Invalid JSON:Unexpected token 'N', ..."ad_norm": NaN,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8888888888888888, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0035555555555555557, | |
| "grad_norm": 0.31538352370262146, | |
| "learning_rate": 1.5833333333333333e-05, | |
| "loss": 1.297884178161621, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0071111111111111115, | |
| "grad_norm": 0.31121939420700073, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 1.3258800506591797, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.010666666666666666, | |
| "grad_norm": 0.28440341353416443, | |
| "learning_rate": 4.9166666666666665e-05, | |
| "loss": 1.3672664642333985, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.014222222222222223, | |
| "grad_norm": 0.282014399766922, | |
| "learning_rate": 6.583333333333334e-05, | |
| "loss": 1.2544504165649415, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 0.26958519220352173, | |
| "learning_rate": 8.25e-05, | |
| "loss": 1.2959243774414062, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.021333333333333333, | |
| "grad_norm": 0.30045831203460693, | |
| "learning_rate": 9.916666666666667e-05, | |
| "loss": 1.2867345809936523, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.024888888888888887, | |
| "grad_norm": 0.28056198358535767, | |
| "learning_rate": 9.999706080734013e-05, | |
| "loss": 1.2609113693237304, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.028444444444444446, | |
| "grad_norm": 0.25958094000816345, | |
| "learning_rate": 9.998761670000927e-05, | |
| "loss": 1.2782093048095704, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 0.2954287528991699, | |
| "learning_rate": 9.997166076639238e-05, | |
| "loss": 1.2719867706298829, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 0.27726614475250244, | |
| "learning_rate": 9.994919508504958e-05, | |
| "loss": 1.3072052001953125, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03911111111111111, | |
| "grad_norm": 0.28370192646980286, | |
| "learning_rate": 9.9920222582558e-05, | |
| "loss": 1.2819059371948243, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.042666666666666665, | |
| "grad_norm": 0.27542224526405334, | |
| "learning_rate": 9.988474703313044e-05, | |
| "loss": 1.2524637222290038, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.04622222222222222, | |
| "grad_norm": 0.2960389256477356, | |
| "learning_rate": 9.984277305812375e-05, | |
| "loss": 1.3067936897277832, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.049777777777777775, | |
| "grad_norm": 0.2984329164028168, | |
| "learning_rate": 9.979430612543684e-05, | |
| "loss": 1.258444118499756, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 0.27751636505126953, | |
| "learning_rate": 9.973935254879832e-05, | |
| "loss": 1.2734871864318849, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05688888888888889, | |
| "grad_norm": 0.2834205627441406, | |
| "learning_rate": 9.967791948694405e-05, | |
| "loss": 1.3134132385253907, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.060444444444444446, | |
| "grad_norm": 0.2821517586708069, | |
| "learning_rate": 9.96100149426846e-05, | |
| "loss": 1.2569908142089843, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 0.284457266330719, | |
| "learning_rate": 9.953564776186273e-05, | |
| "loss": 1.2625727653503418, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.06755555555555555, | |
| "grad_norm": 0.2937893569469452, | |
| "learning_rate": 9.9454827632201e-05, | |
| "loss": 1.338636589050293, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 0.3055609464645386, | |
| "learning_rate": 9.936756508203982e-05, | |
| "loss": 1.3083559989929199, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07466666666666667, | |
| "grad_norm": 0.2724494934082031, | |
| "learning_rate": 9.927387147896589e-05, | |
| "loss": 1.3003810882568358, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.07822222222222222, | |
| "grad_norm": 0.26778730750083923, | |
| "learning_rate": 9.917375902833146e-05, | |
| "loss": 1.285653018951416, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.08177777777777778, | |
| "grad_norm": 0.28135982155799866, | |
| "learning_rate": 9.906724077166418e-05, | |
| "loss": 1.308763599395752, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.08533333333333333, | |
| "grad_norm": 0.32560646533966064, | |
| "learning_rate": 9.895433058496835e-05, | |
| "loss": 1.2624423027038574, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 0.27822455763816833, | |
| "learning_rate": 9.883504317691725e-05, | |
| "loss": 1.2910588264465332, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09244444444444444, | |
| "grad_norm": 0.27096736431121826, | |
| "learning_rate": 9.870939408693707e-05, | |
| "loss": 1.3021769523620605, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 0.277794748544693, | |
| "learning_rate": 9.857739968318253e-05, | |
| "loss": 1.3017699241638183, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.09955555555555555, | |
| "grad_norm": 0.3811008632183075, | |
| "learning_rate": 9.843907716040481e-05, | |
| "loss": 1.3006772994995117, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.10311111111111111, | |
| "grad_norm": 0.2823365330696106, | |
| "learning_rate": 9.829444453771137e-05, | |
| "loss": 1.3110570907592773, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 0.29975101351737976, | |
| "learning_rate": 9.814352065621885e-05, | |
| "loss": 1.2665305137634277, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.11022222222222222, | |
| "grad_norm": 0.28466543555259705, | |
| "learning_rate": 9.798632517659846e-05, | |
| "loss": 1.2834184646606446, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.11377777777777778, | |
| "grad_norm": 0.2719884514808655, | |
| "learning_rate": 9.782287857651498e-05, | |
| "loss": 1.3005459785461426, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.11733333333333333, | |
| "grad_norm": 0.30884769558906555, | |
| "learning_rate": 9.765320214795901e-05, | |
| "loss": 1.3151857376098632, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.12088888888888889, | |
| "grad_norm": 0.28063106536865234, | |
| "learning_rate": 9.747731799447343e-05, | |
| "loss": 1.3190623283386231, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 0.30677250027656555, | |
| "learning_rate": 9.72952490282739e-05, | |
| "loss": 1.3142582893371582, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.28910356760025024, | |
| "learning_rate": 9.710701896726411e-05, | |
| "loss": 1.2787860870361327, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.13155555555555556, | |
| "grad_norm": 0.2909628748893738, | |
| "learning_rate": 9.691265233194615e-05, | |
| "loss": 1.2816056251525878, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.1351111111111111, | |
| "grad_norm": 0.27913960814476013, | |
| "learning_rate": 9.67121744422262e-05, | |
| "loss": 1.2716171264648437, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.13866666666666666, | |
| "grad_norm": 0.3015570342540741, | |
| "learning_rate": 9.650561141411615e-05, | |
| "loss": 1.3473497390747071, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 0.282624214887619, | |
| "learning_rate": 9.629299015633151e-05, | |
| "loss": 1.3115547180175782, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.14577777777777778, | |
| "grad_norm": 0.28449124097824097, | |
| "learning_rate": 9.607433836678598e-05, | |
| "loss": 1.2522377014160155, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.14933333333333335, | |
| "grad_norm": 0.30534765124320984, | |
| "learning_rate": 9.584968452898342e-05, | |
| "loss": 1.3076560974121094, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.15288888888888888, | |
| "grad_norm": 0.3203866183757782, | |
| "learning_rate": 9.561905790830716e-05, | |
| "loss": 1.2956053733825683, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.15644444444444444, | |
| "grad_norm": 0.2609277665615082, | |
| "learning_rate": 9.538248854820778e-05, | |
| "loss": 1.2799407958984375, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.2910199761390686, | |
| "learning_rate": 9.514000726628928e-05, | |
| "loss": 1.2915283203125, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.16355555555555557, | |
| "grad_norm": 0.2923009395599365, | |
| "learning_rate": 9.489164565029458e-05, | |
| "loss": 1.2836386680603027, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.1671111111111111, | |
| "grad_norm": 0.301127165555954, | |
| "learning_rate": 9.463743605399063e-05, | |
| "loss": 1.3104005813598634, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.17066666666666666, | |
| "grad_norm": 0.2957231402397156, | |
| "learning_rate": 9.437741159295363e-05, | |
| "loss": 1.266425323486328, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.17422222222222222, | |
| "grad_norm": 0.2988809049129486, | |
| "learning_rate": 9.411160614025521e-05, | |
| "loss": 1.2832436561584473, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.29408636689186096, | |
| "learning_rate": 9.384005432204974e-05, | |
| "loss": 1.3093104362487793, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "eval_loss": 1.2944837808609009, | |
| "eval_runtime": 2889.2194, | |
| "eval_samples_per_second": 1.731, | |
| "eval_steps_per_second": 1.731, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18133333333333335, | |
| "grad_norm": 0.2915937602519989, | |
| "learning_rate": 9.356279151306375e-05, | |
| "loss": 1.3089990615844727, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.18488888888888888, | |
| "grad_norm": 0.2755884528160095, | |
| "learning_rate": 9.327985383198751e-05, | |
| "loss": 1.274155044555664, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.18844444444444444, | |
| "grad_norm": 0.29075562953948975, | |
| "learning_rate": 9.299127813677009e-05, | |
| "loss": 1.320622444152832, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.26899510622024536, | |
| "learning_rate": 9.269710201981779e-05, | |
| "loss": 1.3017861366271972, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 0.31122830510139465, | |
| "learning_rate": 9.239736380309707e-05, | |
| "loss": 1.3248827934265137, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1991111111111111, | |
| "grad_norm": 0.29395928978919983, | |
| "learning_rate": 9.209210253314243e-05, | |
| "loss": 1.3025559425354003, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.20266666666666666, | |
| "grad_norm": 0.2973605990409851, | |
| "learning_rate": 9.178135797596978e-05, | |
| "loss": 1.3009254455566406, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.20622222222222222, | |
| "grad_norm": 0.3183933198451996, | |
| "learning_rate": 9.146517061189624e-05, | |
| "loss": 1.316084098815918, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.20977777777777779, | |
| "grad_norm": 0.3000907897949219, | |
| "learning_rate": 9.114358163026677e-05, | |
| "loss": 1.3586406707763672, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 0.27883148193359375, | |
| "learning_rate": 9.081663292408851e-05, | |
| "loss": 1.2584849357604981, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.21688888888888888, | |
| "grad_norm": 0.3001721501350403, | |
| "learning_rate": 9.048436708457347e-05, | |
| "loss": 1.314585018157959, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.22044444444444444, | |
| "grad_norm": 0.32649344205856323, | |
| "learning_rate": 9.014682739559013e-05, | |
| "loss": 1.2489622116088868, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 0.28912609815597534, | |
| "learning_rate": 8.980405782802497e-05, | |
| "loss": 1.2546935081481934, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.22755555555555557, | |
| "grad_norm": 0.32853057980537415, | |
| "learning_rate": 8.945610303405441e-05, | |
| "loss": 1.2884913444519044, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 0.2821575701236725, | |
| "learning_rate": 8.910300834132807e-05, | |
| "loss": 1.2917396545410156, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.23466666666666666, | |
| "grad_norm": 0.31761616468429565, | |
| "learning_rate": 8.874481974706389e-05, | |
| "loss": 1.292926025390625, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.23822222222222222, | |
| "grad_norm": 0.283069908618927, | |
| "learning_rate": 8.838158391205626e-05, | |
| "loss": 1.271681022644043, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.24177777777777779, | |
| "grad_norm": 0.2957214415073395, | |
| "learning_rate": 8.801334815459748e-05, | |
| "loss": 1.2740598678588868, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.24533333333333332, | |
| "grad_norm": 0.282450407743454, | |
| "learning_rate": 8.764016044431367e-05, | |
| "loss": 1.3171983718872071, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 0.28648215532302856, | |
| "learning_rate": 8.726206939591589e-05, | |
| "loss": 1.3046676635742187, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.25244444444444447, | |
| "grad_norm": 0.28918009996414185, | |
| "learning_rate": 8.687912426286708e-05, | |
| "loss": 1.298368263244629, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.2839282155036926, | |
| "learning_rate": 8.649137493096595e-05, | |
| "loss": 1.289079761505127, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.25955555555555554, | |
| "grad_norm": 0.3144363760948181, | |
| "learning_rate": 8.609887191184833e-05, | |
| "loss": 1.2597214698791503, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.26311111111111113, | |
| "grad_norm": 0.3045108914375305, | |
| "learning_rate": 8.570166633640716e-05, | |
| "loss": 1.2796483039855957, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.2583145201206207, | |
| "learning_rate": 8.529980994813171e-05, | |
| "loss": 1.2942163467407226, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2702222222222222, | |
| "grad_norm": 0.3217857778072357, | |
| "learning_rate": 8.489335509636693e-05, | |
| "loss": 1.317829418182373, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.2737777777777778, | |
| "grad_norm": 0.28687581419944763, | |
| "learning_rate": 8.448235472949413e-05, | |
| "loss": 1.294327449798584, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.2773333333333333, | |
| "grad_norm": 0.2994561791419983, | |
| "learning_rate": 8.40668623880333e-05, | |
| "loss": 1.2718279838562012, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.2808888888888889, | |
| "grad_norm": 0.29878973960876465, | |
| "learning_rate": 8.364693219766858e-05, | |
| "loss": 1.2716640472412108, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 0.2995190918445587, | |
| "learning_rate": 8.322261886219723e-05, | |
| "loss": 1.2568064689636231, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.3141971826553345, | |
| "learning_rate": 8.27939776564036e-05, | |
| "loss": 1.2757036209106445, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.29155555555555557, | |
| "grad_norm": 0.2948372960090637, | |
| "learning_rate": 8.236106441885844e-05, | |
| "loss": 1.2733377456665038, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.2951111111111111, | |
| "grad_norm": 0.3033512234687805, | |
| "learning_rate": 8.192393554464485e-05, | |
| "loss": 1.2756839752197267, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.2986666666666667, | |
| "grad_norm": 0.31936532258987427, | |
| "learning_rate": 8.14826479780119e-05, | |
| "loss": 1.3197004318237304, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 0.27577313780784607, | |
| "learning_rate": 8.10372592049564e-05, | |
| "loss": 1.2805109977722169, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.30577777777777776, | |
| "grad_norm": 0.29502543807029724, | |
| "learning_rate": 8.05878272457344e-05, | |
| "loss": 1.2883221626281738, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.30933333333333335, | |
| "grad_norm": 0.29035353660583496, | |
| "learning_rate": 8.013441064730293e-05, | |
| "loss": 1.3223690032958983, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.3128888888888889, | |
| "grad_norm": 0.3090946674346924, | |
| "learning_rate": 7.96770684756931e-05, | |
| "loss": 1.3194878578186036, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.3164444444444444, | |
| "grad_norm": 0.30516624450683594, | |
| "learning_rate": 7.921586030831571e-05, | |
| "loss": 1.3071480751037599, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.29030412435531616, | |
| "learning_rate": 7.875084622620016e-05, | |
| "loss": 1.3239538192749023, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.32355555555555554, | |
| "grad_norm": 0.29187288880348206, | |
| "learning_rate": 7.82820868061677e-05, | |
| "loss": 1.3000661849975585, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.32711111111111113, | |
| "grad_norm": 0.31834492087364197, | |
| "learning_rate": 7.780964311294028e-05, | |
| "loss": 1.2818709373474122, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.33066666666666666, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.733357669118561e-05, | |
| "loss": 1.3621267318725585, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.3342222222222222, | |
| "grad_norm": 0.31786996126174927, | |
| "learning_rate": 7.685394955749987e-05, | |
| "loss": 1.297238826751709, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 0.318629652261734, | |
| "learning_rate": 7.637082419232889e-05, | |
| "loss": 1.3202865600585938, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.3413333333333333, | |
| "grad_norm": 0.2845008671283722, | |
| "learning_rate": 7.588426353182885e-05, | |
| "loss": 1.3001307487487792, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.3448888888888889, | |
| "grad_norm": 0.33042728900909424, | |
| "learning_rate": 7.539433095966773e-05, | |
| "loss": 1.3215258598327637, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.34844444444444445, | |
| "grad_norm": 0.313645601272583, | |
| "learning_rate": 7.490109029876829e-05, | |
| "loss": 1.2881192207336425, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.30073919892311096, | |
| "learning_rate": 7.440460580299409e-05, | |
| "loss": 1.3003701210021972, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.29454582929611206, | |
| "learning_rate": 7.390494214877905e-05, | |
| "loss": 1.2804080963134765, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "eval_loss": NaN, | |
| "eval_runtime": 2840.2001, | |
| "eval_samples_per_second": 1.76, | |
| "eval_steps_per_second": 1.76, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3591111111111111, | |
| "grad_norm": 0.31132370233535767, | |
| "learning_rate": 7.340216442670229e-05, | |
| "loss": 1.2707467079162598, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.3626666666666667, | |
| "grad_norm": 0.29832929372787476, | |
| "learning_rate": 7.289633813300871e-05, | |
| "loss": 1.2995878219604493, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.3662222222222222, | |
| "grad_norm": 0.28491711616516113, | |
| "learning_rate": 7.238752916107696e-05, | |
| "loss": 1.2775651931762695, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.36977777777777776, | |
| "grad_norm": 0.29927778244018555, | |
| "learning_rate": 7.187580379283561e-05, | |
| "loss": 1.265756607055664, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 0.2922050952911377, | |
| "learning_rate": 7.136122869012858e-05, | |
| "loss": 1.2872845649719238, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.3768888888888889, | |
| "grad_norm": 0.3260011672973633, | |
| "learning_rate": 7.08438708860313e-05, | |
| "loss": 1.2806350708007812, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.3804444444444444, | |
| "grad_norm": 0.2833918631076813, | |
| "learning_rate": 7.032379777611833e-05, | |
| "loss": 1.3034561157226563, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.2914232313632965, | |
| "learning_rate": 6.98010771096838e-05, | |
| "loss": 1.3088299751281738, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.38755555555555554, | |
| "grad_norm": 0.28914394974708557, | |
| "learning_rate": 6.927577698091586e-05, | |
| "loss": 1.3301461219787598, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 0.33827468752861023, | |
| "learning_rate": 6.87479658200261e-05, | |
| "loss": 1.3212644577026367, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.39466666666666667, | |
| "grad_norm": 0.30936911702156067, | |
| "learning_rate": 6.821771238433516e-05, | |
| "loss": 1.342518138885498, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.3982222222222222, | |
| "grad_norm": 0.31862136721611023, | |
| "learning_rate": 6.768508574931592e-05, | |
| "loss": 1.2782557487487793, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.4017777777777778, | |
| "grad_norm": 0.3259771466255188, | |
| "learning_rate": 6.715015529959503e-05, | |
| "loss": 1.2817959785461426, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.4053333333333333, | |
| "grad_norm": 0.2901138961315155, | |
| "learning_rate": 6.661299071991434e-05, | |
| "loss": 1.3025086402893067, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 0.33082666993141174, | |
| "learning_rate": 6.607366198605305e-05, | |
| "loss": 1.3207198143005372, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.41244444444444445, | |
| "grad_norm": 0.24227800965309143, | |
| "learning_rate": 6.553223935571211e-05, | |
| "loss": 1.2632332801818849, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.2865552008152008, | |
| "learning_rate": 6.498879335936186e-05, | |
| "loss": 1.3484232902526856, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.41955555555555557, | |
| "grad_norm": 0.29671794176101685, | |
| "learning_rate": 6.444339479105406e-05, | |
| "loss": 1.2677297592163086, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.4231111111111111, | |
| "grad_norm": 0.3012571632862091, | |
| "learning_rate": 6.389611469919965e-05, | |
| "loss": 1.267910099029541, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.2888393700122833, | |
| "learning_rate": 6.334702437731336e-05, | |
| "loss": 1.26025447845459, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.43022222222222223, | |
| "grad_norm": 0.28179189562797546, | |
| "learning_rate": 6.27961953547264e-05, | |
| "loss": 1.2963777542114259, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.43377777777777776, | |
| "grad_norm": 0.2751828134059906, | |
| "learning_rate": 6.224369938726846e-05, | |
| "loss": 1.2986252784729004, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.43733333333333335, | |
| "grad_norm": 0.30413609743118286, | |
| "learning_rate": 6.168960844792002e-05, | |
| "loss": 1.2818167686462403, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.4408888888888889, | |
| "grad_norm": 0.29557210206985474, | |
| "learning_rate": 6.113399471743669e-05, | |
| "loss": 1.2678013801574708, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.2809475362300873, | |
| "learning_rate": 6.057693057494625e-05, | |
| "loss": 1.3133273124694824, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.29118746519088745, | |
| "learning_rate": 6.001848858851983e-05, | |
| "loss": 1.3000117301940919, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.45155555555555554, | |
| "grad_norm": 0.2959536910057068, | |
| "learning_rate": 5.9458741505718676e-05, | |
| "loss": 1.2923731803894043, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.45511111111111113, | |
| "grad_norm": 0.26924166083335876, | |
| "learning_rate": 5.889776224411733e-05, | |
| "loss": 1.3184330940246582, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.45866666666666667, | |
| "grad_norm": 0.28626126050949097, | |
| "learning_rate": 5.83356238818048e-05, | |
| "loss": 1.2908548355102538, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.777239964786468e-05, | |
| "loss": 1.3559614181518556, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.4657777777777778, | |
| "grad_norm": 0.3147507607936859, | |
| "learning_rate": 5.720816291283584e-05, | |
| "loss": 1.2960132598876952, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.4693333333333333, | |
| "grad_norm": 0.3098738193511963, | |
| "learning_rate": 5.664298717915434e-05, | |
| "loss": 1.306876277923584, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.4728888888888889, | |
| "grad_norm": 0.33010831475257874, | |
| "learning_rate": 5.607694607157854e-05, | |
| "loss": 1.3001605987548828, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.47644444444444445, | |
| "grad_norm": 0.3010432720184326, | |
| "learning_rate": 5.551011332759796e-05, | |
| "loss": 1.2979639053344727, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.2937035858631134, | |
| "learning_rate": 5.4942562787827656e-05, | |
| "loss": 1.2940792083740233, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.48355555555555557, | |
| "grad_norm": 0.326015442609787, | |
| "learning_rate": 5.4374368386389066e-05, | |
| "loss": 1.2941905021667481, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.4871111111111111, | |
| "grad_norm": 0.3160575032234192, | |
| "learning_rate": 5.380560414127871e-05, | |
| "loss": 1.2940587043762206, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.49066666666666664, | |
| "grad_norm": 0.29464566707611084, | |
| "learning_rate": 5.323634414472594e-05, | |
| "loss": 1.2751301765441894, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.49422222222222223, | |
| "grad_norm": 0.3078067898750305, | |
| "learning_rate": 5.266666255354107e-05, | |
| "loss": 1.2807303428649903, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 0.28077632188796997, | |
| "learning_rate": 5.2096633579455025e-05, | |
| "loss": 1.2641204833984374, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.5013333333333333, | |
| "grad_norm": 0.300456702709198, | |
| "learning_rate": 5.1526331479451886e-05, | |
| "loss": 1.2676878929138184, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.5048888888888889, | |
| "grad_norm": 0.3112626373767853, | |
| "learning_rate": 5.09558305460955e-05, | |
| "loss": 1.299578857421875, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.5084444444444445, | |
| "grad_norm": 0.3101966977119446, | |
| "learning_rate": 5.038520509785151e-05, | |
| "loss": 1.2908551216125488, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.3191506862640381, | |
| "learning_rate": 4.9814529469405956e-05, | |
| "loss": 1.327328872680664, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 0.28775665163993835, | |
| "learning_rate": 4.924387800198178e-05, | |
| "loss": 1.3289175987243653, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.5191111111111111, | |
| "grad_norm": 0.28782543540000916, | |
| "learning_rate": 4.8673325033654546e-05, | |
| "loss": 1.2873039245605469, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.5226666666666666, | |
| "grad_norm": 0.3069140911102295, | |
| "learning_rate": 4.810294488966839e-05, | |
| "loss": 1.3148150444030762, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.5262222222222223, | |
| "grad_norm": 0.3069920837879181, | |
| "learning_rate": 4.753281187275388e-05, | |
| "loss": 1.3092363357543946, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.5297777777777778, | |
| "grad_norm": 0.29772457480430603, | |
| "learning_rate": 4.696300025344862e-05, | |
| "loss": 1.2847169876098632, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.33965858817100525, | |
| "learning_rate": 4.6393584260422155e-05, | |
| "loss": 1.2921103477478026, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "eval_loss": 1.294020652770996, | |
| "eval_runtime": 3801.2842, | |
| "eval_samples_per_second": 1.315, | |
| "eval_steps_per_second": 1.315, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5368888888888889, | |
| "grad_norm": 0.2884041666984558, | |
| "learning_rate": 4.582463807080623e-05, | |
| "loss": 1.2746758460998535, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.5404444444444444, | |
| "grad_norm": 0.3089088201522827, | |
| "learning_rate": 4.5256235800531934e-05, | |
| "loss": 1.276390266418457, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.29441338777542114, | |
| "learning_rate": 4.468845149467457e-05, | |
| "loss": 1.2559406280517578, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.5475555555555556, | |
| "grad_norm": 0.30819985270500183, | |
| "learning_rate": 4.4121359117808034e-05, | |
| "loss": 1.3224390029907227, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 0.2932080626487732, | |
| "learning_rate": 4.355503254436943e-05, | |
| "loss": 1.2737662315368652, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.5546666666666666, | |
| "grad_norm": 0.29400452971458435, | |
| "learning_rate": 4.298954554903558e-05, | |
| "loss": 1.2591508865356444, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.5582222222222222, | |
| "grad_norm": 0.29733583331108093, | |
| "learning_rate": 4.242497179711249e-05, | |
| "loss": 1.3279075622558594, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.5617777777777778, | |
| "grad_norm": 0.3186342418193817, | |
| "learning_rate": 4.1861384834939046e-05, | |
| "loss": 1.3201321601867675, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.5653333333333334, | |
| "grad_norm": 0.3540149927139282, | |
| "learning_rate": 4.129885808030625e-05, | |
| "loss": 1.3074044227600097, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 0.28959041833877563, | |
| "learning_rate": 4.07374648128931e-05, | |
| "loss": 1.2987191200256347, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.5724444444444444, | |
| "grad_norm": 0.28861257433891296, | |
| "learning_rate": 4.017727816472067e-05, | |
| "loss": 1.288304901123047, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.29670071601867676, | |
| "learning_rate": 3.961837111062511e-05, | |
| "loss": 1.2712695121765136, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.5795555555555556, | |
| "grad_norm": 0.27656856179237366, | |
| "learning_rate": 3.90608164587515e-05, | |
| "loss": 1.2713235855102538, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.5831111111111111, | |
| "grad_norm": 0.2939528226852417, | |
| "learning_rate": 3.8504686841069035e-05, | |
| "loss": 1.296719741821289, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 0.29121825098991394, | |
| "learning_rate": 3.795005470390946e-05, | |
| "loss": 1.2787345886230468, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.5902222222222222, | |
| "grad_norm": 0.2960679233074188, | |
| "learning_rate": 3.739699229852957e-05, | |
| "loss": 1.237907600402832, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.5937777777777777, | |
| "grad_norm": 0.2961101233959198, | |
| "learning_rate": 3.6845571671699085e-05, | |
| "loss": 1.2625959396362305, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.5973333333333334, | |
| "grad_norm": 0.30379098653793335, | |
| "learning_rate": 3.629586465631519e-05, | |
| "loss": 1.280212116241455, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.6008888888888889, | |
| "grad_norm": 0.2864052653312683, | |
| "learning_rate": 3.574794286204507e-05, | |
| "loss": 1.2951812744140625, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 0.2769223749637604, | |
| "learning_rate": 3.520187766599719e-05, | |
| "loss": 1.275238037109375, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.2863476574420929, | |
| "learning_rate": 3.4657740203423346e-05, | |
| "loss": 1.2970293045043946, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.6115555555555555, | |
| "grad_norm": 0.3228801488876343, | |
| "learning_rate": 3.411560135845172e-05, | |
| "loss": 1.2682680130004882, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.6151111111111112, | |
| "grad_norm": 0.31453782320022583, | |
| "learning_rate": 3.3575531754853085e-05, | |
| "loss": 1.319148635864258, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.6186666666666667, | |
| "grad_norm": 0.3097997009754181, | |
| "learning_rate": 3.303760174684058e-05, | |
| "loss": 1.2992384910583497, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.3050277531147003, | |
| "learning_rate": 3.2501881409904847e-05, | |
| "loss": 1.2543633460998536, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6257777777777778, | |
| "grad_norm": 0.3190324604511261, | |
| "learning_rate": 3.1968440531685366e-05, | |
| "loss": 1.2791267395019532, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.6293333333333333, | |
| "grad_norm": 0.3011898398399353, | |
| "learning_rate": 3.143734860287928e-05, | |
| "loss": 1.2916112899780274, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.6328888888888888, | |
| "grad_norm": 0.2937512993812561, | |
| "learning_rate": 3.090867480818891e-05, | |
| "loss": 1.294296646118164, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.6364444444444445, | |
| "grad_norm": 0.31571245193481445, | |
| "learning_rate": 3.038248801730922e-05, | |
| "loss": 1.278724765777588, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.2860446572303772, | |
| "learning_rate": 2.9858856775956155e-05, | |
| "loss": 1.3208050727844238, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.6435555555555555, | |
| "grad_norm": 0.30255869030952454, | |
| "learning_rate": 2.9337849296937374e-05, | |
| "loss": 1.2606579780578613, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.6471111111111111, | |
| "grad_norm": 0.2926134169101715, | |
| "learning_rate": 2.8819533451266124e-05, | |
| "loss": 1.2770811080932618, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.6506666666666666, | |
| "grad_norm": 0.30504968762397766, | |
| "learning_rate": 2.830397675931991e-05, | |
| "loss": 1.2516894340515137, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.6542222222222223, | |
| "grad_norm": 0.3088359534740448, | |
| "learning_rate": 2.779124638204455e-05, | |
| "loss": 1.3035507202148438, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 0.3114239275455475, | |
| "learning_rate": 2.7281409112205324e-05, | |
| "loss": 1.3040590286254883, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.6613333333333333, | |
| "grad_norm": 0.29068320989608765, | |
| "learning_rate": 2.6774531365685916e-05, | |
| "loss": 1.3151864051818847, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.6648888888888889, | |
| "grad_norm": 0.29378166794776917, | |
| "learning_rate": 2.6270679172836454e-05, | |
| "loss": 1.2733539581298827, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.6684444444444444, | |
| "grad_norm": 0.28618156909942627, | |
| "learning_rate": 2.5769918169871852e-05, | |
| "loss": 1.2936923027038574, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.3030966818332672, | |
| "learning_rate": 2.5272313590321517e-05, | |
| "loss": 1.2442930221557618, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 0.3269818127155304, | |
| "learning_rate": 2.4777930256531335e-05, | |
| "loss": 1.255967617034912, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.6791111111111111, | |
| "grad_norm": 0.27831149101257324, | |
| "learning_rate": 2.4286832571219426e-05, | |
| "loss": 1.2981350898742676, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.6826666666666666, | |
| "grad_norm": 0.3000192940235138, | |
| "learning_rate": 2.379908450908651e-05, | |
| "loss": 1.3234101295471192, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.6862222222222222, | |
| "grad_norm": 0.2968016564846039, | |
| "learning_rate": 2.331474960848188e-05, | |
| "loss": 1.2768921852111816, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.6897777777777778, | |
| "grad_norm": 0.30178749561309814, | |
| "learning_rate": 2.283389096312648e-05, | |
| "loss": 1.26753568649292, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 0.286243736743927, | |
| "learning_rate": 2.2356571213893616e-05, | |
| "loss": 1.2724492073059082, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.6968888888888889, | |
| "grad_norm": 0.2989359200000763, | |
| "learning_rate": 2.1882852540648972e-05, | |
| "loss": 1.269171905517578, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.7004444444444444, | |
| "grad_norm": 0.30486729741096497, | |
| "learning_rate": 2.1412796654150337e-05, | |
| "loss": 1.267188549041748, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.2730155885219574, | |
| "learning_rate": 2.0946464788008773e-05, | |
| "loss": 1.2734877586364746, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.7075555555555556, | |
| "grad_norm": 0.295740008354187, | |
| "learning_rate": 2.0483917690711674e-05, | |
| "loss": 1.3122530937194825, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.3011320233345032, | |
| "learning_rate": 2.0025215617709226e-05, | |
| "loss": 1.2803524017333985, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "eval_loss": NaN, | |
| "eval_runtime": 3243.1447, | |
| "eval_samples_per_second": 1.542, | |
| "eval_steps_per_second": 1.542, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7146666666666667, | |
| "grad_norm": 0.2972392439842224, | |
| "learning_rate": 1.95704183235649e-05, | |
| "loss": 1.2797088623046875, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.7182222222222222, | |
| "grad_norm": 0.296792596578598, | |
| "learning_rate": 1.9119585054171414e-05, | |
| "loss": 1.2741374969482422, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.7217777777777777, | |
| "grad_norm": 0.290913850069046, | |
| "learning_rate": 1.8672774539032727e-05, | |
| "loss": 1.3360743522644043, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.7253333333333334, | |
| "grad_norm": 0.29572999477386475, | |
| "learning_rate": 1.8230044983613537e-05, | |
| "loss": 1.2794650077819825, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 0.32962796092033386, | |
| "learning_rate": 1.7791454061756784e-05, | |
| "loss": 1.3176474571228027, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.7324444444444445, | |
| "grad_norm": 0.291505366563797, | |
| "learning_rate": 1.7357058908170703e-05, | |
| "loss": 1.2892326354980468, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.29348224401474, | |
| "learning_rate": 1.6926916110985835e-05, | |
| "loss": 1.3315817832946777, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.7395555555555555, | |
| "grad_norm": 0.28669995069503784, | |
| "learning_rate": 1.6501081704383402e-05, | |
| "loss": 1.3001320838928223, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.7431111111111111, | |
| "grad_norm": 0.28902319073677063, | |
| "learning_rate": 1.6079611161295793e-05, | |
| "loss": 1.2577714920043945, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 0.29345715045928955, | |
| "learning_rate": 1.5662559386180243e-05, | |
| "loss": 1.2727604866027833, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.7502222222222222, | |
| "grad_norm": 0.2846791446208954, | |
| "learning_rate": 1.5249980707866352e-05, | |
| "loss": 1.332273006439209, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.7537777777777778, | |
| "grad_norm": 0.2810782492160797, | |
| "learning_rate": 1.4841928872478894e-05, | |
| "loss": 1.293062114715576, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.7573333333333333, | |
| "grad_norm": 0.3269287645816803, | |
| "learning_rate": 1.4438457036436227e-05, | |
| "loss": 1.2866110801696777, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.7608888888888888, | |
| "grad_norm": 0.3067913055419922, | |
| "learning_rate": 1.4039617759525792e-05, | |
| "loss": 1.2902913093566895, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 0.3035895824432373, | |
| "learning_rate": 1.3645462998057168e-05, | |
| "loss": 1.2704421043395997, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.30175936222076416, | |
| "learning_rate": 1.3256044098093717e-05, | |
| "loss": 1.3288631439208984, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.7715555555555556, | |
| "grad_norm": 0.2766067385673523, | |
| "learning_rate": 1.2871411788763937e-05, | |
| "loss": 1.2526258468627929, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.7751111111111111, | |
| "grad_norm": 0.2870315909385681, | |
| "learning_rate": 1.2491616175652898e-05, | |
| "loss": 1.3256899833679199, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.7786666666666666, | |
| "grad_norm": 0.287153959274292, | |
| "learning_rate": 1.2116706734275163e-05, | |
| "loss": 1.32543888092041, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 0.2858220636844635, | |
| "learning_rate": 1.1746732303629571e-05, | |
| "loss": 1.2777803421020508, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.7857777777777778, | |
| "grad_norm": 0.269578754901886, | |
| "learning_rate": 1.1381741079837138e-05, | |
| "loss": 1.2386326789855957, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.7893333333333333, | |
| "grad_norm": 0.28385013341903687, | |
| "learning_rate": 1.1021780609862515e-05, | |
| "loss": 1.2936376571655273, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.7928888888888889, | |
| "grad_norm": 0.3042142689228058, | |
| "learning_rate": 1.0666897785320185e-05, | |
| "loss": 1.2731390953063966, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.7964444444444444, | |
| "grad_norm": 0.3042246103286743, | |
| "learning_rate": 1.0317138836365886e-05, | |
| "loss": 1.280541229248047, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.3210007846355438, | |
| "learning_rate": 9.972549325674286e-06, | |
| "loss": 1.2910484313964843, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8035555555555556, | |
| "grad_norm": 0.3148663640022278, | |
| "learning_rate": 9.633174142503615e-06, | |
| "loss": 1.2725947380065918, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.8071111111111111, | |
| "grad_norm": 0.33494219183921814, | |
| "learning_rate": 9.299057496847952e-06, | |
| "loss": 1.2815696716308593, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.8106666666666666, | |
| "grad_norm": 0.29393038153648376, | |
| "learning_rate": 8.970242913678051e-06, | |
| "loss": 1.2812414169311523, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.8142222222222222, | |
| "grad_norm": 0.318511039018631, | |
| "learning_rate": 8.646773227271455e-06, | |
| "loss": 1.3076900482177733, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 0.28812065720558167, | |
| "learning_rate": 8.32869057563242e-06, | |
| "loss": 1.256617546081543, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.8213333333333334, | |
| "grad_norm": 0.2959403097629547, | |
| "learning_rate": 8.016036395002762e-06, | |
| "loss": 1.277919387817383, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.8248888888888889, | |
| "grad_norm": 0.2908424139022827, | |
| "learning_rate": 7.708851414463896e-06, | |
| "loss": 1.3131205558776855, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.8284444444444444, | |
| "grad_norm": 0.30460283160209656, | |
| "learning_rate": 7.407175650631231e-06, | |
| "loss": 1.3334476470947265, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.3278179168701172, | |
| "learning_rate": 7.111048402441123e-06, | |
| "loss": 1.3268613815307617, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 0.29847270250320435, | |
| "learning_rate": 6.82050824603157e-06, | |
| "loss": 1.278648853302002, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.8391111111111111, | |
| "grad_norm": 0.28696680068969727, | |
| "learning_rate": 6.535593029716846e-06, | |
| "loss": 1.2773630142211914, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.8426666666666667, | |
| "grad_norm": 0.3610166907310486, | |
| "learning_rate": 6.256339869057126e-06, | |
| "loss": 1.2533862113952636, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.8462222222222222, | |
| "grad_norm": 0.3022354245185852, | |
| "learning_rate": 5.982785142023422e-06, | |
| "loss": 1.2814712524414062, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.8497777777777777, | |
| "grad_norm": 0.31899362802505493, | |
| "learning_rate": 5.714964484258728e-06, | |
| "loss": 1.2975558280944823, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.28837141394615173, | |
| "learning_rate": 5.45291278443576e-06, | |
| "loss": 1.2620952606201172, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.8568888888888889, | |
| "grad_norm": 0.25447338819503784, | |
| "learning_rate": 5.19666417971208e-06, | |
| "loss": 1.2767372131347656, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.8604444444444445, | |
| "grad_norm": 0.2946532666683197, | |
| "learning_rate": 4.946252051283057e-06, | |
| "loss": 1.29830961227417, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.3278992474079132, | |
| "learning_rate": 4.701709020033412e-06, | |
| "loss": 1.2811607360839843, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.8675555555555555, | |
| "grad_norm": 0.31154683232307434, | |
| "learning_rate": 4.463066942287631e-06, | |
| "loss": 1.2984890937805176, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 0.270233690738678, | |
| "learning_rate": 4.230356905660149e-06, | |
| "loss": 1.254354190826416, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.8746666666666667, | |
| "grad_norm": 0.29535987973213196, | |
| "learning_rate": 4.003609225005617e-06, | |
| "loss": 1.2947762489318848, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.8782222222222222, | |
| "grad_norm": 0.2988264560699463, | |
| "learning_rate": 3.7828534384697523e-06, | |
| "loss": 1.279798984527588, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.8817777777777778, | |
| "grad_norm": 0.31089723110198975, | |
| "learning_rate": 3.568118303641521e-06, | |
| "loss": 1.262180233001709, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.8853333333333333, | |
| "grad_norm": 0.29605379700660706, | |
| "learning_rate": 3.359431793806844e-06, | |
| "loss": 1.3392036437988282, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.30679917335510254, | |
| "learning_rate": 3.156821094304635e-06, | |
| "loss": 1.3280159950256347, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "eval_loss": 1.288467526435852, | |
| "eval_runtime": 3649.1261, | |
| "eval_samples_per_second": 1.37, | |
| "eval_steps_per_second": 1.37, | |
| "step": 5000 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 5625, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.4749510844416e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |