{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 12652, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 13.192614390744945, "learning_rate": 4.3780025284450067e-07, "loss": 0.6539, "step": 20 }, { "epoch": 0.01, "grad_norm": 10.26668883317148, "learning_rate": 6.881163084702909e-07, "loss": 0.6351, "step": 40 }, { "epoch": 0.01, "grad_norm": 8.093364927344371, "learning_rate": 9.38432364096081e-07, "loss": 0.621, "step": 60 }, { "epoch": 0.01, "grad_norm": 6.435663053769238, "learning_rate": 1.188748419721871e-06, "loss": 0.5702, "step": 80 }, { "epoch": 0.02, "grad_norm": 7.962395022671904, "learning_rate": 1.4390644753476612e-06, "loss": 0.5317, "step": 100 }, { "epoch": 0.02, "grad_norm": 6.601932601034603, "learning_rate": 1.6893805309734515e-06, "loss": 0.5138, "step": 120 }, { "epoch": 0.02, "grad_norm": 5.565400471086017, "learning_rate": 1.9396965865992414e-06, "loss": 0.4772, "step": 140 }, { "epoch": 0.03, "grad_norm": 10.719063890493263, "learning_rate": 2.1900126422250318e-06, "loss": 0.4828, "step": 160 }, { "epoch": 0.03, "grad_norm": 5.456905246118198, "learning_rate": 2.4403286978508217e-06, "loss": 0.4612, "step": 180 }, { "epoch": 0.03, "grad_norm": 5.074136053502421, "learning_rate": 2.6906447534766116e-06, "loss": 0.4614, "step": 200 }, { "epoch": 0.03, "grad_norm": 5.739808379905823, "learning_rate": 2.940960809102403e-06, "loss": 0.4455, "step": 220 }, { "epoch": 0.04, "grad_norm": 8.929223129466571, "learning_rate": 3.1912768647281927e-06, "loss": 0.4215, "step": 240 }, { "epoch": 0.04, "grad_norm": 5.727490359048043, "learning_rate": 3.4415929203539826e-06, "loss": 0.4544, "step": 260 }, { "epoch": 0.04, "grad_norm": 4.692889967552893, "learning_rate": 3.6919089759797726e-06, "loss": 0.4362, "step": 280 }, { "epoch": 0.05, "grad_norm": 5.492004651950497, "learning_rate": 3.942225031605563e-06, "loss": 0.4102, "step": 300 }, { "epoch": 0.05, "grad_norm": 4.905463477586154, "learning_rate": 4.192541087231353e-06, "loss": 0.4163, "step": 320 }, { "epoch": 0.05, "grad_norm": 6.042815351357798, "learning_rate": 4.442857142857143e-06, "loss": 0.4043, "step": 340 }, { "epoch": 0.06, "grad_norm": 3.966783411781683, "learning_rate": 4.693173198482934e-06, "loss": 0.4134, "step": 360 }, { "epoch": 0.06, "grad_norm": 6.15182506494119, "learning_rate": 4.943489254108724e-06, "loss": 0.4095, "step": 380 }, { "epoch": 0.06, "grad_norm": 5.556343872513135, "learning_rate": 5.193805309734513e-06, "loss": 0.3952, "step": 400 }, { "epoch": 0.07, "grad_norm": 6.168472163196086, "learning_rate": 5.4441213653603045e-06, "loss": 0.4, "step": 420 }, { "epoch": 0.07, "grad_norm": 7.710878876344402, "learning_rate": 5.6944374209860944e-06, "loss": 0.4158, "step": 440 }, { "epoch": 0.07, "grad_norm": 10.149729727919098, "learning_rate": 5.944753476611884e-06, "loss": 0.4008, "step": 460 }, { "epoch": 0.08, "grad_norm": 5.276529583928554, "learning_rate": 6.195069532237674e-06, "loss": 0.3693, "step": 480 }, { "epoch": 0.08, "grad_norm": 5.4922316048860305, "learning_rate": 6.445385587863464e-06, "loss": 0.3655, "step": 500 }, { "epoch": 0.08, "grad_norm": 4.834486637416974, "learning_rate": 6.695701643489254e-06, "loss": 0.4071, "step": 520 }, { "epoch": 0.09, "grad_norm": 5.109668221648755, "learning_rate": 6.946017699115044e-06, "loss": 0.3611, "step": 540 }, { "epoch": 0.09, "grad_norm": 4.962197451653283, "learning_rate": 7.196333754740835e-06, "loss": 0.3617, "step": 560 }, { "epoch": 0.09, "grad_norm": 5.725555872687781, "learning_rate": 7.4466498103666256e-06, "loss": 0.3938, "step": 580 }, { "epoch": 0.09, "grad_norm": 21.199977745188423, "learning_rate": 7.696965865992416e-06, "loss": 0.3753, "step": 600 }, { "epoch": 0.1, "grad_norm": 6.822065444923147, "learning_rate": 7.947281921618205e-06, "loss": 0.3732, "step": 620 }, { "epoch": 0.1, "grad_norm": 5.562639041834479, "learning_rate": 8.197597977243996e-06, "loss": 0.3641, "step": 640 }, { "epoch": 0.1, "grad_norm": 6.746641948539073, "learning_rate": 8.447914032869787e-06, "loss": 0.3608, "step": 660 }, { "epoch": 0.11, "grad_norm": 5.083732630988663, "learning_rate": 8.698230088495576e-06, "loss": 0.3489, "step": 680 }, { "epoch": 0.11, "grad_norm": 5.36345764003458, "learning_rate": 8.948546144121367e-06, "loss": 0.3495, "step": 700 }, { "epoch": 0.11, "grad_norm": 5.94599708094623, "learning_rate": 9.198862199747156e-06, "loss": 0.3568, "step": 720 }, { "epoch": 0.12, "grad_norm": 5.59596830741628, "learning_rate": 9.449178255372947e-06, "loss": 0.3424, "step": 740 }, { "epoch": 0.12, "grad_norm": 4.687091250082153, "learning_rate": 9.699494310998736e-06, "loss": 0.3373, "step": 760 }, { "epoch": 0.12, "grad_norm": 4.305364261329056, "learning_rate": 9.949810366624526e-06, "loss": 0.3863, "step": 780 }, { "epoch": 0.13, "grad_norm": 5.439101060195295, "learning_rate": 1.0200126422250315e-05, "loss": 0.3417, "step": 800 }, { "epoch": 0.13, "grad_norm": 5.395373389869628, "learning_rate": 1.0450442477876108e-05, "loss": 0.3375, "step": 820 }, { "epoch": 0.13, "grad_norm": 4.781731874205922, "learning_rate": 1.0700758533501895e-05, "loss": 0.3296, "step": 840 }, { "epoch": 0.14, "grad_norm": 5.456546617203758, "learning_rate": 1.0951074589127688e-05, "loss": 0.3595, "step": 860 }, { "epoch": 0.14, "grad_norm": 5.3912349632260455, "learning_rate": 1.1201390644753475e-05, "loss": 0.3597, "step": 880 }, { "epoch": 0.14, "grad_norm": 4.404539817994699, "learning_rate": 1.1451706700379268e-05, "loss": 0.3314, "step": 900 }, { "epoch": 0.15, "grad_norm": 7.289102012469942, "learning_rate": 1.1702022756005057e-05, "loss": 0.3406, "step": 920 }, { "epoch": 0.15, "grad_norm": 5.786352926324861, "learning_rate": 1.1952338811630847e-05, "loss": 0.3509, "step": 940 }, { "epoch": 0.15, "grad_norm": 4.131405105795528, "learning_rate": 1.220265486725664e-05, "loss": 0.3459, "step": 960 }, { "epoch": 0.15, "grad_norm": 5.036101661631649, "learning_rate": 1.2452970922882427e-05, "loss": 0.3327, "step": 980 }, { "epoch": 0.16, "grad_norm": 4.9031568362075895, "learning_rate": 1.270328697850822e-05, "loss": 0.359, "step": 1000 }, { "epoch": 0.16, "grad_norm": 6.680058126596432, "learning_rate": 1.2953603034134009e-05, "loss": 0.3421, "step": 1020 }, { "epoch": 0.16, "grad_norm": 6.016889325334172, "learning_rate": 1.32039190897598e-05, "loss": 0.3374, "step": 1040 }, { "epoch": 0.17, "grad_norm": 6.844742376412928, "learning_rate": 1.3454235145385589e-05, "loss": 0.3237, "step": 1060 }, { "epoch": 0.17, "grad_norm": 5.969902845962769, "learning_rate": 1.370455120101138e-05, "loss": 0.3237, "step": 1080 }, { "epoch": 0.17, "grad_norm": 6.2950164096609305, "learning_rate": 1.3954867256637168e-05, "loss": 0.3168, "step": 1100 }, { "epoch": 0.18, "grad_norm": 5.2335271803475445, "learning_rate": 1.4205183312262961e-05, "loss": 0.3396, "step": 1120 }, { "epoch": 0.18, "grad_norm": 6.225398358080941, "learning_rate": 1.4455499367888748e-05, "loss": 0.333, "step": 1140 }, { "epoch": 0.18, "grad_norm": 5.3208556045070114, "learning_rate": 1.470581542351454e-05, "loss": 0.3284, "step": 1160 }, { "epoch": 0.19, "grad_norm": 4.7849250581905665, "learning_rate": 1.495613147914033e-05, "loss": 0.3304, "step": 1180 }, { "epoch": 0.19, "grad_norm": 5.184459116726432, "learning_rate": 1.520644753476612e-05, "loss": 0.3462, "step": 1200 }, { "epoch": 0.19, "grad_norm": 4.19686707405711, "learning_rate": 1.545676359039191e-05, "loss": 0.3325, "step": 1220 }, { "epoch": 0.2, "grad_norm": 4.450896337868275, "learning_rate": 1.57070796460177e-05, "loss": 0.3057, "step": 1240 }, { "epoch": 0.2, "grad_norm": 4.50018119447218, "learning_rate": 1.5957395701643493e-05, "loss": 0.3288, "step": 1260 }, { "epoch": 0.2, "grad_norm": 5.166275173654794, "learning_rate": 1.620771175726928e-05, "loss": 0.3167, "step": 1280 }, { "epoch": 0.21, "grad_norm": 5.056697958587676, "learning_rate": 1.645802781289507e-05, "loss": 0.3146, "step": 1300 }, { "epoch": 0.21, "grad_norm": 4.3805271657086795, "learning_rate": 1.670834386852086e-05, "loss": 0.3441, "step": 1320 }, { "epoch": 0.21, "grad_norm": 4.627520106605279, "learning_rate": 1.6958659924146653e-05, "loss": 0.3166, "step": 1340 }, { "epoch": 0.21, "grad_norm": 4.706945749353132, "learning_rate": 1.7208975979772438e-05, "loss": 0.3371, "step": 1360 }, { "epoch": 0.22, "grad_norm": 5.9372079877822665, "learning_rate": 1.745929203539823e-05, "loss": 0.3343, "step": 1380 }, { "epoch": 0.22, "grad_norm": 4.745434424046423, "learning_rate": 1.7709608091024023e-05, "loss": 0.3075, "step": 1400 }, { "epoch": 0.22, "grad_norm": 4.670681702157214, "learning_rate": 1.7959924146649812e-05, "loss": 0.3278, "step": 1420 }, { "epoch": 0.23, "grad_norm": 4.609368793030779, "learning_rate": 1.82102402022756e-05, "loss": 0.3193, "step": 1440 }, { "epoch": 0.23, "grad_norm": 3.8953514957120987, "learning_rate": 1.846055625790139e-05, "loss": 0.3181, "step": 1460 }, { "epoch": 0.23, "grad_norm": 4.048645810385149, "learning_rate": 1.8710872313527183e-05, "loss": 0.3178, "step": 1480 }, { "epoch": 0.24, "grad_norm": 4.039461038727622, "learning_rate": 1.8961188369152972e-05, "loss": 0.3101, "step": 1500 }, { "epoch": 0.24, "grad_norm": 4.865092063730574, "learning_rate": 1.9211504424778764e-05, "loss": 0.3145, "step": 1520 }, { "epoch": 0.24, "grad_norm": 4.320223981560967, "learning_rate": 1.9461820480404553e-05, "loss": 0.3188, "step": 1540 }, { "epoch": 0.25, "grad_norm": 4.24607218830772, "learning_rate": 1.9712136536030343e-05, "loss": 0.3316, "step": 1560 }, { "epoch": 0.25, "grad_norm": 6.659098557418422, "learning_rate": 1.996245259165613e-05, "loss": 0.3103, "step": 1580 }, { "epoch": 0.25, "grad_norm": 4.4003401278033545, "learning_rate": 1.9999982293200938e-05, "loss": 0.3316, "step": 1600 }, { "epoch": 0.26, "grad_norm": 4.466923266959081, "learning_rate": 1.999992108459333e-05, "loss": 0.3149, "step": 1620 }, { "epoch": 0.26, "grad_norm": 4.3370197217130935, "learning_rate": 1.9999816155842287e-05, "loss": 0.3267, "step": 1640 }, { "epoch": 0.26, "grad_norm": 4.928522761323794, "learning_rate": 1.9999667507406614e-05, "loss": 0.3334, "step": 1660 }, { "epoch": 0.27, "grad_norm": 21.27678616799708, "learning_rate": 1.9999475139936266e-05, "loss": 0.31, "step": 1680 }, { "epoch": 0.27, "grad_norm": 9.092160010538011, "learning_rate": 1.9999239054272376e-05, "loss": 0.3219, "step": 1700 }, { "epoch": 0.27, "grad_norm": 4.287462168712924, "learning_rate": 1.9998959251447223e-05, "loss": 0.2765, "step": 1720 }, { "epoch": 0.28, "grad_norm": 4.200006739293095, "learning_rate": 1.9998635732684236e-05, "loss": 0.3187, "step": 1740 }, { "epoch": 0.28, "grad_norm": 4.182543197928632, "learning_rate": 1.9998268499398e-05, "loss": 0.3038, "step": 1760 }, { "epoch": 0.28, "grad_norm": 4.860519421610365, "learning_rate": 1.999785755319424e-05, "loss": 0.3195, "step": 1780 }, { "epoch": 0.28, "grad_norm": 4.6816897176843, "learning_rate": 1.9997402895869806e-05, "loss": 0.3122, "step": 1800 }, { "epoch": 0.29, "grad_norm": 4.765263778254334, "learning_rate": 1.9996904529412684e-05, "loss": 0.3077, "step": 1820 }, { "epoch": 0.29, "grad_norm": 3.8426131141111215, "learning_rate": 1.999636245600198e-05, "loss": 0.3042, "step": 1840 }, { "epoch": 0.29, "grad_norm": 7.697235889469334, "learning_rate": 1.9995776678007892e-05, "loss": 0.3219, "step": 1860 }, { "epoch": 0.3, "grad_norm": 5.3692702870620455, "learning_rate": 1.9995147197991732e-05, "loss": 0.3319, "step": 1880 }, { "epoch": 0.3, "grad_norm": 6.216633345418337, "learning_rate": 1.9994474018705895e-05, "loss": 0.3059, "step": 1900 }, { "epoch": 0.3, "grad_norm": 5.783165856537296, "learning_rate": 1.9993757143093847e-05, "loss": 0.3011, "step": 1920 }, { "epoch": 0.31, "grad_norm": 4.96412779585292, "learning_rate": 1.999299657429011e-05, "loss": 0.3267, "step": 1940 }, { "epoch": 0.31, "grad_norm": 4.584783821625622, "learning_rate": 1.9992192315620268e-05, "loss": 0.3186, "step": 1960 }, { "epoch": 0.31, "grad_norm": 3.966800361004438, "learning_rate": 1.9991344370600926e-05, "loss": 0.3037, "step": 1980 }, { "epoch": 0.32, "grad_norm": 5.549652693595451, "learning_rate": 1.9990452742939716e-05, "loss": 0.3031, "step": 2000 }, { "epoch": 0.32, "grad_norm": 4.092592964025307, "learning_rate": 1.9989517436535264e-05, "loss": 0.299, "step": 2020 }, { "epoch": 0.32, "grad_norm": 7.4143606594255385, "learning_rate": 1.9988538455477186e-05, "loss": 0.3212, "step": 2040 }, { "epoch": 0.33, "grad_norm": 4.429279288399913, "learning_rate": 1.9987515804046065e-05, "loss": 0.2983, "step": 2060 }, { "epoch": 0.33, "grad_norm": 4.168171471024823, "learning_rate": 1.9986449486713425e-05, "loss": 0.2905, "step": 2080 }, { "epoch": 0.33, "grad_norm": 4.023053730559626, "learning_rate": 1.998533950814173e-05, "loss": 0.3202, "step": 2100 }, { "epoch": 0.34, "grad_norm": 4.589556797878434, "learning_rate": 1.998418587318434e-05, "loss": 0.2929, "step": 2120 }, { "epoch": 0.34, "grad_norm": 4.736297519231011, "learning_rate": 1.9982988586885513e-05, "loss": 0.3192, "step": 2140 }, { "epoch": 0.34, "grad_norm": 3.2489559257150242, "learning_rate": 1.9981747654480363e-05, "loss": 0.32, "step": 2160 }, { "epoch": 0.34, "grad_norm": 4.901187690508077, "learning_rate": 1.9980463081394853e-05, "loss": 0.2987, "step": 2180 }, { "epoch": 0.35, "grad_norm": 5.700830285351671, "learning_rate": 1.9979134873245754e-05, "loss": 0.2866, "step": 2200 }, { "epoch": 0.35, "grad_norm": 5.280958720644886, "learning_rate": 1.9977763035840647e-05, "loss": 0.308, "step": 2220 }, { "epoch": 0.35, "grad_norm": 3.658172362891215, "learning_rate": 1.9976347575177864e-05, "loss": 0.3134, "step": 2240 }, { "epoch": 0.36, "grad_norm": 3.649278564224255, "learning_rate": 1.9974888497446493e-05, "loss": 0.2728, "step": 2260 }, { "epoch": 0.36, "grad_norm": 4.213688731917936, "learning_rate": 1.9973385809026328e-05, "loss": 0.2954, "step": 2280 }, { "epoch": 0.36, "grad_norm": 4.290870493890129, "learning_rate": 1.997183951648785e-05, "loss": 0.3041, "step": 2300 }, { "epoch": 0.37, "grad_norm": 4.497146973593326, "learning_rate": 1.9970249626592207e-05, "loss": 0.3039, "step": 2320 }, { "epoch": 0.37, "grad_norm": 3.980598125657535, "learning_rate": 1.9968616146291173e-05, "loss": 0.2876, "step": 2340 }, { "epoch": 0.37, "grad_norm": 4.845633060759088, "learning_rate": 1.9966939082727113e-05, "loss": 0.2859, "step": 2360 }, { "epoch": 0.38, "grad_norm": 5.0017279500150655, "learning_rate": 1.9965218443232964e-05, "loss": 0.2949, "step": 2380 }, { "epoch": 0.38, "grad_norm": 4.812496829341488, "learning_rate": 1.9963454235332197e-05, "loss": 0.3043, "step": 2400 }, { "epoch": 0.38, "grad_norm": 4.196149632421477, "learning_rate": 1.996164646673879e-05, "loss": 0.2917, "step": 2420 }, { "epoch": 0.39, "grad_norm": 3.8759357184009477, "learning_rate": 1.9959795145357187e-05, "loss": 0.2788, "step": 2440 }, { "epoch": 0.39, "grad_norm": 3.840684986809998, "learning_rate": 1.995790027928226e-05, "loss": 0.2877, "step": 2460 }, { "epoch": 0.39, "grad_norm": 5.2172036078456605, "learning_rate": 1.9955961876799288e-05, "loss": 0.283, "step": 2480 }, { "epoch": 0.4, "grad_norm": 5.656431416505367, "learning_rate": 1.995397994638391e-05, "loss": 0.2867, "step": 2500 }, { "epoch": 0.4, "grad_norm": 3.9781055866495945, "learning_rate": 1.9951954496702084e-05, "loss": 0.2842, "step": 2520 }, { "epoch": 0.4, "grad_norm": 3.6294521746502917, "learning_rate": 1.994988553661007e-05, "loss": 0.2672, "step": 2540 }, { "epoch": 0.4, "grad_norm": 4.51151201839679, "learning_rate": 1.9947773075154352e-05, "loss": 0.2904, "step": 2560 }, { "epoch": 0.41, "grad_norm": 4.985296947393056, "learning_rate": 1.9945617121571655e-05, "loss": 0.2999, "step": 2580 }, { "epoch": 0.41, "grad_norm": 3.674752134930168, "learning_rate": 1.9943417685288848e-05, "loss": 0.2785, "step": 2600 }, { "epoch": 0.41, "grad_norm": 4.797788651028868, "learning_rate": 1.9941174775922932e-05, "loss": 0.2983, "step": 2620 }, { "epoch": 0.42, "grad_norm": 4.079696177221718, "learning_rate": 1.9938888403281006e-05, "loss": 0.2777, "step": 2640 }, { "epoch": 0.42, "grad_norm": 3.3610527639517755, "learning_rate": 1.9936558577360198e-05, "loss": 0.2956, "step": 2660 }, { "epoch": 0.42, "grad_norm": 7.188852464890547, "learning_rate": 1.993418530834764e-05, "loss": 0.2829, "step": 2680 }, { "epoch": 0.43, "grad_norm": 3.360659331617603, "learning_rate": 1.993176860662041e-05, "loss": 0.292, "step": 2700 }, { "epoch": 0.43, "grad_norm": 4.107038754867596, "learning_rate": 1.9929308482745514e-05, "loss": 0.2694, "step": 2720 }, { "epoch": 0.43, "grad_norm": 3.9890648859407647, "learning_rate": 1.9926804947479808e-05, "loss": 0.2962, "step": 2740 }, { "epoch": 0.44, "grad_norm": 4.103890708256313, "learning_rate": 1.9924258011769957e-05, "loss": 0.305, "step": 2760 }, { "epoch": 0.44, "grad_norm": 3.6344296373375182, "learning_rate": 1.9921667686752412e-05, "loss": 0.2868, "step": 2780 }, { "epoch": 0.44, "grad_norm": 4.132484812446097, "learning_rate": 1.9919033983753325e-05, "loss": 0.3088, "step": 2800 }, { "epoch": 0.45, "grad_norm": 4.602803346333572, "learning_rate": 1.991635691428853e-05, "loss": 0.2908, "step": 2820 }, { "epoch": 0.45, "grad_norm": 3.553835454417207, "learning_rate": 1.9913636490063475e-05, "loss": 0.2959, "step": 2840 }, { "epoch": 0.45, "grad_norm": 3.3951346647945857, "learning_rate": 1.991087272297318e-05, "loss": 0.2857, "step": 2860 }, { "epoch": 0.46, "grad_norm": 5.5117950542313014, "learning_rate": 1.9908065625102174e-05, "loss": 0.3072, "step": 2880 }, { "epoch": 0.46, "grad_norm": 4.238784802315611, "learning_rate": 1.9905215208724454e-05, "loss": 0.2781, "step": 2900 }, { "epoch": 0.46, "grad_norm": 3.5164479288590287, "learning_rate": 1.990232148630343e-05, "loss": 0.2794, "step": 2920 }, { "epoch": 0.46, "grad_norm": 4.085534858193557, "learning_rate": 1.9899384470491854e-05, "loss": 0.2858, "step": 2940 }, { "epoch": 0.47, "grad_norm": 3.5369416675799563, "learning_rate": 1.98964041741318e-05, "loss": 0.2672, "step": 2960 }, { "epoch": 0.47, "grad_norm": 4.60995552157379, "learning_rate": 1.989338061025456e-05, "loss": 0.2868, "step": 2980 }, { "epoch": 0.47, "grad_norm": 3.3514909358709444, "learning_rate": 1.989031379208063e-05, "loss": 0.2723, "step": 3000 }, { "epoch": 0.48, "grad_norm": 3.923203323163403, "learning_rate": 1.9887203733019632e-05, "loss": 0.28, "step": 3020 }, { "epoch": 0.48, "grad_norm": 3.429027033234547, "learning_rate": 1.9884050446670256e-05, "loss": 0.2952, "step": 3040 }, { "epoch": 0.48, "grad_norm": 3.5801457858104397, "learning_rate": 1.9880853946820197e-05, "loss": 0.2804, "step": 3060 }, { "epoch": 0.49, "grad_norm": 4.712617033863597, "learning_rate": 1.9877614247446116e-05, "loss": 0.2892, "step": 3080 }, { "epoch": 0.49, "grad_norm": 4.933700451637125, "learning_rate": 1.987433136271354e-05, "loss": 0.2717, "step": 3100 }, { "epoch": 0.49, "grad_norm": 4.790425988609337, "learning_rate": 1.9871005306976846e-05, "loss": 0.2695, "step": 3120 }, { "epoch": 0.5, "grad_norm": 3.7466637075374045, "learning_rate": 1.9867636094779166e-05, "loss": 0.2624, "step": 3140 }, { "epoch": 0.5, "grad_norm": 3.50365965112591, "learning_rate": 1.9864223740852334e-05, "loss": 0.2844, "step": 3160 }, { "epoch": 0.5, "grad_norm": 4.569034860813186, "learning_rate": 1.9860768260116815e-05, "loss": 0.2905, "step": 3180 }, { "epoch": 0.51, "grad_norm": 4.170649316306094, "learning_rate": 1.9857269667681655e-05, "loss": 0.2674, "step": 3200 }, { "epoch": 0.51, "grad_norm": 4.434905247240572, "learning_rate": 1.98537279788444e-05, "loss": 0.2794, "step": 3220 }, { "epoch": 0.51, "grad_norm": 3.9321442768019623, "learning_rate": 1.9850143209091034e-05, "loss": 0.2881, "step": 3240 }, { "epoch": 0.52, "grad_norm": 5.363885077307175, "learning_rate": 1.9846515374095914e-05, "loss": 0.2858, "step": 3260 }, { "epoch": 0.52, "grad_norm": 4.191427182787896, "learning_rate": 1.98428444897217e-05, "loss": 0.2816, "step": 3280 }, { "epoch": 0.52, "grad_norm": 4.006135526500501, "learning_rate": 1.983913057201928e-05, "loss": 0.2755, "step": 3300 }, { "epoch": 0.52, "grad_norm": 3.8000033328393075, "learning_rate": 1.9835373637227703e-05, "loss": 0.2733, "step": 3320 }, { "epoch": 0.53, "grad_norm": 5.333992420638842, "learning_rate": 1.9831573701774123e-05, "loss": 0.2779, "step": 3340 }, { "epoch": 0.53, "grad_norm": 4.252307489826801, "learning_rate": 1.9827730782273703e-05, "loss": 0.2592, "step": 3360 }, { "epoch": 0.53, "grad_norm": 3.525549279021786, "learning_rate": 1.982384489552955e-05, "loss": 0.2739, "step": 3380 }, { "epoch": 0.54, "grad_norm": 7.6522667312835475, "learning_rate": 1.9819916058532657e-05, "loss": 0.2816, "step": 3400 }, { "epoch": 0.54, "grad_norm": 11.523203122616478, "learning_rate": 1.98159442884618e-05, "loss": 0.2715, "step": 3420 }, { "epoch": 0.54, "grad_norm": 4.787543449503492, "learning_rate": 1.9811929602683497e-05, "loss": 0.2618, "step": 3440 }, { "epoch": 0.55, "grad_norm": 5.207304138221566, "learning_rate": 1.9807872018751904e-05, "loss": 0.254, "step": 3460 }, { "epoch": 0.55, "grad_norm": 3.4061625058197453, "learning_rate": 1.9803771554408745e-05, "loss": 0.2526, "step": 3480 }, { "epoch": 0.55, "grad_norm": 4.380956899549788, "learning_rate": 1.9799628227583248e-05, "loss": 0.2797, "step": 3500 }, { "epoch": 0.56, "grad_norm": 4.127558290565574, "learning_rate": 1.9795442056392054e-05, "loss": 0.2712, "step": 3520 }, { "epoch": 0.56, "grad_norm": 4.10767355386799, "learning_rate": 1.9791213059139132e-05, "loss": 0.2832, "step": 3540 }, { "epoch": 0.56, "grad_norm": 3.662712320155948, "learning_rate": 1.978694125431572e-05, "loss": 0.2663, "step": 3560 }, { "epoch": 0.57, "grad_norm": 3.8370942587808607, "learning_rate": 1.978262666060022e-05, "loss": 0.2796, "step": 3580 }, { "epoch": 0.57, "grad_norm": 3.3559744723580143, "learning_rate": 1.9778269296858138e-05, "loss": 0.2758, "step": 3600 }, { "epoch": 0.57, "grad_norm": 4.912706342639339, "learning_rate": 1.977386918214198e-05, "loss": 0.2691, "step": 3620 }, { "epoch": 0.58, "grad_norm": 3.827769733993231, "learning_rate": 1.9769426335691194e-05, "loss": 0.2888, "step": 3640 }, { "epoch": 0.58, "grad_norm": 8.974780882814423, "learning_rate": 1.9764940776932057e-05, "loss": 0.279, "step": 3660 }, { "epoch": 0.58, "grad_norm": 7.288399652923296, "learning_rate": 1.9760412525477615e-05, "loss": 0.27, "step": 3680 }, { "epoch": 0.58, "grad_norm": 5.013042224656741, "learning_rate": 1.9755841601127587e-05, "loss": 0.2635, "step": 3700 }, { "epoch": 0.59, "grad_norm": 4.113083011382156, "learning_rate": 1.9751228023868275e-05, "loss": 0.2742, "step": 3720 }, { "epoch": 0.59, "grad_norm": 3.8810313877329516, "learning_rate": 1.974657181387248e-05, "loss": 0.2608, "step": 3740 }, { "epoch": 0.59, "grad_norm": 4.321527293044816, "learning_rate": 1.974187299149942e-05, "loss": 0.2642, "step": 3760 }, { "epoch": 0.6, "grad_norm": 4.507987668552812, "learning_rate": 1.973713157729462e-05, "loss": 0.2839, "step": 3780 }, { "epoch": 0.6, "grad_norm": 4.112442322300466, "learning_rate": 1.9732347591989863e-05, "loss": 0.2767, "step": 3800 }, { "epoch": 0.6, "grad_norm": 3.9228516840537075, "learning_rate": 1.972752105650304e-05, "loss": 0.2516, "step": 3820 }, { "epoch": 0.61, "grad_norm": 3.9874693579378344, "learning_rate": 1.972265199193813e-05, "loss": 0.2875, "step": 3840 }, { "epoch": 0.61, "grad_norm": 5.032079024162868, "learning_rate": 1.9717740419585033e-05, "loss": 0.266, "step": 3860 }, { "epoch": 0.61, "grad_norm": 3.5996549440957164, "learning_rate": 1.9712786360919543e-05, "loss": 0.2548, "step": 3880 }, { "epoch": 0.62, "grad_norm": 4.683794515218512, "learning_rate": 1.9707789837603205e-05, "loss": 0.2701, "step": 3900 }, { "epoch": 0.62, "grad_norm": 4.52037247566233, "learning_rate": 1.9702750871483248e-05, "loss": 0.261, "step": 3920 }, { "epoch": 0.62, "grad_norm": 4.5693573304091535, "learning_rate": 1.9697669484592487e-05, "loss": 0.2568, "step": 3940 }, { "epoch": 0.63, "grad_norm": 3.48060438103784, "learning_rate": 1.9692545699149212e-05, "loss": 0.2545, "step": 3960 }, { "epoch": 0.63, "grad_norm": 4.719616538813032, "learning_rate": 1.9687379537557107e-05, "loss": 0.2676, "step": 3980 }, { "epoch": 0.63, "grad_norm": 4.593894281521921, "learning_rate": 1.9682171022405133e-05, "loss": 0.2803, "step": 4000 }, { "epoch": 0.64, "grad_norm": 3.2767058377982203, "learning_rate": 1.967692017646746e-05, "loss": 0.258, "step": 4020 }, { "epoch": 0.64, "grad_norm": 4.102789530001537, "learning_rate": 1.9671627022703333e-05, "loss": 0.2804, "step": 4040 }, { "epoch": 0.64, "grad_norm": 3.747557671264137, "learning_rate": 1.9666291584256995e-05, "loss": 0.275, "step": 4060 }, { "epoch": 0.64, "grad_norm": 4.527128041668808, "learning_rate": 1.9660913884457572e-05, "loss": 0.2637, "step": 4080 }, { "epoch": 0.65, "grad_norm": 6.26616500848599, "learning_rate": 1.965549394681899e-05, "loss": 0.2769, "step": 4100 }, { "epoch": 0.65, "grad_norm": 4.208333585128525, "learning_rate": 1.9650031795039847e-05, "loss": 0.2506, "step": 4120 }, { "epoch": 0.65, "grad_norm": 3.770191763293591, "learning_rate": 1.9644527453003326e-05, "loss": 0.2716, "step": 4140 }, { "epoch": 0.66, "grad_norm": 8.625265366162292, "learning_rate": 1.9638980944777085e-05, "loss": 0.2831, "step": 4160 }, { "epoch": 0.66, "grad_norm": 3.9442544557070285, "learning_rate": 1.9633392294613155e-05, "loss": 0.2777, "step": 4180 }, { "epoch": 0.66, "grad_norm": 3.273256865914117, "learning_rate": 1.962776152694783e-05, "loss": 0.2507, "step": 4200 }, { "epoch": 0.67, "grad_norm": 3.7930612759054445, "learning_rate": 1.9622088666401566e-05, "loss": 0.2545, "step": 4220 }, { "epoch": 0.67, "grad_norm": 3.605998255572159, "learning_rate": 1.9616373737778864e-05, "loss": 0.2718, "step": 4240 }, { "epoch": 0.67, "grad_norm": 3.3060364912064966, "learning_rate": 1.961061676606817e-05, "loss": 0.2615, "step": 4260 }, { "epoch": 0.68, "grad_norm": 3.180581756735597, "learning_rate": 1.9604817776441762e-05, "loss": 0.2475, "step": 4280 }, { "epoch": 0.68, "grad_norm": 4.109627023373317, "learning_rate": 1.9598976794255647e-05, "loss": 0.2781, "step": 4300 }, { "epoch": 0.68, "grad_norm": 3.8246626162863007, "learning_rate": 1.9593093845049435e-05, "loss": 0.2797, "step": 4320 }, { "epoch": 0.69, "grad_norm": 3.880680350794998, "learning_rate": 1.9587168954546233e-05, "loss": 0.2728, "step": 4340 }, { "epoch": 0.69, "grad_norm": 3.504579545295932, "learning_rate": 1.9581202148652555e-05, "loss": 0.2517, "step": 4360 }, { "epoch": 0.69, "grad_norm": 3.251503083301662, "learning_rate": 1.957519345345817e-05, "loss": 0.234, "step": 4380 }, { "epoch": 0.7, "grad_norm": 3.8245770887559467, "learning_rate": 1.9569142895236014e-05, "loss": 0.2691, "step": 4400 }, { "epoch": 0.7, "grad_norm": 3.7480270281848163, "learning_rate": 1.9563050500442067e-05, "loss": 0.2563, "step": 4420 }, { "epoch": 0.7, "grad_norm": 3.6499757321739805, "learning_rate": 1.9556916295715248e-05, "loss": 0.2599, "step": 4440 }, { "epoch": 0.71, "grad_norm": 4.289401570798809, "learning_rate": 1.955074030787727e-05, "loss": 0.2494, "step": 4460 }, { "epoch": 0.71, "grad_norm": 4.463444688322885, "learning_rate": 1.9544522563932567e-05, "loss": 0.2503, "step": 4480 }, { "epoch": 0.71, "grad_norm": 3.4818768633740076, "learning_rate": 1.953826309106813e-05, "loss": 0.2569, "step": 4500 }, { "epoch": 0.71, "grad_norm": 4.094470710503364, "learning_rate": 1.9531961916653416e-05, "loss": 0.2722, "step": 4520 }, { "epoch": 0.72, "grad_norm": 4.500248160407271, "learning_rate": 1.9525619068240227e-05, "loss": 0.2525, "step": 4540 }, { "epoch": 0.72, "grad_norm": 3.482370390112968, "learning_rate": 1.951923457356258e-05, "loss": 0.2534, "step": 4560 }, { "epoch": 0.72, "grad_norm": 5.274501845715762, "learning_rate": 1.9512808460536586e-05, "loss": 0.2454, "step": 4580 }, { "epoch": 0.73, "grad_norm": 4.177646944576608, "learning_rate": 1.9506340757260332e-05, "loss": 0.2474, "step": 4600 }, { "epoch": 0.73, "grad_norm": 3.56021633905292, "learning_rate": 1.9499831492013772e-05, "loss": 0.2744, "step": 4620 }, { "epoch": 0.73, "grad_norm": 3.485877700901466, "learning_rate": 1.9493280693258565e-05, "loss": 0.2741, "step": 4640 }, { "epoch": 0.74, "grad_norm": 4.495079741486709, "learning_rate": 1.9486688389637993e-05, "loss": 0.2399, "step": 4660 }, { "epoch": 0.74, "grad_norm": 3.879830579696546, "learning_rate": 1.9480054609976815e-05, "loss": 0.2719, "step": 4680 }, { "epoch": 0.74, "grad_norm": 3.683250176597202, "learning_rate": 1.9473379383281136e-05, "loss": 0.2647, "step": 4700 }, { "epoch": 0.75, "grad_norm": 3.4564441591881585, "learning_rate": 1.9466662738738295e-05, "loss": 0.2614, "step": 4720 }, { "epoch": 0.75, "grad_norm": 3.5430190851406516, "learning_rate": 1.945990470571672e-05, "loss": 0.2509, "step": 4740 }, { "epoch": 0.75, "grad_norm": 3.594438702478659, "learning_rate": 1.945310531376582e-05, "loss": 0.249, "step": 4760 }, { "epoch": 0.76, "grad_norm": 3.765747448208518, "learning_rate": 1.944626459261585e-05, "loss": 0.2503, "step": 4780 }, { "epoch": 0.76, "grad_norm": 4.2901200899686405, "learning_rate": 1.9439382572177755e-05, "loss": 0.2424, "step": 4800 }, { "epoch": 0.76, "grad_norm": 3.476240164568998, "learning_rate": 1.9432459282543085e-05, "loss": 0.264, "step": 4820 }, { "epoch": 0.77, "grad_norm": 3.670630970694611, "learning_rate": 1.942549475398382e-05, "loss": 0.2663, "step": 4840 }, { "epoch": 0.77, "grad_norm": 3.7957980665515034, "learning_rate": 1.941848901695227e-05, "loss": 0.2621, "step": 4860 }, { "epoch": 0.77, "grad_norm": 4.793895081633612, "learning_rate": 1.9411442102080914e-05, "loss": 0.2498, "step": 4880 }, { "epoch": 0.77, "grad_norm": 3.464785397972565, "learning_rate": 1.9404354040182302e-05, "loss": 0.242, "step": 4900 }, { "epoch": 0.78, "grad_norm": 3.486787653538589, "learning_rate": 1.9397224862248875e-05, "loss": 0.2623, "step": 4920 }, { "epoch": 0.78, "grad_norm": 4.089389487964841, "learning_rate": 1.939005459945287e-05, "loss": 0.2543, "step": 4940 }, { "epoch": 0.78, "grad_norm": 3.1762519290562414, "learning_rate": 1.9382843283146158e-05, "loss": 0.2654, "step": 4960 }, { "epoch": 0.79, "grad_norm": 3.835699443072975, "learning_rate": 1.937559094486012e-05, "loss": 0.2743, "step": 4980 }, { "epoch": 0.79, "grad_norm": 3.8852244202589334, "learning_rate": 1.936829761630551e-05, "loss": 0.2629, "step": 5000 }, { "epoch": 0.79, "grad_norm": 3.1906813658574835, "learning_rate": 1.93609633293723e-05, "loss": 0.2417, "step": 5020 }, { "epoch": 0.8, "grad_norm": 2.3269103292999986, "learning_rate": 1.9353588116129556e-05, "loss": 0.218, "step": 5040 }, { "epoch": 0.8, "grad_norm": 3.2694392219817927, "learning_rate": 1.9346172008825302e-05, "loss": 0.2321, "step": 5060 }, { "epoch": 0.8, "grad_norm": 2.7636752394800634, "learning_rate": 1.9338715039886357e-05, "loss": 0.2112, "step": 5080 }, { "epoch": 0.81, "grad_norm": 1.7461733533221422, "learning_rate": 1.9331217241918223e-05, "loss": 0.2065, "step": 5100 }, { "epoch": 0.81, "grad_norm": 2.1112769700310166, "learning_rate": 1.9323678647704908e-05, "loss": 0.2178, "step": 5120 }, { "epoch": 0.81, "grad_norm": 1.928506961128503, "learning_rate": 1.9316099290208812e-05, "loss": 0.2222, "step": 5140 }, { "epoch": 0.82, "grad_norm": 2.289733371787946, "learning_rate": 1.9308479202570575e-05, "loss": 0.2331, "step": 5160 }, { "epoch": 0.82, "grad_norm": 1.997736424764338, "learning_rate": 1.9300818418108923e-05, "loss": 0.2133, "step": 5180 }, { "epoch": 0.82, "grad_norm": 2.260897208845383, "learning_rate": 1.9293116970320528e-05, "loss": 0.2092, "step": 5200 }, { "epoch": 0.83, "grad_norm": 2.01842381348837, "learning_rate": 1.928537489287986e-05, "loss": 0.2189, "step": 5220 }, { "epoch": 0.83, "grad_norm": 2.377313290393669, "learning_rate": 1.927759221963905e-05, "loss": 0.2374, "step": 5240 }, { "epoch": 0.83, "grad_norm": 2.431465602229034, "learning_rate": 1.926976898462773e-05, "loss": 0.2219, "step": 5260 }, { "epoch": 0.83, "grad_norm": 2.272111328930043, "learning_rate": 1.9261905222052882e-05, "loss": 0.2182, "step": 5280 }, { "epoch": 0.84, "grad_norm": 2.1252795764788788, "learning_rate": 1.9254000966298702e-05, "loss": 0.2219, "step": 5300 }, { "epoch": 0.84, "grad_norm": 2.1286718396965143, "learning_rate": 1.924605625192643e-05, "loss": 0.2181, "step": 5320 }, { "epoch": 0.84, "grad_norm": 2.263405074029898, "learning_rate": 1.923807111367423e-05, "loss": 0.2082, "step": 5340 }, { "epoch": 0.85, "grad_norm": 2.349777405930586, "learning_rate": 1.9230045586457004e-05, "loss": 0.2148, "step": 5360 }, { "epoch": 0.85, "grad_norm": 1.951720991944945, "learning_rate": 1.9221979705366256e-05, "loss": 0.2177, "step": 5380 }, { "epoch": 0.85, "grad_norm": 2.6396096890364915, "learning_rate": 1.9213873505669944e-05, "loss": 0.224, "step": 5400 }, { "epoch": 0.86, "grad_norm": 2.340017372784151, "learning_rate": 1.9205727022812307e-05, "loss": 0.2149, "step": 5420 }, { "epoch": 0.86, "grad_norm": 2.489578364113247, "learning_rate": 1.9197540292413734e-05, "loss": 0.2229, "step": 5440 }, { "epoch": 0.86, "grad_norm": 1.9693131625322577, "learning_rate": 1.9189313350270585e-05, "loss": 0.2114, "step": 5460 }, { "epoch": 0.87, "grad_norm": 2.3178735109826505, "learning_rate": 1.9181046232355053e-05, "loss": 0.212, "step": 5480 }, { "epoch": 0.87, "grad_norm": 2.1928905009683133, "learning_rate": 1.9172738974814993e-05, "loss": 0.2176, "step": 5500 }, { "epoch": 0.87, "grad_norm": 2.188333080240406, "learning_rate": 1.9164391613973778e-05, "loss": 0.2077, "step": 5520 }, { "epoch": 0.88, "grad_norm": 2.3037514030133517, "learning_rate": 1.9156004186330123e-05, "loss": 0.2121, "step": 5540 }, { "epoch": 0.88, "grad_norm": 5.37450956724229, "learning_rate": 1.914757672855794e-05, "loss": 0.2095, "step": 5560 }, { "epoch": 0.88, "grad_norm": 2.040655690264704, "learning_rate": 1.9139109277506173e-05, "loss": 0.1993, "step": 5580 }, { "epoch": 0.89, "grad_norm": 2.0548550497572076, "learning_rate": 1.9130601870198633e-05, "loss": 0.2203, "step": 5600 }, { "epoch": 0.89, "grad_norm": 2.8075751525385924, "learning_rate": 1.912205454383384e-05, "loss": 0.2255, "step": 5620 }, { "epoch": 0.89, "grad_norm": 2.213380252712336, "learning_rate": 1.9113467335784855e-05, "loss": 0.2192, "step": 5640 }, { "epoch": 0.89, "grad_norm": 2.7606591076219633, "learning_rate": 1.9104840283599136e-05, "loss": 0.2245, "step": 5660 }, { "epoch": 0.9, "grad_norm": 2.1216745722948787, "learning_rate": 1.909617342499834e-05, "loss": 0.2071, "step": 5680 }, { "epoch": 0.9, "grad_norm": 2.2785095057805065, "learning_rate": 1.908746679787819e-05, "loss": 0.2172, "step": 5700 }, { "epoch": 0.9, "grad_norm": 2.2981748471055576, "learning_rate": 1.907872044030829e-05, "loss": 0.2099, "step": 5720 }, { "epoch": 0.91, "grad_norm": 1.960812811667782, "learning_rate": 1.9069934390531962e-05, "loss": 0.2146, "step": 5740 }, { "epoch": 0.91, "grad_norm": 3.794717686019522, "learning_rate": 1.90611086869661e-05, "loss": 0.208, "step": 5760 }, { "epoch": 0.91, "grad_norm": 2.6008278665808033, "learning_rate": 1.9052243368200958e-05, "loss": 0.2134, "step": 5780 }, { "epoch": 0.92, "grad_norm": 2.4705009528511277, "learning_rate": 1.9043338473000025e-05, "loss": 0.2253, "step": 5800 }, { "epoch": 0.92, "grad_norm": 2.401514124088343, "learning_rate": 1.9034394040299827e-05, "loss": 0.2173, "step": 5820 }, { "epoch": 0.92, "grad_norm": 3.1101268070286685, "learning_rate": 1.9025410109209777e-05, "loss": 0.2133, "step": 5840 }, { "epoch": 0.93, "grad_norm": 2.7159957966217187, "learning_rate": 1.9016386719011982e-05, "loss": 0.2022, "step": 5860 }, { "epoch": 0.93, "grad_norm": 2.385988014350966, "learning_rate": 1.90073239091611e-05, "loss": 0.2164, "step": 5880 }, { "epoch": 0.93, "grad_norm": 2.1587806998998325, "learning_rate": 1.899822171928413e-05, "loss": 0.203, "step": 5900 }, { "epoch": 0.94, "grad_norm": 2.2767020052526497, "learning_rate": 1.8989080189180278e-05, "loss": 0.2216, "step": 5920 }, { "epoch": 0.94, "grad_norm": 3.090507426402487, "learning_rate": 1.8979899358820756e-05, "loss": 0.2045, "step": 5940 }, { "epoch": 0.94, "grad_norm": 1.777654599684145, "learning_rate": 1.8970679268348617e-05, "loss": 0.2017, "step": 5960 }, { "epoch": 0.95, "grad_norm": 2.188623412353316, "learning_rate": 1.8961419958078577e-05, "loss": 0.2122, "step": 5980 }, { "epoch": 0.95, "grad_norm": 2.262035920826766, "learning_rate": 1.8952121468496842e-05, "loss": 0.2149, "step": 6000 }, { "epoch": 0.95, "grad_norm": 2.1025895229154403, "learning_rate": 1.894278384026093e-05, "loss": 0.2075, "step": 6020 }, { "epoch": 0.95, "grad_norm": 2.350054387214, "learning_rate": 1.893340711419949e-05, "loss": 0.2026, "step": 6040 }, { "epoch": 0.96, "grad_norm": 2.5056109405602704, "learning_rate": 1.8923991331312125e-05, "loss": 0.2003, "step": 6060 }, { "epoch": 0.96, "grad_norm": 2.202062147526405, "learning_rate": 1.891453653276921e-05, "loss": 0.2078, "step": 6080 }, { "epoch": 0.96, "grad_norm": 2.2246885746094445, "learning_rate": 1.8905042759911734e-05, "loss": 0.2065, "step": 6100 }, { "epoch": 0.97, "grad_norm": 2.3053321066228554, "learning_rate": 1.8895510054251074e-05, "loss": 0.2266, "step": 6120 }, { "epoch": 0.97, "grad_norm": 2.988781297520643, "learning_rate": 1.888593845746886e-05, "loss": 0.2168, "step": 6140 }, { "epoch": 0.97, "grad_norm": 2.4575329679216167, "learning_rate": 1.887632801141676e-05, "loss": 0.1994, "step": 6160 }, { "epoch": 0.98, "grad_norm": 1.974119765257406, "learning_rate": 1.886667875811632e-05, "loss": 0.2009, "step": 6180 }, { "epoch": 0.98, "grad_norm": 2.250505072807692, "learning_rate": 1.885699073975877e-05, "loss": 0.2011, "step": 6200 }, { "epoch": 0.98, "grad_norm": 3.4317109345564774, "learning_rate": 1.8847263998704822e-05, "loss": 0.214, "step": 6220 }, { "epoch": 0.99, "grad_norm": 1.9321307572015396, "learning_rate": 1.883749857748453e-05, "loss": 0.2121, "step": 6240 }, { "epoch": 0.99, "grad_norm": 2.44399716037142, "learning_rate": 1.8827694518797058e-05, "loss": 0.2121, "step": 6260 }, { "epoch": 0.99, "grad_norm": 3.70267918474469, "learning_rate": 1.881785186551051e-05, "loss": 0.218, "step": 6280 }, { "epoch": 1.0, "grad_norm": 2.2164340760047763, "learning_rate": 1.880797066066176e-05, "loss": 0.2018, "step": 6300 }, { "epoch": 1.0, "grad_norm": 2.0350434314114065, "learning_rate": 1.8798050947456237e-05, "loss": 0.2146, "step": 6320 }, { "epoch": 1.0, "grad_norm": 2.2542825844152614, "learning_rate": 1.8788092769267742e-05, "loss": 0.1966, "step": 6340 }, { "epoch": 1.01, "grad_norm": 2.540485628816784, "learning_rate": 1.877809616963828e-05, "loss": 0.181, "step": 6360 }, { "epoch": 1.01, "grad_norm": 2.0965630152825794, "learning_rate": 1.8768061192277835e-05, "loss": 0.1768, "step": 6380 }, { "epoch": 1.01, "grad_norm": 1.832162295211555, "learning_rate": 1.8757987881064214e-05, "loss": 0.1841, "step": 6400 }, { "epoch": 1.01, "grad_norm": 2.3417686734318877, "learning_rate": 1.8747876280042826e-05, "loss": 0.1748, "step": 6420 }, { "epoch": 1.02, "grad_norm": 2.3993985614578657, "learning_rate": 1.8737726433426505e-05, "loss": 0.1806, "step": 6440 }, { "epoch": 1.02, "grad_norm": 2.0282614216820862, "learning_rate": 1.872753838559532e-05, "loss": 0.1948, "step": 6460 }, { "epoch": 1.02, "grad_norm": 2.8165988300739553, "learning_rate": 1.8717312181096363e-05, "loss": 0.175, "step": 6480 }, { "epoch": 1.03, "grad_norm": 2.2451015768881493, "learning_rate": 1.870704786464357e-05, "loss": 0.1798, "step": 6500 }, { "epoch": 1.03, "grad_norm": 2.1815849986963296, "learning_rate": 1.869674548111753e-05, "loss": 0.1831, "step": 6520 }, { "epoch": 1.03, "grad_norm": 2.1640382620765846, "learning_rate": 1.8686405075565258e-05, "loss": 0.1829, "step": 6540 }, { "epoch": 1.04, "grad_norm": 2.9145132743989945, "learning_rate": 1.8676026693200048e-05, "loss": 0.1747, "step": 6560 }, { "epoch": 1.04, "grad_norm": 2.045497051914788, "learning_rate": 1.8665610379401226e-05, "loss": 0.1753, "step": 6580 }, { "epoch": 1.04, "grad_norm": 1.7634716575081997, "learning_rate": 1.8655156179713974e-05, "loss": 0.1814, "step": 6600 }, { "epoch": 1.05, "grad_norm": 1.7599637579437575, "learning_rate": 1.8644664139849144e-05, "loss": 0.1911, "step": 6620 }, { "epoch": 1.05, "grad_norm": 2.120306666344395, "learning_rate": 1.863413430568303e-05, "loss": 0.1875, "step": 6640 }, { "epoch": 1.05, "grad_norm": 1.639791731170639, "learning_rate": 1.8623566723257188e-05, "loss": 0.1784, "step": 6660 }, { "epoch": 1.06, "grad_norm": 2.0062873480341943, "learning_rate": 1.8612961438778225e-05, "loss": 0.1764, "step": 6680 }, { "epoch": 1.06, "grad_norm": 2.4050323576148234, "learning_rate": 1.8602318498617592e-05, "loss": 0.1782, "step": 6700 }, { "epoch": 1.06, "grad_norm": 1.897036816703736, "learning_rate": 1.8591637949311408e-05, "loss": 0.1789, "step": 6720 }, { "epoch": 1.07, "grad_norm": 3.1448228404692387, "learning_rate": 1.8580919837560224e-05, "loss": 0.1657, "step": 6740 }, { "epoch": 1.07, "grad_norm": 1.979007580704552, "learning_rate": 1.8570164210228826e-05, "loss": 0.1773, "step": 6760 }, { "epoch": 1.07, "grad_norm": 1.8928740814253107, "learning_rate": 1.8559371114346058e-05, "loss": 0.182, "step": 6780 }, { "epoch": 1.07, "grad_norm": 2.1321312791081413, "learning_rate": 1.854854059710457e-05, "loss": 0.182, "step": 6800 }, { "epoch": 1.08, "grad_norm": 2.488188785290195, "learning_rate": 1.8537672705860653e-05, "loss": 0.1871, "step": 6820 }, { "epoch": 1.08, "grad_norm": 2.512962685869269, "learning_rate": 1.8526767488134015e-05, "loss": 0.1796, "step": 6840 }, { "epoch": 1.08, "grad_norm": 2.457219103013576, "learning_rate": 1.8515824991607562e-05, "loss": 0.1768, "step": 6860 }, { "epoch": 1.09, "grad_norm": 2.0058477265819272, "learning_rate": 1.850484526412721e-05, "loss": 0.1878, "step": 6880 }, { "epoch": 1.09, "grad_norm": 1.968329732936287, "learning_rate": 1.8493828353701666e-05, "loss": 0.175, "step": 6900 }, { "epoch": 1.09, "grad_norm": 2.2137462446632177, "learning_rate": 1.8482774308502218e-05, "loss": 0.1814, "step": 6920 }, { "epoch": 1.1, "grad_norm": 2.1543171277309003, "learning_rate": 1.8471683176862517e-05, "loss": 0.1886, "step": 6940 }, { "epoch": 1.1, "grad_norm": 4.139111281755878, "learning_rate": 1.8460555007278392e-05, "loss": 0.175, "step": 6960 }, { "epoch": 1.1, "grad_norm": 2.054321604592727, "learning_rate": 1.844938984840761e-05, "loss": 0.1768, "step": 6980 }, { "epoch": 1.11, "grad_norm": 2.256285682459467, "learning_rate": 1.843818774906967e-05, "loss": 0.1786, "step": 7000 }, { "epoch": 1.11, "grad_norm": 2.5864624228248627, "learning_rate": 1.8426948758245588e-05, "loss": 0.1789, "step": 7020 }, { "epoch": 1.11, "grad_norm": 2.2452602266240786, "learning_rate": 1.8415672925077706e-05, "loss": 0.1743, "step": 7040 }, { "epoch": 1.12, "grad_norm": 1.905290268237584, "learning_rate": 1.8404360298869443e-05, "loss": 0.1789, "step": 7060 }, { "epoch": 1.12, "grad_norm": 2.1523717750344824, "learning_rate": 1.8393010929085106e-05, "loss": 0.1716, "step": 7080 }, { "epoch": 1.12, "grad_norm": 1.8556344369896958, "learning_rate": 1.8381624865349644e-05, "loss": 0.1814, "step": 7100 }, { "epoch": 1.13, "grad_norm": 2.078547958125406, "learning_rate": 1.837020215744847e-05, "loss": 0.1774, "step": 7120 }, { "epoch": 1.13, "grad_norm": 2.029658706613511, "learning_rate": 1.8358742855327222e-05, "loss": 0.1699, "step": 7140 }, { "epoch": 1.13, "grad_norm": 2.2998633634367316, "learning_rate": 1.8347247009091528e-05, "loss": 0.1762, "step": 7160 }, { "epoch": 1.13, "grad_norm": 2.2901708634113227, "learning_rate": 1.8335714669006818e-05, "loss": 0.1682, "step": 7180 }, { "epoch": 1.14, "grad_norm": 1.6706682645288262, "learning_rate": 1.8324145885498092e-05, "loss": 0.1797, "step": 7200 }, { "epoch": 1.14, "grad_norm": 1.9131372823635615, "learning_rate": 1.8312540709149696e-05, "loss": 0.1777, "step": 7220 }, { "epoch": 1.14, "grad_norm": 2.07909984239155, "learning_rate": 1.8300899190705098e-05, "loss": 0.1752, "step": 7240 }, { "epoch": 1.15, "grad_norm": 2.1760276778157857, "learning_rate": 1.828922138106668e-05, "loss": 0.1847, "step": 7260 }, { "epoch": 1.15, "grad_norm": 2.247820116102049, "learning_rate": 1.8277507331295495e-05, "loss": 0.1899, "step": 7280 }, { "epoch": 1.15, "grad_norm": 1.8558783470602038, "learning_rate": 1.8265757092611075e-05, "loss": 0.176, "step": 7300 }, { "epoch": 1.16, "grad_norm": 2.198791896123805, "learning_rate": 1.8253970716391166e-05, "loss": 0.1781, "step": 7320 }, { "epoch": 1.16, "grad_norm": 2.0878033811578196, "learning_rate": 1.8242148254171532e-05, "loss": 0.186, "step": 7340 }, { "epoch": 1.16, "grad_norm": 2.0609054181173714, "learning_rate": 1.8230289757645737e-05, "loss": 0.1835, "step": 7360 }, { "epoch": 1.17, "grad_norm": 2.458294711914135, "learning_rate": 1.8218395278664876e-05, "loss": 0.1746, "step": 7380 }, { "epoch": 1.17, "grad_norm": 2.657085063045043, "learning_rate": 1.8206464869237405e-05, "loss": 0.2005, "step": 7400 }, { "epoch": 1.17, "grad_norm": 2.4744370061009566, "learning_rate": 1.819449858152887e-05, "loss": 0.1793, "step": 7420 }, { "epoch": 1.18, "grad_norm": 2.149043105341473, "learning_rate": 1.8182496467861694e-05, "loss": 0.1652, "step": 7440 }, { "epoch": 1.18, "grad_norm": 1.9846578681762732, "learning_rate": 1.8170458580714957e-05, "loss": 0.1877, "step": 7460 }, { "epoch": 1.18, "grad_norm": 1.8277088571957387, "learning_rate": 1.815838497272415e-05, "loss": 0.183, "step": 7480 }, { "epoch": 1.19, "grad_norm": 1.9586834966474451, "learning_rate": 1.814627569668096e-05, "loss": 0.1766, "step": 7500 }, { "epoch": 1.19, "grad_norm": 2.2511207120669097, "learning_rate": 1.8134130805533027e-05, "loss": 0.1767, "step": 7520 }, { "epoch": 1.19, "grad_norm": 1.5549651238311348, "learning_rate": 1.8121950352383714e-05, "loss": 0.1955, "step": 7540 }, { "epoch": 1.2, "grad_norm": 2.5865251040638744, "learning_rate": 1.810973439049189e-05, "loss": 0.1819, "step": 7560 }, { "epoch": 1.2, "grad_norm": 2.399656899087326, "learning_rate": 1.809748297327167e-05, "loss": 0.1851, "step": 7580 }, { "epoch": 1.2, "grad_norm": 2.5607074887948773, "learning_rate": 1.8085196154292215e-05, "loss": 0.1793, "step": 7600 }, { "epoch": 1.2, "grad_norm": 2.251395428592012, "learning_rate": 1.8072873987277463e-05, "loss": 0.167, "step": 7620 }, { "epoch": 1.21, "grad_norm": 1.9489827901942562, "learning_rate": 1.8060516526105924e-05, "loss": 0.1884, "step": 7640 }, { "epoch": 1.21, "grad_norm": 2.0637425234570808, "learning_rate": 1.804812382481042e-05, "loss": 0.1876, "step": 7660 }, { "epoch": 1.21, "grad_norm": 1.9347382212160897, "learning_rate": 1.8035695937577863e-05, "loss": 0.1726, "step": 7680 }, { "epoch": 1.22, "grad_norm": 2.4247952482951596, "learning_rate": 1.8023232918749026e-05, "loss": 0.1779, "step": 7700 }, { "epoch": 1.22, "grad_norm": 2.3176200478415843, "learning_rate": 1.8010734822818278e-05, "loss": 0.1837, "step": 7720 }, { "epoch": 1.22, "grad_norm": 2.680156700051065, "learning_rate": 1.7998201704433374e-05, "loss": 0.1815, "step": 7740 }, { "epoch": 1.23, "grad_norm": 1.7853843493153718, "learning_rate": 1.7985633618395197e-05, "loss": 0.1793, "step": 7760 }, { "epoch": 1.23, "grad_norm": 2.064374130022476, "learning_rate": 1.7973030619657535e-05, "loss": 0.18, "step": 7780 }, { "epoch": 1.23, "grad_norm": 2.694008332010934, "learning_rate": 1.7960392763326813e-05, "loss": 0.1787, "step": 7800 }, { "epoch": 1.24, "grad_norm": 1.8953685050923013, "learning_rate": 1.794772010466189e-05, "loss": 0.1811, "step": 7820 }, { "epoch": 1.24, "grad_norm": 2.3442305008755753, "learning_rate": 1.7935012699073787e-05, "loss": 0.1924, "step": 7840 }, { "epoch": 1.24, "grad_norm": 2.4097855590450936, "learning_rate": 1.7922270602125464e-05, "loss": 0.171, "step": 7860 }, { "epoch": 1.25, "grad_norm": 1.8537246377229875, "learning_rate": 1.7909493869531555e-05, "loss": 0.1838, "step": 7880 }, { "epoch": 1.25, "grad_norm": 2.0039152922249697, "learning_rate": 1.789668255715815e-05, "loss": 0.1847, "step": 7900 }, { "epoch": 1.25, "grad_norm": 1.529179165379362, "learning_rate": 1.7883836721022534e-05, "loss": 0.1728, "step": 7920 }, { "epoch": 1.26, "grad_norm": 1.9763886145297185, "learning_rate": 1.7870956417292945e-05, "loss": 0.1688, "step": 7940 }, { "epoch": 1.26, "grad_norm": 1.9066521892575954, "learning_rate": 1.7858041702288335e-05, "loss": 0.153, "step": 7960 }, { "epoch": 1.26, "grad_norm": 1.7450878138703785, "learning_rate": 1.784509263247811e-05, "loss": 0.1893, "step": 7980 }, { "epoch": 1.26, "grad_norm": 1.9620586700844385, "learning_rate": 1.7832109264481904e-05, "loss": 0.185, "step": 8000 }, { "epoch": 1.27, "grad_norm": 1.8923898451681622, "learning_rate": 1.7819091655069314e-05, "loss": 0.1754, "step": 8020 }, { "epoch": 1.27, "grad_norm": 2.0884487166296872, "learning_rate": 1.7806039861159653e-05, "loss": 0.1765, "step": 8040 }, { "epoch": 1.27, "grad_norm": 2.222016415072701, "learning_rate": 1.7792953939821702e-05, "loss": 0.1788, "step": 8060 }, { "epoch": 1.28, "grad_norm": 2.009439484553775, "learning_rate": 1.7779833948273482e-05, "loss": 0.1811, "step": 8080 }, { "epoch": 1.28, "grad_norm": 2.785344314111028, "learning_rate": 1.7766679943881966e-05, "loss": 0.1701, "step": 8100 }, { "epoch": 1.28, "grad_norm": 2.4016974843695453, "learning_rate": 1.775349198416286e-05, "loss": 0.1809, "step": 8120 }, { "epoch": 1.29, "grad_norm": 5.230470573305321, "learning_rate": 1.774027012678033e-05, "loss": 0.1701, "step": 8140 }, { "epoch": 1.29, "grad_norm": 2.373672415463304, "learning_rate": 1.7727014429546762e-05, "loss": 0.1832, "step": 8160 }, { "epoch": 1.29, "grad_norm": 3.3302260379998776, "learning_rate": 1.7713724950422516e-05, "loss": 0.1756, "step": 8180 }, { "epoch": 1.3, "grad_norm": 2.1278369142809925, "learning_rate": 1.770040174751565e-05, "loss": 0.1751, "step": 8200 }, { "epoch": 1.3, "grad_norm": 2.1056194640732526, "learning_rate": 1.7687044879081685e-05, "loss": 0.1733, "step": 8220 }, { "epoch": 1.3, "grad_norm": 1.9218053417452738, "learning_rate": 1.7673654403523336e-05, "loss": 0.1787, "step": 8240 }, { "epoch": 1.31, "grad_norm": 2.015545274207546, "learning_rate": 1.766023037939028e-05, "loss": 0.175, "step": 8260 }, { "epoch": 1.31, "grad_norm": 2.6233040008294544, "learning_rate": 1.7646772865378873e-05, "loss": 0.1662, "step": 8280 }, { "epoch": 1.31, "grad_norm": 1.8332410774716108, "learning_rate": 1.7633281920331906e-05, "loss": 0.1838, "step": 8300 }, { "epoch": 1.32, "grad_norm": 1.9451280989055022, "learning_rate": 1.761975760323835e-05, "loss": 0.1736, "step": 8320 }, { "epoch": 1.32, "grad_norm": 2.3255879960964476, "learning_rate": 1.76061999732331e-05, "loss": 0.1656, "step": 8340 }, { "epoch": 1.32, "grad_norm": 2.2343648630192363, "learning_rate": 1.7592609089596685e-05, "loss": 0.1779, "step": 8360 }, { "epoch": 1.32, "grad_norm": 3.1163417653612955, "learning_rate": 1.7578985011755077e-05, "loss": 0.1715, "step": 8380 }, { "epoch": 1.33, "grad_norm": 2.225480415684875, "learning_rate": 1.7565327799279354e-05, "loss": 0.1716, "step": 8400 }, { "epoch": 1.33, "grad_norm": 2.626122370963583, "learning_rate": 1.7551637511885494e-05, "loss": 0.163, "step": 8420 }, { "epoch": 1.33, "grad_norm": 2.338476937693113, "learning_rate": 1.7537914209434085e-05, "loss": 0.1827, "step": 8440 }, { "epoch": 1.34, "grad_norm": 2.932171635630059, "learning_rate": 1.752415795193008e-05, "loss": 0.1637, "step": 8460 }, { "epoch": 1.34, "grad_norm": 2.168193633641859, "learning_rate": 1.7510368799522514e-05, "loss": 0.1843, "step": 8480 }, { "epoch": 1.34, "grad_norm": 2.0903535310262886, "learning_rate": 1.7496546812504273e-05, "loss": 0.179, "step": 8500 }, { "epoch": 1.35, "grad_norm": 1.873644784081685, "learning_rate": 1.7482692051311805e-05, "loss": 0.1814, "step": 8520 }, { "epoch": 1.35, "grad_norm": 1.9178052782056696, "learning_rate": 1.7468804576524853e-05, "loss": 0.1722, "step": 8540 }, { "epoch": 1.35, "grad_norm": 2.2334060168419163, "learning_rate": 1.7454884448866212e-05, "loss": 0.1878, "step": 8560 }, { "epoch": 1.36, "grad_norm": 2.1292565293829493, "learning_rate": 1.7440931729201448e-05, "loss": 0.1795, "step": 8580 }, { "epoch": 1.36, "grad_norm": 2.423860978511733, "learning_rate": 1.7426946478538626e-05, "loss": 0.1675, "step": 8600 }, { "epoch": 1.36, "grad_norm": 2.1047220409931877, "learning_rate": 1.741292875802807e-05, "loss": 0.1734, "step": 8620 }, { "epoch": 1.37, "grad_norm": 2.1785335425805776, "learning_rate": 1.7398878628962062e-05, "loss": 0.1808, "step": 8640 }, { "epoch": 1.37, "grad_norm": 1.854826356210005, "learning_rate": 1.7384796152774602e-05, "loss": 0.1833, "step": 8660 }, { "epoch": 1.37, "grad_norm": 2.428632402856712, "learning_rate": 1.737068139104111e-05, "loss": 0.1736, "step": 8680 }, { "epoch": 1.38, "grad_norm": 2.5679580432934412, "learning_rate": 1.7356534405478197e-05, "loss": 0.1775, "step": 8700 }, { "epoch": 1.38, "grad_norm": 2.2642899361494124, "learning_rate": 1.7342355257943354e-05, "loss": 0.1696, "step": 8720 }, { "epoch": 1.38, "grad_norm": 2.5767432931886947, "learning_rate": 1.732814401043471e-05, "loss": 0.186, "step": 8740 }, { "epoch": 1.38, "grad_norm": 1.893321045994624, "learning_rate": 1.7313900725090744e-05, "loss": 0.1652, "step": 8760 }, { "epoch": 1.39, "grad_norm": 1.7371063719634066, "learning_rate": 1.7299625464190025e-05, "loss": 0.1743, "step": 8780 }, { "epoch": 1.39, "grad_norm": 2.3410836884450994, "learning_rate": 1.7285318290150934e-05, "loss": 0.1749, "step": 8800 }, { "epoch": 1.39, "grad_norm": 2.0076324738418667, "learning_rate": 1.727097926553139e-05, "loss": 0.171, "step": 8820 }, { "epoch": 1.4, "grad_norm": 2.114082790821531, "learning_rate": 1.7256608453028577e-05, "loss": 0.1826, "step": 8840 }, { "epoch": 1.4, "grad_norm": 1.8298354013347164, "learning_rate": 1.7242205915478677e-05, "loss": 0.1694, "step": 8860 }, { "epoch": 1.4, "grad_norm": 2.074394146818531, "learning_rate": 1.722777171585658e-05, "loss": 0.1709, "step": 8880 }, { "epoch": 1.41, "grad_norm": 2.1343987124286183, "learning_rate": 1.721330591727562e-05, "loss": 0.174, "step": 8900 }, { "epoch": 1.41, "grad_norm": 2.113311146861362, "learning_rate": 1.7198808582987313e-05, "loss": 0.1725, "step": 8920 }, { "epoch": 1.41, "grad_norm": 2.201652738347589, "learning_rate": 1.718427977638104e-05, "loss": 0.1761, "step": 8940 }, { "epoch": 1.42, "grad_norm": 2.0551908043411773, "learning_rate": 1.7169719560983817e-05, "loss": 0.1771, "step": 8960 }, { "epoch": 1.42, "grad_norm": 1.4177831102654264, "learning_rate": 1.7155128000459967e-05, "loss": 0.1611, "step": 8980 }, { "epoch": 1.42, "grad_norm": 1.974977060981519, "learning_rate": 1.71405051586109e-05, "loss": 0.1701, "step": 9000 }, { "epoch": 1.43, "grad_norm": 2.394531449135901, "learning_rate": 1.7125851099374784e-05, "loss": 0.1841, "step": 9020 }, { "epoch": 1.43, "grad_norm": 2.017565126183319, "learning_rate": 1.7111165886826288e-05, "loss": 0.1659, "step": 9040 }, { "epoch": 1.43, "grad_norm": 1.9783212428806802, "learning_rate": 1.70964495851763e-05, "loss": 0.1807, "step": 9060 }, { "epoch": 1.44, "grad_norm": 2.0181529434741, "learning_rate": 1.708170225877165e-05, "loss": 0.195, "step": 9080 }, { "epoch": 1.44, "grad_norm": 2.2209977061102237, "learning_rate": 1.7066923972094804e-05, "loss": 0.1798, "step": 9100 }, { "epoch": 1.44, "grad_norm": 2.132481028750825, "learning_rate": 1.705211478976363e-05, "loss": 0.171, "step": 9120 }, { "epoch": 1.44, "grad_norm": 1.6746882362029545, "learning_rate": 1.7037274776531064e-05, "loss": 0.1917, "step": 9140 }, { "epoch": 1.45, "grad_norm": 2.185315703710572, "learning_rate": 1.702240399728486e-05, "loss": 0.1655, "step": 9160 }, { "epoch": 1.45, "grad_norm": 2.4321099376111084, "learning_rate": 1.7007502517047293e-05, "loss": 0.1778, "step": 9180 }, { "epoch": 1.45, "grad_norm": 2.0789938809670385, "learning_rate": 1.6992570400974876e-05, "loss": 0.1795, "step": 9200 }, { "epoch": 1.46, "grad_norm": 2.0820901085870127, "learning_rate": 1.6977607714358085e-05, "loss": 0.1691, "step": 9220 }, { "epoch": 1.46, "grad_norm": 2.0783905777418834, "learning_rate": 1.6962614522621047e-05, "loss": 0.1758, "step": 9240 }, { "epoch": 1.46, "grad_norm": 2.0755554604762887, "learning_rate": 1.69475908913213e-05, "loss": 0.1745, "step": 9260 }, { "epoch": 1.47, "grad_norm": 3.1531529103037226, "learning_rate": 1.693253688614945e-05, "loss": 0.1754, "step": 9280 }, { "epoch": 1.47, "grad_norm": 2.2993858263736064, "learning_rate": 1.6917452572928936e-05, "loss": 0.166, "step": 9300 }, { "epoch": 1.47, "grad_norm": 1.75447681551487, "learning_rate": 1.69023380176157e-05, "loss": 0.178, "step": 9320 }, { "epoch": 1.48, "grad_norm": 1.973995547249979, "learning_rate": 1.688719328629793e-05, "loss": 0.1799, "step": 9340 }, { "epoch": 1.48, "grad_norm": 2.4126605558582397, "learning_rate": 1.687201844519575e-05, "loss": 0.1711, "step": 9360 }, { "epoch": 1.48, "grad_norm": 1.9841098157654595, "learning_rate": 1.685681356066094e-05, "loss": 0.1767, "step": 9380 }, { "epoch": 1.49, "grad_norm": 2.1398055464198746, "learning_rate": 1.684157869917665e-05, "loss": 0.1727, "step": 9400 }, { "epoch": 1.49, "grad_norm": 1.9307719901389893, "learning_rate": 1.6826313927357096e-05, "loss": 0.1772, "step": 9420 }, { "epoch": 1.49, "grad_norm": 1.8311039706909333, "learning_rate": 1.681101931194729e-05, "loss": 0.172, "step": 9440 }, { "epoch": 1.5, "grad_norm": 2.1030690065273228, "learning_rate": 1.6795694919822713e-05, "loss": 0.1681, "step": 9460 }, { "epoch": 1.5, "grad_norm": 2.295817125082063, "learning_rate": 1.6780340817989067e-05, "loss": 0.1756, "step": 9480 }, { "epoch": 1.5, "grad_norm": 2.126975655764791, "learning_rate": 1.6764957073581937e-05, "loss": 0.1705, "step": 9500 }, { "epoch": 1.5, "grad_norm": 2.1165992510815443, "learning_rate": 1.6749543753866544e-05, "loss": 0.1769, "step": 9520 }, { "epoch": 1.51, "grad_norm": 2.2911343620991875, "learning_rate": 1.6734100926237405e-05, "loss": 0.1786, "step": 9540 }, { "epoch": 1.51, "grad_norm": 1.9750133369092064, "learning_rate": 1.6718628658218078e-05, "loss": 0.1787, "step": 9560 }, { "epoch": 1.51, "grad_norm": 2.1096785715440003, "learning_rate": 1.670312701746083e-05, "loss": 0.1746, "step": 9580 }, { "epoch": 1.52, "grad_norm": 2.0647784277246126, "learning_rate": 1.6687596071746376e-05, "loss": 0.173, "step": 9600 }, { "epoch": 1.52, "grad_norm": 1.7141270975686904, "learning_rate": 1.667203588898356e-05, "loss": 0.17, "step": 9620 }, { "epoch": 1.52, "grad_norm": 2.1624948490500433, "learning_rate": 1.665644653720906e-05, "loss": 0.1746, "step": 9640 }, { "epoch": 1.53, "grad_norm": 2.2815311206414797, "learning_rate": 1.6640828084587104e-05, "loss": 0.1811, "step": 9660 }, { "epoch": 1.53, "grad_norm": 1.9683000565100015, "learning_rate": 1.662518059940916e-05, "loss": 0.1883, "step": 9680 }, { "epoch": 1.53, "grad_norm": 2.099153179909315, "learning_rate": 1.6609504150093634e-05, "loss": 0.1563, "step": 9700 }, { "epoch": 1.54, "grad_norm": 2.3648488796091542, "learning_rate": 1.659379880518559e-05, "loss": 0.1711, "step": 9720 }, { "epoch": 1.54, "grad_norm": 1.9477130219057668, "learning_rate": 1.6578064633356426e-05, "loss": 0.1807, "step": 9740 }, { "epoch": 1.54, "grad_norm": 1.8763997892785833, "learning_rate": 1.6562301703403588e-05, "loss": 0.17, "step": 9760 }, { "epoch": 1.55, "grad_norm": 1.9473897115797032, "learning_rate": 1.654651008425027e-05, "loss": 0.1695, "step": 9780 }, { "epoch": 1.55, "grad_norm": 1.94275021614424, "learning_rate": 1.653068984494511e-05, "loss": 0.1786, "step": 9800 }, { "epoch": 1.55, "grad_norm": 2.591445266572519, "learning_rate": 1.6514841054661884e-05, "loss": 0.185, "step": 9820 }, { "epoch": 1.56, "grad_norm": 2.2087296424190948, "learning_rate": 1.64989637826992e-05, "loss": 0.1913, "step": 9840 }, { "epoch": 1.56, "grad_norm": 2.0227388921292886, "learning_rate": 1.6483058098480214e-05, "loss": 0.1771, "step": 9860 }, { "epoch": 1.56, "grad_norm": 1.941226451998698, "learning_rate": 1.646712407155231e-05, "loss": 0.1604, "step": 9880 }, { "epoch": 1.56, "grad_norm": 1.799624851657898, "learning_rate": 1.64511617715868e-05, "loss": 0.1703, "step": 9900 }, { "epoch": 1.57, "grad_norm": 2.1026662884608873, "learning_rate": 1.6435171268378617e-05, "loss": 0.1699, "step": 9920 }, { "epoch": 1.57, "grad_norm": 2.037064549070868, "learning_rate": 1.641915263184601e-05, "loss": 0.1749, "step": 9940 }, { "epoch": 1.57, "grad_norm": 2.289972742743045, "learning_rate": 1.6403105932030253e-05, "loss": 0.171, "step": 9960 }, { "epoch": 1.58, "grad_norm": 1.7397923448674204, "learning_rate": 1.638703123909531e-05, "loss": 0.1722, "step": 9980 }, { "epoch": 1.58, "grad_norm": 2.2809191947314327, "learning_rate": 1.6370928623327557e-05, "loss": 0.1789, "step": 10000 }, { "epoch": 1.58, "grad_norm": 5.837716808009581, "learning_rate": 1.635479815513546e-05, "loss": 0.1688, "step": 10020 }, { "epoch": 1.59, "grad_norm": 1.9385163502489462, "learning_rate": 1.6338639905049256e-05, "loss": 0.1684, "step": 10040 }, { "epoch": 1.59, "grad_norm": 2.4083879559228265, "learning_rate": 1.6322453943720677e-05, "loss": 0.1834, "step": 10060 }, { "epoch": 1.59, "grad_norm": 1.8784457651634563, "learning_rate": 1.6306240341922616e-05, "loss": 0.1726, "step": 10080 }, { "epoch": 1.6, "grad_norm": 2.5015152154155134, "learning_rate": 1.628999917054882e-05, "loss": 0.175, "step": 10100 }, { "epoch": 1.6, "grad_norm": 2.3462998824591317, "learning_rate": 1.627373050061358e-05, "loss": 0.1836, "step": 10120 }, { "epoch": 1.6, "grad_norm": 6.491573153186662, "learning_rate": 1.625743440325143e-05, "loss": 0.1708, "step": 10140 }, { "epoch": 1.61, "grad_norm": 1.8735636613758684, "learning_rate": 1.6241110949716837e-05, "loss": 0.1644, "step": 10160 }, { "epoch": 1.61, "grad_norm": 2.1199720026362776, "learning_rate": 1.6224760211383867e-05, "loss": 0.1745, "step": 10180 }, { "epoch": 1.61, "grad_norm": 2.26663511712426, "learning_rate": 1.6208382259745902e-05, "loss": 0.1722, "step": 10200 }, { "epoch": 1.62, "grad_norm": 1.6274143037047173, "learning_rate": 1.6191977166415303e-05, "loss": 0.1689, "step": 10220 }, { "epoch": 1.62, "grad_norm": 1.7970519932361564, "learning_rate": 1.617554500312311e-05, "loss": 0.186, "step": 10240 }, { "epoch": 1.62, "grad_norm": 2.241357629833042, "learning_rate": 1.6159085841718732e-05, "loss": 0.1691, "step": 10260 }, { "epoch": 1.63, "grad_norm": 1.6347966162730332, "learning_rate": 1.614259975416963e-05, "loss": 0.1689, "step": 10280 }, { "epoch": 1.63, "grad_norm": 1.9003545555776973, "learning_rate": 1.612608681256098e-05, "loss": 0.1831, "step": 10300 }, { "epoch": 1.63, "grad_norm": 2.224301766553686, "learning_rate": 1.61095470890954e-05, "loss": 0.1809, "step": 10320 }, { "epoch": 1.63, "grad_norm": 2.0418622318986057, "learning_rate": 1.609298065609259e-05, "loss": 0.1732, "step": 10340 }, { "epoch": 1.64, "grad_norm": 1.8847041460989982, "learning_rate": 1.607638758598906e-05, "loss": 0.18, "step": 10360 }, { "epoch": 1.64, "grad_norm": 2.6034376013955765, "learning_rate": 1.6059767951337775e-05, "loss": 0.1659, "step": 10380 }, { "epoch": 1.64, "grad_norm": 2.717869407722093, "learning_rate": 1.6043121824807853e-05, "loss": 0.1778, "step": 10400 }, { "epoch": 1.65, "grad_norm": 1.9310171511622796, "learning_rate": 1.6026449279184252e-05, "loss": 0.1742, "step": 10420 }, { "epoch": 1.65, "grad_norm": 2.1672348842377542, "learning_rate": 1.6009750387367446e-05, "loss": 0.1751, "step": 10440 }, { "epoch": 1.65, "grad_norm": 2.2155157882826932, "learning_rate": 1.5993025222373107e-05, "loss": 0.1716, "step": 10460 }, { "epoch": 1.66, "grad_norm": 2.12252169432859, "learning_rate": 1.5976273857331788e-05, "loss": 0.1743, "step": 10480 }, { "epoch": 1.66, "grad_norm": 1.8928313189284733, "learning_rate": 1.59594963654886e-05, "loss": 0.1666, "step": 10500 }, { "epoch": 1.66, "grad_norm": 2.1209231144675917, "learning_rate": 1.594269282020289e-05, "loss": 0.1654, "step": 10520 }, { "epoch": 1.67, "grad_norm": 1.7234563029310586, "learning_rate": 1.592586329494793e-05, "loss": 0.1732, "step": 10540 }, { "epoch": 1.67, "grad_norm": 1.8992491192039225, "learning_rate": 1.590900786331058e-05, "loss": 0.1843, "step": 10560 }, { "epoch": 1.67, "grad_norm": 2.255292073046707, "learning_rate": 1.5892126598990988e-05, "loss": 0.1663, "step": 10580 }, { "epoch": 1.68, "grad_norm": 2.0226484844460937, "learning_rate": 1.587521957580224e-05, "loss": 0.1742, "step": 10600 }, { "epoch": 1.68, "grad_norm": 1.9769425460270078, "learning_rate": 1.5858286867670067e-05, "loss": 0.1699, "step": 10620 }, { "epoch": 1.68, "grad_norm": 2.0941096893627047, "learning_rate": 1.584132854863249e-05, "loss": 0.1689, "step": 10640 }, { "epoch": 1.69, "grad_norm": 3.594450243853988, "learning_rate": 1.5824344692839528e-05, "loss": 0.1592, "step": 10660 }, { "epoch": 1.69, "grad_norm": 2.7600105322761266, "learning_rate": 1.5807335374552863e-05, "loss": 0.1696, "step": 10680 }, { "epoch": 1.69, "grad_norm": 1.9277553383287402, "learning_rate": 1.5790300668145488e-05, "loss": 0.1567, "step": 10700 }, { "epoch": 1.69, "grad_norm": 1.8061002015824392, "learning_rate": 1.577324064810143e-05, "loss": 0.1714, "step": 10720 }, { "epoch": 1.7, "grad_norm": 2.075080879197821, "learning_rate": 1.575615538901539e-05, "loss": 0.1651, "step": 10740 }, { "epoch": 1.7, "grad_norm": 1.9389319511933993, "learning_rate": 1.573904496559242e-05, "loss": 0.1664, "step": 10760 }, { "epoch": 1.7, "grad_norm": 1.989109668636887, "learning_rate": 1.5721909452647604e-05, "loss": 0.1644, "step": 10780 }, { "epoch": 1.71, "grad_norm": 1.802713809407511, "learning_rate": 1.570474892510575e-05, "loss": 0.158, "step": 10800 }, { "epoch": 1.71, "grad_norm": 2.1683529625057183, "learning_rate": 1.5687563458001015e-05, "loss": 0.1731, "step": 10820 }, { "epoch": 1.71, "grad_norm": 1.7930535855304202, "learning_rate": 1.5670353126476615e-05, "loss": 0.1721, "step": 10840 }, { "epoch": 1.72, "grad_norm": 2.8089113864956756, "learning_rate": 1.565311800578449e-05, "loss": 0.1581, "step": 10860 }, { "epoch": 1.72, "grad_norm": 1.698115339142883, "learning_rate": 1.5635858171284962e-05, "loss": 0.1604, "step": 10880 }, { "epoch": 1.72, "grad_norm": 2.003861038263119, "learning_rate": 1.561857369844642e-05, "loss": 0.1651, "step": 10900 }, { "epoch": 1.73, "grad_norm": 1.9604631435332058, "learning_rate": 1.5601264662844976e-05, "loss": 0.166, "step": 10920 }, { "epoch": 1.73, "grad_norm": 1.8966802933890046, "learning_rate": 1.5583931140164156e-05, "loss": 0.1504, "step": 10940 }, { "epoch": 1.73, "grad_norm": 1.8798424813043781, "learning_rate": 1.556657320619454e-05, "loss": 0.1712, "step": 10960 }, { "epoch": 1.74, "grad_norm": 1.8480005568507623, "learning_rate": 1.5549190936833452e-05, "loss": 0.166, "step": 10980 }, { "epoch": 1.74, "grad_norm": 1.9730818789739197, "learning_rate": 1.553178440808463e-05, "loss": 0.1766, "step": 11000 }, { "epoch": 1.74, "grad_norm": 2.055586585535215, "learning_rate": 1.5514353696057872e-05, "loss": 0.1684, "step": 11020 }, { "epoch": 1.75, "grad_norm": 2.375617097692436, "learning_rate": 1.5496898876968733e-05, "loss": 0.1774, "step": 11040 }, { "epoch": 1.75, "grad_norm": 1.8908606768542418, "learning_rate": 1.5479420027138157e-05, "loss": 0.1727, "step": 11060 }, { "epoch": 1.75, "grad_norm": 1.584681873586119, "learning_rate": 1.5461917222992176e-05, "loss": 0.1554, "step": 11080 }, { "epoch": 1.75, "grad_norm": 1.9010909888431742, "learning_rate": 1.5444390541061557e-05, "loss": 0.172, "step": 11100 }, { "epoch": 1.76, "grad_norm": 1.6694865055184347, "learning_rate": 1.5426840057981474e-05, "loss": 0.1679, "step": 11120 }, { "epoch": 1.76, "grad_norm": 2.168091256744236, "learning_rate": 1.5409265850491172e-05, "loss": 0.1581, "step": 11140 }, { "epoch": 1.76, "grad_norm": 1.6447757348933234, "learning_rate": 1.539166799543363e-05, "loss": 0.1566, "step": 11160 }, { "epoch": 1.77, "grad_norm": 1.8720093081886822, "learning_rate": 1.5374046569755216e-05, "loss": 0.1755, "step": 11180 }, { "epoch": 1.77, "grad_norm": 1.8424853497444225, "learning_rate": 1.5356401650505376e-05, "loss": 0.1672, "step": 11200 }, { "epoch": 1.77, "grad_norm": 1.7513883396920575, "learning_rate": 1.533873331483627e-05, "loss": 0.1612, "step": 11220 }, { "epoch": 1.78, "grad_norm": 2.8850905765482815, "learning_rate": 1.5321041640002455e-05, "loss": 0.17, "step": 11240 }, { "epoch": 1.78, "grad_norm": 1.5977457110067312, "learning_rate": 1.5303326703360534e-05, "loss": 0.168, "step": 11260 }, { "epoch": 1.78, "grad_norm": 1.839271796672507, "learning_rate": 1.5285588582368814e-05, "loss": 0.1622, "step": 11280 }, { "epoch": 1.79, "grad_norm": 2.0851069593552998, "learning_rate": 1.526782735458699e-05, "loss": 0.1616, "step": 11300 }, { "epoch": 1.79, "grad_norm": 1.8161832282859776, "learning_rate": 1.5250043097675773e-05, "loss": 0.16, "step": 11320 }, { "epoch": 1.79, "grad_norm": 2.1227652074250916, "learning_rate": 1.5232235889396589e-05, "loss": 0.1734, "step": 11340 }, { "epoch": 1.8, "grad_norm": 2.054825114584178, "learning_rate": 1.5214405807611212e-05, "loss": 0.1693, "step": 11360 }, { "epoch": 1.8, "grad_norm": 1.9053808638546215, "learning_rate": 1.5196552930281414e-05, "loss": 0.1555, "step": 11380 }, { "epoch": 1.8, "grad_norm": 2.2568013920238466, "learning_rate": 1.517867733546866e-05, "loss": 0.1777, "step": 11400 }, { "epoch": 1.81, "grad_norm": 1.755591383734745, "learning_rate": 1.516077910133374e-05, "loss": 0.1646, "step": 11420 }, { "epoch": 1.81, "grad_norm": 2.040664452818391, "learning_rate": 1.5142858306136432e-05, "loss": 0.1562, "step": 11440 }, { "epoch": 1.81, "grad_norm": 2.9163542625602634, "learning_rate": 1.5124915028235168e-05, "loss": 0.1671, "step": 11460 }, { "epoch": 1.81, "grad_norm": 1.745198816269922, "learning_rate": 1.5106949346086675e-05, "loss": 0.1645, "step": 11480 }, { "epoch": 1.82, "grad_norm": 3.185139698456198, "learning_rate": 1.5088961338245656e-05, "loss": 0.1645, "step": 11500 }, { "epoch": 1.82, "grad_norm": 1.8737538329143142, "learning_rate": 1.5070951083364413e-05, "loss": 0.1701, "step": 11520 }, { "epoch": 1.82, "grad_norm": 2.28486597548739, "learning_rate": 1.5052918660192548e-05, "loss": 0.1551, "step": 11540 }, { "epoch": 1.83, "grad_norm": 1.9377161951346134, "learning_rate": 1.5034864147576574e-05, "loss": 0.1708, "step": 11560 }, { "epoch": 1.83, "grad_norm": 2.0573176386356713, "learning_rate": 1.5016787624459602e-05, "loss": 0.1605, "step": 11580 }, { "epoch": 1.83, "grad_norm": 1.8884308117753286, "learning_rate": 1.499868916988097e-05, "loss": 0.1635, "step": 11600 }, { "epoch": 1.84, "grad_norm": 2.1599042583369026, "learning_rate": 1.4980568862975921e-05, "loss": 0.1624, "step": 11620 }, { "epoch": 1.84, "grad_norm": 1.9050485545701388, "learning_rate": 1.4962426782975251e-05, "loss": 0.1686, "step": 11640 }, { "epoch": 1.84, "grad_norm": 2.0206229765693404, "learning_rate": 1.4944263009204945e-05, "loss": 0.1609, "step": 11660 }, { "epoch": 1.85, "grad_norm": 2.1626323284609077, "learning_rate": 1.4926077621085858e-05, "loss": 0.1663, "step": 11680 }, { "epoch": 1.85, "grad_norm": 1.7733761949067968, "learning_rate": 1.4907870698133342e-05, "loss": 0.1647, "step": 11700 }, { "epoch": 1.85, "grad_norm": 1.7803734747589461, "learning_rate": 1.4889642319956916e-05, "loss": 0.1704, "step": 11720 }, { "epoch": 1.86, "grad_norm": 2.1492586543303975, "learning_rate": 1.4871392566259912e-05, "loss": 0.1607, "step": 11740 }, { "epoch": 1.86, "grad_norm": 2.00354928746823, "learning_rate": 1.4853121516839119e-05, "loss": 0.1649, "step": 11760 }, { "epoch": 1.86, "grad_norm": 1.6161133520624342, "learning_rate": 1.4834829251584452e-05, "loss": 0.161, "step": 11780 }, { "epoch": 1.87, "grad_norm": 3.956943883806942, "learning_rate": 1.4816515850478586e-05, "loss": 0.1652, "step": 11800 }, { "epoch": 1.87, "grad_norm": 2.2931018381308457, "learning_rate": 1.4798181393596612e-05, "loss": 0.1662, "step": 11820 }, { "epoch": 1.87, "grad_norm": 2.174850511346493, "learning_rate": 1.4779825961105685e-05, "loss": 0.164, "step": 11840 }, { "epoch": 1.87, "grad_norm": 1.5470178711542586, "learning_rate": 1.4761449633264679e-05, "loss": 0.1702, "step": 11860 }, { "epoch": 1.88, "grad_norm": 3.3308654391418204, "learning_rate": 1.4743052490423835e-05, "loss": 0.1618, "step": 11880 }, { "epoch": 1.88, "grad_norm": 1.9720383082261177, "learning_rate": 1.4724634613024404e-05, "loss": 0.1586, "step": 11900 }, { "epoch": 1.88, "grad_norm": 1.883967887969211, "learning_rate": 1.4706196081598298e-05, "loss": 0.1696, "step": 11920 }, { "epoch": 1.89, "grad_norm": 1.6976412060172419, "learning_rate": 1.4687736976767737e-05, "loss": 0.1594, "step": 11940 }, { "epoch": 1.89, "grad_norm": 1.529332108170923, "learning_rate": 1.4669257379244905e-05, "loss": 0.1619, "step": 11960 }, { "epoch": 1.89, "grad_norm": 1.7970574991179633, "learning_rate": 1.465075736983158e-05, "loss": 0.1584, "step": 11980 }, { "epoch": 1.9, "grad_norm": 1.8625966219833416, "learning_rate": 1.46322370294188e-05, "loss": 0.1642, "step": 12000 }, { "epoch": 1.9, "grad_norm": 1.99642977340393, "learning_rate": 1.4613696438986493e-05, "loss": 0.1576, "step": 12020 }, { "epoch": 1.9, "grad_norm": 2.5428182671668944, "learning_rate": 1.4595135679603135e-05, "loss": 0.1741, "step": 12040 }, { "epoch": 1.91, "grad_norm": 3.8472437913445336, "learning_rate": 1.457655483242539e-05, "loss": 0.1713, "step": 12060 }, { "epoch": 1.91, "grad_norm": 2.0144352913407935, "learning_rate": 1.4557953978697748e-05, "loss": 0.1688, "step": 12080 }, { "epoch": 1.91, "grad_norm": 1.792402123900783, "learning_rate": 1.4539333199752189e-05, "loss": 0.166, "step": 12100 }, { "epoch": 1.92, "grad_norm": 2.141224129916374, "learning_rate": 1.4520692577007808e-05, "loss": 0.1626, "step": 12120 }, { "epoch": 1.92, "grad_norm": 1.73495865357573, "learning_rate": 1.4502032191970468e-05, "loss": 0.1599, "step": 12140 }, { "epoch": 1.92, "grad_norm": 10.641224758817746, "learning_rate": 1.4483352126232446e-05, "loss": 0.178, "step": 12160 }, { "epoch": 1.93, "grad_norm": 1.8420610612571067, "learning_rate": 1.4464652461472068e-05, "loss": 0.1648, "step": 12180 }, { "epoch": 1.93, "grad_norm": 1.8800775441958055, "learning_rate": 1.4445933279453358e-05, "loss": 0.1696, "step": 12200 }, { "epoch": 1.93, "grad_norm": 2.151331613378997, "learning_rate": 1.4427194662025678e-05, "loss": 0.1803, "step": 12220 }, { "epoch": 1.93, "grad_norm": 1.9226187530668073, "learning_rate": 1.4408436691123373e-05, "loss": 0.1635, "step": 12240 }, { "epoch": 1.94, "grad_norm": 1.869667544978119, "learning_rate": 1.4389659448765408e-05, "loss": 0.1662, "step": 12260 }, { "epoch": 1.94, "grad_norm": 2.123839678653108, "learning_rate": 1.437086301705502e-05, "loss": 0.1607, "step": 12280 }, { "epoch": 1.94, "grad_norm": 1.8970609470082447, "learning_rate": 1.4352047478179341e-05, "loss": 0.1682, "step": 12300 }, { "epoch": 1.95, "grad_norm": 1.8846339362592277, "learning_rate": 1.4333212914409055e-05, "loss": 0.1681, "step": 12320 }, { "epoch": 1.95, "grad_norm": 2.596536724951098, "learning_rate": 1.4314359408098029e-05, "loss": 0.164, "step": 12340 }, { "epoch": 1.95, "grad_norm": 1.983728679020799, "learning_rate": 1.4295487041682956e-05, "loss": 0.1657, "step": 12360 }, { "epoch": 1.96, "grad_norm": 2.0578280622839555, "learning_rate": 1.4276595897682996e-05, "loss": 0.1788, "step": 12380 }, { "epoch": 1.96, "grad_norm": 1.8225653464241232, "learning_rate": 1.425768605869942e-05, "loss": 0.1864, "step": 12400 }, { "epoch": 1.96, "grad_norm": 1.8018042581666063, "learning_rate": 1.4238757607415225e-05, "loss": 0.1708, "step": 12420 }, { "epoch": 1.97, "grad_norm": 1.6471439004511217, "learning_rate": 1.421981062659481e-05, "loss": 0.1608, "step": 12440 }, { "epoch": 1.97, "grad_norm": 1.785407584963669, "learning_rate": 1.420084519908358e-05, "loss": 0.1659, "step": 12460 }, { "epoch": 1.97, "grad_norm": 2.4554236266808225, "learning_rate": 1.4181861407807606e-05, "loss": 0.1675, "step": 12480 }, { "epoch": 1.98, "grad_norm": 1.872120044188974, "learning_rate": 1.4162859335773253e-05, "loss": 0.1814, "step": 12500 }, { "epoch": 1.98, "grad_norm": 1.8127393893298422, "learning_rate": 1.4143839066066813e-05, "loss": 0.1773, "step": 12520 }, { "epoch": 1.98, "grad_norm": 2.011729890366605, "learning_rate": 1.4124800681854152e-05, "loss": 0.156, "step": 12540 }, { "epoch": 1.99, "grad_norm": 1.6619766845812105, "learning_rate": 1.410574426638034e-05, "loss": 0.1576, "step": 12560 }, { "epoch": 1.99, "grad_norm": 2.1915362586421234, "learning_rate": 1.4086669902969292e-05, "loss": 0.1736, "step": 12580 }, { "epoch": 1.99, "grad_norm": 2.067978259195529, "learning_rate": 1.4067577675023391e-05, "loss": 0.163, "step": 12600 }, { "epoch": 1.99, "grad_norm": 1.7953546477600242, "learning_rate": 1.4048467666023144e-05, "loss": 0.1686, "step": 12620 }, { "epoch": 2.0, "grad_norm": 2.297945149563059, "learning_rate": 1.4029339959526795e-05, "loss": 0.1594, "step": 12640 } ], "logging_steps": 20, "max_steps": 31630, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }