| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3327048907618942, | |
| "eval_steps": 375, | |
| "global_step": 1125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0002957376806772393, | |
| "grad_norm": 5.918768405914307, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3931, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0005914753613544786, | |
| "grad_norm": 6.164484024047852, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5398, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0008872130420317178, | |
| "grad_norm": 5.500333309173584, | |
| "learning_rate": 6e-05, | |
| "loss": 1.4837, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0011829507227089572, | |
| "grad_norm": 6.977994441986084, | |
| "learning_rate": 8e-05, | |
| "loss": 1.5093, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0014786884033861965, | |
| "grad_norm": 10.189373016357422, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3296, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0017744260840634356, | |
| "grad_norm": 8.993965148925781, | |
| "learning_rate": 0.00012, | |
| "loss": 0.9724, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.002070163764740675, | |
| "grad_norm": 2.4879002571105957, | |
| "learning_rate": 0.00014, | |
| "loss": 0.6618, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0023659014454179145, | |
| "grad_norm": 3.1456634998321533, | |
| "learning_rate": 0.00016, | |
| "loss": 0.5917, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0026616391260951538, | |
| "grad_norm": 2.20988392829895, | |
| "learning_rate": 0.00018, | |
| "loss": 0.5093, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.002957376806772393, | |
| "grad_norm": 1.189419150352478, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5016, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.003253114487449632, | |
| "grad_norm": 0.7363479733467102, | |
| "learning_rate": 0.00019999977772170748, | |
| "loss": 0.4919, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0035488521681268713, | |
| "grad_norm": 0.8657339215278625, | |
| "learning_rate": 0.00019999911088781805, | |
| "loss": 0.5377, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0038445898488041106, | |
| "grad_norm": 0.5705428123474121, | |
| "learning_rate": 0.0001999979995012962, | |
| "loss": 0.5056, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00414032752948135, | |
| "grad_norm": 0.6129299402236938, | |
| "learning_rate": 0.00019999644356708261, | |
| "loss": 0.4554, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.00443606521015859, | |
| "grad_norm": 0.4173682928085327, | |
| "learning_rate": 0.00019999444309209432, | |
| "loss": 0.3437, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.004731802890835829, | |
| "grad_norm": 0.39718082547187805, | |
| "learning_rate": 0.0001999919980852246, | |
| "loss": 0.5607, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.005027540571513068, | |
| "grad_norm": 0.2396816462278366, | |
| "learning_rate": 0.00019998910855734288, | |
| "loss": 0.4337, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0053232782521903076, | |
| "grad_norm": 0.36948394775390625, | |
| "learning_rate": 0.0001999857745212947, | |
| "loss": 0.4386, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.005619015932867547, | |
| "grad_norm": 0.4336908161640167, | |
| "learning_rate": 0.00019998199599190178, | |
| "loss": 0.4447, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.005914753613544786, | |
| "grad_norm": 0.3717949390411377, | |
| "learning_rate": 0.0001999777729859618, | |
| "loss": 0.3468, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0062104912942220255, | |
| "grad_norm": 0.2210528403520584, | |
| "learning_rate": 0.00019997310552224846, | |
| "loss": 0.3254, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.006506228974899264, | |
| "grad_norm": 0.373976469039917, | |
| "learning_rate": 0.00019996799362151122, | |
| "loss": 0.3819, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.006801966655576503, | |
| "grad_norm": 0.4056873917579651, | |
| "learning_rate": 0.00019996243730647538, | |
| "loss": 0.395, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0070977043362537426, | |
| "grad_norm": 0.2301822304725647, | |
| "learning_rate": 0.00019995643660184191, | |
| "loss": 0.4185, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.007393442016930982, | |
| "grad_norm": 0.22644546627998352, | |
| "learning_rate": 0.00019994999153428737, | |
| "loss": 0.3096, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.007689179697608221, | |
| "grad_norm": 0.3260557949542999, | |
| "learning_rate": 0.00019994310213246368, | |
| "loss": 0.3562, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.007984917378285461, | |
| "grad_norm": 0.2493218034505844, | |
| "learning_rate": 0.00019993576842699816, | |
| "loss": 0.4506, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0082806550589627, | |
| "grad_norm": 0.21800749003887177, | |
| "learning_rate": 0.0001999279904504933, | |
| "loss": 0.3406, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.00857639273963994, | |
| "grad_norm": 0.2646995186805725, | |
| "learning_rate": 0.00019991976823752653, | |
| "loss": 0.3676, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.00887213042031718, | |
| "grad_norm": 0.2616923749446869, | |
| "learning_rate": 0.00019991110182465032, | |
| "loss": 0.3685, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.009167868100994419, | |
| "grad_norm": 0.2577691078186035, | |
| "learning_rate": 0.00019990199125039174, | |
| "loss": 0.3006, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.009463605781671658, | |
| "grad_norm": 0.26562878489494324, | |
| "learning_rate": 0.00019989243655525247, | |
| "loss": 0.3532, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.009759343462348897, | |
| "grad_norm": 0.26207393407821655, | |
| "learning_rate": 0.00019988243778170853, | |
| "loss": 0.3386, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.010055081143026137, | |
| "grad_norm": 0.2501125931739807, | |
| "learning_rate": 0.0001998719949742101, | |
| "loss": 0.2988, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.010350818823703376, | |
| "grad_norm": 0.3094285726547241, | |
| "learning_rate": 0.0001998611081791814, | |
| "loss": 0.3979, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.010646556504380615, | |
| "grad_norm": 0.2816741466522217, | |
| "learning_rate": 0.00019984977744502038, | |
| "loss": 0.3338, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.010942294185057854, | |
| "grad_norm": 0.33405667543411255, | |
| "learning_rate": 0.00019983800282209857, | |
| "loss": 0.4424, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.011238031865735094, | |
| "grad_norm": 0.26445043087005615, | |
| "learning_rate": 0.00019982578436276082, | |
| "loss": 0.3255, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.011533769546412333, | |
| "grad_norm": 0.2719035744667053, | |
| "learning_rate": 0.00019981312212132512, | |
| "loss": 0.3665, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.011829507227089572, | |
| "grad_norm": 0.41192904114723206, | |
| "learning_rate": 0.00019980001615408228, | |
| "loss": 0.3978, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.012125244907766812, | |
| "grad_norm": 0.30053630471229553, | |
| "learning_rate": 0.00019978646651929572, | |
| "loss": 0.3416, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.012420982588444051, | |
| "grad_norm": 0.32080647349357605, | |
| "learning_rate": 0.00019977247327720128, | |
| "loss": 0.3517, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.012716720269121289, | |
| "grad_norm": 0.3234342038631439, | |
| "learning_rate": 0.0001997580364900068, | |
| "loss": 0.2878, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.013012457949798528, | |
| "grad_norm": 0.33844128251075745, | |
| "learning_rate": 0.000199743156221892, | |
| "loss": 0.2822, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.013308195630475767, | |
| "grad_norm": 0.36423033475875854, | |
| "learning_rate": 0.00019972783253900808, | |
| "loss": 0.3715, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.013603933311153006, | |
| "grad_norm": 0.34381452202796936, | |
| "learning_rate": 0.00019971206550947748, | |
| "loss": 0.3983, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.013899670991830246, | |
| "grad_norm": 0.3789922595024109, | |
| "learning_rate": 0.00019969585520339354, | |
| "loss": 0.2693, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.014195408672507485, | |
| "grad_norm": 0.4301562011241913, | |
| "learning_rate": 0.0001996792016928203, | |
| "loss": 0.3752, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.014491146353184724, | |
| "grad_norm": 0.4085782766342163, | |
| "learning_rate": 0.00019966210505179197, | |
| "loss": 0.3779, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.014786884033861964, | |
| "grad_norm": 0.7027013897895813, | |
| "learning_rate": 0.00019964456535631286, | |
| "loss": 0.4713, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015082621714539203, | |
| "grad_norm": 0.12943406403064728, | |
| "learning_rate": 0.0001996265826843568, | |
| "loss": 0.1727, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.015378359395216442, | |
| "grad_norm": 0.20666781067848206, | |
| "learning_rate": 0.00019960815711586696, | |
| "loss": 0.2999, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01567409707589368, | |
| "grad_norm": 0.18581970036029816, | |
| "learning_rate": 0.00019958928873275539, | |
| "loss": 0.3003, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.015969834756570923, | |
| "grad_norm": 0.1757449209690094, | |
| "learning_rate": 0.00019956997761890277, | |
| "loss": 0.2614, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01626557243724816, | |
| "grad_norm": 0.1774774044752121, | |
| "learning_rate": 0.00019955022386015792, | |
| "loss": 0.3129, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0165613101179254, | |
| "grad_norm": 0.17331115901470184, | |
| "learning_rate": 0.00019953002754433743, | |
| "loss": 0.3113, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01685704779860264, | |
| "grad_norm": 0.18873649835586548, | |
| "learning_rate": 0.00019950938876122542, | |
| "loss": 0.2683, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01715278547927988, | |
| "grad_norm": 0.15520928800106049, | |
| "learning_rate": 0.00019948830760257291, | |
| "loss": 0.2173, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.017448523159957118, | |
| "grad_norm": 0.14979037642478943, | |
| "learning_rate": 0.0001994667841620976, | |
| "loss": 0.23, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01774426084063436, | |
| "grad_norm": 0.15720731019973755, | |
| "learning_rate": 0.00019944481853548335, | |
| "loss": 0.2606, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.018039998521311596, | |
| "grad_norm": 0.18836985528469086, | |
| "learning_rate": 0.00019942241082037982, | |
| "loss": 0.34, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.018335736201988837, | |
| "grad_norm": 0.17339695990085602, | |
| "learning_rate": 0.00019939956111640197, | |
| "loss": 0.2927, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.018631473882666075, | |
| "grad_norm": 0.16999205946922302, | |
| "learning_rate": 0.00019937626952512964, | |
| "loss": 0.2591, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.018927211563343316, | |
| "grad_norm": 0.1601104885339737, | |
| "learning_rate": 0.0001993525361501072, | |
| "loss": 0.2929, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.019222949244020553, | |
| "grad_norm": 0.17350880801677704, | |
| "learning_rate": 0.00019932836109684286, | |
| "loss": 0.2854, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.019518686924697794, | |
| "grad_norm": 0.18091264367103577, | |
| "learning_rate": 0.00019930374447280845, | |
| "loss": 0.2839, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.019814424605375032, | |
| "grad_norm": 0.1761292666196823, | |
| "learning_rate": 0.00019927868638743875, | |
| "loss": 0.2545, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.020110162286052273, | |
| "grad_norm": 0.18896149098873138, | |
| "learning_rate": 0.0001992531869521312, | |
| "loss": 0.2782, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02040589996672951, | |
| "grad_norm": 0.1839849352836609, | |
| "learning_rate": 0.00019922724628024515, | |
| "loss": 0.2596, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.02070163764740675, | |
| "grad_norm": 0.18426676094532013, | |
| "learning_rate": 0.0001992008644871016, | |
| "loss": 0.2921, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02099737532808399, | |
| "grad_norm": 0.23805294930934906, | |
| "learning_rate": 0.00019917404168998256, | |
| "loss": 0.3308, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.02129311300876123, | |
| "grad_norm": 0.18368969857692719, | |
| "learning_rate": 0.0001991467780081305, | |
| "loss": 0.3328, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.021588850689438468, | |
| "grad_norm": 0.20601886510849, | |
| "learning_rate": 0.00019911907356274795, | |
| "loss": 0.3049, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02188458837011571, | |
| "grad_norm": 0.21643660962581635, | |
| "learning_rate": 0.00019909092847699683, | |
| "loss": 0.3097, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.022180326050792946, | |
| "grad_norm": 0.1990053504705429, | |
| "learning_rate": 0.00019906234287599798, | |
| "loss": 0.3245, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.022476063731470187, | |
| "grad_norm": 0.18081244826316833, | |
| "learning_rate": 0.00019903331688683057, | |
| "loss": 0.2234, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.022771801412147425, | |
| "grad_norm": 0.19501319527626038, | |
| "learning_rate": 0.00019900385063853154, | |
| "loss": 0.2852, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.023067539092824666, | |
| "grad_norm": 0.19676417112350464, | |
| "learning_rate": 0.00019897394426209505, | |
| "loss": 0.1825, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.023363276773501904, | |
| "grad_norm": 0.22049355506896973, | |
| "learning_rate": 0.00019894359789047187, | |
| "loss": 0.2708, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.023659014454179145, | |
| "grad_norm": 0.24137771129608154, | |
| "learning_rate": 0.00019891281165856873, | |
| "loss": 0.317, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.023954752134856382, | |
| "grad_norm": 0.2464693933725357, | |
| "learning_rate": 0.00019888158570324795, | |
| "loss": 0.281, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.024250489815533623, | |
| "grad_norm": 0.2559899091720581, | |
| "learning_rate": 0.0001988499201633265, | |
| "loss": 0.2977, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.02454622749621086, | |
| "grad_norm": 0.23781578242778778, | |
| "learning_rate": 0.00019881781517957562, | |
| "loss": 0.2156, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.024841965176888102, | |
| "grad_norm": 0.25769996643066406, | |
| "learning_rate": 0.0001987852708947202, | |
| "loss": 0.2588, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.02513770285756534, | |
| "grad_norm": 0.24173754453659058, | |
| "learning_rate": 0.00019875228745343794, | |
| "loss": 0.2689, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.025433440538242577, | |
| "grad_norm": 0.23814626038074493, | |
| "learning_rate": 0.0001987188650023589, | |
| "loss": 0.2404, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.025729178218919818, | |
| "grad_norm": 0.2124136984348297, | |
| "learning_rate": 0.0001986850036900648, | |
| "loss": 0.2472, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.026024915899597056, | |
| "grad_norm": 0.24799956381320953, | |
| "learning_rate": 0.00019865070366708836, | |
| "loss": 0.2429, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.026320653580274297, | |
| "grad_norm": 0.2826499938964844, | |
| "learning_rate": 0.00019861596508591255, | |
| "loss": 0.2906, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.026616391260951534, | |
| "grad_norm": 0.25857439637184143, | |
| "learning_rate": 0.00019858078810097002, | |
| "loss": 0.2139, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.026912128941628775, | |
| "grad_norm": 0.25622865557670593, | |
| "learning_rate": 0.00019854517286864245, | |
| "loss": 0.266, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.027207866622306013, | |
| "grad_norm": 0.2770698368549347, | |
| "learning_rate": 0.0001985091195472596, | |
| "loss": 0.254, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.027503604302983254, | |
| "grad_norm": 0.31867358088493347, | |
| "learning_rate": 0.0001984726282970989, | |
| "loss": 0.2851, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02779934198366049, | |
| "grad_norm": 0.35994890332221985, | |
| "learning_rate": 0.0001984356992803847, | |
| "loss": 0.321, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.028095079664337733, | |
| "grad_norm": 0.337107390165329, | |
| "learning_rate": 0.00019839833266128724, | |
| "loss": 0.2248, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.02839081734501497, | |
| "grad_norm": 0.34881696105003357, | |
| "learning_rate": 0.00019836052860592237, | |
| "loss": 0.2786, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02868655502569221, | |
| "grad_norm": 0.3969687521457672, | |
| "learning_rate": 0.0001983222872823505, | |
| "loss": 0.3405, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.02898229270636945, | |
| "grad_norm": 0.38202965259552, | |
| "learning_rate": 0.00019828360886057594, | |
| "loss": 0.275, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02927803038704669, | |
| "grad_norm": 0.4236074686050415, | |
| "learning_rate": 0.00019824449351254616, | |
| "loss": 0.2973, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.029573768067723927, | |
| "grad_norm": 0.6627792716026306, | |
| "learning_rate": 0.00019820494141215104, | |
| "loss": 0.3571, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02986950574840117, | |
| "grad_norm": 0.14578954875469208, | |
| "learning_rate": 0.000198164952735222, | |
| "loss": 0.1857, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.030165243429078406, | |
| "grad_norm": 0.1730443388223648, | |
| "learning_rate": 0.00019812452765953135, | |
| "loss": 0.2189, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.030460981109755647, | |
| "grad_norm": 0.20889230072498322, | |
| "learning_rate": 0.00019808366636479147, | |
| "loss": 0.264, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.030756718790432885, | |
| "grad_norm": 0.19660377502441406, | |
| "learning_rate": 0.00019804236903265388, | |
| "loss": 0.3058, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.031052456471110126, | |
| "grad_norm": 0.16153180599212646, | |
| "learning_rate": 0.00019800063584670863, | |
| "loss": 0.2465, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.03134819415178736, | |
| "grad_norm": 0.17819426953792572, | |
| "learning_rate": 0.00019795846699248332, | |
| "loss": 0.2406, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.031643931832464604, | |
| "grad_norm": 0.16484011709690094, | |
| "learning_rate": 0.00019791586265744237, | |
| "loss": 0.211, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.031939669513141845, | |
| "grad_norm": 0.19172953069210052, | |
| "learning_rate": 0.00019787282303098617, | |
| "loss": 0.2686, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.03223540719381908, | |
| "grad_norm": 0.17775307595729828, | |
| "learning_rate": 0.0001978293483044502, | |
| "loss": 0.2464, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.03253114487449632, | |
| "grad_norm": 0.17548003792762756, | |
| "learning_rate": 0.00019778543867110426, | |
| "loss": 0.2489, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03282688255517356, | |
| "grad_norm": 0.17837557196617126, | |
| "learning_rate": 0.00019774109432615147, | |
| "loss": 0.2305, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.0331226202358508, | |
| "grad_norm": 0.21005579829216003, | |
| "learning_rate": 0.00019769631546672756, | |
| "loss": 0.3109, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.03341835791652804, | |
| "grad_norm": 0.1606261134147644, | |
| "learning_rate": 0.00019765110229189988, | |
| "loss": 0.2297, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.03371409559720528, | |
| "grad_norm": 0.16127589344978333, | |
| "learning_rate": 0.00019760545500266657, | |
| "loss": 0.2121, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.03400983327788252, | |
| "grad_norm": 0.18138477206230164, | |
| "learning_rate": 0.00019755937380195568, | |
| "loss": 0.2556, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.03430557095855976, | |
| "grad_norm": 0.19160465896129608, | |
| "learning_rate": 0.00019751285889462423, | |
| "loss": 0.2534, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.034601308639236994, | |
| "grad_norm": 0.19009210169315338, | |
| "learning_rate": 0.0001974659104874573, | |
| "loss": 0.2289, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.034897046319914235, | |
| "grad_norm": 0.1938568651676178, | |
| "learning_rate": 0.0001974185287891671, | |
| "loss": 0.2483, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.035192784000591476, | |
| "grad_norm": 0.19120760262012482, | |
| "learning_rate": 0.0001973707140103921, | |
| "loss": 0.2522, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.03548852168126872, | |
| "grad_norm": 0.20597144961357117, | |
| "learning_rate": 0.00019732246636369605, | |
| "loss": 0.2776, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03578425936194595, | |
| "grad_norm": 0.19169169664382935, | |
| "learning_rate": 0.00019727378606356703, | |
| "loss": 0.2821, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.03607999704262319, | |
| "grad_norm": 0.24905505776405334, | |
| "learning_rate": 0.00019722467332641656, | |
| "loss": 0.3392, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.03637573472330043, | |
| "grad_norm": 0.21315141022205353, | |
| "learning_rate": 0.00019717512837057855, | |
| "loss": 0.2514, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.036671472403977674, | |
| "grad_norm": 0.20183053612709045, | |
| "learning_rate": 0.0001971251514163083, | |
| "loss": 0.2522, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.03696721008465491, | |
| "grad_norm": 0.21825973689556122, | |
| "learning_rate": 0.0001970747426857817, | |
| "loss": 0.2249, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03726294776533215, | |
| "grad_norm": 0.19961951673030853, | |
| "learning_rate": 0.00019702390240309404, | |
| "loss": 0.2081, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.03755868544600939, | |
| "grad_norm": 0.24127651751041412, | |
| "learning_rate": 0.0001969726307942592, | |
| "loss": 0.2543, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03785442312668663, | |
| "grad_norm": 0.18847639858722687, | |
| "learning_rate": 0.00019692092808720846, | |
| "loss": 0.1804, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.038150160807363866, | |
| "grad_norm": 0.2245255559682846, | |
| "learning_rate": 0.0001968687945117896, | |
| "loss": 0.2615, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.03844589848804111, | |
| "grad_norm": 0.2330857813358307, | |
| "learning_rate": 0.00019681623029976588, | |
| "loss": 0.24, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.03874163616871835, | |
| "grad_norm": 0.2308180183172226, | |
| "learning_rate": 0.00019676323568481498, | |
| "loss": 0.2735, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.03903737384939559, | |
| "grad_norm": 0.23675177991390228, | |
| "learning_rate": 0.00019670981090252792, | |
| "loss": 0.273, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.03933311153007282, | |
| "grad_norm": 0.2156166285276413, | |
| "learning_rate": 0.00019665595619040808, | |
| "loss": 0.2038, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.039628849210750064, | |
| "grad_norm": 0.22249513864517212, | |
| "learning_rate": 0.0001966016717878702, | |
| "loss": 0.2396, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.039924586891427305, | |
| "grad_norm": 0.25741225481033325, | |
| "learning_rate": 0.00019654695793623907, | |
| "loss": 0.2921, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.040220324572104546, | |
| "grad_norm": 0.25070619583129883, | |
| "learning_rate": 0.0001964918148787488, | |
| "loss": 0.2107, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.04051606225278178, | |
| "grad_norm": 0.2761012017726898, | |
| "learning_rate": 0.00019643624286054144, | |
| "loss": 0.2443, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.04081179993345902, | |
| "grad_norm": 0.22816140949726105, | |
| "learning_rate": 0.00019638024212866606, | |
| "loss": 0.1939, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.04110753761413626, | |
| "grad_norm": 0.2465653121471405, | |
| "learning_rate": 0.0001963238129320776, | |
| "loss": 0.2535, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.0414032752948135, | |
| "grad_norm": 0.38914790749549866, | |
| "learning_rate": 0.00019626695552163578, | |
| "loss": 0.3959, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.04169901297549074, | |
| "grad_norm": 0.2719482481479645, | |
| "learning_rate": 0.00019620967015010395, | |
| "loss": 0.273, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.04199475065616798, | |
| "grad_norm": 0.26919126510620117, | |
| "learning_rate": 0.00019615195707214803, | |
| "loss": 0.2197, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.04229048833684522, | |
| "grad_norm": 0.2746446132659912, | |
| "learning_rate": 0.0001960938165443353, | |
| "loss": 0.251, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.04258622601752246, | |
| "grad_norm": 0.31786006689071655, | |
| "learning_rate": 0.00019603524882513327, | |
| "loss": 0.2585, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.042881963698199695, | |
| "grad_norm": 0.3586949110031128, | |
| "learning_rate": 0.0001959762541749086, | |
| "loss": 0.2868, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.043177701378876936, | |
| "grad_norm": 0.3204839527606964, | |
| "learning_rate": 0.00019591683285592593, | |
| "loss": 0.2449, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.04347343905955418, | |
| "grad_norm": 0.3199440836906433, | |
| "learning_rate": 0.00019585698513234663, | |
| "loss": 0.238, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.04376917674023142, | |
| "grad_norm": 0.460660457611084, | |
| "learning_rate": 0.0001957967112702277, | |
| "loss": 0.2707, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.04406491442090865, | |
| "grad_norm": 0.3981362581253052, | |
| "learning_rate": 0.00019573601153752052, | |
| "loss": 0.2259, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.04436065210158589, | |
| "grad_norm": 0.5862023234367371, | |
| "learning_rate": 0.00019567488620406983, | |
| "loss": 0.3031, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.044656389782263134, | |
| "grad_norm": 0.14805294573307037, | |
| "learning_rate": 0.00019561333554161224, | |
| "loss": 0.1515, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.044952127462940375, | |
| "grad_norm": 0.2211844027042389, | |
| "learning_rate": 0.0001955513598237753, | |
| "loss": 0.2199, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.04524786514361761, | |
| "grad_norm": 0.20431561768054962, | |
| "learning_rate": 0.00019548895932607621, | |
| "loss": 0.2077, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.04554360282429485, | |
| "grad_norm": 0.1585177332162857, | |
| "learning_rate": 0.00019542613432592038, | |
| "loss": 0.1702, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.04583934050497209, | |
| "grad_norm": 0.21959953010082245, | |
| "learning_rate": 0.00019536288510260056, | |
| "loss": 0.3018, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.04613507818564933, | |
| "grad_norm": 0.18285472691059113, | |
| "learning_rate": 0.00019529921193729534, | |
| "loss": 0.2621, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.046430815866326566, | |
| "grad_norm": 0.19817829132080078, | |
| "learning_rate": 0.00019523511511306793, | |
| "loss": 0.2542, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.04672655354700381, | |
| "grad_norm": 0.2109287828207016, | |
| "learning_rate": 0.000195170594914865, | |
| "loss": 0.2398, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.04702229122768105, | |
| "grad_norm": 0.20559227466583252, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.2302, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.04731802890835829, | |
| "grad_norm": 0.2374628186225891, | |
| "learning_rate": 0.00019504028554572864, | |
| "loss": 0.2729, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.047613766589035524, | |
| "grad_norm": 0.17768986523151398, | |
| "learning_rate": 0.00019497449695409408, | |
| "loss": 0.2692, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.047909504269712765, | |
| "grad_norm": 0.1690852791070938, | |
| "learning_rate": 0.00019490828614707916, | |
| "loss": 0.2147, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.048205241950390006, | |
| "grad_norm": 0.20444796979427338, | |
| "learning_rate": 0.00019484165341902845, | |
| "loss": 0.2317, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.04850097963106725, | |
| "grad_norm": 0.2058713436126709, | |
| "learning_rate": 0.00019477459906616206, | |
| "loss": 0.1968, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.04879671731174448, | |
| "grad_norm": 0.20283836126327515, | |
| "learning_rate": 0.00019470712338657458, | |
| "loss": 0.2275, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.04909245499242172, | |
| "grad_norm": 0.18474236130714417, | |
| "learning_rate": 0.0001946392266802336, | |
| "loss": 0.2602, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.04938819267309896, | |
| "grad_norm": 0.17107436060905457, | |
| "learning_rate": 0.0001945709092489783, | |
| "loss": 0.2182, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.049683930353776204, | |
| "grad_norm": 0.18627148866653442, | |
| "learning_rate": 0.00019450217139651844, | |
| "loss": 0.1782, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.04997966803445344, | |
| "grad_norm": 0.19079387187957764, | |
| "learning_rate": 0.0001944330134284326, | |
| "loss": 0.2239, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.05027540571513068, | |
| "grad_norm": 0.22388561069965363, | |
| "learning_rate": 0.00019436343565216711, | |
| "loss": 0.2257, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.05057114339580792, | |
| "grad_norm": 0.20720115303993225, | |
| "learning_rate": 0.00019429343837703455, | |
| "loss": 0.2546, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.050866881076485154, | |
| "grad_norm": 0.19800806045532227, | |
| "learning_rate": 0.0001942230219142124, | |
| "loss": 0.1958, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.051162618757162395, | |
| "grad_norm": 0.22758133709430695, | |
| "learning_rate": 0.0001941521865767417, | |
| "loss": 0.2499, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.051458356437839636, | |
| "grad_norm": 0.20841172337532043, | |
| "learning_rate": 0.0001940809326795256, | |
| "loss": 0.2369, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.05175409411851688, | |
| "grad_norm": 0.20555779337882996, | |
| "learning_rate": 0.000194009260539328, | |
| "loss": 0.2435, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.05204983179919411, | |
| "grad_norm": 0.2264283448457718, | |
| "learning_rate": 0.0001939371704747721, | |
| "loss": 0.2603, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.05234556947987135, | |
| "grad_norm": 0.22826440632343292, | |
| "learning_rate": 0.00019386466280633906, | |
| "loss": 0.2574, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.052641307160548594, | |
| "grad_norm": 0.2327851504087448, | |
| "learning_rate": 0.00019379173785636646, | |
| "loss": 0.2125, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.052937044841225835, | |
| "grad_norm": 0.25769078731536865, | |
| "learning_rate": 0.000193718395949047, | |
| "loss": 0.2594, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.05323278252190307, | |
| "grad_norm": 0.24143129587173462, | |
| "learning_rate": 0.00019364463741042694, | |
| "loss": 0.2285, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.05352852020258031, | |
| "grad_norm": 0.21483063697814941, | |
| "learning_rate": 0.00019357046256840473, | |
| "loss": 0.1883, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.05382425788325755, | |
| "grad_norm": 0.2387738972902298, | |
| "learning_rate": 0.00019349587175272948, | |
| "loss": 0.2697, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.05411999556393479, | |
| "grad_norm": 0.2653301954269409, | |
| "learning_rate": 0.0001934208652949996, | |
| "loss": 0.2403, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.054415733244612026, | |
| "grad_norm": 0.28791022300720215, | |
| "learning_rate": 0.00019334544352866127, | |
| "loss": 0.268, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.05471147092528927, | |
| "grad_norm": 0.2307739108800888, | |
| "learning_rate": 0.00019326960678900688, | |
| "loss": 0.2088, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.05500720860596651, | |
| "grad_norm": 0.2275734543800354, | |
| "learning_rate": 0.00019319335541317361, | |
| "loss": 0.1935, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.05530294628664375, | |
| "grad_norm": 0.23581242561340332, | |
| "learning_rate": 0.00019311668974014208, | |
| "loss": 0.2037, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.05559868396732098, | |
| "grad_norm": 0.25955161452293396, | |
| "learning_rate": 0.00019303961011073447, | |
| "loss": 0.2216, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.055894421647998224, | |
| "grad_norm": 0.262517511844635, | |
| "learning_rate": 0.00019296211686761346, | |
| "loss": 0.2143, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.056190159328675465, | |
| "grad_norm": 0.3077569305896759, | |
| "learning_rate": 0.00019288421035528028, | |
| "loss": 0.2356, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.056485897009352706, | |
| "grad_norm": 0.26791897416114807, | |
| "learning_rate": 0.00019280589092007352, | |
| "loss": 0.1884, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.05678163469002994, | |
| "grad_norm": 0.27822402119636536, | |
| "learning_rate": 0.00019272715891016735, | |
| "loss": 0.2118, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.05707737237070718, | |
| "grad_norm": 0.2592899203300476, | |
| "learning_rate": 0.00019264801467557007, | |
| "loss": 0.1402, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.05737311005138442, | |
| "grad_norm": 0.37037453055381775, | |
| "learning_rate": 0.00019256845856812266, | |
| "loss": 0.3472, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.057668847732061664, | |
| "grad_norm": 0.30506840348243713, | |
| "learning_rate": 0.000192488490941497, | |
| "loss": 0.1907, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.0579645854127389, | |
| "grad_norm": 0.3820033669471741, | |
| "learning_rate": 0.00019240811215119448, | |
| "loss": 0.2557, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.05826032309341614, | |
| "grad_norm": 0.40331634879112244, | |
| "learning_rate": 0.00019232732255454422, | |
| "loss": 0.2579, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.05855606077409338, | |
| "grad_norm": 0.4118432104587555, | |
| "learning_rate": 0.00019224612251070175, | |
| "loss": 0.2316, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.05885179845477062, | |
| "grad_norm": 0.4977587163448334, | |
| "learning_rate": 0.0001921645123806472, | |
| "loss": 0.2964, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.059147536135447855, | |
| "grad_norm": 0.5816993713378906, | |
| "learning_rate": 0.0001920824925271838, | |
| "loss": 0.2308, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.059443273816125096, | |
| "grad_norm": 0.1586022526025772, | |
| "learning_rate": 0.0001920000633149362, | |
| "loss": 0.1695, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.05973901149680234, | |
| "grad_norm": 0.17268988490104675, | |
| "learning_rate": 0.00019191722511034884, | |
| "loss": 0.1737, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.06003474917747958, | |
| "grad_norm": 0.17306619882583618, | |
| "learning_rate": 0.00019183397828168448, | |
| "loss": 0.1928, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.06033048685815681, | |
| "grad_norm": 0.18508224189281464, | |
| "learning_rate": 0.00019175032319902234, | |
| "loss": 0.2293, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.06062622453883405, | |
| "grad_norm": 0.15483330190181732, | |
| "learning_rate": 0.00019166626023425662, | |
| "loss": 0.1827, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.060921962219511294, | |
| "grad_norm": 0.190659299492836, | |
| "learning_rate": 0.00019158178976109476, | |
| "loss": 0.2827, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.061217699900188535, | |
| "grad_norm": 0.16438104212284088, | |
| "learning_rate": 0.0001914969121550558, | |
| "loss": 0.2186, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.06151343758086577, | |
| "grad_norm": 0.17598582804203033, | |
| "learning_rate": 0.00019141162779346874, | |
| "loss": 0.1971, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.06180917526154301, | |
| "grad_norm": 0.18765893578529358, | |
| "learning_rate": 0.00019132593705547082, | |
| "loss": 0.1996, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.06210491294222025, | |
| "grad_norm": 0.21454760432243347, | |
| "learning_rate": 0.00019123984032200586, | |
| "loss": 0.2427, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.06240065062289749, | |
| "grad_norm": 0.21339420974254608, | |
| "learning_rate": 0.00019115333797582254, | |
| "loss": 0.2539, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.06269638830357473, | |
| "grad_norm": 0.18333065509796143, | |
| "learning_rate": 0.00019106643040147278, | |
| "loss": 0.2447, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.06299212598425197, | |
| "grad_norm": 0.19549886882305145, | |
| "learning_rate": 0.00019097911798530987, | |
| "loss": 0.2451, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.06328786366492921, | |
| "grad_norm": 0.19182579219341278, | |
| "learning_rate": 0.00019089140111548696, | |
| "loss": 0.2058, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.06358360134560645, | |
| "grad_norm": 0.19501961767673492, | |
| "learning_rate": 0.00019080328018195513, | |
| "loss": 0.2365, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.06387933902628369, | |
| "grad_norm": 0.19913062453269958, | |
| "learning_rate": 0.0001907147555764618, | |
| "loss": 0.2204, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.06417507670696093, | |
| "grad_norm": 0.1938629001379013, | |
| "learning_rate": 0.00019062582769254895, | |
| "loss": 0.2554, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.06447081438763816, | |
| "grad_norm": 0.20500774681568146, | |
| "learning_rate": 0.00019053649692555135, | |
| "loss": 0.2719, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.0647665520683154, | |
| "grad_norm": 0.19601133465766907, | |
| "learning_rate": 0.00019044676367259476, | |
| "loss": 0.2176, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.06506228974899264, | |
| "grad_norm": 0.21002735197544098, | |
| "learning_rate": 0.00019035662833259432, | |
| "loss": 0.2987, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.06535802742966988, | |
| "grad_norm": 0.19716577231884003, | |
| "learning_rate": 0.00019026609130625257, | |
| "loss": 0.2103, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.06565376511034712, | |
| "grad_norm": 0.20321358740329742, | |
| "learning_rate": 0.00019017515299605788, | |
| "loss": 0.2373, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.06594950279102436, | |
| "grad_norm": 0.21129213273525238, | |
| "learning_rate": 0.00019008381380628247, | |
| "loss": 0.2284, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.0662452404717016, | |
| "grad_norm": 0.20547550916671753, | |
| "learning_rate": 0.00018999207414298067, | |
| "loss": 0.2162, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.06654097815237885, | |
| "grad_norm": 0.19809496402740479, | |
| "learning_rate": 0.00018989993441398726, | |
| "loss": 0.2176, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.06683671583305607, | |
| "grad_norm": 0.23077711462974548, | |
| "learning_rate": 0.00018980739502891546, | |
| "loss": 0.2389, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.06713245351373331, | |
| "grad_norm": 0.24560177326202393, | |
| "learning_rate": 0.0001897144563991552, | |
| "loss": 0.2347, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.06742819119441056, | |
| "grad_norm": 0.24137191474437714, | |
| "learning_rate": 0.00018962111893787128, | |
| "loss": 0.2191, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.0677239288750878, | |
| "grad_norm": 0.23104238510131836, | |
| "learning_rate": 0.00018952738306000151, | |
| "loss": 0.2207, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.06801966655576504, | |
| "grad_norm": 0.22218018770217896, | |
| "learning_rate": 0.00018943324918225494, | |
| "loss": 0.21, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.06831540423644228, | |
| "grad_norm": 0.21070846915245056, | |
| "learning_rate": 0.0001893387177231099, | |
| "loss": 0.165, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.06861114191711952, | |
| "grad_norm": 0.2444663941860199, | |
| "learning_rate": 0.0001892437891028122, | |
| "loss": 0.2336, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.06890687959779676, | |
| "grad_norm": 0.2369096875190735, | |
| "learning_rate": 0.0001891484637433733, | |
| "loss": 0.2232, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.06920261727847399, | |
| "grad_norm": 0.2586706280708313, | |
| "learning_rate": 0.00018905274206856837, | |
| "loss": 0.2468, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.06949835495915123, | |
| "grad_norm": 0.30194905400276184, | |
| "learning_rate": 0.00018895662450393438, | |
| "loss": 0.254, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.06979409263982847, | |
| "grad_norm": 0.2761548161506653, | |
| "learning_rate": 0.00018886011147676833, | |
| "loss": 0.2953, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.07008983032050571, | |
| "grad_norm": 0.23956750333309174, | |
| "learning_rate": 0.00018876320341612522, | |
| "loss": 0.2249, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.07038556800118295, | |
| "grad_norm": 0.2570217549800873, | |
| "learning_rate": 0.00018866590075281624, | |
| "loss": 0.2153, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.0706813056818602, | |
| "grad_norm": 0.2857867181301117, | |
| "learning_rate": 0.00018856820391940674, | |
| "loss": 0.3025, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.07097704336253743, | |
| "grad_norm": 0.25232836604118347, | |
| "learning_rate": 0.00018847011335021449, | |
| "loss": 0.201, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.07127278104321468, | |
| "grad_norm": 0.28288912773132324, | |
| "learning_rate": 0.00018837162948130752, | |
| "loss": 0.2385, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.0715685187238919, | |
| "grad_norm": 0.3228309154510498, | |
| "learning_rate": 0.00018827275275050233, | |
| "loss": 0.2761, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.07186425640456914, | |
| "grad_norm": 0.2894141972064972, | |
| "learning_rate": 0.00018817348359736203, | |
| "loss": 0.2615, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.07215999408524638, | |
| "grad_norm": 0.3046562373638153, | |
| "learning_rate": 0.00018807382246319412, | |
| "loss": 0.237, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.07245573176592363, | |
| "grad_norm": 0.3270159661769867, | |
| "learning_rate": 0.00018797376979104872, | |
| "loss": 0.209, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.07275146944660087, | |
| "grad_norm": 0.2986398935317993, | |
| "learning_rate": 0.00018787332602571662, | |
| "loss": 0.2032, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.07304720712727811, | |
| "grad_norm": 0.3203541040420532, | |
| "learning_rate": 0.00018777249161372713, | |
| "loss": 0.2154, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.07334294480795535, | |
| "grad_norm": 0.3694257438182831, | |
| "learning_rate": 0.00018767126700334634, | |
| "loss": 0.1959, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.07363868248863259, | |
| "grad_norm": 0.6173413991928101, | |
| "learning_rate": 0.0001875696526445749, | |
| "loss": 0.1682, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.07393442016930982, | |
| "grad_norm": 0.5118496417999268, | |
| "learning_rate": 0.0001874676489891461, | |
| "loss": 0.276, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.07423015784998706, | |
| "grad_norm": 0.15977254509925842, | |
| "learning_rate": 0.00018736525649052394, | |
| "loss": 0.2124, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.0745258955306643, | |
| "grad_norm": 0.15199963748455048, | |
| "learning_rate": 0.00018726247560390099, | |
| "loss": 0.1718, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.07482163321134154, | |
| "grad_norm": 0.16801315546035767, | |
| "learning_rate": 0.00018715930678619644, | |
| "loss": 0.1961, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.07511737089201878, | |
| "grad_norm": 0.17173269391059875, | |
| "learning_rate": 0.00018705575049605413, | |
| "loss": 0.2053, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.07541310857269602, | |
| "grad_norm": 0.19244447350502014, | |
| "learning_rate": 0.00018695180719384029, | |
| "loss": 0.2841, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.07570884625337326, | |
| "grad_norm": 0.1731448471546173, | |
| "learning_rate": 0.00018684747734164177, | |
| "loss": 0.2246, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.07600458393405049, | |
| "grad_norm": 0.1740599274635315, | |
| "learning_rate": 0.00018674276140326376, | |
| "loss": 0.229, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.07630032161472773, | |
| "grad_norm": 0.15664459764957428, | |
| "learning_rate": 0.00018663765984422786, | |
| "loss": 0.2015, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.07659605929540497, | |
| "grad_norm": 0.18093277513980865, | |
| "learning_rate": 0.00018653217313177004, | |
| "loss": 0.2371, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.07689179697608221, | |
| "grad_norm": 0.15789927542209625, | |
| "learning_rate": 0.00018642630173483832, | |
| "loss": 0.19, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.07718753465675945, | |
| "grad_norm": 0.1597701758146286, | |
| "learning_rate": 0.00018632004612409103, | |
| "loss": 0.2096, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.0774832723374367, | |
| "grad_norm": 0.17183595895767212, | |
| "learning_rate": 0.00018621340677189453, | |
| "loss": 0.2333, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.07777901001811394, | |
| "grad_norm": 0.17315109074115753, | |
| "learning_rate": 0.00018610638415232097, | |
| "loss": 0.2215, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.07807474769879118, | |
| "grad_norm": 0.17420059442520142, | |
| "learning_rate": 0.00018599897874114652, | |
| "loss": 0.2216, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.0783704853794684, | |
| "grad_norm": 0.19842545688152313, | |
| "learning_rate": 0.00018589119101584898, | |
| "loss": 0.212, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.07866622306014565, | |
| "grad_norm": 0.18231399357318878, | |
| "learning_rate": 0.00018578302145560584, | |
| "loss": 0.2284, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.07896196074082289, | |
| "grad_norm": 0.1868034154176712, | |
| "learning_rate": 0.00018567447054129195, | |
| "loss": 0.2458, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.07925769842150013, | |
| "grad_norm": 0.19086702167987823, | |
| "learning_rate": 0.00018556553875547754, | |
| "loss": 0.2335, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.07955343610217737, | |
| "grad_norm": 0.2202424705028534, | |
| "learning_rate": 0.00018545622658242607, | |
| "loss": 0.282, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.07984917378285461, | |
| "grad_norm": 0.20102711021900177, | |
| "learning_rate": 0.00018534653450809197, | |
| "loss": 0.2376, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.08014491146353185, | |
| "grad_norm": 0.2291051596403122, | |
| "learning_rate": 0.00018523646302011867, | |
| "loss": 0.2448, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.08044064914420909, | |
| "grad_norm": 0.20590786635875702, | |
| "learning_rate": 0.00018512601260783606, | |
| "loss": 0.2235, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.08073638682488632, | |
| "grad_norm": 0.18361449241638184, | |
| "learning_rate": 0.00018501518376225887, | |
| "loss": 0.1882, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.08103212450556356, | |
| "grad_norm": 0.20607051253318787, | |
| "learning_rate": 0.00018490397697608395, | |
| "loss": 0.216, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.0813278621862408, | |
| "grad_norm": 0.20951667428016663, | |
| "learning_rate": 0.0001847923927436884, | |
| "loss": 0.2283, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.08162359986691804, | |
| "grad_norm": 0.23473979532718658, | |
| "learning_rate": 0.00018468043156112728, | |
| "loss": 0.2582, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.08191933754759528, | |
| "grad_norm": 0.1933351308107376, | |
| "learning_rate": 0.0001845680939261314, | |
| "loss": 0.1938, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.08221507522827252, | |
| "grad_norm": 0.23549608886241913, | |
| "learning_rate": 0.00018445538033810515, | |
| "loss": 0.2166, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.08251081290894977, | |
| "grad_norm": 0.2569220960140228, | |
| "learning_rate": 0.00018434229129812418, | |
| "loss": 0.2697, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.082806550589627, | |
| "grad_norm": 0.22212399542331696, | |
| "learning_rate": 0.0001842288273089332, | |
| "loss": 0.26, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.08310228827030423, | |
| "grad_norm": 0.21168506145477295, | |
| "learning_rate": 0.00018411498887494396, | |
| "loss": 0.2008, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.08339802595098147, | |
| "grad_norm": 0.23749881982803345, | |
| "learning_rate": 0.00018400077650223263, | |
| "loss": 0.2163, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.08369376363165872, | |
| "grad_norm": 0.23951789736747742, | |
| "learning_rate": 0.0001838861906985379, | |
| "loss": 0.2235, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.08398950131233596, | |
| "grad_norm": 0.23467004299163818, | |
| "learning_rate": 0.00018377123197325842, | |
| "loss": 0.2119, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.0842852389930132, | |
| "grad_norm": 0.25748157501220703, | |
| "learning_rate": 0.00018365590083745085, | |
| "loss": 0.2498, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.08458097667369044, | |
| "grad_norm": 0.2640005946159363, | |
| "learning_rate": 0.00018354019780382735, | |
| "loss": 0.2297, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.08487671435436768, | |
| "grad_norm": 0.22059418261051178, | |
| "learning_rate": 0.0001834241233867533, | |
| "loss": 0.1856, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.08517245203504492, | |
| "grad_norm": 0.2318236380815506, | |
| "learning_rate": 0.00018330767810224524, | |
| "loss": 0.186, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.08546818971572215, | |
| "grad_norm": 0.22884918749332428, | |
| "learning_rate": 0.0001831908624679683, | |
| "loss": 0.2019, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.08576392739639939, | |
| "grad_norm": 0.26514625549316406, | |
| "learning_rate": 0.0001830736770032341, | |
| "loss": 0.2338, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.08605966507707663, | |
| "grad_norm": 0.2545234262943268, | |
| "learning_rate": 0.0001829561222289984, | |
| "loss": 0.2051, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.08635540275775387, | |
| "grad_norm": 0.31720927357673645, | |
| "learning_rate": 0.00018283819866785853, | |
| "loss": 0.2127, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.08665114043843111, | |
| "grad_norm": 0.24507926404476166, | |
| "learning_rate": 0.0001827199068440516, | |
| "loss": 0.1481, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.08694687811910835, | |
| "grad_norm": 0.2604232728481293, | |
| "learning_rate": 0.00018260124728345162, | |
| "loss": 0.1804, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.0872426157997856, | |
| "grad_norm": 0.2785932123661041, | |
| "learning_rate": 0.00018248222051356754, | |
| "loss": 0.2188, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.08753835348046284, | |
| "grad_norm": 0.3473379611968994, | |
| "learning_rate": 0.00018236282706354063, | |
| "loss": 0.2223, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.08783409116114006, | |
| "grad_norm": 0.35401713848114014, | |
| "learning_rate": 0.00018224306746414238, | |
| "loss": 0.2602, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.0881298288418173, | |
| "grad_norm": 0.4033154249191284, | |
| "learning_rate": 0.00018212294224777197, | |
| "loss": 0.2324, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.08842556652249454, | |
| "grad_norm": 0.5243091583251953, | |
| "learning_rate": 0.00018200245194845399, | |
| "loss": 0.3447, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.08872130420317179, | |
| "grad_norm": 0.5621427297592163, | |
| "learning_rate": 0.00018188159710183594, | |
| "loss": 0.2627, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08901704188384903, | |
| "grad_norm": 0.1358180195093155, | |
| "learning_rate": 0.000181760378245186, | |
| "loss": 0.1579, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.08931277956452627, | |
| "grad_norm": 0.16979822516441345, | |
| "learning_rate": 0.00018163879591739067, | |
| "loss": 0.2157, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.08960851724520351, | |
| "grad_norm": 0.18489453196525574, | |
| "learning_rate": 0.0001815168506589521, | |
| "loss": 0.2149, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.08990425492588075, | |
| "grad_norm": 0.19990918040275574, | |
| "learning_rate": 0.000181394543011986, | |
| "loss": 0.254, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.09019999260655798, | |
| "grad_norm": 0.17698073387145996, | |
| "learning_rate": 0.00018127187352021907, | |
| "loss": 0.2131, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.09049573028723522, | |
| "grad_norm": 0.1820111721754074, | |
| "learning_rate": 0.0001811488427289866, | |
| "loss": 0.2372, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.09079146796791246, | |
| "grad_norm": 0.1698734015226364, | |
| "learning_rate": 0.00018102545118523007, | |
| "loss": 0.2199, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.0910872056485897, | |
| "grad_norm": 0.17691490054130554, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.2474, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.09138294332926694, | |
| "grad_norm": 0.1919829100370407, | |
| "learning_rate": 0.00018077758803592718, | |
| "loss": 0.2447, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.09167868100994418, | |
| "grad_norm": 0.4307519495487213, | |
| "learning_rate": 0.00018065311753227273, | |
| "loss": 0.2342, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.09197441869062142, | |
| "grad_norm": 0.18838980793952942, | |
| "learning_rate": 0.0001805282884798732, | |
| "loss": 0.1829, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.09227015637129866, | |
| "grad_norm": 0.1960785984992981, | |
| "learning_rate": 0.00018040310143366446, | |
| "loss": 0.2279, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.09256589405197589, | |
| "grad_norm": 0.1780913770198822, | |
| "learning_rate": 0.00018027755695017368, | |
| "loss": 0.222, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.09286163173265313, | |
| "grad_norm": 0.1910991370677948, | |
| "learning_rate": 0.00018015165558751717, | |
| "loss": 0.2378, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.09315736941333037, | |
| "grad_norm": 0.17035876214504242, | |
| "learning_rate": 0.00018002539790539773, | |
| "loss": 0.1834, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.09345310709400761, | |
| "grad_norm": 0.18461062014102936, | |
| "learning_rate": 0.00017989878446510215, | |
| "loss": 0.2255, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.09374884477468486, | |
| "grad_norm": 0.18789273500442505, | |
| "learning_rate": 0.00017977181582949888, | |
| "loss": 0.2236, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.0940445824553621, | |
| "grad_norm": 0.18466150760650635, | |
| "learning_rate": 0.0001796444925630353, | |
| "loss": 0.1987, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.09434032013603934, | |
| "grad_norm": 0.2048148661851883, | |
| "learning_rate": 0.00017951681523173542, | |
| "loss": 0.2059, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.09463605781671658, | |
| "grad_norm": 0.2097977101802826, | |
| "learning_rate": 0.0001793887844031972, | |
| "loss": 0.231, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.0949317954973938, | |
| "grad_norm": 0.20636926591396332, | |
| "learning_rate": 0.00017926040064659014, | |
| "loss": 0.2326, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.09522753317807105, | |
| "grad_norm": 0.2094830870628357, | |
| "learning_rate": 0.0001791316645326526, | |
| "loss": 0.2059, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.09552327085874829, | |
| "grad_norm": 0.22407394647598267, | |
| "learning_rate": 0.00017900257663368963, | |
| "loss": 0.226, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.09581900853942553, | |
| "grad_norm": 0.19603627920150757, | |
| "learning_rate": 0.0001788731375235698, | |
| "loss": 0.2188, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.09611474622010277, | |
| "grad_norm": 0.23884060978889465, | |
| "learning_rate": 0.00017874334777772327, | |
| "loss": 0.2372, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.09641048390078001, | |
| "grad_norm": 0.22886501252651215, | |
| "learning_rate": 0.00017861320797313892, | |
| "loss": 0.2336, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.09670622158145725, | |
| "grad_norm": 0.22485503554344177, | |
| "learning_rate": 0.0001784827186883618, | |
| "loss": 0.2339, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.0970019592621345, | |
| "grad_norm": 0.21763861179351807, | |
| "learning_rate": 0.00017835188050349064, | |
| "loss": 0.2002, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.09729769694281172, | |
| "grad_norm": 0.2242451012134552, | |
| "learning_rate": 0.00017822069400017516, | |
| "loss": 0.2347, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.09759343462348896, | |
| "grad_norm": 0.24562625586986542, | |
| "learning_rate": 0.00017808915976161362, | |
| "loss": 0.2339, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0978891723041662, | |
| "grad_norm": 0.2314700335264206, | |
| "learning_rate": 0.00017795727837255015, | |
| "loss": 0.2708, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.09818490998484344, | |
| "grad_norm": 0.26339125633239746, | |
| "learning_rate": 0.00017782505041927216, | |
| "loss": 0.2557, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.09848064766552068, | |
| "grad_norm": 0.22803835570812225, | |
| "learning_rate": 0.00017769247648960774, | |
| "loss": 0.196, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.09877638534619793, | |
| "grad_norm": 0.22814923524856567, | |
| "learning_rate": 0.00017755955717292296, | |
| "loss": 0.1687, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.09907212302687517, | |
| "grad_norm": 0.25240692496299744, | |
| "learning_rate": 0.00017742629306011944, | |
| "loss": 0.2064, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.09936786070755241, | |
| "grad_norm": 0.23173846304416656, | |
| "learning_rate": 0.00017729268474363154, | |
| "loss": 0.1797, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.09966359838822964, | |
| "grad_norm": 0.2367689609527588, | |
| "learning_rate": 0.0001771587328174239, | |
| "loss": 0.1796, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.09995933606890688, | |
| "grad_norm": 0.2612314522266388, | |
| "learning_rate": 0.0001770244378769885, | |
| "loss": 0.2132, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.10025507374958412, | |
| "grad_norm": 0.2403004765510559, | |
| "learning_rate": 0.0001768898005193425, | |
| "loss": 0.1919, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.10055081143026136, | |
| "grad_norm": 0.26991987228393555, | |
| "learning_rate": 0.000176754821343025, | |
| "loss": 0.2262, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1008465491109386, | |
| "grad_norm": 0.28544318675994873, | |
| "learning_rate": 0.0001766195009480949, | |
| "loss": 0.2229, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.10114228679161584, | |
| "grad_norm": 0.2717716693878174, | |
| "learning_rate": 0.0001764838399361279, | |
| "loss": 0.2146, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.10143802447229308, | |
| "grad_norm": 0.2891225218772888, | |
| "learning_rate": 0.00017634783891021393, | |
| "loss": 0.2462, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.10173376215297031, | |
| "grad_norm": 0.3248477578163147, | |
| "learning_rate": 0.00017621149847495458, | |
| "loss": 0.2475, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.10202949983364755, | |
| "grad_norm": 0.3657064735889435, | |
| "learning_rate": 0.00017607481923646016, | |
| "loss": 0.28, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.10232523751432479, | |
| "grad_norm": 0.3362722396850586, | |
| "learning_rate": 0.0001759378018023473, | |
| "loss": 0.2462, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.10262097519500203, | |
| "grad_norm": 0.3996289074420929, | |
| "learning_rate": 0.00017580044678173592, | |
| "loss": 0.2258, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.10291671287567927, | |
| "grad_norm": 0.36670979857444763, | |
| "learning_rate": 0.00017566275478524693, | |
| "loss": 0.2466, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.10321245055635651, | |
| "grad_norm": 0.48458173871040344, | |
| "learning_rate": 0.0001755247264249991, | |
| "loss": 0.3022, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.10350818823703375, | |
| "grad_norm": 0.49845030903816223, | |
| "learning_rate": 0.0001753863623146066, | |
| "loss": 0.2424, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.103803925917711, | |
| "grad_norm": 0.11847740411758423, | |
| "learning_rate": 0.00017524766306917618, | |
| "loss": 0.1449, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.10409966359838822, | |
| "grad_norm": 0.14606598019599915, | |
| "learning_rate": 0.0001751086293053045, | |
| "loss": 0.1713, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.10439540127906546, | |
| "grad_norm": 0.1666148602962494, | |
| "learning_rate": 0.0001749692616410753, | |
| "loss": 0.2063, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.1046911389597427, | |
| "grad_norm": 0.1671825349330902, | |
| "learning_rate": 0.00017482956069605668, | |
| "loss": 0.2283, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.10498687664041995, | |
| "grad_norm": 0.17825308442115784, | |
| "learning_rate": 0.00017468952709129846, | |
| "loss": 0.2379, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.10528261432109719, | |
| "grad_norm": 0.1678950935602188, | |
| "learning_rate": 0.00017454916144932922, | |
| "loss": 0.2227, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.10557835200177443, | |
| "grad_norm": 0.16522592306137085, | |
| "learning_rate": 0.0001744084643941536, | |
| "loss": 0.2099, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.10587408968245167, | |
| "grad_norm": 0.17215241491794586, | |
| "learning_rate": 0.00017426743655124974, | |
| "loss": 0.2612, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.10616982736312891, | |
| "grad_norm": 0.17087267339229584, | |
| "learning_rate": 0.0001741260785475661, | |
| "loss": 0.2301, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.10646556504380614, | |
| "grad_norm": 0.1661638766527176, | |
| "learning_rate": 0.00017398439101151905, | |
| "loss": 0.2174, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.10676130272448338, | |
| "grad_norm": 0.17075875401496887, | |
| "learning_rate": 0.00017384237457298987, | |
| "loss": 0.1701, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.10705704040516062, | |
| "grad_norm": 0.1797911673784256, | |
| "learning_rate": 0.00017370002986332193, | |
| "loss": 0.2192, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.10735277808583786, | |
| "grad_norm": 0.17848677933216095, | |
| "learning_rate": 0.00017355735751531807, | |
| "loss": 0.2193, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.1076485157665151, | |
| "grad_norm": 0.17944815754890442, | |
| "learning_rate": 0.00017341435816323756, | |
| "loss": 0.2008, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.10794425344719234, | |
| "grad_norm": 0.17322401702404022, | |
| "learning_rate": 0.00017327103244279348, | |
| "loss": 0.2084, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.10823999112786958, | |
| "grad_norm": 0.19021335244178772, | |
| "learning_rate": 0.00017312738099114973, | |
| "loss": 0.2398, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.10853572880854682, | |
| "grad_norm": 0.18541230261325836, | |
| "learning_rate": 0.00017298340444691835, | |
| "loss": 0.2618, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.10883146648922405, | |
| "grad_norm": 0.18366391956806183, | |
| "learning_rate": 0.00017283910345015647, | |
| "loss": 0.2356, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.10912720416990129, | |
| "grad_norm": 0.1941191852092743, | |
| "learning_rate": 0.0001726944786423637, | |
| "loss": 0.2228, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.10942294185057853, | |
| "grad_norm": 0.16953422129154205, | |
| "learning_rate": 0.00017254953066647913, | |
| "loss": 0.1555, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.10971867953125577, | |
| "grad_norm": 0.20326456427574158, | |
| "learning_rate": 0.00017240426016687863, | |
| "loss": 0.2514, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.11001441721193302, | |
| "grad_norm": 0.1932348906993866, | |
| "learning_rate": 0.00017225866778937165, | |
| "loss": 0.2068, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.11031015489261026, | |
| "grad_norm": 0.21190403401851654, | |
| "learning_rate": 0.00017211275418119876, | |
| "loss": 0.2196, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.1106058925732875, | |
| "grad_norm": 0.20619221031665802, | |
| "learning_rate": 0.0001719665199910285, | |
| "loss": 0.2368, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.11090163025396474, | |
| "grad_norm": 0.1878051608800888, | |
| "learning_rate": 0.00017181996586895454, | |
| "loss": 0.1801, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.11090163025396474, | |
| "eval_loss": 0.21186766028404236, | |
| "eval_runtime": 323.3922, | |
| "eval_samples_per_second": 17.61, | |
| "eval_steps_per_second": 8.807, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.11119736793464197, | |
| "grad_norm": 0.22650651633739471, | |
| "learning_rate": 0.00017167309246649297, | |
| "loss": 0.2119, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.11149310561531921, | |
| "grad_norm": 0.21036145091056824, | |
| "learning_rate": 0.0001715259004365791, | |
| "loss": 0.2579, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.11178884329599645, | |
| "grad_norm": 0.2401810735464096, | |
| "learning_rate": 0.00017137839043356484, | |
| "loss": 0.2931, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.11208458097667369, | |
| "grad_norm": 0.22352933883666992, | |
| "learning_rate": 0.00017123056311321562, | |
| "loss": 0.213, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.11238031865735093, | |
| "grad_norm": 0.24567389488220215, | |
| "learning_rate": 0.0001710824191327075, | |
| "loss": 0.2738, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.11267605633802817, | |
| "grad_norm": 0.24588455259799957, | |
| "learning_rate": 0.00017093395915062428, | |
| "loss": 0.2252, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.11297179401870541, | |
| "grad_norm": 0.21331793069839478, | |
| "learning_rate": 0.00017078518382695465, | |
| "loss": 0.165, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.11326753169938265, | |
| "grad_norm": 0.2419130951166153, | |
| "learning_rate": 0.00017063609382308908, | |
| "loss": 0.2355, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.11356326938005988, | |
| "grad_norm": 0.25924235582351685, | |
| "learning_rate": 0.00017048668980181698, | |
| "loss": 0.233, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.11385900706073712, | |
| "grad_norm": 0.26302260160446167, | |
| "learning_rate": 0.00017033697242732377, | |
| "loss": 0.2379, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.11415474474141436, | |
| "grad_norm": 0.24101543426513672, | |
| "learning_rate": 0.0001701869423651879, | |
| "loss": 0.1956, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.1144504824220916, | |
| "grad_norm": 0.26949170231819153, | |
| "learning_rate": 0.00017003660028237793, | |
| "loss": 0.2365, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.11474622010276884, | |
| "grad_norm": 0.22656166553497314, | |
| "learning_rate": 0.00016988594684724947, | |
| "loss": 0.1488, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.11504195778344609, | |
| "grad_norm": 0.25537586212158203, | |
| "learning_rate": 0.00016973498272954222, | |
| "loss": 0.2277, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.11533769546412333, | |
| "grad_norm": 0.2855624556541443, | |
| "learning_rate": 0.00016958370860037717, | |
| "loss": 0.2155, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.11563343314480057, | |
| "grad_norm": 0.2733490467071533, | |
| "learning_rate": 0.00016943212513225345, | |
| "loss": 0.2189, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.1159291708254778, | |
| "grad_norm": 0.2860839366912842, | |
| "learning_rate": 0.00016928023299904533, | |
| "loss": 0.276, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.11622490850615504, | |
| "grad_norm": 0.29845502972602844, | |
| "learning_rate": 0.0001691280328759992, | |
| "loss": 0.2542, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.11652064618683228, | |
| "grad_norm": 0.3229489326477051, | |
| "learning_rate": 0.00016897552543973084, | |
| "loss": 0.239, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.11681638386750952, | |
| "grad_norm": 0.31158486008644104, | |
| "learning_rate": 0.00016882271136822206, | |
| "loss": 0.2078, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.11711212154818676, | |
| "grad_norm": 0.33740320801734924, | |
| "learning_rate": 0.0001686695913408179, | |
| "loss": 0.2092, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.117407859228864, | |
| "grad_norm": 0.42870861291885376, | |
| "learning_rate": 0.0001685161660382235, | |
| "loss": 0.3145, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.11770359690954124, | |
| "grad_norm": 0.41335102915763855, | |
| "learning_rate": 0.00016836243614250113, | |
| "loss": 0.2852, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.11799933459021848, | |
| "grad_norm": 0.3501723110675812, | |
| "learning_rate": 0.00016820840233706719, | |
| "loss": 0.2022, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.11829507227089571, | |
| "grad_norm": 0.511377215385437, | |
| "learning_rate": 0.0001680540653066891, | |
| "loss": 0.3176, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.11859080995157295, | |
| "grad_norm": 0.13668859004974365, | |
| "learning_rate": 0.00016789942573748232, | |
| "loss": 0.1905, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.11888654763225019, | |
| "grad_norm": 0.15192629396915436, | |
| "learning_rate": 0.0001677444843169072, | |
| "loss": 0.2082, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.11918228531292743, | |
| "grad_norm": 0.16673530638217926, | |
| "learning_rate": 0.00016758924173376603, | |
| "loss": 0.2217, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.11947802299360467, | |
| "grad_norm": 0.14643211662769318, | |
| "learning_rate": 0.0001674336986781999, | |
| "loss": 0.1797, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.11977376067428191, | |
| "grad_norm": 0.1829608678817749, | |
| "learning_rate": 0.00016727785584168581, | |
| "loss": 0.2738, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.12006949835495916, | |
| "grad_norm": 0.1723003387451172, | |
| "learning_rate": 0.0001671217139170333, | |
| "loss": 0.2193, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.1203652360356364, | |
| "grad_norm": 0.175453320145607, | |
| "learning_rate": 0.00016696527359838154, | |
| "loss": 0.2252, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.12066097371631362, | |
| "grad_norm": 0.18448546528816223, | |
| "learning_rate": 0.00016680853558119632, | |
| "loss": 0.2511, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.12095671139699087, | |
| "grad_norm": 0.18752612173557281, | |
| "learning_rate": 0.0001666515005622668, | |
| "loss": 0.2624, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.1212524490776681, | |
| "grad_norm": 0.15819008648395538, | |
| "learning_rate": 0.0001664941692397025, | |
| "loss": 0.1884, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.12154818675834535, | |
| "grad_norm": 0.17365534603595734, | |
| "learning_rate": 0.00016633654231293013, | |
| "loss": 0.2065, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.12184392443902259, | |
| "grad_norm": 0.1719895303249359, | |
| "learning_rate": 0.00016617862048269065, | |
| "loss": 0.2239, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.12213966211969983, | |
| "grad_norm": 0.1667676866054535, | |
| "learning_rate": 0.00016602040445103588, | |
| "loss": 0.1775, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.12243539980037707, | |
| "grad_norm": 0.17780280113220215, | |
| "learning_rate": 0.00016586189492132566, | |
| "loss": 0.2271, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.12273113748105431, | |
| "grad_norm": 0.16043958067893982, | |
| "learning_rate": 0.00016570309259822453, | |
| "loss": 0.19, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.12302687516173154, | |
| "grad_norm": 0.2024742215871811, | |
| "learning_rate": 0.0001655439981876987, | |
| "loss": 0.207, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.12332261284240878, | |
| "grad_norm": 0.1686304360628128, | |
| "learning_rate": 0.00016538461239701277, | |
| "loss": 0.1874, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.12361835052308602, | |
| "grad_norm": 0.19449354708194733, | |
| "learning_rate": 0.00016522493593472683, | |
| "loss": 0.2119, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.12391408820376326, | |
| "grad_norm": 0.16912192106246948, | |
| "learning_rate": 0.0001650649695106931, | |
| "loss": 0.1543, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.1242098258844405, | |
| "grad_norm": 0.20457497239112854, | |
| "learning_rate": 0.00016490471383605288, | |
| "loss": 0.206, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.12450556356511774, | |
| "grad_norm": 0.16567295789718628, | |
| "learning_rate": 0.00016474416962323325, | |
| "loss": 0.1518, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.12480130124579498, | |
| "grad_norm": 0.19134259223937988, | |
| "learning_rate": 0.00016458333758594414, | |
| "loss": 0.207, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.12509703892647223, | |
| "grad_norm": 0.22228240966796875, | |
| "learning_rate": 0.00016442221843917496, | |
| "loss": 0.2208, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.12539277660714945, | |
| "grad_norm": 0.23446987569332123, | |
| "learning_rate": 0.00016426081289919143, | |
| "loss": 0.2229, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.1256885142878267, | |
| "grad_norm": 0.20911334455013275, | |
| "learning_rate": 0.0001640991216835326, | |
| "loss": 0.1946, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.12598425196850394, | |
| "grad_norm": 0.21853110194206238, | |
| "learning_rate": 0.00016393714551100734, | |
| "loss": 0.2157, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.12627998964918116, | |
| "grad_norm": 0.23242445290088654, | |
| "learning_rate": 0.0001637748851016914, | |
| "loss": 0.2161, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.12657572732985842, | |
| "grad_norm": 0.24394045770168304, | |
| "learning_rate": 0.00016361234117692413, | |
| "loss": 0.232, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.12687146501053564, | |
| "grad_norm": 0.21284179389476776, | |
| "learning_rate": 0.00016344951445930526, | |
| "loss": 0.162, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.1271672026912129, | |
| "grad_norm": 0.24403221905231476, | |
| "learning_rate": 0.0001632864056726917, | |
| "loss": 0.1994, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.12746294037189013, | |
| "grad_norm": 0.2400125414133072, | |
| "learning_rate": 0.00016312301554219426, | |
| "loss": 0.2004, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.12775867805256738, | |
| "grad_norm": 0.2092212736606598, | |
| "learning_rate": 0.00016295934479417453, | |
| "loss": 0.1618, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.1280544157332446, | |
| "grad_norm": 0.21888083219528198, | |
| "learning_rate": 0.00016279539415624164, | |
| "loss": 0.181, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.12835015341392186, | |
| "grad_norm": 0.23019088804721832, | |
| "learning_rate": 0.0001626311643572489, | |
| "loss": 0.1816, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.1286458910945991, | |
| "grad_norm": 0.23748983442783356, | |
| "learning_rate": 0.00016246665612729074, | |
| "loss": 0.1869, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.12894162877527632, | |
| "grad_norm": 0.23818852007389069, | |
| "learning_rate": 0.00016230187019769928, | |
| "loss": 0.1959, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.12923736645595357, | |
| "grad_norm": 0.27259862422943115, | |
| "learning_rate": 0.00016213680730104124, | |
| "loss": 0.2183, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.1295331041366308, | |
| "grad_norm": 0.2296111136674881, | |
| "learning_rate": 0.0001619714681711146, | |
| "loss": 0.1615, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.12982884181730805, | |
| "grad_norm": 0.2418239265680313, | |
| "learning_rate": 0.00016180585354294536, | |
| "loss": 0.2222, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.13012457949798528, | |
| "grad_norm": 0.2754133641719818, | |
| "learning_rate": 0.00016163996415278424, | |
| "loss": 0.2433, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.13042031717866254, | |
| "grad_norm": 0.31593406200408936, | |
| "learning_rate": 0.00016147380073810346, | |
| "loss": 0.2492, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.13071605485933976, | |
| "grad_norm": 0.30033108592033386, | |
| "learning_rate": 0.0001613073640375934, | |
| "loss": 0.2539, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.131011792540017, | |
| "grad_norm": 0.2534523606300354, | |
| "learning_rate": 0.00016114065479115946, | |
| "loss": 0.1382, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.13130753022069425, | |
| "grad_norm": 0.31533321738243103, | |
| "learning_rate": 0.00016097367373991842, | |
| "loss": 0.2456, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.13160326790137147, | |
| "grad_norm": 0.3007992208003998, | |
| "learning_rate": 0.00016080642162619565, | |
| "loss": 0.2018, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.13189900558204873, | |
| "grad_norm": 0.3328956961631775, | |
| "learning_rate": 0.0001606388991935214, | |
| "loss": 0.2181, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.13219474326272596, | |
| "grad_norm": 0.34528401494026184, | |
| "learning_rate": 0.0001604711071866277, | |
| "loss": 0.2354, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.1324904809434032, | |
| "grad_norm": 0.3941013216972351, | |
| "learning_rate": 0.00016030304635144494, | |
| "loss": 0.2796, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.13278621862408044, | |
| "grad_norm": 0.36703160405158997, | |
| "learning_rate": 0.00016013471743509862, | |
| "loss": 0.2258, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.1330819563047577, | |
| "grad_norm": 0.44615501165390015, | |
| "learning_rate": 0.00015996612118590603, | |
| "loss": 0.1753, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.13337769398543492, | |
| "grad_norm": 0.09873225539922714, | |
| "learning_rate": 0.00015979725835337294, | |
| "loss": 0.1385, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.13367343166611215, | |
| "grad_norm": 0.13112150132656097, | |
| "learning_rate": 0.00015962812968819016, | |
| "loss": 0.2112, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.1339691693467894, | |
| "grad_norm": 0.1437048316001892, | |
| "learning_rate": 0.0001594587359422303, | |
| "loss": 0.1976, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.13426490702746663, | |
| "grad_norm": 0.14785504341125488, | |
| "learning_rate": 0.0001592890778685444, | |
| "loss": 0.1953, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.13456064470814388, | |
| "grad_norm": 0.15337800979614258, | |
| "learning_rate": 0.00015911915622135862, | |
| "loss": 0.1938, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.1348563823888211, | |
| "grad_norm": 0.16650818288326263, | |
| "learning_rate": 0.00015894897175607086, | |
| "loss": 0.1915, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.13515212006949837, | |
| "grad_norm": 0.15892527997493744, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.1972, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.1354478577501756, | |
| "grad_norm": 0.16916155815124512, | |
| "learning_rate": 0.00015860781739861928, | |
| "loss": 0.2193, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.13574359543085282, | |
| "grad_norm": 0.1601080596446991, | |
| "learning_rate": 0.00015843684902307962, | |
| "loss": 0.1921, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.13603933311153008, | |
| "grad_norm": 0.16756726801395416, | |
| "learning_rate": 0.00015826562086267956, | |
| "loss": 0.241, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.1363350707922073, | |
| "grad_norm": 0.16822628676891327, | |
| "learning_rate": 0.00015809413367862512, | |
| "loss": 0.2028, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.13663080847288456, | |
| "grad_norm": 0.15600813925266266, | |
| "learning_rate": 0.00015792238823327388, | |
| "loss": 0.1639, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.13692654615356178, | |
| "grad_norm": 0.16066166758537292, | |
| "learning_rate": 0.00015775038529013152, | |
| "loss": 0.197, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.13722228383423904, | |
| "grad_norm": 0.181250661611557, | |
| "learning_rate": 0.0001575781256138485, | |
| "loss": 0.2464, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.13751802151491627, | |
| "grad_norm": 0.19740843772888184, | |
| "learning_rate": 0.00015740560997021648, | |
| "loss": 0.2688, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.13781375919559352, | |
| "grad_norm": 0.178134024143219, | |
| "learning_rate": 0.00015723283912616513, | |
| "loss": 0.2315, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.13810949687627075, | |
| "grad_norm": 0.16021274030208588, | |
| "learning_rate": 0.00015705981384975866, | |
| "loss": 0.1819, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.13840523455694798, | |
| "grad_norm": 0.20027248561382294, | |
| "learning_rate": 0.0001568865349101923, | |
| "loss": 0.2789, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.13870097223762523, | |
| "grad_norm": 0.1927415281534195, | |
| "learning_rate": 0.00015671300307778898, | |
| "loss": 0.2, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.13899670991830246, | |
| "grad_norm": 0.1823057234287262, | |
| "learning_rate": 0.00015653921912399589, | |
| "loss": 0.1549, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.1392924475989797, | |
| "grad_norm": 0.19187289476394653, | |
| "learning_rate": 0.00015636518382138107, | |
| "loss": 0.1975, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.13958818527965694, | |
| "grad_norm": 0.21980108320713043, | |
| "learning_rate": 0.0001561908979436299, | |
| "loss": 0.2242, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.1398839229603342, | |
| "grad_norm": 0.21112766861915588, | |
| "learning_rate": 0.00015601636226554168, | |
| "loss": 0.2254, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.14017966064101142, | |
| "grad_norm": 0.1739622801542282, | |
| "learning_rate": 0.00015584157756302634, | |
| "loss": 0.1647, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.14047539832168865, | |
| "grad_norm": 0.2054637223482132, | |
| "learning_rate": 0.0001556665446131007, | |
| "loss": 0.1974, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.1407711360023659, | |
| "grad_norm": 0.2365804761648178, | |
| "learning_rate": 0.00015549126419388536, | |
| "loss": 0.2244, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.14106687368304313, | |
| "grad_norm": 0.20265153050422668, | |
| "learning_rate": 0.0001553157370846009, | |
| "loss": 0.1727, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.1413626113637204, | |
| "grad_norm": 0.1899023801088333, | |
| "learning_rate": 0.00015513996406556465, | |
| "loss": 0.1573, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.1416583490443976, | |
| "grad_norm": 0.25980237126350403, | |
| "learning_rate": 0.00015496394591818716, | |
| "loss": 0.2644, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.14195408672507487, | |
| "grad_norm": 0.20405927300453186, | |
| "learning_rate": 0.0001547876834249687, | |
| "loss": 0.1915, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1422498244057521, | |
| "grad_norm": 0.22452029585838318, | |
| "learning_rate": 0.00015461117736949577, | |
| "loss": 0.2066, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.14254556208642935, | |
| "grad_norm": 0.25543588399887085, | |
| "learning_rate": 0.00015443442853643762, | |
| "loss": 0.2462, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.14284129976710658, | |
| "grad_norm": 0.26937565207481384, | |
| "learning_rate": 0.00015425743771154294, | |
| "loss": 0.2779, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.1431370374477838, | |
| "grad_norm": 0.24261003732681274, | |
| "learning_rate": 0.00015408020568163602, | |
| "loss": 0.1922, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.14343277512846106, | |
| "grad_norm": 0.23764987289905548, | |
| "learning_rate": 0.00015390273323461352, | |
| "loss": 0.1766, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.1437285128091383, | |
| "grad_norm": 0.26802945137023926, | |
| "learning_rate": 0.0001537250211594409, | |
| "loss": 0.2005, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.14402425048981554, | |
| "grad_norm": 0.2354622781276703, | |
| "learning_rate": 0.0001535470702461489, | |
| "loss": 0.1849, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.14431998817049277, | |
| "grad_norm": 0.2462967336177826, | |
| "learning_rate": 0.00015336888128583, | |
| "loss": 0.1995, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.14461572585117002, | |
| "grad_norm": 0.2529776394367218, | |
| "learning_rate": 0.000153190455070635, | |
| "loss": 0.1994, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.14491146353184725, | |
| "grad_norm": 0.27993232011795044, | |
| "learning_rate": 0.00015301179239376938, | |
| "loss": 0.2212, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.14520720121252448, | |
| "grad_norm": 0.2503523528575897, | |
| "learning_rate": 0.00015283289404948976, | |
| "loss": 0.1711, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.14550293889320173, | |
| "grad_norm": 0.2650829255580902, | |
| "learning_rate": 0.0001526537608331006, | |
| "loss": 0.1991, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.14579867657387896, | |
| "grad_norm": 0.2573229670524597, | |
| "learning_rate": 0.00015247439354095041, | |
| "loss": 0.1629, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.14609441425455622, | |
| "grad_norm": 0.2993515431880951, | |
| "learning_rate": 0.00015229479297042823, | |
| "loss": 0.2513, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.14639015193523344, | |
| "grad_norm": 0.2918243706226349, | |
| "learning_rate": 0.00015211495991996027, | |
| "loss": 0.1799, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.1466858896159107, | |
| "grad_norm": 0.32243257761001587, | |
| "learning_rate": 0.0001519348951890062, | |
| "loss": 0.2384, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.14698162729658792, | |
| "grad_norm": 0.32650378346443176, | |
| "learning_rate": 0.0001517545995780556, | |
| "loss": 0.2419, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.14727736497726518, | |
| "grad_norm": 0.28698667883872986, | |
| "learning_rate": 0.00015157407388862452, | |
| "loss": 0.1575, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.1475731026579424, | |
| "grad_norm": 0.5042946934700012, | |
| "learning_rate": 0.00015139331892325179, | |
| "loss": 0.3989, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.14786884033861963, | |
| "grad_norm": 0.6159984469413757, | |
| "learning_rate": 0.0001512123354854955, | |
| "loss": 0.3383, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1481645780192969, | |
| "grad_norm": 0.12425012141466141, | |
| "learning_rate": 0.0001510311243799295, | |
| "loss": 0.1775, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.14846031569997412, | |
| "grad_norm": 0.12108102440834045, | |
| "learning_rate": 0.00015084968641213958, | |
| "loss": 0.1404, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.14875605338065137, | |
| "grad_norm": 0.12233590334653854, | |
| "learning_rate": 0.00015066802238872023, | |
| "loss": 0.1668, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.1490517910613286, | |
| "grad_norm": 0.1354806274175644, | |
| "learning_rate": 0.0001504861331172709, | |
| "loss": 0.1827, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.14934752874200585, | |
| "grad_norm": 0.1641916036605835, | |
| "learning_rate": 0.0001503040194063922, | |
| "loss": 0.1819, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.14964326642268308, | |
| "grad_norm": 0.1794111579656601, | |
| "learning_rate": 0.00015012168206568268, | |
| "loss": 0.2923, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.1499390041033603, | |
| "grad_norm": 0.1664586067199707, | |
| "learning_rate": 0.00014993912190573505, | |
| "loss": 0.222, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.15023474178403756, | |
| "grad_norm": 0.16165000200271606, | |
| "learning_rate": 0.00014975633973813242, | |
| "loss": 0.2264, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.1505304794647148, | |
| "grad_norm": 0.17169785499572754, | |
| "learning_rate": 0.00014957333637544503, | |
| "loss": 0.2151, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.15082621714539204, | |
| "grad_norm": 0.1526673138141632, | |
| "learning_rate": 0.00014939011263122634, | |
| "loss": 0.1745, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.15112195482606927, | |
| "grad_norm": 0.15362168848514557, | |
| "learning_rate": 0.0001492066693200096, | |
| "loss": 0.1649, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.15141769250674653, | |
| "grad_norm": 0.16641461849212646, | |
| "learning_rate": 0.00014902300725730413, | |
| "loss": 0.2128, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.15171343018742375, | |
| "grad_norm": 0.15893928706645966, | |
| "learning_rate": 0.00014883912725959167, | |
| "loss": 0.1949, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.15200916786810098, | |
| "grad_norm": 0.17531882226467133, | |
| "learning_rate": 0.00014865503014432292, | |
| "loss": 0.2354, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.15230490554877824, | |
| "grad_norm": 0.20174053311347961, | |
| "learning_rate": 0.00014847071672991367, | |
| "loss": 0.2515, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.15260064322945546, | |
| "grad_norm": 0.17568667232990265, | |
| "learning_rate": 0.0001482861878357414, | |
| "loss": 0.2216, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.15289638091013272, | |
| "grad_norm": 0.17869949340820312, | |
| "learning_rate": 0.00014810144428214144, | |
| "loss": 0.2044, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.15319211859080994, | |
| "grad_norm": 0.18138998746871948, | |
| "learning_rate": 0.0001479164868904034, | |
| "loss": 0.1876, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.1534878562714872, | |
| "grad_norm": 0.17516851425170898, | |
| "learning_rate": 0.00014773131648276758, | |
| "loss": 0.1834, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.15378359395216443, | |
| "grad_norm": 0.1887725293636322, | |
| "learning_rate": 0.00014754593388242117, | |
| "loss": 0.1885, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.15407933163284168, | |
| "grad_norm": 0.18247954547405243, | |
| "learning_rate": 0.0001473603399134948, | |
| "loss": 0.1621, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.1543750693135189, | |
| "grad_norm": 0.1879930943250656, | |
| "learning_rate": 0.0001471745354010586, | |
| "loss": 0.2155, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.15467080699419614, | |
| "grad_norm": 0.16984637081623077, | |
| "learning_rate": 0.00014698852117111884, | |
| "loss": 0.1724, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.1549665446748734, | |
| "grad_norm": 0.1831916868686676, | |
| "learning_rate": 0.000146802298050614, | |
| "loss": 0.157, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.15526228235555062, | |
| "grad_norm": 0.20019155740737915, | |
| "learning_rate": 0.0001466158668674112, | |
| "loss": 0.1544, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.15555802003622787, | |
| "grad_norm": 0.1930214762687683, | |
| "learning_rate": 0.00014642922845030257, | |
| "loss": 0.1891, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.1558537577169051, | |
| "grad_norm": 0.26559117436408997, | |
| "learning_rate": 0.0001462423836290015, | |
| "loss": 0.289, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.15614949539758236, | |
| "grad_norm": 0.2079695165157318, | |
| "learning_rate": 0.00014605533323413887, | |
| "loss": 0.1474, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.15644523307825958, | |
| "grad_norm": 0.23225319385528564, | |
| "learning_rate": 0.00014586807809725962, | |
| "loss": 0.177, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.1567409707589368, | |
| "grad_norm": 0.2178119421005249, | |
| "learning_rate": 0.00014568061905081875, | |
| "loss": 0.2066, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.15703670843961406, | |
| "grad_norm": 0.2345496565103531, | |
| "learning_rate": 0.00014549295692817778, | |
| "loss": 0.2144, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.1573324461202913, | |
| "grad_norm": 0.24489574134349823, | |
| "learning_rate": 0.00014530509256360102, | |
| "loss": 0.2046, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.15762818380096855, | |
| "grad_norm": 0.23358239233493805, | |
| "learning_rate": 0.00014511702679225193, | |
| "loss": 0.2251, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.15792392148164577, | |
| "grad_norm": 0.26981320977211, | |
| "learning_rate": 0.0001449287604501893, | |
| "loss": 0.2222, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.15821965916232303, | |
| "grad_norm": 0.24023444950580597, | |
| "learning_rate": 0.00014474029437436348, | |
| "loss": 0.2155, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.15851539684300026, | |
| "grad_norm": 0.19078750908374786, | |
| "learning_rate": 0.00014455162940261285, | |
| "loss": 0.1276, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.1588111345236775, | |
| "grad_norm": 0.22319287061691284, | |
| "learning_rate": 0.0001443627663736599, | |
| "loss": 0.2082, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.15910687220435474, | |
| "grad_norm": 0.24529913067817688, | |
| "learning_rate": 0.00014417370612710778, | |
| "loss": 0.177, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.15940260988503197, | |
| "grad_norm": 0.26183465123176575, | |
| "learning_rate": 0.00014398444950343623, | |
| "loss": 0.1725, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.15969834756570922, | |
| "grad_norm": 0.2594473361968994, | |
| "learning_rate": 0.00014379499734399798, | |
| "loss": 0.2477, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.15999408524638645, | |
| "grad_norm": 0.2427864521741867, | |
| "learning_rate": 0.0001436053504910151, | |
| "loss": 0.1737, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.1602898229270637, | |
| "grad_norm": 0.30450040102005005, | |
| "learning_rate": 0.0001434155097875752, | |
| "loss": 0.2318, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.16058556060774093, | |
| "grad_norm": 0.3049463927745819, | |
| "learning_rate": 0.00014322547607762762, | |
| "loss": 0.2495, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.16088129828841818, | |
| "grad_norm": 0.27216774225234985, | |
| "learning_rate": 0.0001430352502059797, | |
| "loss": 0.1733, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.1611770359690954, | |
| "grad_norm": 0.3504975140094757, | |
| "learning_rate": 0.0001428448330182931, | |
| "loss": 0.209, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.16147277364977264, | |
| "grad_norm": 0.3599260151386261, | |
| "learning_rate": 0.00014265422536107993, | |
| "loss": 0.2185, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.1617685113304499, | |
| "grad_norm": 0.31020569801330566, | |
| "learning_rate": 0.00014246342808169914, | |
| "loss": 0.2292, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.16206424901112712, | |
| "grad_norm": 0.3614603281021118, | |
| "learning_rate": 0.00014227244202835257, | |
| "loss": 0.1979, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.16235998669180438, | |
| "grad_norm": 0.44161731004714966, | |
| "learning_rate": 0.0001420812680500813, | |
| "loss": 0.1933, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.1626557243724816, | |
| "grad_norm": 0.6861674189567566, | |
| "learning_rate": 0.00014188990699676184, | |
| "loss": 0.3917, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.16295146205315886, | |
| "grad_norm": 0.10810007899999619, | |
| "learning_rate": 0.00014169835971910238, | |
| "loss": 0.1231, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.16324719973383608, | |
| "grad_norm": 0.1477546989917755, | |
| "learning_rate": 0.0001415066270686389, | |
| "loss": 0.1923, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.16354293741451334, | |
| "grad_norm": 0.14868442714214325, | |
| "learning_rate": 0.00014131470989773158, | |
| "loss": 0.1956, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.16383867509519057, | |
| "grad_norm": 0.1274333894252777, | |
| "learning_rate": 0.0001411226090595608, | |
| "loss": 0.1461, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.1641344127758678, | |
| "grad_norm": 0.1790267527103424, | |
| "learning_rate": 0.00014093032540812348, | |
| "loss": 0.2404, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.16443015045654505, | |
| "grad_norm": 0.1788930892944336, | |
| "learning_rate": 0.0001407378597982293, | |
| "loss": 0.2194, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.16472588813722228, | |
| "grad_norm": 0.20187751948833466, | |
| "learning_rate": 0.00014054521308549673, | |
| "loss": 0.2107, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.16502162581789953, | |
| "grad_norm": 0.16323895752429962, | |
| "learning_rate": 0.0001403523861263495, | |
| "loss": 0.1895, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.16531736349857676, | |
| "grad_norm": 0.14459684491157532, | |
| "learning_rate": 0.00014015937977801256, | |
| "loss": 0.1856, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.165613101179254, | |
| "grad_norm": 0.16795897483825684, | |
| "learning_rate": 0.00013996619489850822, | |
| "loss": 0.2049, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.16590883885993124, | |
| "grad_norm": 0.16099309921264648, | |
| "learning_rate": 0.00013977283234665273, | |
| "loss": 0.2114, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.16620457654060847, | |
| "grad_norm": 0.1711655557155609, | |
| "learning_rate": 0.00013957929298205195, | |
| "loss": 0.1889, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.16650031422128572, | |
| "grad_norm": 0.20089977979660034, | |
| "learning_rate": 0.00013938557766509792, | |
| "loss": 0.2932, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.16679605190196295, | |
| "grad_norm": 0.18003849685192108, | |
| "learning_rate": 0.0001391916872569648, | |
| "loss": 0.1676, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.1670917895826402, | |
| "grad_norm": 0.19262616336345673, | |
| "learning_rate": 0.00013899762261960518, | |
| "loss": 0.1983, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.16738752726331743, | |
| "grad_norm": 0.17830273509025574, | |
| "learning_rate": 0.0001388033846157462, | |
| "loss": 0.1977, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.1676832649439947, | |
| "grad_norm": 0.18847207725048065, | |
| "learning_rate": 0.0001386089741088857, | |
| "loss": 0.2294, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.1679790026246719, | |
| "grad_norm": 0.18042877316474915, | |
| "learning_rate": 0.00013841439196328836, | |
| "loss": 0.1961, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.16827474030534917, | |
| "grad_norm": 0.18396486341953278, | |
| "learning_rate": 0.00013821963904398193, | |
| "loss": 0.1872, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.1685704779860264, | |
| "grad_norm": 0.2012127786874771, | |
| "learning_rate": 0.00013802471621675338, | |
| "loss": 0.194, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.16886621566670362, | |
| "grad_norm": 0.18555906414985657, | |
| "learning_rate": 0.00013782962434814492, | |
| "loss": 0.1971, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.16916195334738088, | |
| "grad_norm": 0.1829305738210678, | |
| "learning_rate": 0.00013763436430545034, | |
| "loss": 0.1687, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.1694576910280581, | |
| "grad_norm": 0.18491117656230927, | |
| "learning_rate": 0.00013743893695671096, | |
| "loss": 0.1691, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.16975342870873536, | |
| "grad_norm": 0.1929929107427597, | |
| "learning_rate": 0.00013724334317071198, | |
| "loss": 0.2112, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.1700491663894126, | |
| "grad_norm": 0.21051141619682312, | |
| "learning_rate": 0.00013704758381697844, | |
| "loss": 0.1932, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.17034490407008984, | |
| "grad_norm": 0.23053379356861115, | |
| "learning_rate": 0.00013685165976577146, | |
| "loss": 0.2552, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.17064064175076707, | |
| "grad_norm": 0.2063068449497223, | |
| "learning_rate": 0.0001366555718880843, | |
| "loss": 0.1479, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.1709363794314443, | |
| "grad_norm": 0.21455182135105133, | |
| "learning_rate": 0.00013645932105563844, | |
| "loss": 0.204, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.17123211711212155, | |
| "grad_norm": 0.23406703770160675, | |
| "learning_rate": 0.00013626290814088005, | |
| "loss": 0.2021, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.17152785479279878, | |
| "grad_norm": 0.2362387478351593, | |
| "learning_rate": 0.00013606633401697557, | |
| "loss": 0.2392, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.17182359247347603, | |
| "grad_norm": 0.19675102829933167, | |
| "learning_rate": 0.00013586959955780824, | |
| "loss": 0.1491, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.17211933015415326, | |
| "grad_norm": 0.25349652767181396, | |
| "learning_rate": 0.00013567270563797398, | |
| "loss": 0.2077, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.17241506783483052, | |
| "grad_norm": 0.24108415842056274, | |
| "learning_rate": 0.00013547565313277776, | |
| "loss": 0.2037, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.17271080551550774, | |
| "grad_norm": 0.2342023402452469, | |
| "learning_rate": 0.00013527844291822948, | |
| "loss": 0.1752, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.173006543196185, | |
| "grad_norm": 0.21363839507102966, | |
| "learning_rate": 0.0001350810758710401, | |
| "loss": 0.1707, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.17330228087686222, | |
| "grad_norm": 0.25588366389274597, | |
| "learning_rate": 0.00013488355286861783, | |
| "loss": 0.2348, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.17359801855753945, | |
| "grad_norm": 0.26161813735961914, | |
| "learning_rate": 0.0001346858747890642, | |
| "loss": 0.2326, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.1738937562382167, | |
| "grad_norm": 0.26976850628852844, | |
| "learning_rate": 0.00013448804251117003, | |
| "loss": 0.2392, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.17418949391889393, | |
| "grad_norm": 0.25815528631210327, | |
| "learning_rate": 0.0001342900569144119, | |
| "loss": 0.2083, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.1744852315995712, | |
| "grad_norm": 0.2357477843761444, | |
| "learning_rate": 0.0001340919188789477, | |
| "loss": 0.1978, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.17478096928024842, | |
| "grad_norm": 0.23640482127666473, | |
| "learning_rate": 0.00013389362928561317, | |
| "loss": 0.1737, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.17507670696092567, | |
| "grad_norm": 0.2584919035434723, | |
| "learning_rate": 0.00013369518901591772, | |
| "loss": 0.18, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.1753724446416029, | |
| "grad_norm": 0.2640099823474884, | |
| "learning_rate": 0.00013349659895204067, | |
| "loss": 0.1936, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.17566818232228013, | |
| "grad_norm": 0.30272793769836426, | |
| "learning_rate": 0.0001332978599768272, | |
| "loss": 0.2426, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.17596392000295738, | |
| "grad_norm": 0.29123789072036743, | |
| "learning_rate": 0.00013309897297378455, | |
| "loss": 0.1799, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.1762596576836346, | |
| "grad_norm": 0.31216293573379517, | |
| "learning_rate": 0.00013289993882707797, | |
| "loss": 0.2074, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.17655539536431186, | |
| "grad_norm": 0.32396575808525085, | |
| "learning_rate": 0.00013270075842152678, | |
| "loss": 0.2328, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.1768511330449891, | |
| "grad_norm": 0.3656167984008789, | |
| "learning_rate": 0.00013250143264260074, | |
| "loss": 0.2458, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.17714687072566634, | |
| "grad_norm": 0.4183952212333679, | |
| "learning_rate": 0.0001323019623764156, | |
| "loss": 0.2968, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.17744260840634357, | |
| "grad_norm": 0.563667356967926, | |
| "learning_rate": 0.00013210234850972964, | |
| "loss": 0.3082, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1777383460870208, | |
| "grad_norm": 0.11673900485038757, | |
| "learning_rate": 0.0001319025919299394, | |
| "loss": 0.1225, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.17803408376769805, | |
| "grad_norm": 0.15502867102622986, | |
| "learning_rate": 0.00013170269352507597, | |
| "loss": 0.1869, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.17832982144837528, | |
| "grad_norm": 0.1288733333349228, | |
| "learning_rate": 0.0001315026541838008, | |
| "loss": 0.1496, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.17862555912905254, | |
| "grad_norm": 0.16827955842018127, | |
| "learning_rate": 0.00013130247479540202, | |
| "loss": 0.2337, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.17892129680972976, | |
| "grad_norm": 0.1500786542892456, | |
| "learning_rate": 0.00013110215624979025, | |
| "loss": 0.1962, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.17921703449040702, | |
| "grad_norm": 0.16338172554969788, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.2002, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.17951277217108424, | |
| "grad_norm": 0.17155061662197113, | |
| "learning_rate": 0.00013070110524965954, | |
| "loss": 0.2127, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.1798085098517615, | |
| "grad_norm": 0.1854557991027832, | |
| "learning_rate": 0.00013050037457803924, | |
| "loss": 0.2304, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.18010424753243873, | |
| "grad_norm": 0.16282689571380615, | |
| "learning_rate": 0.0001302995083149953, | |
| "loss": 0.1854, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.18039998521311595, | |
| "grad_norm": 0.1648624688386917, | |
| "learning_rate": 0.0001300985073534919, | |
| "loss": 0.2144, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.1806957228937932, | |
| "grad_norm": 0.15584151446819305, | |
| "learning_rate": 0.00012989737258709203, | |
| "loss": 0.1721, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.18099146057447044, | |
| "grad_norm": 0.16049699485301971, | |
| "learning_rate": 0.00012969610490995358, | |
| "loss": 0.165, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.1812871982551477, | |
| "grad_norm": 0.17223642766475677, | |
| "learning_rate": 0.00012949470521682528, | |
| "loss": 0.1902, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.18158293593582492, | |
| "grad_norm": 0.18705599009990692, | |
| "learning_rate": 0.0001292931744030427, | |
| "loss": 0.2414, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.18187867361650217, | |
| "grad_norm": 0.21196113526821136, | |
| "learning_rate": 0.0001290915133645243, | |
| "loss": 0.2609, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.1821744112971794, | |
| "grad_norm": 0.1778162270784378, | |
| "learning_rate": 0.00012888972299776754, | |
| "loss": 0.1849, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.18247014897785663, | |
| "grad_norm": 0.17998971045017242, | |
| "learning_rate": 0.00012868780419984482, | |
| "loss": 0.1704, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.18276588665853388, | |
| "grad_norm": 0.19640791416168213, | |
| "learning_rate": 0.00012848575786839943, | |
| "loss": 0.2088, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.1830616243392111, | |
| "grad_norm": 0.19208578765392303, | |
| "learning_rate": 0.0001282835849016416, | |
| "loss": 0.1754, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.18335736201988836, | |
| "grad_norm": 0.1946994960308075, | |
| "learning_rate": 0.00012808128619834461, | |
| "loss": 0.1852, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.1836530997005656, | |
| "grad_norm": 0.1998986154794693, | |
| "learning_rate": 0.0001278788626578407, | |
| "loss": 0.2074, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.18394883738124285, | |
| "grad_norm": 0.1936517357826233, | |
| "learning_rate": 0.00012767631518001698, | |
| "loss": 0.1938, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.18424457506192007, | |
| "grad_norm": 0.2083950638771057, | |
| "learning_rate": 0.00012747364466531163, | |
| "loss": 0.1993, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.18454031274259733, | |
| "grad_norm": 0.20452117919921875, | |
| "learning_rate": 0.00012727085201470973, | |
| "loss": 0.1721, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.18483605042327456, | |
| "grad_norm": 0.23634770512580872, | |
| "learning_rate": 0.00012706793812973941, | |
| "loss": 0.2541, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.18513178810395178, | |
| "grad_norm": 0.24558795988559723, | |
| "learning_rate": 0.0001268649039124677, | |
| "loss": 0.2825, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.18542752578462904, | |
| "grad_norm": 0.2423345446586609, | |
| "learning_rate": 0.00012666175026549662, | |
| "loss": 0.2449, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.18572326346530627, | |
| "grad_norm": 0.20482169091701508, | |
| "learning_rate": 0.000126458478091959, | |
| "loss": 0.2034, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.18601900114598352, | |
| "grad_norm": 0.2271885871887207, | |
| "learning_rate": 0.00012625508829551473, | |
| "loss": 0.2243, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.18631473882666075, | |
| "grad_norm": 0.2068728655576706, | |
| "learning_rate": 0.00012605158178034654, | |
| "loss": 0.1797, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.186610476507338, | |
| "grad_norm": 0.2342994064092636, | |
| "learning_rate": 0.00012584795945115603, | |
| "loss": 0.2005, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.18690621418801523, | |
| "grad_norm": 0.22803792357444763, | |
| "learning_rate": 0.0001256442222131597, | |
| "loss": 0.2167, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.18720195186869246, | |
| "grad_norm": 0.23853746056556702, | |
| "learning_rate": 0.0001254403709720848, | |
| "loss": 0.2077, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.1874976895493697, | |
| "grad_norm": 0.2190757393836975, | |
| "learning_rate": 0.0001252364066341655, | |
| "loss": 0.1949, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.18779342723004694, | |
| "grad_norm": 0.22064319252967834, | |
| "learning_rate": 0.00012503233010613865, | |
| "loss": 0.1563, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.1880891649107242, | |
| "grad_norm": 0.22475862503051758, | |
| "learning_rate": 0.00012482814229523997, | |
| "loss": 0.164, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.18838490259140142, | |
| "grad_norm": 0.23672211170196533, | |
| "learning_rate": 0.00012462384410919975, | |
| "loss": 0.1975, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.18868064027207868, | |
| "grad_norm": 0.22147880494594574, | |
| "learning_rate": 0.00012441943645623903, | |
| "loss": 0.1799, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.1889763779527559, | |
| "grad_norm": 0.23822712898254395, | |
| "learning_rate": 0.00012421492024506555, | |
| "loss": 0.197, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.18927211563343316, | |
| "grad_norm": 0.2708362638950348, | |
| "learning_rate": 0.00012401029638486953, | |
| "loss": 0.2175, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.18956785331411038, | |
| "grad_norm": 0.23913449048995972, | |
| "learning_rate": 0.0001238055657853198, | |
| "loss": 0.1567, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.1898635909947876, | |
| "grad_norm": 0.2158363163471222, | |
| "learning_rate": 0.00012360072935655982, | |
| "loss": 0.1398, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.19015932867546487, | |
| "grad_norm": 0.2530500888824463, | |
| "learning_rate": 0.00012339578800920332, | |
| "loss": 0.1711, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.1904550663561421, | |
| "grad_norm": 0.2804064154624939, | |
| "learning_rate": 0.00012319074265433063, | |
| "loss": 0.1629, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.19075080403681935, | |
| "grad_norm": 0.38679811358451843, | |
| "learning_rate": 0.00012298559420348437, | |
| "loss": 0.2072, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.19104654171749658, | |
| "grad_norm": 0.25860467553138733, | |
| "learning_rate": 0.00012278034356866545, | |
| "loss": 0.1584, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.19134227939817383, | |
| "grad_norm": 0.42005908489227295, | |
| "learning_rate": 0.00012257499166232907, | |
| "loss": 0.2906, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.19163801707885106, | |
| "grad_norm": 0.3620118498802185, | |
| "learning_rate": 0.0001223695393973807, | |
| "loss": 0.2341, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.19193375475952829, | |
| "grad_norm": 0.43819329142570496, | |
| "learning_rate": 0.0001221639876871719, | |
| "loss": 0.237, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.19222949244020554, | |
| "grad_norm": 0.6452730894088745, | |
| "learning_rate": 0.0001219583374454963, | |
| "loss": 0.2783, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.19252523012088277, | |
| "grad_norm": 0.13952232897281647, | |
| "learning_rate": 0.00012175258958658564, | |
| "loss": 0.1729, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.19282096780156002, | |
| "grad_norm": 0.16192123293876648, | |
| "learning_rate": 0.00012154674502510555, | |
| "loss": 0.2102, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.19311670548223725, | |
| "grad_norm": 0.1605851650238037, | |
| "learning_rate": 0.00012134080467615159, | |
| "loss": 0.1727, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.1934124431629145, | |
| "grad_norm": 0.1601710319519043, | |
| "learning_rate": 0.00012113476945524513, | |
| "loss": 0.1528, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.19370818084359173, | |
| "grad_norm": 0.1721370667219162, | |
| "learning_rate": 0.00012092864027832933, | |
| "loss": 0.1819, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.194003918524269, | |
| "grad_norm": 0.16613177955150604, | |
| "learning_rate": 0.000120722418061765, | |
| "loss": 0.2218, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.19429965620494621, | |
| "grad_norm": 0.17850197851657867, | |
| "learning_rate": 0.0001205161037223266, | |
| "loss": 0.2397, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.19459539388562344, | |
| "grad_norm": 0.18693998456001282, | |
| "learning_rate": 0.00012030969817719808, | |
| "loss": 0.238, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.1948911315663007, | |
| "grad_norm": 0.17822137475013733, | |
| "learning_rate": 0.00012010320234396894, | |
| "loss": 0.1988, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.19518686924697792, | |
| "grad_norm": 0.23123416304588318, | |
| "learning_rate": 0.00011989661714062999, | |
| "loss": 0.2258, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.19548260692765518, | |
| "grad_norm": 0.1795627623796463, | |
| "learning_rate": 0.0001196899434855693, | |
| "loss": 0.1911, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.1957783446083324, | |
| "grad_norm": 0.1633031815290451, | |
| "learning_rate": 0.00011948318229756827, | |
| "loss": 0.1737, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.19607408228900966, | |
| "grad_norm": 0.1819830685853958, | |
| "learning_rate": 0.00011927633449579735, | |
| "loss": 0.179, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.1963698199696869, | |
| "grad_norm": 0.18557539582252502, | |
| "learning_rate": 0.0001190694009998121, | |
| "loss": 0.2491, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.19666555765036411, | |
| "grad_norm": 0.19799882173538208, | |
| "learning_rate": 0.00011886238272954897, | |
| "loss": 0.2407, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.19696129533104137, | |
| "grad_norm": 0.19097891449928284, | |
| "learning_rate": 0.00011865528060532127, | |
| "loss": 0.2252, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.1972570330117186, | |
| "grad_norm": 0.170382559299469, | |
| "learning_rate": 0.0001184480955478152, | |
| "loss": 0.1763, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.19755277069239585, | |
| "grad_norm": 0.18122434616088867, | |
| "learning_rate": 0.00011824082847808558, | |
| "loss": 0.1891, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.19784850837307308, | |
| "grad_norm": 0.20141829550266266, | |
| "learning_rate": 0.00011803348031755179, | |
| "loss": 0.1939, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.19814424605375033, | |
| "grad_norm": 0.19852298498153687, | |
| "learning_rate": 0.0001178260519879937, | |
| "loss": 0.2078, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.19843998373442756, | |
| "grad_norm": 0.18741367757320404, | |
| "learning_rate": 0.00011761854441154767, | |
| "loss": 0.1668, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.19873572141510482, | |
| "grad_norm": 0.18189528584480286, | |
| "learning_rate": 0.00011741095851070228, | |
| "loss": 0.1611, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.19903145909578204, | |
| "grad_norm": 0.1981499046087265, | |
| "learning_rate": 0.00011720329520829429, | |
| "loss": 0.2201, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.19932719677645927, | |
| "grad_norm": 0.19393107295036316, | |
| "learning_rate": 0.0001169955554275046, | |
| "loss": 0.193, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.19962293445713652, | |
| "grad_norm": 0.19245851039886475, | |
| "learning_rate": 0.0001167877400918541, | |
| "loss": 0.1546, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.19991867213781375, | |
| "grad_norm": 0.2009516805410385, | |
| "learning_rate": 0.00011657985012519952, | |
| "loss": 0.1963, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.200214409818491, | |
| "grad_norm": 0.19694218039512634, | |
| "learning_rate": 0.00011637188645172944, | |
| "loss": 0.1624, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.20051014749916823, | |
| "grad_norm": 0.21188542246818542, | |
| "learning_rate": 0.00011616384999596006, | |
| "loss": 0.1885, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.2008058851798455, | |
| "grad_norm": 0.18305402994155884, | |
| "learning_rate": 0.00011595574168273111, | |
| "loss": 0.1353, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.20110162286052272, | |
| "grad_norm": 0.23467271029949188, | |
| "learning_rate": 0.0001157475624372018, | |
| "loss": 0.2274, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.20139736054119994, | |
| "grad_norm": 0.21421073377132416, | |
| "learning_rate": 0.0001155393131848467, | |
| "loss": 0.1823, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.2016930982218772, | |
| "grad_norm": 0.23332397639751434, | |
| "learning_rate": 0.00011533099485145155, | |
| "loss": 0.2598, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.20198883590255443, | |
| "grad_norm": 0.21139463782310486, | |
| "learning_rate": 0.00011512260836310924, | |
| "loss": 0.1666, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.20228457358323168, | |
| "grad_norm": 0.25425291061401367, | |
| "learning_rate": 0.00011491415464621562, | |
| "loss": 0.2299, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.2025803112639089, | |
| "grad_norm": 0.26843997836112976, | |
| "learning_rate": 0.00011470563462746541, | |
| "loss": 0.2056, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.20287604894458616, | |
| "grad_norm": 0.26034682989120483, | |
| "learning_rate": 0.00011449704923384812, | |
| "loss": 0.2092, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.2031717866252634, | |
| "grad_norm": 0.25974202156066895, | |
| "learning_rate": 0.00011428839939264382, | |
| "loss": 0.1912, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.20346752430594062, | |
| "grad_norm": 0.2530285716056824, | |
| "learning_rate": 0.0001140796860314191, | |
| "loss": 0.1965, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.20376326198661787, | |
| "grad_norm": 0.2534104883670807, | |
| "learning_rate": 0.00011387091007802297, | |
| "loss": 0.1936, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.2040589996672951, | |
| "grad_norm": 0.24163153767585754, | |
| "learning_rate": 0.0001136620724605827, | |
| "loss": 0.1617, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.20435473734797235, | |
| "grad_norm": 0.24182729423046112, | |
| "learning_rate": 0.00011345317410749964, | |
| "loss": 0.1716, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.20465047502864958, | |
| "grad_norm": 0.26009559631347656, | |
| "learning_rate": 0.00011324421594744516, | |
| "loss": 0.1832, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.20494621270932684, | |
| "grad_norm": 0.24706415832042694, | |
| "learning_rate": 0.00011303519890935656, | |
| "loss": 0.1604, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.20524195039000406, | |
| "grad_norm": 0.3163825571537018, | |
| "learning_rate": 0.00011282612392243286, | |
| "loss": 0.2144, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.20553768807068132, | |
| "grad_norm": 0.25135427713394165, | |
| "learning_rate": 0.00011261699191613066, | |
| "loss": 0.1179, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.20583342575135855, | |
| "grad_norm": 0.31126582622528076, | |
| "learning_rate": 0.00011240780382016005, | |
| "loss": 0.1907, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.20612916343203577, | |
| "grad_norm": 0.34292182326316833, | |
| "learning_rate": 0.00011219856056448051, | |
| "loss": 0.2255, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.20642490111271303, | |
| "grad_norm": 0.3281469941139221, | |
| "learning_rate": 0.00011198926307929664, | |
| "loss": 0.1755, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.20672063879339025, | |
| "grad_norm": 0.4065954387187958, | |
| "learning_rate": 0.00011177991229505431, | |
| "loss": 0.2255, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.2070163764740675, | |
| "grad_norm": 0.5669332146644592, | |
| "learning_rate": 0.00011157050914243614, | |
| "loss": 0.3493, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.20731211415474474, | |
| "grad_norm": 0.10655932128429413, | |
| "learning_rate": 0.00011136105455235766, | |
| "loss": 0.1189, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.207607851835422, | |
| "grad_norm": 0.12350799888372421, | |
| "learning_rate": 0.00011115154945596305, | |
| "loss": 0.1384, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.20790358951609922, | |
| "grad_norm": 0.16289597749710083, | |
| "learning_rate": 0.00011094199478462095, | |
| "loss": 0.1981, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.20819932719677645, | |
| "grad_norm": 0.16098204255104065, | |
| "learning_rate": 0.00011073239146992054, | |
| "loss": 0.1988, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.2084950648774537, | |
| "grad_norm": 0.17264235019683838, | |
| "learning_rate": 0.00011052274044366711, | |
| "loss": 0.1934, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.20879080255813093, | |
| "grad_norm": 0.18324457108974457, | |
| "learning_rate": 0.00011031304263787812, | |
| "loss": 0.2202, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.20908654023880818, | |
| "grad_norm": 0.1545696258544922, | |
| "learning_rate": 0.00011010329898477891, | |
| "loss": 0.1787, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.2093822779194854, | |
| "grad_norm": 0.1769752949476242, | |
| "learning_rate": 0.0001098935104167988, | |
| "loss": 0.1949, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.20967801560016266, | |
| "grad_norm": 0.1559659242630005, | |
| "learning_rate": 0.00010968367786656663, | |
| "loss": 0.1562, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.2099737532808399, | |
| "grad_norm": 0.1613316386938095, | |
| "learning_rate": 0.00010947380226690684, | |
| "loss": 0.1836, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.21026949096151715, | |
| "grad_norm": 0.18986111879348755, | |
| "learning_rate": 0.00010926388455083522, | |
| "loss": 0.241, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.21056522864219437, | |
| "grad_norm": 0.17520417273044586, | |
| "learning_rate": 0.00010905392565155477, | |
| "loss": 0.181, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.2108609663228716, | |
| "grad_norm": 0.17421351373195648, | |
| "learning_rate": 0.00010884392650245165, | |
| "loss": 0.2037, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.21115670400354886, | |
| "grad_norm": 0.17320917546749115, | |
| "learning_rate": 0.00010863388803709089, | |
| "loss": 0.18, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.21145244168422608, | |
| "grad_norm": 0.19977597892284393, | |
| "learning_rate": 0.00010842381118921232, | |
| "loss": 0.2179, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.21174817936490334, | |
| "grad_norm": 0.19101393222808838, | |
| "learning_rate": 0.00010821369689272638, | |
| "loss": 0.2022, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.21204391704558057, | |
| "grad_norm": 0.22305834293365479, | |
| "learning_rate": 0.00010800354608171003, | |
| "loss": 0.2826, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.21233965472625782, | |
| "grad_norm": 0.1698966771364212, | |
| "learning_rate": 0.00010779335969040252, | |
| "loss": 0.1619, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.21263539240693505, | |
| "grad_norm": 0.1868032068014145, | |
| "learning_rate": 0.00010758313865320134, | |
| "loss": 0.1907, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.21293113008761227, | |
| "grad_norm": 0.1948738545179367, | |
| "learning_rate": 0.00010737288390465792, | |
| "loss": 0.2101, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.21322686776828953, | |
| "grad_norm": 0.18316994607448578, | |
| "learning_rate": 0.00010716259637947357, | |
| "loss": 0.1733, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.21352260544896676, | |
| "grad_norm": 0.18635033071041107, | |
| "learning_rate": 0.00010695227701249537, | |
| "loss": 0.177, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.213818343129644, | |
| "grad_norm": 0.22876830399036407, | |
| "learning_rate": 0.00010674192673871191, | |
| "loss": 0.1832, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.21411408081032124, | |
| "grad_norm": 0.17134256660938263, | |
| "learning_rate": 0.00010653154649324917, | |
| "loss": 0.1529, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.2144098184909985, | |
| "grad_norm": 0.2593710422515869, | |
| "learning_rate": 0.00010632113721136636, | |
| "loss": 0.2832, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.21470555617167572, | |
| "grad_norm": 0.24183620512485504, | |
| "learning_rate": 0.00010611069982845183, | |
| "loss": 0.2632, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.21500129385235298, | |
| "grad_norm": 0.23827098309993744, | |
| "learning_rate": 0.00010590023528001884, | |
| "loss": 0.227, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.2152970315330302, | |
| "grad_norm": 0.20484110713005066, | |
| "learning_rate": 0.00010568974450170139, | |
| "loss": 0.2038, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.21559276921370743, | |
| "grad_norm": 0.22389110922813416, | |
| "learning_rate": 0.00010547922842925008, | |
| "loss": 0.2181, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.21588850689438469, | |
| "grad_norm": 0.22689417004585266, | |
| "learning_rate": 0.00010526868799852796, | |
| "loss": 0.2301, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.2161842445750619, | |
| "grad_norm": 0.21914827823638916, | |
| "learning_rate": 0.0001050581241455064, | |
| "loss": 0.1857, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.21647998225573917, | |
| "grad_norm": 0.2330751121044159, | |
| "learning_rate": 0.00010484753780626089, | |
| "loss": 0.1797, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.2167757199364164, | |
| "grad_norm": 0.22893856465816498, | |
| "learning_rate": 0.00010463692991696685, | |
| "loss": 0.2063, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.21707145761709365, | |
| "grad_norm": 0.24380573630332947, | |
| "learning_rate": 0.00010442630141389549, | |
| "loss": 0.1874, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.21736719529777088, | |
| "grad_norm": 0.24543069303035736, | |
| "learning_rate": 0.00010421565323340971, | |
| "loss": 0.1931, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.2176629329784481, | |
| "grad_norm": 0.22604742646217346, | |
| "learning_rate": 0.00010400498631195992, | |
| "loss": 0.143, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.21795867065912536, | |
| "grad_norm": 0.2889672815799713, | |
| "learning_rate": 0.00010379430158607975, | |
| "loss": 0.2563, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.21825440833980259, | |
| "grad_norm": 0.24721504747867584, | |
| "learning_rate": 0.000103583599992382, | |
| "loss": 0.1586, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.21855014602047984, | |
| "grad_norm": 0.23907063901424408, | |
| "learning_rate": 0.0001033728824675545, | |
| "loss": 0.1921, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.21884588370115707, | |
| "grad_norm": 0.22747616469860077, | |
| "learning_rate": 0.0001031621499483559, | |
| "loss": 0.1351, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.21914162138183432, | |
| "grad_norm": 0.3032531142234802, | |
| "learning_rate": 0.00010295140337161146, | |
| "loss": 0.2032, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.21943735906251155, | |
| "grad_norm": 0.24150900542736053, | |
| "learning_rate": 0.00010274064367420897, | |
| "loss": 0.1577, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.2197330967431888, | |
| "grad_norm": 0.29281744360923767, | |
| "learning_rate": 0.00010252987179309459, | |
| "loss": 0.2267, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.22002883442386603, | |
| "grad_norm": 0.2816901206970215, | |
| "learning_rate": 0.00010231908866526851, | |
| "loss": 0.2086, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.22032457210454326, | |
| "grad_norm": 0.343010276556015, | |
| "learning_rate": 0.00010210829522778111, | |
| "loss": 0.224, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.22062030978522051, | |
| "grad_norm": 0.31681838631629944, | |
| "learning_rate": 0.00010189749241772844, | |
| "loss": 0.2536, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.22091604746589774, | |
| "grad_norm": 0.2629532516002655, | |
| "learning_rate": 0.00010168668117224825, | |
| "loss": 0.1457, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.221211785146575, | |
| "grad_norm": 0.33355703949928284, | |
| "learning_rate": 0.00010147586242851585, | |
| "loss": 0.2115, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.22150752282725222, | |
| "grad_norm": 0.3667524755001068, | |
| "learning_rate": 0.00010126503712373982, | |
| "loss": 0.2341, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.22180326050792948, | |
| "grad_norm": 0.5130096673965454, | |
| "learning_rate": 0.00010105420619515798, | |
| "loss": 0.2986, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.22180326050792948, | |
| "eval_loss": 0.194038987159729, | |
| "eval_runtime": 323.8656, | |
| "eval_samples_per_second": 17.584, | |
| "eval_steps_per_second": 8.794, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2220989981886067, | |
| "grad_norm": 0.115847647190094, | |
| "learning_rate": 0.00010084337058003303, | |
| "loss": 0.1349, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.22239473586928393, | |
| "grad_norm": 0.13945455849170685, | |
| "learning_rate": 0.00010063253121564868, | |
| "loss": 0.1831, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.2226904735499612, | |
| "grad_norm": 0.1549556851387024, | |
| "learning_rate": 0.00010042168903930514, | |
| "loss": 0.1844, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.22298621123063841, | |
| "grad_norm": 0.1443348079919815, | |
| "learning_rate": 0.00010021084498831522, | |
| "loss": 0.1744, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.22328194891131567, | |
| "grad_norm": 0.17714332044124603, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2593, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.2235776865919929, | |
| "grad_norm": 0.16148702800273895, | |
| "learning_rate": 9.97891550116848e-05, | |
| "loss": 0.2155, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.22387342427267015, | |
| "grad_norm": 0.16650299727916718, | |
| "learning_rate": 9.957831096069488e-05, | |
| "loss": 0.2036, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.22416916195334738, | |
| "grad_norm": 0.15004287660121918, | |
| "learning_rate": 9.936746878435136e-05, | |
| "loss": 0.1661, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.22446489963402463, | |
| "grad_norm": 0.15887637436389923, | |
| "learning_rate": 9.915662941996699e-05, | |
| "loss": 0.1838, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.22476063731470186, | |
| "grad_norm": 0.17734317481517792, | |
| "learning_rate": 9.894579380484204e-05, | |
| "loss": 0.2173, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.2250563749953791, | |
| "grad_norm": 0.17889343202114105, | |
| "learning_rate": 9.873496287626019e-05, | |
| "loss": 0.1961, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.22535211267605634, | |
| "grad_norm": 0.1788489818572998, | |
| "learning_rate": 9.852413757148417e-05, | |
| "loss": 0.2122, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.22564785035673357, | |
| "grad_norm": 0.1814776510000229, | |
| "learning_rate": 9.831331882775178e-05, | |
| "loss": 0.2257, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.22594358803741083, | |
| "grad_norm": 0.2010183185338974, | |
| "learning_rate": 9.81025075822716e-05, | |
| "loss": 0.2178, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.22623932571808805, | |
| "grad_norm": 0.1668553501367569, | |
| "learning_rate": 9.789170477221891e-05, | |
| "loss": 0.1643, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.2265350633987653, | |
| "grad_norm": 0.18353386223316193, | |
| "learning_rate": 9.76809113347315e-05, | |
| "loss": 0.191, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.22683080107944253, | |
| "grad_norm": 0.19008232653141022, | |
| "learning_rate": 9.747012820690543e-05, | |
| "loss": 0.2259, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.22712653876011976, | |
| "grad_norm": 0.1797637939453125, | |
| "learning_rate": 9.725935632579104e-05, | |
| "loss": 0.1842, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.22742227644079702, | |
| "grad_norm": 0.1997433602809906, | |
| "learning_rate": 9.704859662838855e-05, | |
| "loss": 0.2003, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.22771801412147424, | |
| "grad_norm": 0.17852815985679626, | |
| "learning_rate": 9.683785005164411e-05, | |
| "loss": 0.1665, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.2280137518021515, | |
| "grad_norm": 0.19264471530914307, | |
| "learning_rate": 9.662711753244551e-05, | |
| "loss": 0.2015, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.22830948948282873, | |
| "grad_norm": 0.20208404958248138, | |
| "learning_rate": 9.641640000761802e-05, | |
| "loss": 0.1836, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.22860522716350598, | |
| "grad_norm": 0.20002898573875427, | |
| "learning_rate": 9.620569841392029e-05, | |
| "loss": 0.1956, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.2289009648441832, | |
| "grad_norm": 0.20128233730793, | |
| "learning_rate": 9.59950136880401e-05, | |
| "loss": 0.218, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.22919670252486044, | |
| "grad_norm": 0.1997966766357422, | |
| "learning_rate": 9.57843467665903e-05, | |
| "loss": 0.1771, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.2294924402055377, | |
| "grad_norm": 0.22592507302761078, | |
| "learning_rate": 9.557369858610453e-05, | |
| "loss": 0.2222, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.22978817788621492, | |
| "grad_norm": 0.18361607193946838, | |
| "learning_rate": 9.53630700830332e-05, | |
| "loss": 0.1476, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.23008391556689217, | |
| "grad_norm": 0.17344044148921967, | |
| "learning_rate": 9.51524621937391e-05, | |
| "loss": 0.1298, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.2303796532475694, | |
| "grad_norm": 0.20982904732227325, | |
| "learning_rate": 9.494187585449358e-05, | |
| "loss": 0.2031, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.23067539092824665, | |
| "grad_norm": 0.20879507064819336, | |
| "learning_rate": 9.473131200147205e-05, | |
| "loss": 0.1622, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.23097112860892388, | |
| "grad_norm": 0.24040478467941284, | |
| "learning_rate": 9.452077157074994e-05, | |
| "loss": 0.2284, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.23126686628960114, | |
| "grad_norm": 0.2203570157289505, | |
| "learning_rate": 9.431025549829862e-05, | |
| "loss": 0.1829, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.23156260397027836, | |
| "grad_norm": 0.2384481579065323, | |
| "learning_rate": 9.409976471998118e-05, | |
| "loss": 0.178, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.2318583416509556, | |
| "grad_norm": 0.2443898469209671, | |
| "learning_rate": 9.388930017154819e-05, | |
| "loss": 0.169, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.23215407933163285, | |
| "grad_norm": 0.27730923891067505, | |
| "learning_rate": 9.367886278863366e-05, | |
| "loss": 0.2722, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.23244981701231007, | |
| "grad_norm": 0.2513563632965088, | |
| "learning_rate": 9.346845350675088e-05, | |
| "loss": 0.1804, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.23274555469298733, | |
| "grad_norm": 0.25606629252433777, | |
| "learning_rate": 9.325807326128814e-05, | |
| "loss": 0.2188, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.23304129237366455, | |
| "grad_norm": 0.24177873134613037, | |
| "learning_rate": 9.304772298750463e-05, | |
| "loss": 0.2054, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.2333370300543418, | |
| "grad_norm": 0.23143790662288666, | |
| "learning_rate": 9.283740362052642e-05, | |
| "loss": 0.1496, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.23363276773501904, | |
| "grad_norm": 0.25107666850090027, | |
| "learning_rate": 9.26271160953421e-05, | |
| "loss": 0.1276, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.23392850541569626, | |
| "grad_norm": 0.25497153401374817, | |
| "learning_rate": 9.241686134679867e-05, | |
| "loss": 0.1567, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.23422424309637352, | |
| "grad_norm": 0.3068770170211792, | |
| "learning_rate": 9.220664030959749e-05, | |
| "loss": 0.1859, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.23451998077705075, | |
| "grad_norm": 0.2575652301311493, | |
| "learning_rate": 9.199645391828999e-05, | |
| "loss": 0.1626, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.234815718457728, | |
| "grad_norm": 0.2657597064971924, | |
| "learning_rate": 9.178630310727365e-05, | |
| "loss": 0.1557, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.23511145613840523, | |
| "grad_norm": 0.3128422200679779, | |
| "learning_rate": 9.157618881078772e-05, | |
| "loss": 0.2013, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.23540719381908248, | |
| "grad_norm": 0.280213326215744, | |
| "learning_rate": 9.136611196290915e-05, | |
| "loss": 0.1513, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.2357029314997597, | |
| "grad_norm": 0.32494768500328064, | |
| "learning_rate": 9.115607349754834e-05, | |
| "loss": 0.164, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.23599866918043697, | |
| "grad_norm": 0.31203532218933105, | |
| "learning_rate": 9.094607434844523e-05, | |
| "loss": 0.1492, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.2362944068611142, | |
| "grad_norm": 0.44517087936401367, | |
| "learning_rate": 9.07361154491648e-05, | |
| "loss": 0.2025, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.23659014454179142, | |
| "grad_norm": 0.5797000527381897, | |
| "learning_rate": 9.052619773309317e-05, | |
| "loss": 0.2612, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.23688588222246867, | |
| "grad_norm": 0.11432012170553207, | |
| "learning_rate": 9.031632213343339e-05, | |
| "loss": 0.1212, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.2371816199031459, | |
| "grad_norm": 0.15115061402320862, | |
| "learning_rate": 9.01064895832012e-05, | |
| "loss": 0.1791, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.23747735758382316, | |
| "grad_norm": 0.1600162535905838, | |
| "learning_rate": 8.98967010152211e-05, | |
| "loss": 0.1696, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.23777309526450038, | |
| "grad_norm": 0.1539337933063507, | |
| "learning_rate": 8.968695736212193e-05, | |
| "loss": 0.1813, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.23806883294517764, | |
| "grad_norm": 0.15929153561592102, | |
| "learning_rate": 8.947725955633294e-05, | |
| "loss": 0.1551, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.23836457062585487, | |
| "grad_norm": 0.14739802479743958, | |
| "learning_rate": 8.926760853007946e-05, | |
| "loss": 0.1555, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.2386603083065321, | |
| "grad_norm": 0.16396670043468475, | |
| "learning_rate": 8.905800521537905e-05, | |
| "loss": 0.1776, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.23895604598720935, | |
| "grad_norm": 0.16819187998771667, | |
| "learning_rate": 8.884845054403699e-05, | |
| "loss": 0.1925, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.23925178366788658, | |
| "grad_norm": 0.17342273890972137, | |
| "learning_rate": 8.863894544764236e-05, | |
| "loss": 0.183, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.23954752134856383, | |
| "grad_norm": 0.18809068202972412, | |
| "learning_rate": 8.84294908575639e-05, | |
| "loss": 0.2289, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.23984325902924106, | |
| "grad_norm": 0.18252570927143097, | |
| "learning_rate": 8.822008770494572e-05, | |
| "loss": 0.2011, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.2401389967099183, | |
| "grad_norm": 0.1622556895017624, | |
| "learning_rate": 8.801073692070337e-05, | |
| "loss": 0.1416, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.24043473439059554, | |
| "grad_norm": 0.17876408994197845, | |
| "learning_rate": 8.780143943551954e-05, | |
| "loss": 0.2014, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.2407304720712728, | |
| "grad_norm": 0.18106475472450256, | |
| "learning_rate": 8.759219617983999e-05, | |
| "loss": 0.2002, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.24102620975195002, | |
| "grad_norm": 0.20041576027870178, | |
| "learning_rate": 8.738300808386935e-05, | |
| "loss": 0.2289, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.24132194743262725, | |
| "grad_norm": 0.17808635532855988, | |
| "learning_rate": 8.717387607756713e-05, | |
| "loss": 0.1851, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.2416176851133045, | |
| "grad_norm": 0.22498534619808197, | |
| "learning_rate": 8.696480109064342e-05, | |
| "loss": 0.2488, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.24191342279398173, | |
| "grad_norm": 0.19668036699295044, | |
| "learning_rate": 8.675578405255485e-05, | |
| "loss": 0.1995, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.24220916047465899, | |
| "grad_norm": 0.19187040627002716, | |
| "learning_rate": 8.654682589250038e-05, | |
| "loss": 0.1759, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.2425048981553362, | |
| "grad_norm": 0.208050936460495, | |
| "learning_rate": 8.633792753941733e-05, | |
| "loss": 0.2253, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.24280063583601347, | |
| "grad_norm": 0.18446066975593567, | |
| "learning_rate": 8.612908992197705e-05, | |
| "loss": 0.1743, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.2430963735166907, | |
| "grad_norm": 0.19328705966472626, | |
| "learning_rate": 8.592031396858093e-05, | |
| "loss": 0.1572, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.24339211119736792, | |
| "grad_norm": 0.21649695932865143, | |
| "learning_rate": 8.571160060735624e-05, | |
| "loss": 0.2124, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.24368784887804518, | |
| "grad_norm": 0.1915416270494461, | |
| "learning_rate": 8.550295076615188e-05, | |
| "loss": 0.169, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.2439835865587224, | |
| "grad_norm": 0.20172414183616638, | |
| "learning_rate": 8.529436537253458e-05, | |
| "loss": 0.1644, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.24427932423939966, | |
| "grad_norm": 0.19948998093605042, | |
| "learning_rate": 8.508584535378439e-05, | |
| "loss": 0.2082, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.24457506192007689, | |
| "grad_norm": 0.22914430499076843, | |
| "learning_rate": 8.487739163689079e-05, | |
| "loss": 0.205, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.24487079960075414, | |
| "grad_norm": 0.21713916957378387, | |
| "learning_rate": 8.466900514854847e-05, | |
| "loss": 0.2202, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.24516653728143137, | |
| "grad_norm": 0.2027435600757599, | |
| "learning_rate": 8.446068681515334e-05, | |
| "loss": 0.1789, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.24546227496210862, | |
| "grad_norm": 0.24722711741924286, | |
| "learning_rate": 8.425243756279824e-05, | |
| "loss": 0.1915, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.24575801264278585, | |
| "grad_norm": 0.2104775607585907, | |
| "learning_rate": 8.404425831726894e-05, | |
| "loss": 0.1414, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.24605375032346308, | |
| "grad_norm": 0.22853389382362366, | |
| "learning_rate": 8.383615000404e-05, | |
| "loss": 0.1666, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.24634948800414033, | |
| "grad_norm": 0.22614425420761108, | |
| "learning_rate": 8.362811354827059e-05, | |
| "loss": 0.1459, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.24664522568481756, | |
| "grad_norm": 0.23260444402694702, | |
| "learning_rate": 8.342014987480047e-05, | |
| "loss": 0.1829, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.24694096336549481, | |
| "grad_norm": 0.27205580472946167, | |
| "learning_rate": 8.321225990814591e-05, | |
| "loss": 0.1895, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.24723670104617204, | |
| "grad_norm": 0.27459388971328735, | |
| "learning_rate": 8.300444457249543e-05, | |
| "loss": 0.222, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.2475324387268493, | |
| "grad_norm": 0.25127366185188293, | |
| "learning_rate": 8.279670479170573e-05, | |
| "loss": 0.1232, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.24782817640752652, | |
| "grad_norm": 0.244548961520195, | |
| "learning_rate": 8.258904148929775e-05, | |
| "loss": 0.2147, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.24812391408820375, | |
| "grad_norm": 0.24851387739181519, | |
| "learning_rate": 8.238145558845235e-05, | |
| "loss": 0.2079, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.248419651768881, | |
| "grad_norm": 0.29407623410224915, | |
| "learning_rate": 8.217394801200631e-05, | |
| "loss": 0.2305, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.24871538944955823, | |
| "grad_norm": 0.2977339029312134, | |
| "learning_rate": 8.196651968244826e-05, | |
| "loss": 0.2305, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.2490111271302355, | |
| "grad_norm": 0.3162849247455597, | |
| "learning_rate": 8.175917152191447e-05, | |
| "loss": 0.2099, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.24930686481091272, | |
| "grad_norm": 0.315044105052948, | |
| "learning_rate": 8.15519044521848e-05, | |
| "loss": 0.2464, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.24960260249158997, | |
| "grad_norm": 0.23404821753501892, | |
| "learning_rate": 8.134471939467874e-05, | |
| "loss": 0.1401, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.2498983401722672, | |
| "grad_norm": 0.28799015283584595, | |
| "learning_rate": 8.113761727045105e-05, | |
| "loss": 0.1694, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.25019407785294445, | |
| "grad_norm": 0.3503269553184509, | |
| "learning_rate": 8.093059900018792e-05, | |
| "loss": 0.2245, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.25048981553362165, | |
| "grad_norm": 0.3673860430717468, | |
| "learning_rate": 8.072366550420266e-05, | |
| "loss": 0.224, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.2507855532142989, | |
| "grad_norm": 0.4725538492202759, | |
| "learning_rate": 8.051681770243175e-05, | |
| "loss": 0.1999, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.25108129089497616, | |
| "grad_norm": 0.454391211271286, | |
| "learning_rate": 8.031005651443073e-05, | |
| "loss": 0.2453, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.2513770285756534, | |
| "grad_norm": 0.6510867476463318, | |
| "learning_rate": 8.010338285937006e-05, | |
| "loss": 0.3214, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2516727662563306, | |
| "grad_norm": 0.1124342530965805, | |
| "learning_rate": 7.989679765603108e-05, | |
| "loss": 0.1517, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.25196850393700787, | |
| "grad_norm": 0.1325543373823166, | |
| "learning_rate": 7.969030182280192e-05, | |
| "loss": 0.177, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.2522642416176851, | |
| "grad_norm": 0.14178207516670227, | |
| "learning_rate": 7.948389627767343e-05, | |
| "loss": 0.1601, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.2525599792983623, | |
| "grad_norm": 0.15501843392848969, | |
| "learning_rate": 7.927758193823501e-05, | |
| "loss": 0.161, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.2528557169790396, | |
| "grad_norm": 0.15361559391021729, | |
| "learning_rate": 7.907135972167069e-05, | |
| "loss": 0.1775, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.25315145465971683, | |
| "grad_norm": 0.1672465205192566, | |
| "learning_rate": 7.88652305447549e-05, | |
| "loss": 0.1796, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.2534471923403941, | |
| "grad_norm": 0.1553036868572235, | |
| "learning_rate": 7.865919532384844e-05, | |
| "loss": 0.1647, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.2537429300210713, | |
| "grad_norm": 0.16316907107830048, | |
| "learning_rate": 7.845325497489449e-05, | |
| "loss": 0.1431, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.25403866770174854, | |
| "grad_norm": 0.1651161015033722, | |
| "learning_rate": 7.82474104134144e-05, | |
| "loss": 0.1587, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.2543344053824258, | |
| "grad_norm": 0.18129472434520721, | |
| "learning_rate": 7.804166255450373e-05, | |
| "loss": 0.1635, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.25463014306310305, | |
| "grad_norm": 0.18252769112586975, | |
| "learning_rate": 7.783601231282812e-05, | |
| "loss": 0.2157, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.25492588074378025, | |
| "grad_norm": 0.17854921519756317, | |
| "learning_rate": 7.763046060261932e-05, | |
| "loss": 0.1804, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.2552216184244575, | |
| "grad_norm": 0.1713121086359024, | |
| "learning_rate": 7.742500833767094e-05, | |
| "loss": 0.1704, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.25551735610513476, | |
| "grad_norm": 0.18181180953979492, | |
| "learning_rate": 7.721965643133458e-05, | |
| "loss": 0.192, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.25581309378581196, | |
| "grad_norm": 0.18746164441108704, | |
| "learning_rate": 7.701440579651566e-05, | |
| "loss": 0.2068, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.2561088314664892, | |
| "grad_norm": 0.19459117949008942, | |
| "learning_rate": 7.680925734566937e-05, | |
| "loss": 0.1758, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.2564045691471665, | |
| "grad_norm": 0.18460817635059357, | |
| "learning_rate": 7.660421199079669e-05, | |
| "loss": 0.1814, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.2567003068278437, | |
| "grad_norm": 0.20752295851707458, | |
| "learning_rate": 7.639927064344022e-05, | |
| "loss": 0.2329, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.2569960445085209, | |
| "grad_norm": 0.2058822363615036, | |
| "learning_rate": 7.619443421468021e-05, | |
| "loss": 0.2083, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.2572917821891982, | |
| "grad_norm": 0.2240809202194214, | |
| "learning_rate": 7.598970361513051e-05, | |
| "loss": 0.2273, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.25758751986987544, | |
| "grad_norm": 0.20064809918403625, | |
| "learning_rate": 7.578507975493448e-05, | |
| "loss": 0.1672, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.25788325755055264, | |
| "grad_norm": 0.2004321813583374, | |
| "learning_rate": 7.558056354376098e-05, | |
| "loss": 0.1797, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.2581789952312299, | |
| "grad_norm": 0.21047692000865936, | |
| "learning_rate": 7.537615589080027e-05, | |
| "loss": 0.1897, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.25847473291190715, | |
| "grad_norm": 0.2111281305551529, | |
| "learning_rate": 7.517185770476006e-05, | |
| "loss": 0.1866, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.2587704705925844, | |
| "grad_norm": 0.208242729306221, | |
| "learning_rate": 7.496766989386136e-05, | |
| "loss": 0.2176, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.2590662082732616, | |
| "grad_norm": 0.24142082035541534, | |
| "learning_rate": 7.476359336583454e-05, | |
| "loss": 0.2557, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.25936194595393885, | |
| "grad_norm": 0.19397041201591492, | |
| "learning_rate": 7.455962902791522e-05, | |
| "loss": 0.1439, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.2596576836346161, | |
| "grad_norm": 0.24474111199378967, | |
| "learning_rate": 7.435577778684033e-05, | |
| "loss": 0.2145, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.2599534213152933, | |
| "grad_norm": 0.2113002985715866, | |
| "learning_rate": 7.415204054884399e-05, | |
| "loss": 0.1806, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.26024915899597056, | |
| "grad_norm": 0.23416808247566223, | |
| "learning_rate": 7.394841821965345e-05, | |
| "loss": 0.2005, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.2605448966766478, | |
| "grad_norm": 0.21154069900512695, | |
| "learning_rate": 7.374491170448525e-05, | |
| "loss": 0.1623, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.2608406343573251, | |
| "grad_norm": 0.20221436023712158, | |
| "learning_rate": 7.3541521908041e-05, | |
| "loss": 0.1428, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.2611363720380023, | |
| "grad_norm": 0.24770595133304596, | |
| "learning_rate": 7.33382497345034e-05, | |
| "loss": 0.1745, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.26143210971867953, | |
| "grad_norm": 0.21654701232910156, | |
| "learning_rate": 7.313509608753231e-05, | |
| "loss": 0.1651, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.2617278473993568, | |
| "grad_norm": 0.21652589738368988, | |
| "learning_rate": 7.293206187026061e-05, | |
| "loss": 0.1332, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.262023585080034, | |
| "grad_norm": 0.29497137665748596, | |
| "learning_rate": 7.27291479852903e-05, | |
| "loss": 0.1639, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.26231932276071124, | |
| "grad_norm": 0.2427954226732254, | |
| "learning_rate": 7.252635533468843e-05, | |
| "loss": 0.1723, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.2626150604413885, | |
| "grad_norm": 0.2615757882595062, | |
| "learning_rate": 7.232368481998309e-05, | |
| "loss": 0.1891, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.26291079812206575, | |
| "grad_norm": 0.21479550004005432, | |
| "learning_rate": 7.212113734215932e-05, | |
| "loss": 0.1295, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.26320653580274295, | |
| "grad_norm": 0.29835671186447144, | |
| "learning_rate": 7.191871380165538e-05, | |
| "loss": 0.2239, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.2635022734834202, | |
| "grad_norm": 0.304277241230011, | |
| "learning_rate": 7.17164150983584e-05, | |
| "loss": 0.1899, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.26379801116409746, | |
| "grad_norm": 0.28030019998550415, | |
| "learning_rate": 7.151424213160061e-05, | |
| "loss": 0.1829, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.2640937488447747, | |
| "grad_norm": 0.26866456866264343, | |
| "learning_rate": 7.131219580015521e-05, | |
| "loss": 0.1586, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.2643894865254519, | |
| "grad_norm": 0.2647477984428406, | |
| "learning_rate": 7.11102770022325e-05, | |
| "loss": 0.1242, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.26468522420612917, | |
| "grad_norm": 0.27126333117485046, | |
| "learning_rate": 7.090848663547574e-05, | |
| "loss": 0.1559, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.2649809618868064, | |
| "grad_norm": 0.2967953681945801, | |
| "learning_rate": 7.070682559695736e-05, | |
| "loss": 0.1913, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.2652766995674836, | |
| "grad_norm": 0.3403211236000061, | |
| "learning_rate": 7.050529478317476e-05, | |
| "loss": 0.181, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.2655724372481609, | |
| "grad_norm": 0.3107486963272095, | |
| "learning_rate": 7.03038950900464e-05, | |
| "loss": 0.1367, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.26586817492883813, | |
| "grad_norm": 0.4827352464199066, | |
| "learning_rate": 7.010262741290798e-05, | |
| "loss": 0.2186, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.2661639126095154, | |
| "grad_norm": 0.5701566934585571, | |
| "learning_rate": 6.990149264650814e-05, | |
| "loss": 0.2813, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2664596502901926, | |
| "grad_norm": 0.10672613233327866, | |
| "learning_rate": 6.970049168500474e-05, | |
| "loss": 0.1362, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.26675538797086984, | |
| "grad_norm": 0.123167484998703, | |
| "learning_rate": 6.94996254219608e-05, | |
| "loss": 0.1512, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.2670511256515471, | |
| "grad_norm": 0.14451348781585693, | |
| "learning_rate": 6.929889475034048e-05, | |
| "loss": 0.2042, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.2673468633322243, | |
| "grad_norm": 0.15311995148658752, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.1774, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.26764260101290155, | |
| "grad_norm": 0.1696614921092987, | |
| "learning_rate": 6.889784375020978e-05, | |
| "loss": 0.2346, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.2679383386935788, | |
| "grad_norm": 0.158672496676445, | |
| "learning_rate": 6.869752520459803e-05, | |
| "loss": 0.1889, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.26823407637425606, | |
| "grad_norm": 0.1748230755329132, | |
| "learning_rate": 6.849734581619918e-05, | |
| "loss": 0.1925, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.26852981405493326, | |
| "grad_norm": 0.1830466389656067, | |
| "learning_rate": 6.829730647492404e-05, | |
| "loss": 0.2056, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.2688255517356105, | |
| "grad_norm": 0.17888867855072021, | |
| "learning_rate": 6.80974080700606e-05, | |
| "loss": 0.194, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.26912128941628777, | |
| "grad_norm": 0.1619090586900711, | |
| "learning_rate": 6.789765149027039e-05, | |
| "loss": 0.1591, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.26941702709696497, | |
| "grad_norm": 0.17818139493465424, | |
| "learning_rate": 6.769803762358443e-05, | |
| "loss": 0.2072, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.2697127647776422, | |
| "grad_norm": 0.17115117609500885, | |
| "learning_rate": 6.749856735739928e-05, | |
| "loss": 0.1528, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.2700085024583195, | |
| "grad_norm": 0.18800678849220276, | |
| "learning_rate": 6.729924157847323e-05, | |
| "loss": 0.2115, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.27030424013899673, | |
| "grad_norm": 0.19144469499588013, | |
| "learning_rate": 6.710006117292209e-05, | |
| "loss": 0.2246, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.27059997781967393, | |
| "grad_norm": 0.17849388718605042, | |
| "learning_rate": 6.690102702621548e-05, | |
| "loss": 0.1629, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.2708957155003512, | |
| "grad_norm": 0.18855208158493042, | |
| "learning_rate": 6.670214002317278e-05, | |
| "loss": 0.1774, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.27119145318102844, | |
| "grad_norm": 0.1926363706588745, | |
| "learning_rate": 6.650340104795932e-05, | |
| "loss": 0.201, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.27148719086170564, | |
| "grad_norm": 0.21725712716579437, | |
| "learning_rate": 6.630481098408228e-05, | |
| "loss": 0.2316, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.2717829285423829, | |
| "grad_norm": 0.2055988311767578, | |
| "learning_rate": 6.610637071438686e-05, | |
| "loss": 0.1953, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.27207866622306015, | |
| "grad_norm": 0.186083123087883, | |
| "learning_rate": 6.590808112105232e-05, | |
| "loss": 0.1441, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2723744039037374, | |
| "grad_norm": 0.2097381055355072, | |
| "learning_rate": 6.570994308558812e-05, | |
| "loss": 0.2146, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.2726701415844146, | |
| "grad_norm": 0.18082082271575928, | |
| "learning_rate": 6.551195748882997e-05, | |
| "loss": 0.1919, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.27296587926509186, | |
| "grad_norm": 0.19363375008106232, | |
| "learning_rate": 6.531412521093586e-05, | |
| "loss": 0.1879, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.2732616169457691, | |
| "grad_norm": 0.2286785989999771, | |
| "learning_rate": 6.51164471313822e-05, | |
| "loss": 0.2411, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.27355735462644637, | |
| "grad_norm": 0.21459421515464783, | |
| "learning_rate": 6.491892412895995e-05, | |
| "loss": 0.1957, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.27385309230712357, | |
| "grad_norm": 0.18594589829444885, | |
| "learning_rate": 6.472155708177052e-05, | |
| "loss": 0.1569, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.2741488299878008, | |
| "grad_norm": 0.20744723081588745, | |
| "learning_rate": 6.452434686722224e-05, | |
| "loss": 0.1756, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.2744445676684781, | |
| "grad_norm": 0.2012493759393692, | |
| "learning_rate": 6.432729436202604e-05, | |
| "loss": 0.1642, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.2747403053491553, | |
| "grad_norm": 0.2502315938472748, | |
| "learning_rate": 6.41304004421918e-05, | |
| "loss": 0.2156, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.27503604302983253, | |
| "grad_norm": 0.22481419146060944, | |
| "learning_rate": 6.393366598302446e-05, | |
| "loss": 0.1745, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.2753317807105098, | |
| "grad_norm": 0.21398960053920746, | |
| "learning_rate": 6.373709185911998e-05, | |
| "loss": 0.1837, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.27562751839118704, | |
| "grad_norm": 0.2531358301639557, | |
| "learning_rate": 6.354067894436155e-05, | |
| "loss": 0.2479, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.27592325607186424, | |
| "grad_norm": 0.2133578509092331, | |
| "learning_rate": 6.334442811191576e-05, | |
| "loss": 0.1834, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.2762189937525415, | |
| "grad_norm": 0.24796530604362488, | |
| "learning_rate": 6.314834023422858e-05, | |
| "loss": 0.1974, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.27651473143321875, | |
| "grad_norm": 0.2332259565591812, | |
| "learning_rate": 6.295241618302156e-05, | |
| "loss": 0.1923, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.27681046911389595, | |
| "grad_norm": 0.221559077501297, | |
| "learning_rate": 6.275665682928803e-05, | |
| "loss": 0.154, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.2771062067945732, | |
| "grad_norm": 0.24034371972084045, | |
| "learning_rate": 6.256106304328905e-05, | |
| "loss": 0.187, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.27740194447525046, | |
| "grad_norm": 0.25638851523399353, | |
| "learning_rate": 6.23656356945497e-05, | |
| "loss": 0.1833, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.2776976821559277, | |
| "grad_norm": 0.25576722621917725, | |
| "learning_rate": 6.21703756518551e-05, | |
| "loss": 0.1888, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.2779934198366049, | |
| "grad_norm": 0.2722214162349701, | |
| "learning_rate": 6.197528378324665e-05, | |
| "loss": 0.1869, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.27828915751728217, | |
| "grad_norm": 0.2651073634624481, | |
| "learning_rate": 6.17803609560181e-05, | |
| "loss": 0.1871, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.2785848951979594, | |
| "grad_norm": 0.24269293248653412, | |
| "learning_rate": 6.158560803671168e-05, | |
| "loss": 0.157, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.2788806328786366, | |
| "grad_norm": 0.24029861390590668, | |
| "learning_rate": 6.139102589111435e-05, | |
| "loss": 0.1162, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.2791763705593139, | |
| "grad_norm": 0.30216261744499207, | |
| "learning_rate": 6.119661538425381e-05, | |
| "loss": 0.1928, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.27947210823999113, | |
| "grad_norm": 0.29744547605514526, | |
| "learning_rate": 6.100237738039484e-05, | |
| "loss": 0.195, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.2797678459206684, | |
| "grad_norm": 0.3087689280509949, | |
| "learning_rate": 6.0808312743035236e-05, | |
| "loss": 0.2018, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.2800635836013456, | |
| "grad_norm": 0.30813395977020264, | |
| "learning_rate": 6.061442233490211e-05, | |
| "loss": 0.1712, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.28035932128202284, | |
| "grad_norm": 0.2932427227497101, | |
| "learning_rate": 6.042070701794806e-05, | |
| "loss": 0.1527, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.2806550589627001, | |
| "grad_norm": 0.3830631673336029, | |
| "learning_rate": 6.0227167653347305e-05, | |
| "loss": 0.1975, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.2809507966433773, | |
| "grad_norm": 0.5851044654846191, | |
| "learning_rate": 6.0033805101491794e-05, | |
| "loss": 0.283, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.28124653432405455, | |
| "grad_norm": 0.12433775514364243, | |
| "learning_rate": 5.98406202219875e-05, | |
| "loss": 0.176, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.2815422720047318, | |
| "grad_norm": 0.13331016898155212, | |
| "learning_rate": 5.964761387365052e-05, | |
| "loss": 0.1551, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.28183800968540906, | |
| "grad_norm": 0.15550710260868073, | |
| "learning_rate": 5.9454786914503255e-05, | |
| "loss": 0.1739, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.28213374736608626, | |
| "grad_norm": 0.1473347395658493, | |
| "learning_rate": 5.926214020177074e-05, | |
| "loss": 0.1637, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.2824294850467635, | |
| "grad_norm": 0.17270837724208832, | |
| "learning_rate": 5.9069674591876534e-05, | |
| "loss": 0.1727, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.2827252227274408, | |
| "grad_norm": 0.14710725843906403, | |
| "learning_rate": 5.887739094043923e-05, | |
| "loss": 0.1256, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.28302096040811797, | |
| "grad_norm": 0.1642570048570633, | |
| "learning_rate": 5.868529010226845e-05, | |
| "loss": 0.1671, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.2833166980887952, | |
| "grad_norm": 0.1761314868927002, | |
| "learning_rate": 5.849337293136112e-05, | |
| "loss": 0.1987, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.2836124357694725, | |
| "grad_norm": 0.15199066698551178, | |
| "learning_rate": 5.830164028089766e-05, | |
| "loss": 0.1716, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.28390817345014974, | |
| "grad_norm": 0.18991252779960632, | |
| "learning_rate": 5.811009300323818e-05, | |
| "loss": 0.2306, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.28420391113082694, | |
| "grad_norm": 0.18644598126411438, | |
| "learning_rate": 5.791873194991872e-05, | |
| "loss": 0.2177, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.2844996488115042, | |
| "grad_norm": 0.17034782469272614, | |
| "learning_rate": 5.7727557971647427e-05, | |
| "loss": 0.1614, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.28479538649218145, | |
| "grad_norm": 0.17145271599292755, | |
| "learning_rate": 5.7536571918300864e-05, | |
| "loss": 0.1853, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.2850911241728587, | |
| "grad_norm": 0.20625253021717072, | |
| "learning_rate": 5.734577463892008e-05, | |
| "loss": 0.2303, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.2853868618535359, | |
| "grad_norm": 0.17384910583496094, | |
| "learning_rate": 5.7155166981706956e-05, | |
| "loss": 0.1725, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.28568259953421316, | |
| "grad_norm": 0.1765337586402893, | |
| "learning_rate": 5.6964749794020354e-05, | |
| "loss": 0.1495, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.2859783372148904, | |
| "grad_norm": 0.22681960463523865, | |
| "learning_rate": 5.6774523922372394e-05, | |
| "loss": 0.2106, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.2862740748955676, | |
| "grad_norm": 0.21203817427158356, | |
| "learning_rate": 5.6584490212424804e-05, | |
| "loss": 0.2547, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.28656981257624486, | |
| "grad_norm": 0.1752469837665558, | |
| "learning_rate": 5.639464950898491e-05, | |
| "loss": 0.1504, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.2868655502569221, | |
| "grad_norm": 0.18426285684108734, | |
| "learning_rate": 5.620500265600206e-05, | |
| "loss": 0.1627, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.2871612879375994, | |
| "grad_norm": 0.20485304296016693, | |
| "learning_rate": 5.601555049656382e-05, | |
| "loss": 0.2002, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.2874570256182766, | |
| "grad_norm": 0.21289728581905365, | |
| "learning_rate": 5.58262938728922e-05, | |
| "loss": 0.2021, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.28775276329895383, | |
| "grad_norm": 0.20441678166389465, | |
| "learning_rate": 5.563723362634008e-05, | |
| "loss": 0.1861, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.2880485009796311, | |
| "grad_norm": 0.22460561990737915, | |
| "learning_rate": 5.544837059738719e-05, | |
| "loss": 0.2161, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.2883442386603083, | |
| "grad_norm": 0.2056421935558319, | |
| "learning_rate": 5.525970562563656e-05, | |
| "loss": 0.1906, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.28863997634098554, | |
| "grad_norm": 0.22834764420986176, | |
| "learning_rate": 5.507123954981073e-05, | |
| "loss": 0.1751, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.2889357140216628, | |
| "grad_norm": 0.19000335037708282, | |
| "learning_rate": 5.488297320774807e-05, | |
| "loss": 0.1728, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.28923145170234005, | |
| "grad_norm": 0.209343820810318, | |
| "learning_rate": 5.4694907436399e-05, | |
| "loss": 0.1645, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.28952718938301725, | |
| "grad_norm": 0.2307262420654297, | |
| "learning_rate": 5.4507043071822284e-05, | |
| "loss": 0.2139, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.2898229270636945, | |
| "grad_norm": 0.2119540274143219, | |
| "learning_rate": 5.431938094918132e-05, | |
| "loss": 0.1719, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.29011866474437176, | |
| "grad_norm": 0.24757258594036102, | |
| "learning_rate": 5.41319219027404e-05, | |
| "loss": 0.2469, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.29041440242504896, | |
| "grad_norm": 0.23471477627754211, | |
| "learning_rate": 5.394466676586114e-05, | |
| "loss": 0.1832, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.2907101401057262, | |
| "grad_norm": 0.22288459539413452, | |
| "learning_rate": 5.375761637099854e-05, | |
| "loss": 0.1564, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.29100587778640347, | |
| "grad_norm": 0.22171498835086823, | |
| "learning_rate": 5.357077154969742e-05, | |
| "loss": 0.1845, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.2913016154670807, | |
| "grad_norm": 0.2288697361946106, | |
| "learning_rate": 5.3384133132588784e-05, | |
| "loss": 0.1793, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.2915973531477579, | |
| "grad_norm": 0.30814412236213684, | |
| "learning_rate": 5.3197701949386e-05, | |
| "loss": 0.2074, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.2918930908284352, | |
| "grad_norm": 0.22758153080940247, | |
| "learning_rate": 5.301147882888116e-05, | |
| "loss": 0.1599, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.29218882850911243, | |
| "grad_norm": 0.23016326129436493, | |
| "learning_rate": 5.28254645989414e-05, | |
| "loss": 0.1528, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.29248456618978963, | |
| "grad_norm": 0.2873390316963196, | |
| "learning_rate": 5.2639660086505226e-05, | |
| "loss": 0.2101, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.2927803038704669, | |
| "grad_norm": 0.2666926681995392, | |
| "learning_rate": 5.2454066117578815e-05, | |
| "loss": 0.1964, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.29307604155114414, | |
| "grad_norm": 0.26695266366004944, | |
| "learning_rate": 5.226868351723244e-05, | |
| "loss": 0.1716, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.2933717792318214, | |
| "grad_norm": 0.3416760265827179, | |
| "learning_rate": 5.2083513109596616e-05, | |
| "loss": 0.2446, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.2936675169124986, | |
| "grad_norm": 0.2576669156551361, | |
| "learning_rate": 5.189855571785859e-05, | |
| "loss": 0.111, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.29396325459317585, | |
| "grad_norm": 0.2883356213569641, | |
| "learning_rate": 5.171381216425863e-05, | |
| "loss": 0.1702, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.2942589922738531, | |
| "grad_norm": 0.3241150379180908, | |
| "learning_rate": 5.152928327008635e-05, | |
| "loss": 0.2342, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.29455472995453036, | |
| "grad_norm": 0.3817709684371948, | |
| "learning_rate": 5.134496985567714e-05, | |
| "loss": 0.2512, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.29485046763520756, | |
| "grad_norm": 0.34549957513809204, | |
| "learning_rate": 5.116087274040837e-05, | |
| "loss": 0.2041, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.2951462053158848, | |
| "grad_norm": 0.39972737431526184, | |
| "learning_rate": 5.0976992742695925e-05, | |
| "loss": 0.2526, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.29544194299656207, | |
| "grad_norm": 0.4173913598060608, | |
| "learning_rate": 5.07933306799904e-05, | |
| "loss": 0.2467, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.29573768067723927, | |
| "grad_norm": 0.58672034740448, | |
| "learning_rate": 5.060988736877366e-05, | |
| "loss": 0.2481, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2960334183579165, | |
| "grad_norm": 0.10767298191785812, | |
| "learning_rate": 5.042666362455498e-05, | |
| "loss": 0.1372, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.2963291560385938, | |
| "grad_norm": 0.14959672093391418, | |
| "learning_rate": 5.024366026186755e-05, | |
| "loss": 0.2057, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.29662489371927103, | |
| "grad_norm": 0.14518089592456818, | |
| "learning_rate": 5.006087809426496e-05, | |
| "loss": 0.1744, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.29692063139994823, | |
| "grad_norm": 0.16306771337985992, | |
| "learning_rate": 4.987831793431731e-05, | |
| "loss": 0.1904, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.2972163690806255, | |
| "grad_norm": 0.18029610812664032, | |
| "learning_rate": 4.9695980593607817e-05, | |
| "loss": 0.2223, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.29751210676130274, | |
| "grad_norm": 0.18586061894893646, | |
| "learning_rate": 4.9513866882729146e-05, | |
| "loss": 0.2306, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.29780784444197994, | |
| "grad_norm": 0.17091114819049835, | |
| "learning_rate": 4.9331977611279777e-05, | |
| "loss": 0.1717, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.2981035821226572, | |
| "grad_norm": 0.15413694083690643, | |
| "learning_rate": 4.9150313587860433e-05, | |
| "loss": 0.1638, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.29839931980333445, | |
| "grad_norm": 0.15122149884700775, | |
| "learning_rate": 4.896887562007054e-05, | |
| "loss": 0.1489, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.2986950574840117, | |
| "grad_norm": 0.16240714490413666, | |
| "learning_rate": 4.8787664514504504e-05, | |
| "loss": 0.1718, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.2989907951646889, | |
| "grad_norm": 0.17468984425067902, | |
| "learning_rate": 4.860668107674823e-05, | |
| "loss": 0.1577, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.29928653284536616, | |
| "grad_norm": 0.19555851817131042, | |
| "learning_rate": 4.8425926111375506e-05, | |
| "loss": 0.2611, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.2995822705260434, | |
| "grad_norm": 0.20301313698291779, | |
| "learning_rate": 4.824540042194443e-05, | |
| "loss": 0.2266, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.2998780082067206, | |
| "grad_norm": 0.1904042363166809, | |
| "learning_rate": 4.8065104810993856e-05, | |
| "loss": 0.1847, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.30017374588739787, | |
| "grad_norm": 0.18135389685630798, | |
| "learning_rate": 4.788504008003978e-05, | |
| "loss": 0.1712, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.3004694835680751, | |
| "grad_norm": 0.2035091519355774, | |
| "learning_rate": 4.770520702957182e-05, | |
| "loss": 0.2459, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.3007652212487524, | |
| "grad_norm": 0.1759434938430786, | |
| "learning_rate": 4.752560645904962e-05, | |
| "loss": 0.1494, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.3010609589294296, | |
| "grad_norm": 0.21353545784950256, | |
| "learning_rate": 4.734623916689941e-05, | |
| "loss": 0.2468, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.30135669661010683, | |
| "grad_norm": 0.19691097736358643, | |
| "learning_rate": 4.716710595051022e-05, | |
| "loss": 0.1913, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.3016524342907841, | |
| "grad_norm": 0.20307397842407227, | |
| "learning_rate": 4.698820760623064e-05, | |
| "loss": 0.2042, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.3019481719714613, | |
| "grad_norm": 0.1996135711669922, | |
| "learning_rate": 4.6809544929365004e-05, | |
| "loss": 0.1734, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.30224390965213854, | |
| "grad_norm": 0.1997922658920288, | |
| "learning_rate": 4.663111871417e-05, | |
| "loss": 0.1601, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.3025396473328158, | |
| "grad_norm": 0.2047778218984604, | |
| "learning_rate": 4.645292975385111e-05, | |
| "loss": 0.1823, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.30283538501349305, | |
| "grad_norm": 0.21905602514743805, | |
| "learning_rate": 4.627497884055912e-05, | |
| "loss": 0.1889, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.30313112269417025, | |
| "grad_norm": 0.22655650973320007, | |
| "learning_rate": 4.609726676538652e-05, | |
| "loss": 0.204, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.3034268603748475, | |
| "grad_norm": 0.2180885523557663, | |
| "learning_rate": 4.591979431836402e-05, | |
| "loss": 0.2277, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.30372259805552476, | |
| "grad_norm": 0.2357274740934372, | |
| "learning_rate": 4.574256228845706e-05, | |
| "loss": 0.1818, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.30401833573620196, | |
| "grad_norm": 0.24533364176750183, | |
| "learning_rate": 4.5565571463562365e-05, | |
| "loss": 0.209, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.3043140734168792, | |
| "grad_norm": 0.22792395949363708, | |
| "learning_rate": 4.5388822630504256e-05, | |
| "loss": 0.2153, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.30460981109755647, | |
| "grad_norm": 0.2345706820487976, | |
| "learning_rate": 4.521231657503132e-05, | |
| "loss": 0.1802, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.3049055487782337, | |
| "grad_norm": 0.23721234500408173, | |
| "learning_rate": 4.503605408181286e-05, | |
| "loss": 0.2189, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.3052012864589109, | |
| "grad_norm": 0.24193932116031647, | |
| "learning_rate": 4.486003593443537e-05, | |
| "loss": 0.2247, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.3054970241395882, | |
| "grad_norm": 0.24470211565494537, | |
| "learning_rate": 4.468426291539914e-05, | |
| "loss": 0.1972, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.30579276182026544, | |
| "grad_norm": 0.22956788539886475, | |
| "learning_rate": 4.4508735806114654e-05, | |
| "loss": 0.1631, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.3060884995009427, | |
| "grad_norm": 0.2537655532360077, | |
| "learning_rate": 4.433345538689929e-05, | |
| "loss": 0.1794, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.3063842371816199, | |
| "grad_norm": 0.25624555349349976, | |
| "learning_rate": 4.415842243697369e-05, | |
| "loss": 0.2303, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.30667997486229714, | |
| "grad_norm": 0.2514943480491638, | |
| "learning_rate": 4.39836377344583e-05, | |
| "loss": 0.2273, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.3069757125429744, | |
| "grad_norm": 0.28939127922058105, | |
| "learning_rate": 4.380910205637012e-05, | |
| "loss": 0.2312, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.3072714502236516, | |
| "grad_norm": 0.26121294498443604, | |
| "learning_rate": 4.363481617861893e-05, | |
| "loss": 0.1898, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.30756718790432885, | |
| "grad_norm": 0.295492023229599, | |
| "learning_rate": 4.346078087600412e-05, | |
| "loss": 0.2393, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.3078629255850061, | |
| "grad_norm": 0.2641611099243164, | |
| "learning_rate": 4.3286996922211034e-05, | |
| "loss": 0.183, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.30815866326568336, | |
| "grad_norm": 0.2948877513408661, | |
| "learning_rate": 4.311346508980772e-05, | |
| "loss": 0.2088, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.30845440094636056, | |
| "grad_norm": 0.28656068444252014, | |
| "learning_rate": 4.2940186150241365e-05, | |
| "loss": 0.2029, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.3087501386270378, | |
| "grad_norm": 0.2820681929588318, | |
| "learning_rate": 4.27671608738349e-05, | |
| "loss": 0.1992, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.3090458763077151, | |
| "grad_norm": 0.34954750537872314, | |
| "learning_rate": 4.2594390029783534e-05, | |
| "loss": 0.186, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.30934161398839227, | |
| "grad_norm": 0.3151059150695801, | |
| "learning_rate": 4.242187438615153e-05, | |
| "loss": 0.1745, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.3096373516690695, | |
| "grad_norm": 0.3306736350059509, | |
| "learning_rate": 4.224961470986849e-05, | |
| "loss": 0.1497, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.3099330893497468, | |
| "grad_norm": 0.3289806842803955, | |
| "learning_rate": 4.207761176672614e-05, | |
| "loss": 0.1829, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.31022882703042404, | |
| "grad_norm": 0.363712340593338, | |
| "learning_rate": 4.190586632137491e-05, | |
| "loss": 0.2085, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.31052456471110124, | |
| "grad_norm": 0.4890635311603546, | |
| "learning_rate": 4.173437913732048e-05, | |
| "loss": 0.2581, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3108203023917785, | |
| "grad_norm": 0.10311943292617798, | |
| "learning_rate": 4.156315097692037e-05, | |
| "loss": 0.1254, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.31111604007245575, | |
| "grad_norm": 0.12436298280954361, | |
| "learning_rate": 4.139218260138074e-05, | |
| "loss": 0.1637, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.31141177775313295, | |
| "grad_norm": 0.1472722887992859, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.2009, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.3117075154338102, | |
| "grad_norm": 0.1530962437391281, | |
| "learning_rate": 4.1051028243929125e-05, | |
| "loss": 0.1974, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.31200325311448746, | |
| "grad_norm": 0.15057165920734406, | |
| "learning_rate": 4.088084377864135e-05, | |
| "loss": 0.1616, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.3122989907951647, | |
| "grad_norm": 0.17087000608444214, | |
| "learning_rate": 4.07109221314556e-05, | |
| "loss": 0.1957, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.3125947284758419, | |
| "grad_norm": 0.17251181602478027, | |
| "learning_rate": 4.054126405776971e-05, | |
| "loss": 0.1952, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.31289046615651916, | |
| "grad_norm": 0.17809037864208221, | |
| "learning_rate": 4.037187031180985e-05, | |
| "loss": 0.22, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.3131862038371964, | |
| "grad_norm": 0.17678867280483246, | |
| "learning_rate": 4.020274164662707e-05, | |
| "loss": 0.1695, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.3134819415178736, | |
| "grad_norm": 0.17473168671131134, | |
| "learning_rate": 4.003387881409397e-05, | |
| "loss": 0.1947, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.3137776791985509, | |
| "grad_norm": 0.16512729227542877, | |
| "learning_rate": 3.986528256490141e-05, | |
| "loss": 0.1541, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.31407341687922813, | |
| "grad_norm": 0.1978088766336441, | |
| "learning_rate": 3.969695364855511e-05, | |
| "loss": 0.2243, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.3143691545599054, | |
| "grad_norm": 0.2224223017692566, | |
| "learning_rate": 3.952889281337235e-05, | |
| "loss": 0.2512, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.3146648922405826, | |
| "grad_norm": 0.19392409920692444, | |
| "learning_rate": 3.93611008064786e-05, | |
| "loss": 0.1894, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.31496062992125984, | |
| "grad_norm": 0.20266327261924744, | |
| "learning_rate": 3.9193578373804364e-05, | |
| "loss": 0.2205, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.3152563676019371, | |
| "grad_norm": 0.1855594664812088, | |
| "learning_rate": 3.90263262600816e-05, | |
| "loss": 0.189, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.31555210528261435, | |
| "grad_norm": 0.1787533313035965, | |
| "learning_rate": 3.88593452088406e-05, | |
| "loss": 0.1558, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.31584784296329155, | |
| "grad_norm": 0.18954086303710938, | |
| "learning_rate": 3.869263596240661e-05, | |
| "loss": 0.1614, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.3161435806439688, | |
| "grad_norm": 0.1829981952905655, | |
| "learning_rate": 3.8526199261896544e-05, | |
| "loss": 0.1646, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.31643931832464606, | |
| "grad_norm": 0.1618681699037552, | |
| "learning_rate": 3.836003584721577e-05, | |
| "loss": 0.1343, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.31673505600532326, | |
| "grad_norm": 0.1964295357465744, | |
| "learning_rate": 3.8194146457054655e-05, | |
| "loss": 0.1592, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.3170307936860005, | |
| "grad_norm": 0.2202250063419342, | |
| "learning_rate": 3.802853182888543e-05, | |
| "loss": 0.1966, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.31732653136667777, | |
| "grad_norm": 0.18789294362068176, | |
| "learning_rate": 3.786319269895877e-05, | |
| "loss": 0.1788, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.317622269047355, | |
| "grad_norm": 0.21173900365829468, | |
| "learning_rate": 3.769812980230074e-05, | |
| "loss": 0.1972, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.3179180067280322, | |
| "grad_norm": 0.19276420772075653, | |
| "learning_rate": 3.7533343872709294e-05, | |
| "loss": 0.1512, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.3182137444087095, | |
| "grad_norm": 0.20392374694347382, | |
| "learning_rate": 3.736883564275112e-05, | |
| "loss": 0.17, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.31850948208938673, | |
| "grad_norm": 0.2147340625524521, | |
| "learning_rate": 3.7204605843758386e-05, | |
| "loss": 0.1964, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.31880521977006393, | |
| "grad_norm": 0.2067640870809555, | |
| "learning_rate": 3.704065520582549e-05, | |
| "loss": 0.1778, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.3191009574507412, | |
| "grad_norm": 0.2175614833831787, | |
| "learning_rate": 3.6876984457805786e-05, | |
| "loss": 0.1743, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.31939669513141844, | |
| "grad_norm": 0.24874760210514069, | |
| "learning_rate": 3.671359432730834e-05, | |
| "loss": 0.2407, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.3196924328120957, | |
| "grad_norm": 0.2241019457578659, | |
| "learning_rate": 3.655048554069478e-05, | |
| "loss": 0.1543, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.3199881704927729, | |
| "grad_norm": 0.21091891825199127, | |
| "learning_rate": 3.638765882307589e-05, | |
| "loss": 0.1665, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.32028390817345015, | |
| "grad_norm": 0.2465236932039261, | |
| "learning_rate": 3.6225114898308634e-05, | |
| "loss": 0.1946, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.3205796458541274, | |
| "grad_norm": 0.21476277709007263, | |
| "learning_rate": 3.6062854488992714e-05, | |
| "loss": 0.1506, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.3208753835348046, | |
| "grad_norm": 0.21354220807552338, | |
| "learning_rate": 3.5900878316467454e-05, | |
| "loss": 0.1596, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.32117112121548186, | |
| "grad_norm": 0.22661133110523224, | |
| "learning_rate": 3.573918710080857e-05, | |
| "loss": 0.1826, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.3214668588961591, | |
| "grad_norm": 0.27455446124076843, | |
| "learning_rate": 3.5577781560825066e-05, | |
| "loss": 0.2736, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.32176259657683637, | |
| "grad_norm": 0.2345777302980423, | |
| "learning_rate": 3.541666241405588e-05, | |
| "loss": 0.1832, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.32205833425751357, | |
| "grad_norm": 0.2570793628692627, | |
| "learning_rate": 3.5255830376766764e-05, | |
| "loss": 0.1677, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.3223540719381908, | |
| "grad_norm": 0.2677852213382721, | |
| "learning_rate": 3.509528616394716e-05, | |
| "loss": 0.1459, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.3226498096188681, | |
| "grad_norm": 0.264311283826828, | |
| "learning_rate": 3.4935030489306883e-05, | |
| "loss": 0.153, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.3229455472995453, | |
| "grad_norm": 0.23239494860172272, | |
| "learning_rate": 3.4775064065273165e-05, | |
| "loss": 0.1069, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.32324128498022253, | |
| "grad_norm": 0.2711178660392761, | |
| "learning_rate": 3.4615387602987236e-05, | |
| "loss": 0.1556, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.3235370226608998, | |
| "grad_norm": 0.2906439006328583, | |
| "learning_rate": 3.445600181230134e-05, | |
| "loss": 0.1422, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.32383276034157704, | |
| "grad_norm": 0.29046574234962463, | |
| "learning_rate": 3.429690740177549e-05, | |
| "loss": 0.1434, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.32412849802225424, | |
| "grad_norm": 0.3531874418258667, | |
| "learning_rate": 3.413810507867436e-05, | |
| "loss": 0.208, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.3244242357029315, | |
| "grad_norm": 0.409625381231308, | |
| "learning_rate": 3.397959554896415e-05, | |
| "loss": 0.3359, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.32471997338360875, | |
| "grad_norm": 0.3413349390029907, | |
| "learning_rate": 3.3821379517309405e-05, | |
| "loss": 0.1659, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.325015711064286, | |
| "grad_norm": 0.40019068121910095, | |
| "learning_rate": 3.3663457687069924e-05, | |
| "loss": 0.1975, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.3253114487449632, | |
| "grad_norm": 0.44681063294410706, | |
| "learning_rate": 3.350583076029754e-05, | |
| "loss": 0.2325, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.32560718642564046, | |
| "grad_norm": 0.10489065200090408, | |
| "learning_rate": 3.334849943773323e-05, | |
| "loss": 0.1379, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.3259029241063177, | |
| "grad_norm": 0.13087670505046844, | |
| "learning_rate": 3.319146441880371e-05, | |
| "loss": 0.1632, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.3261986617869949, | |
| "grad_norm": 0.14324639737606049, | |
| "learning_rate": 3.3034726401618444e-05, | |
| "loss": 0.1943, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.32649439946767217, | |
| "grad_norm": 0.13718920946121216, | |
| "learning_rate": 3.28782860829667e-05, | |
| "loss": 0.1442, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.3267901371483494, | |
| "grad_norm": 0.17169281840324402, | |
| "learning_rate": 3.272214415831418e-05, | |
| "loss": 0.1917, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.3270858748290267, | |
| "grad_norm": 0.15564316511154175, | |
| "learning_rate": 3.2566301321800085e-05, | |
| "loss": 0.1558, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.3273816125097039, | |
| "grad_norm": 0.1527709662914276, | |
| "learning_rate": 3.241075826623401e-05, | |
| "loss": 0.2016, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.32767735019038113, | |
| "grad_norm": 0.18245702981948853, | |
| "learning_rate": 3.225551568309284e-05, | |
| "loss": 0.1997, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.3279730878710584, | |
| "grad_norm": 0.14350612461566925, | |
| "learning_rate": 3.210057426251773e-05, | |
| "loss": 0.1301, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.3282688255517356, | |
| "grad_norm": 0.1906651109457016, | |
| "learning_rate": 3.1945934693310896e-05, | |
| "loss": 0.1982, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.32856456323241284, | |
| "grad_norm": 0.21341390907764435, | |
| "learning_rate": 3.179159766293282e-05, | |
| "loss": 0.2742, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.3288603009130901, | |
| "grad_norm": 0.18190768361091614, | |
| "learning_rate": 3.163756385749889e-05, | |
| "loss": 0.1947, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.32915603859376735, | |
| "grad_norm": 0.1745702177286148, | |
| "learning_rate": 3.148383396177653e-05, | |
| "loss": 0.171, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.32945177627444455, | |
| "grad_norm": 0.18562626838684082, | |
| "learning_rate": 3.133040865918213e-05, | |
| "loss": 0.1952, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.3297475139551218, | |
| "grad_norm": 0.18852019309997559, | |
| "learning_rate": 3.117728863177796e-05, | |
| "loss": 0.1815, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.33004325163579906, | |
| "grad_norm": 0.194679394364357, | |
| "learning_rate": 3.102447456026919e-05, | |
| "loss": 0.1826, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.33033898931647626, | |
| "grad_norm": 0.21376194059848785, | |
| "learning_rate": 3.0871967124000834e-05, | |
| "loss": 0.2233, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.3306347269971535, | |
| "grad_norm": 0.2095218151807785, | |
| "learning_rate": 3.0719767000954714e-05, | |
| "loss": 0.1997, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.33093046467783077, | |
| "grad_norm": 0.19964182376861572, | |
| "learning_rate": 3.056787486774656e-05, | |
| "loss": 0.1715, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.331226202358508, | |
| "grad_norm": 0.19055882096290588, | |
| "learning_rate": 3.041629139962283e-05, | |
| "loss": 0.1629, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.3315219400391852, | |
| "grad_norm": 0.17358700931072235, | |
| "learning_rate": 3.0265017270457775e-05, | |
| "loss": 0.11, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.3318176777198625, | |
| "grad_norm": 0.1570686399936676, | |
| "learning_rate": 3.0114053152750556e-05, | |
| "loss": 0.1093, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.33211341540053974, | |
| "grad_norm": 0.2210582047700882, | |
| "learning_rate": 2.9963399717622077e-05, | |
| "loss": 0.1951, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.33240915308121693, | |
| "grad_norm": 0.19304311275482178, | |
| "learning_rate": 2.98130576348121e-05, | |
| "loss": 0.1603, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.3327048907618942, | |
| "grad_norm": 0.23825351893901825, | |
| "learning_rate": 2.966302757267625e-05, | |
| "loss": 0.2385, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.3327048907618942, | |
| "eval_loss": 0.18209633231163025, | |
| "eval_runtime": 323.5083, | |
| "eval_samples_per_second": 17.604, | |
| "eval_steps_per_second": 8.803, | |
| "step": 1125 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 375, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4886246305696317e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |