| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997655182451062, | |
| "eval_steps": 500, | |
| "global_step": 966, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010349539526346855, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 2.061855670103093e-06, | |
| "loss": 1.7433, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005174769763173427, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 1.0309278350515464e-05, | |
| "loss": 1.7444, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010349539526346855, | |
| "grad_norm": 1.0, | |
| "learning_rate": 2.0618556701030927e-05, | |
| "loss": 1.7363, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015524309289520283, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.0927835051546395e-05, | |
| "loss": 1.7131, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02069907905269371, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 4.1237113402061855e-05, | |
| "loss": 1.7097, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02587384881586714, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 5.1546391752577315e-05, | |
| "loss": 1.6876, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.031048618579040566, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 6.185567010309279e-05, | |
| "loss": 1.552, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.036223388342213995, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.216494845360825e-05, | |
| "loss": 1.5049, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04139815810538742, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 8.247422680412371e-05, | |
| "loss": 1.4914, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04657292786856085, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 9.278350515463918e-05, | |
| "loss": 1.4412, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05174769763173428, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00010309278350515463, | |
| "loss": 1.4186, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0569224673949077, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0001134020618556701, | |
| "loss": 1.4002, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06209723715808113, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00012371134020618558, | |
| "loss": 1.3963, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06727200692125455, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00013402061855670103, | |
| "loss": 1.3791, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07244677668442799, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.0001443298969072165, | |
| "loss": 1.3596, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07762154644760141, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 0.00015463917525773197, | |
| "loss": 1.3376, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08279631621077484, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 0.00016494845360824742, | |
| "loss": 1.3465, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08797108597394827, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 0.0001752577319587629, | |
| "loss": 1.3416, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0931458557371217, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 0.00018556701030927837, | |
| "loss": 1.329, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09832062550029512, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.00019587628865979381, | |
| "loss": 1.3077, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10349539526346856, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.00019999411876613578, | |
| "loss": 1.3115, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10867016502664198, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 0.00019995818039769561, | |
| "loss": 1.3185, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1138449347898154, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00019988958274085367, | |
| "loss": 1.2811, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11901970455298884, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 0.0001997883482084895, | |
| "loss": 1.3017, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12419447431616226, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 0.0001996545098769134, | |
| "loss": 1.2927, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1293692440793357, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 0.0001994881114750593, | |
| "loss": 1.3015, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1345440138425091, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 0.00019928920737019733, | |
| "loss": 1.2854, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13971878360568254, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.00019905786255017044, | |
| "loss": 1.2995, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.14489355336885598, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00019879415260216088, | |
| "loss": 1.2739, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1500683231320294, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00019849816368799355, | |
| "loss": 1.2802, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.15524309289520283, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 0.0001981699925159847, | |
| "loss": 1.3024, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16041786265837626, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 0.0001978097463093441, | |
| "loss": 1.2642, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.16559263242154967, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 0.00019741754277114232, | |
| "loss": 1.2516, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1707674021847231, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 0.00019699351004585352, | |
| "loss": 1.2675, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.17594217194789655, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 0.00019653778667748695, | |
| "loss": 1.2646, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18111694171106996, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.00019605052156432043, | |
| "loss": 1.2773, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1862917114742434, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 0.000195531873910251, | |
| "loss": 1.2653, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19146648123741683, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.00019498201317277828, | |
| "loss": 1.2572, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.19664125100059024, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 0.0001944011190076376, | |
| "loss": 1.2737, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20181602076376368, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.00019378938121010128, | |
| "loss": 1.2848, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2069907905269371, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 0.00019314699965296696, | |
| "loss": 1.2814, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21216556029011052, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.000192474184221253, | |
| "loss": 1.2685, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.21734033005328396, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 0.00019177115474362322, | |
| "loss": 1.2611, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2225150998164574, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 0.000191038140920562, | |
| "loss": 1.2541, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2276898695796308, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.0001902753822493248, | |
| "loss": 1.2657, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23286463934280424, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.00018948312794568673, | |
| "loss": 1.2687, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.23803940910597768, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 0.0001886616368625171, | |
| "loss": 1.2779, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2432141788691511, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 0.00018781117740520386, | |
| "loss": 1.2647, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.24838894863232452, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 0.00018693202744395827, | |
| "loss": 1.244, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.25356371839549796, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 0.00018602447422302597, | |
| "loss": 1.2408, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2587384881586714, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 0.00018508881426683614, | |
| "loss": 1.2875, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2639132579218448, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 1.259, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2690880276850182, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 0.00018313440606301763, | |
| "loss": 1.2646, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27426279744819165, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 0.00018211629637824515, | |
| "loss": 1.2452, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2794375672113651, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 0.00018107135687529043, | |
| "loss": 1.2608, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2846123369745385, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00017999992896673665, | |
| "loss": 1.2508, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.28978710673771196, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 0.00017890236271971122, | |
| "loss": 1.2413, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.29496187650088534, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.00017777901674150828, | |
| "loss": 1.269, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3001366462640588, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 0.00017663025806242123, | |
| "loss": 1.2453, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3053114160272322, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 0.00017545646201582303, | |
| "loss": 1.2591, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.31048618579040566, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 0.00017425801211553373, | |
| "loss": 1.2555, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3156609555535791, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 0.0001730352999305152, | |
| "loss": 1.2656, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.32083572531675253, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.00017178872495693398, | |
| "loss": 1.2608, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3260104950799259, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 0.00017051869448763408, | |
| "loss": 1.2787, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.33118526484309935, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 0.00016922562347906238, | |
| "loss": 1.2887, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3363600346062728, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 0.00016790993441569022, | |
| "loss": 1.2624, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3415348043694462, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 0.00016657205717197495, | |
| "loss": 1.2616, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.34670957413261966, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 0.00016521242887190764, | |
| "loss": 1.2643, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3518843438957931, | |
| "grad_norm": 0.123046875, | |
| "learning_rate": 0.0001638314937461915, | |
| "loss": 1.2767, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3570591136589665, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00016242970298709867, | |
| "loss": 1.2648, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3622338834221399, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 0.00016100751460105243, | |
| "loss": 1.2409, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.36740865318531335, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 0.000159565393258983, | |
| "loss": 1.2392, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3725834229484868, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 0.00015810381014450556, | |
| "loss": 1.2709, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3777581927116602, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 0.00015662324279997115, | |
| "loss": 1.2556, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.38293296247483366, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.000155124174970439, | |
| "loss": 1.2623, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.38810773223800704, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 0.00015360709644562297, | |
| "loss": 1.2489, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3932825020011805, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 0.0001520725028998629, | |
| "loss": 1.2771, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3984572717643539, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.00015052089573017283, | |
| "loss": 1.2523, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.40363204152752735, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.00014895278189242016, | |
| "loss": 1.241, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4088068112907008, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00014736867373568794, | |
| "loss": 1.2578, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4139815810538742, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 0.00014576908883487548, | |
| "loss": 1.2625, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4191563508170476, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 0.0001441545498215912, | |
| "loss": 1.2606, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.42433112058022104, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00014252558421339354, | |
| "loss": 1.2321, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4295058903433945, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 0.00014088272424143546, | |
| "loss": 1.2494, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4346806601065679, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 0.0001392265066765686, | |
| "loss": 1.2588, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.43985542986974135, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 0.00013755747265396467, | |
| "loss": 1.2534, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4450301996329148, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 0.00013587616749631036, | |
| "loss": 1.2393, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4502049693960882, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 0.00013418314053563455, | |
| "loss": 1.2526, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4553797391592616, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 0.00013247894493382506, | |
| "loss": 1.2459, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.46055450892243505, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 0.00013076413750189467, | |
| "loss": 1.2582, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4657292786856085, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.00012903927851805414, | |
| "loss": 1.2537, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4709040484487819, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 0.00012730493154465311, | |
| "loss": 1.2601, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.47607881821195536, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.0001255616632440475, | |
| "loss": 1.2612, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.48125358797512874, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 0.00012381004319345402, | |
| "loss": 1.2555, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4864283577383022, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00012205064369885291, | |
| "loss": 1.247, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4916031275014756, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 0.00012028403960799821, | |
| "loss": 1.2467, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.49677789726464905, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 0.00011851080812259838, | |
| "loss": 1.2369, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5019526670278225, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 0.00011673152860972724, | |
| "loss": 1.2644, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5071274367909959, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.00011494678241252781, | |
| "loss": 1.2345, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5123022065541694, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00011315715266027014, | |
| "loss": 1.2556, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5174769763173428, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 0.00011136322407782602, | |
| "loss": 1.2403, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5226517460805162, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 0.00010956558279462187, | |
| "loss": 1.2345, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5278265158436896, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00010776481615313315, | |
| "loss": 1.2298, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.533001285606863, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 0.00010596151251698199, | |
| "loss": 1.2475, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5381760553700364, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 0.0001041562610787017, | |
| "loss": 1.2542, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5433508251332099, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00010234965166723003, | |
| "loss": 1.2534, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5485255948963833, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.00010054227455519458, | |
| "loss": 1.2598, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5537003646595567, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 9.873472026605346e-05, | |
| "loss": 1.2551, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5588751344227302, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 9.692757938115376e-05, | |
| "loss": 1.2553, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5640499041859036, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 9.512144234677116e-05, | |
| "loss": 1.2584, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.569224673949077, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 9.331689928119381e-05, | |
| "loss": 1.2492, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5743994437122505, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 9.151453978191311e-05, | |
| "loss": 1.2315, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5795742134754239, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 8.9714952732985e-05, | |
| "loss": 1.2255, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5847489832385974, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 8.791872611262393e-05, | |
| "loss": 1.2269, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5899237530017707, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 8.612644680109319e-05, | |
| "loss": 1.2481, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5950985227649441, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 8.433870038895377e-05, | |
| "loss": 1.2304, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6002732925281176, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 8.255607098573454e-05, | |
| "loss": 1.2507, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.605448062291291, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 8.077914102908661e-05, | |
| "loss": 1.2652, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6106228320544644, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 7.90084910944835e-05, | |
| "loss": 1.2474, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6157976018176379, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 7.724469970553027e-05, | |
| "loss": 1.2387, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6209723715808113, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 7.548834314494255e-05, | |
| "loss": 1.236, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6261471413439847, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 7.37399952662581e-05, | |
| "loss": 1.2304, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6313219111071582, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 7.200022730634187e-05, | |
| "loss": 1.229, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6364966808703316, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 7.026960769874605e-05, | |
| "loss": 1.2483, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6416714506335051, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 6.85487018879861e-05, | |
| "loss": 1.2515, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6468462203966785, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 6.683807214479323e-05, | |
| "loss": 1.2388, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6520209901598518, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 6.513827738240407e-05, | |
| "loss": 1.2316, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6571957599230253, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 6.344987297394713e-05, | |
| "loss": 1.2409, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6623705296861987, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 6.177341057098607e-05, | |
| "loss": 1.2661, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6675452994493721, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 6.010943792327875e-05, | |
| "loss": 1.2272, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6727200692125456, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 1.2615, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.677894838975719, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 5.682113231116557e-05, | |
| "loss": 1.2653, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6830696087388924, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 5.519787373327725e-05, | |
| "loss": 1.2355, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6882443785020659, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 5.358925333264403e-05, | |
| "loss": 1.2322, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6934191482652393, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 5.1995796693038844e-05, | |
| "loss": 1.2549, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6985939180284128, | |
| "grad_norm": 0.0927734375, | |
| "learning_rate": 5.0418024443786395e-05, | |
| "loss": 1.2155, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7037686877915862, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 4.885645208965779e-05, | |
| "loss": 1.2497, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7089434575547596, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 4.7311589842440415e-05, | |
| "loss": 1.2692, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.714118227317933, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 4.578394245423626e-05, | |
| "loss": 1.2593, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7192929970811064, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 4.427400905254483e-05, | |
| "loss": 1.2458, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7244677668442798, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 4.278228297718307e-05, | |
| "loss": 1.2285, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7296425366074533, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 4.1309251619097154e-05, | |
| "loss": 1.243, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7348173063706267, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 3.985539626111708e-05, | |
| "loss": 1.2492, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7399920761338001, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 3.842119192070762e-05, | |
| "loss": 1.2401, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7451668458969736, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 3.700710719476594e-05, | |
| "loss": 1.2563, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.750341615660147, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 3.561360410651713e-05, | |
| "loss": 1.2479, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7555163854233204, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 3.424113795455779e-05, | |
| "loss": 1.2366, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7606911551864939, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 3.289015716409631e-05, | |
| "loss": 1.2184, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7658659249496673, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 3.156110314043933e-05, | |
| "loss": 1.2454, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7710406947128408, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 3.025441012477126e-05, | |
| "loss": 1.2323, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7762154644760141, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 2.89705050522753e-05, | |
| "loss": 1.2344, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7813902342391875, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 2.7709807412640743e-05, | |
| "loss": 1.2428, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.786565004002361, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 2.6472729113003615e-05, | |
| "loss": 1.2585, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7917397737655344, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 2.5259674343364104e-05, | |
| "loss": 1.2397, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7969145435287078, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 2.4071039444526046e-05, | |
| "loss": 1.2581, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8020893132918813, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 2.290721277860064e-05, | |
| "loss": 1.2582, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8072640830550547, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 2.176857460211693e-05, | |
| "loss": 1.2389, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8124388528182281, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 2.0655496941780994e-05, | |
| "loss": 1.245, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8176136225814016, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 1.9568343472923524e-05, | |
| "loss": 1.2398, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.822788392344575, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 1.8507469400676735e-05, | |
| "loss": 1.2408, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8279631621077485, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 1.7473221343918055e-05, | |
| "loss": 1.2355, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8331379318709219, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 1.6465937222019745e-05, | |
| "loss": 1.2552, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8383127016340952, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 1.5485946144440354e-05, | |
| "loss": 1.2461, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8434874713972687, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 1.4533568303195333e-05, | |
| "loss": 1.2449, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8486622411604421, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 1.3609114868240536e-05, | |
| "loss": 1.2424, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8538370109236155, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 1.2712887885803937e-05, | |
| "loss": 1.2218, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.859011780686789, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 1.184518017969829e-05, | |
| "loss": 1.2556, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8641865504499624, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 1.1006275255646681e-05, | |
| "loss": 1.2493, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8693613202131358, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 1.0196447208653193e-05, | |
| "loss": 1.2509, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8745360899763093, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 9.415960633447674e-06, | |
| "loss": 1.2322, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.8797108597394827, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 8.665070538035036e-06, | |
| "loss": 1.2422, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8848856295026561, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 7.944022260376416e-06, | |
| "loss": 1.238, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8900603992658296, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 7.253051388230248e-06, | |
| "loss": 1.25, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.895235169029003, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 6.592383682178593e-06, | |
| "loss": 1.2485, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9004099387921763, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 5.96223500186458e-06, | |
| "loss": 1.2503, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9055847085553498, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 5.362811235464726e-06, | |
| "loss": 1.2552, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9107594783185232, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 4.794308232419065e-06, | |
| "loss": 1.2271, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9159342480816967, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 4.256911739441505e-06, | |
| "loss": 1.2295, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9211090178448701, | |
| "grad_norm": 0.068359375, | |
| "learning_rate": 3.7507973398307584e-06, | |
| "loss": 1.2511, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9262837876080435, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 3.276130396102217e-06, | |
| "loss": 1.2527, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.931458557371217, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 2.8330659959589946e-06, | |
| "loss": 1.2444, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9366333271343904, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 2.421748901620369e-06, | |
| "loss": 1.2495, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9418080968975638, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 2.0423135025235294e-06, | |
| "loss": 1.2631, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9469828666607373, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 1.6948837714146793e-06, | |
| "loss": 1.2336, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9521576364239107, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 1.3795732238434556e-06, | |
| "loss": 1.28, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9573324061870842, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 1.0964848810740113e-06, | |
| "loss": 1.2429, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9625071759502575, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 8.457112364250064e-07, | |
| "loss": 1.2278, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9676819457134309, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 6.273342250492254e-07, | |
| "loss": 1.2675, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9728567154766044, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 4.4142519716300344e-07, | |
| "loss": 1.2405, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9780314852397778, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 2.8804489473395203e-07, | |
| "loss": 1.2512, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9832062550029512, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 1.6724343163483368e-07, | |
| "loss": 1.2416, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9883810247661247, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 7.906027726981568e-08, | |
| "loss": 1.2292, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.9935557945292981, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 2.3524243678685332e-08, | |
| "loss": 1.2187, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9987305642924715, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 6.534761230914122e-10, | |
| "loss": 1.2557, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.9997655182451062, | |
| "eval_loss": 1.2492531538009644, | |
| "eval_runtime": 2450.1313, | |
| "eval_samples_per_second": 5.585, | |
| "eval_steps_per_second": 5.585, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.9997655182451062, | |
| "step": 966, | |
| "total_flos": 3.041046347630248e+18, | |
| "train_loss": 1.1233504201067654, | |
| "train_runtime": 77534.6649, | |
| "train_samples_per_second": 1.595, | |
| "train_steps_per_second": 0.012 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 966, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 20, | |
| "total_flos": 3.041046347630248e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |