | |
| "step": 4244 | |
| }, | |
| { | |
| "epoch": 6.840579710144928, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.9165134664170263e-07, | |
| "loss": 0.1824, | |
| "step": 4248 | |
| }, | |
| { | |
| "epoch": 6.847020933977456, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.6065455098264195e-07, | |
| "loss": 0.1787, | |
| "step": 4252 | |
| }, | |
| { | |
| "epoch": 6.853462157809984, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 3.309339664513078e-07, | |
| "loss": 0.1622, | |
| "step": 4256 | |
| }, | |
| { | |
| "epoch": 6.859903381642512, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 3.024898465220582e-07, | |
| "loss": 0.1756, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 6.8663446054750406, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.7532243378285966e-07, | |
| "loss": 0.1465, | |
| "step": 4264 | |
| }, | |
| { | |
| "epoch": 6.872785829307569, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 2.494319599331718e-07, | |
| "loss": 0.1519, | |
| "step": 4268 | |
| }, | |
| { | |
| "epoch": 6.879227053140097, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 2.2481864578194898e-07, | |
| "loss": 0.175, | |
| "step": 4272 | |
| }, | |
| { | |
| "epoch": 6.885668276972625, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 2.0148270124582533e-07, | |
| "loss": 0.161, | |
| "step": 4276 | |
| }, | |
| { | |
| "epoch": 6.892109500805153, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.7942432534728268e-07, | |
| "loss": 0.1582, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 6.898550724637682, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.5864370621293531e-07, | |
| "loss": 0.1852, | |
| "step": 4284 | |
| }, | |
| { | |
| "epoch": 6.90499194847021, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 1.3914102107193127e-07, | |
| "loss": 0.1738, | |
| "step": 4288 | |
| }, | |
| { | |
| "epoch": 6.911433172302738, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 1.2091643625452008e-07, | |
| "loss": 0.149, | |
| "step": 4292 | |
| }, | |
| { | |
| "epoch": 6.917874396135265, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 1.0397010719050414e-07, | |
| "loss": 0.1963, | |
| "step": 4296 | |
| }, | |
| { | |
| "epoch": 6.9243156199677935, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 8.830217840800624e-08, | |
| "loss": 0.182, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 6.930756843800322, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 7.391278353223728e-08, | |
| "loss": 0.1708, | |
| "step": 4304 | |
| }, | |
| { | |
| "epoch": 6.93719806763285, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.080204528426391e-08, | |
| "loss": 0.1593, | |
| "step": 4308 | |
| }, | |
| { | |
| "epoch": 6.943639291465378, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 4.897007548010923e-08, | |
| "loss": 0.1506, | |
| "step": 4312 | |
| }, | |
| { | |
| "epoch": 6.950080515297906, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.841697502963703e-08, | |
| "loss": 0.152, | |
| "step": 4316 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.9142833935819065e-08, | |
| "loss": 0.168, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 6.962962962962963, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 2.1147731293919002e-08, | |
| "loss": 0.1745, | |
| "step": 4324 | |
| }, | |
| { | |
| "epoch": 6.969404186795491, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.4431735290809654e-08, | |
| "loss": 0.1848, | |
| "step": 4328 | |
| }, | |
| { | |
| "epoch": 6.975845410628019, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 8.994903204390113e-09, | |
| "loss": 0.1714, | |
| "step": 4332 | |
| }, | |
| { | |
| "epoch": 6.982286634460547, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 4.837281403119453e-09, | |
| "loss": 0.2034, | |
| "step": 4336 | |
| }, | |
| { | |
| "epoch": 6.988727858293076, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 1.958905345600392e-09, | |
| "loss": 0.1877, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 6.995169082125604, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 3.597995803128473e-10, | |
| "loss": 0.1574, | |
| "step": 4344 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 4347, | |
| "total_flos": 2.864063510520791e+18, | |
| "train_loss": 0.36376489625186625, | |
| "train_runtime": 17966.6164, | |
| "train_samples_per_second": 7.736, | |
| "train_steps_per_second": 0.242 | |
| } | |
| ], | |
| "logging_steps": 4, | |
| "max_steps": 4347, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.864063510520791e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |