{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.4999858203495, "eval_steps": 500, "global_step": 11019, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00045374881599918323, "grad_norm": 3.5486392974853516, "learning_rate": 6.042296072507553e-07, "loss": 0.3781, "step": 10 }, { "epoch": 0.0009074976319983665, "grad_norm": 2.2130320072174072, "learning_rate": 1.2084592145015106e-06, "loss": 0.3327, "step": 20 }, { "epoch": 0.0013612464479975498, "grad_norm": 1.3342896699905396, "learning_rate": 1.8126888217522659e-06, "loss": 0.2722, "step": 30 }, { "epoch": 0.001814995263996733, "grad_norm": 1.0985041856765747, "learning_rate": 2.4169184290030213e-06, "loss": 0.1678, "step": 40 }, { "epoch": 0.002268744079995916, "grad_norm": 0.8793349266052246, "learning_rate": 3.0211480362537765e-06, "loss": 0.1129, "step": 50 }, { "epoch": 0.0027224928959950997, "grad_norm": 0.6948434710502625, "learning_rate": 3.6253776435045317e-06, "loss": 0.0963, "step": 60 }, { "epoch": 0.0031762417119942828, "grad_norm": 0.6138444542884827, "learning_rate": 4.229607250755287e-06, "loss": 0.0921, "step": 70 }, { "epoch": 0.003629990527993466, "grad_norm": 0.860797643661499, "learning_rate": 4.833836858006043e-06, "loss": 0.0774, "step": 80 }, { "epoch": 0.004083739343992649, "grad_norm": 0.610936164855957, "learning_rate": 5.438066465256799e-06, "loss": 0.0693, "step": 90 }, { "epoch": 0.004537488159991832, "grad_norm": 0.5868649482727051, "learning_rate": 6.042296072507553e-06, "loss": 0.0603, "step": 100 }, { "epoch": 0.004991236975991016, "grad_norm": 0.49356886744499207, "learning_rate": 6.646525679758309e-06, "loss": 0.0752, "step": 110 }, { "epoch": 0.005444985791990199, "grad_norm": 0.5456549525260925, "learning_rate": 7.2507552870090635e-06, "loss": 0.0571, "step": 120 }, { "epoch": 0.005898734607989382, "grad_norm": 0.6095007061958313, "learning_rate": 7.85498489425982e-06, "loss": 0.0552, "step": 130 }, { "epoch": 0.0063524834239885655, "grad_norm": 0.42237332463264465, "learning_rate": 8.459214501510575e-06, "loss": 0.0639, "step": 140 }, { "epoch": 0.006806232239987749, "grad_norm": 0.571368932723999, "learning_rate": 9.06344410876133e-06, "loss": 0.0534, "step": 150 }, { "epoch": 0.007259981055986932, "grad_norm": 0.5768399834632874, "learning_rate": 9.667673716012085e-06, "loss": 0.0546, "step": 160 }, { "epoch": 0.007713729871986115, "grad_norm": 0.5476046204566956, "learning_rate": 1.0271903323262842e-05, "loss": 0.042, "step": 170 }, { "epoch": 0.008167478687985298, "grad_norm": 0.4237188696861267, "learning_rate": 1.0876132930513597e-05, "loss": 0.0466, "step": 180 }, { "epoch": 0.008621227503984481, "grad_norm": 0.4303552210330963, "learning_rate": 1.1480362537764351e-05, "loss": 0.0555, "step": 190 }, { "epoch": 0.009074976319983665, "grad_norm": 0.4543503522872925, "learning_rate": 1.2084592145015106e-05, "loss": 0.0418, "step": 200 }, { "epoch": 0.009528725135982848, "grad_norm": 0.47898203134536743, "learning_rate": 1.2688821752265863e-05, "loss": 0.0428, "step": 210 }, { "epoch": 0.009982473951982032, "grad_norm": 0.417222797870636, "learning_rate": 1.3293051359516618e-05, "loss": 0.0396, "step": 220 }, { "epoch": 0.010436222767981215, "grad_norm": 0.4032108187675476, "learning_rate": 1.3897280966767372e-05, "loss": 0.0391, "step": 230 }, { "epoch": 0.010889971583980399, "grad_norm": 0.3091837465763092, "learning_rate": 1.4501510574018127e-05, "loss": 0.0426, "step": 240 }, { "epoch": 0.01134372039997958, "grad_norm": 0.32662463188171387, "learning_rate": 1.5105740181268884e-05, "loss": 0.038, "step": 250 }, { "epoch": 0.011797469215978764, "grad_norm": 0.4448952078819275, "learning_rate": 1.570996978851964e-05, "loss": 0.0477, "step": 260 }, { "epoch": 0.012251218031977948, "grad_norm": 0.4139327108860016, "learning_rate": 1.6314199395770393e-05, "loss": 0.0469, "step": 270 }, { "epoch": 0.012704966847977131, "grad_norm": 0.3968771994113922, "learning_rate": 1.691842900302115e-05, "loss": 0.0454, "step": 280 }, { "epoch": 0.013158715663976315, "grad_norm": 0.4650236666202545, "learning_rate": 1.7522658610271906e-05, "loss": 0.0451, "step": 290 }, { "epoch": 0.013612464479975498, "grad_norm": 0.4143680930137634, "learning_rate": 1.812688821752266e-05, "loss": 0.0377, "step": 300 }, { "epoch": 0.014066213295974682, "grad_norm": 0.37959203124046326, "learning_rate": 1.8731117824773413e-05, "loss": 0.0432, "step": 310 }, { "epoch": 0.014519962111973863, "grad_norm": 0.43344926834106445, "learning_rate": 1.933534743202417e-05, "loss": 0.0402, "step": 320 }, { "epoch": 0.014973710927973047, "grad_norm": 0.32157906889915466, "learning_rate": 1.9939577039274927e-05, "loss": 0.0372, "step": 330 }, { "epoch": 0.01542745974397223, "grad_norm": 0.3956120014190674, "learning_rate": 1.9999965008575334e-05, "loss": 0.0409, "step": 340 }, { "epoch": 0.015881208559971414, "grad_norm": 0.4085632264614105, "learning_rate": 1.999984405087852e-05, "loss": 0.0374, "step": 350 }, { "epoch": 0.016334957375970596, "grad_norm": 0.3496035933494568, "learning_rate": 1.999963669596147e-05, "loss": 0.0326, "step": 360 }, { "epoch": 0.01678870619196978, "grad_norm": 0.3280976116657257, "learning_rate": 1.9999342945615705e-05, "loss": 0.04, "step": 370 }, { "epoch": 0.017242455007968963, "grad_norm": 0.35505032539367676, "learning_rate": 1.9998962802379185e-05, "loss": 0.0336, "step": 380 }, { "epoch": 0.017696203823968148, "grad_norm": 0.36404356360435486, "learning_rate": 1.9998496269536293e-05, "loss": 0.0354, "step": 390 }, { "epoch": 0.01814995263996733, "grad_norm": 0.3447716534137726, "learning_rate": 1.9997943351117804e-05, "loss": 0.0368, "step": 400 }, { "epoch": 0.018603701455966515, "grad_norm": 0.26203060150146484, "learning_rate": 1.9997304051900853e-05, "loss": 0.0337, "step": 410 }, { "epoch": 0.019057450271965697, "grad_norm": 0.32277223467826843, "learning_rate": 1.9996578377408897e-05, "loss": 0.0332, "step": 420 }, { "epoch": 0.01951119908796488, "grad_norm": 0.43078309297561646, "learning_rate": 1.9995766333911663e-05, "loss": 0.0297, "step": 430 }, { "epoch": 0.019964947903964064, "grad_norm": 0.30445027351379395, "learning_rate": 1.999486792842508e-05, "loss": 0.0389, "step": 440 }, { "epoch": 0.020418696719963245, "grad_norm": 0.44035622477531433, "learning_rate": 1.999388316871125e-05, "loss": 0.0375, "step": 450 }, { "epoch": 0.02087244553596243, "grad_norm": 0.372060626745224, "learning_rate": 1.9992812063278354e-05, "loss": 0.0331, "step": 460 }, { "epoch": 0.021326194351961612, "grad_norm": 0.27705034613609314, "learning_rate": 1.9991654621380593e-05, "loss": 0.0248, "step": 470 }, { "epoch": 0.021779943167960798, "grad_norm": 0.2888132631778717, "learning_rate": 1.9990410853018094e-05, "loss": 0.0361, "step": 480 }, { "epoch": 0.02223369198395998, "grad_norm": 0.3042416274547577, "learning_rate": 1.998908076893684e-05, "loss": 0.029, "step": 490 }, { "epoch": 0.02268744079995916, "grad_norm": 0.24554918706417084, "learning_rate": 1.9987664380628566e-05, "loss": 0.0362, "step": 500 }, { "epoch": 0.023141189615958346, "grad_norm": 0.25402751564979553, "learning_rate": 1.9986161700330668e-05, "loss": 0.0296, "step": 510 }, { "epoch": 0.023594938431957528, "grad_norm": 0.3481331467628479, "learning_rate": 1.998457274102608e-05, "loss": 0.0329, "step": 520 }, { "epoch": 0.024048687247956713, "grad_norm": 0.31753596663475037, "learning_rate": 1.9982897516443194e-05, "loss": 0.0339, "step": 530 }, { "epoch": 0.024502436063955895, "grad_norm": 0.35809817910194397, "learning_rate": 1.9981136041055703e-05, "loss": 0.0353, "step": 540 }, { "epoch": 0.02495618487995508, "grad_norm": 0.3065260946750641, "learning_rate": 1.99792883300825e-05, "loss": 0.0353, "step": 550 }, { "epoch": 0.025409933695954262, "grad_norm": 0.34474602341651917, "learning_rate": 1.997735439948755e-05, "loss": 0.0336, "step": 560 }, { "epoch": 0.025863682511953444, "grad_norm": 0.269667387008667, "learning_rate": 1.997533426597973e-05, "loss": 0.032, "step": 570 }, { "epoch": 0.02631743132795263, "grad_norm": 0.39037978649139404, "learning_rate": 1.9973227947012713e-05, "loss": 0.0278, "step": 580 }, { "epoch": 0.02677118014395181, "grad_norm": 0.3307776153087616, "learning_rate": 1.9971035460784783e-05, "loss": 0.0344, "step": 590 }, { "epoch": 0.027224928959950996, "grad_norm": 0.2869550585746765, "learning_rate": 1.9968756826238713e-05, "loss": 0.0327, "step": 600 }, { "epoch": 0.027678677775950178, "grad_norm": 0.25692203640937805, "learning_rate": 1.9966392063061573e-05, "loss": 0.0346, "step": 610 }, { "epoch": 0.028132426591949363, "grad_norm": 0.21003565192222595, "learning_rate": 1.9963941191684585e-05, "loss": 0.0287, "step": 620 }, { "epoch": 0.028586175407948545, "grad_norm": 0.23621736466884613, "learning_rate": 1.9961404233282926e-05, "loss": 0.0283, "step": 630 }, { "epoch": 0.029039924223947727, "grad_norm": 0.33250758051872253, "learning_rate": 1.995878120977555e-05, "loss": 0.031, "step": 640 }, { "epoch": 0.029493673039946912, "grad_norm": 0.24574914574623108, "learning_rate": 1.9956072143825006e-05, "loss": 0.03, "step": 650 }, { "epoch": 0.029947421855946094, "grad_norm": 0.28682562708854675, "learning_rate": 1.9953277058837237e-05, "loss": 0.0304, "step": 660 }, { "epoch": 0.03040117067194528, "grad_norm": 0.4012982249259949, "learning_rate": 1.9950395978961376e-05, "loss": 0.0347, "step": 670 }, { "epoch": 0.03085491948794446, "grad_norm": 0.3115476369857788, "learning_rate": 1.9947428929089536e-05, "loss": 0.0289, "step": 680 }, { "epoch": 0.031308668303943646, "grad_norm": 0.24223241209983826, "learning_rate": 1.9944375934856606e-05, "loss": 0.0269, "step": 690 }, { "epoch": 0.03176241711994283, "grad_norm": 0.20952603220939636, "learning_rate": 1.9941237022640024e-05, "loss": 0.0266, "step": 700 }, { "epoch": 0.03221616593594201, "grad_norm": 0.269159197807312, "learning_rate": 1.9938012219559536e-05, "loss": 0.0256, "step": 710 }, { "epoch": 0.03266991475194119, "grad_norm": 0.21880276501178741, "learning_rate": 1.9934701553476983e-05, "loss": 0.0321, "step": 720 }, { "epoch": 0.03312366356794038, "grad_norm": 0.29239243268966675, "learning_rate": 1.993130505299604e-05, "loss": 0.0295, "step": 730 }, { "epoch": 0.03357741238393956, "grad_norm": 0.2072737067937851, "learning_rate": 1.9927822747461987e-05, "loss": 0.0262, "step": 740 }, { "epoch": 0.03403116119993874, "grad_norm": 0.23652520775794983, "learning_rate": 1.9924254666961446e-05, "loss": 0.0247, "step": 750 }, { "epoch": 0.034484910015937925, "grad_norm": 0.2358781099319458, "learning_rate": 1.9920600842322123e-05, "loss": 0.0211, "step": 760 }, { "epoch": 0.034938658831937114, "grad_norm": 0.2626686990261078, "learning_rate": 1.9916861305112536e-05, "loss": 0.0278, "step": 770 }, { "epoch": 0.035392407647936296, "grad_norm": 0.30486413836479187, "learning_rate": 1.9913036087641756e-05, "loss": 0.029, "step": 780 }, { "epoch": 0.03584615646393548, "grad_norm": 0.22027461230754852, "learning_rate": 1.9909125222959106e-05, "loss": 0.0249, "step": 790 }, { "epoch": 0.03629990527993466, "grad_norm": 0.36433184146881104, "learning_rate": 1.9905128744853903e-05, "loss": 0.0262, "step": 800 }, { "epoch": 0.03675365409593384, "grad_norm": 0.2639385759830475, "learning_rate": 1.9901046687855142e-05, "loss": 0.0332, "step": 810 }, { "epoch": 0.03720740291193303, "grad_norm": 0.2315681129693985, "learning_rate": 1.9896879087231212e-05, "loss": 0.0314, "step": 820 }, { "epoch": 0.03766115172793221, "grad_norm": 0.18850234150886536, "learning_rate": 1.989262597898959e-05, "loss": 0.0245, "step": 830 }, { "epoch": 0.03811490054393139, "grad_norm": 0.19689935445785522, "learning_rate": 1.9888287399876514e-05, "loss": 0.0262, "step": 840 }, { "epoch": 0.038568649359930575, "grad_norm": 0.25434932112693787, "learning_rate": 1.9883863387376688e-05, "loss": 0.0252, "step": 850 }, { "epoch": 0.03902239817592976, "grad_norm": 0.2297980934381485, "learning_rate": 1.9879353979712953e-05, "loss": 0.0289, "step": 860 }, { "epoch": 0.039476146991928945, "grad_norm": 0.18525287508964539, "learning_rate": 1.987475921584594e-05, "loss": 0.0214, "step": 870 }, { "epoch": 0.03992989580792813, "grad_norm": 0.2375829815864563, "learning_rate": 1.987007913547375e-05, "loss": 0.0281, "step": 880 }, { "epoch": 0.04038364462392731, "grad_norm": 0.2593577802181244, "learning_rate": 1.9865313779031607e-05, "loss": 0.0292, "step": 890 }, { "epoch": 0.04083739343992649, "grad_norm": 0.26970919966697693, "learning_rate": 1.986046318769151e-05, "loss": 0.0301, "step": 900 }, { "epoch": 0.04129114225592568, "grad_norm": 0.20702382922172546, "learning_rate": 1.9855527403361874e-05, "loss": 0.0282, "step": 910 }, { "epoch": 0.04174489107192486, "grad_norm": 0.23546156287193298, "learning_rate": 1.9850506468687164e-05, "loss": 0.0269, "step": 920 }, { "epoch": 0.04219863988792404, "grad_norm": 0.18576021492481232, "learning_rate": 1.9845400427047542e-05, "loss": 0.022, "step": 930 }, { "epoch": 0.042652388703923225, "grad_norm": 0.2637688219547272, "learning_rate": 1.9840209322558476e-05, "loss": 0.029, "step": 940 }, { "epoch": 0.043106137519922406, "grad_norm": 0.6882661581039429, "learning_rate": 1.983493320007036e-05, "loss": 0.0246, "step": 950 }, { "epoch": 0.043559886335921595, "grad_norm": 0.28765228390693665, "learning_rate": 1.9829572105168137e-05, "loss": 0.0308, "step": 960 }, { "epoch": 0.04401363515192078, "grad_norm": 0.2787792980670929, "learning_rate": 1.9824126084170907e-05, "loss": 0.0271, "step": 970 }, { "epoch": 0.04446738396791996, "grad_norm": 0.2273041307926178, "learning_rate": 1.9818595184131505e-05, "loss": 0.0296, "step": 980 }, { "epoch": 0.04492113278391914, "grad_norm": 0.26197728514671326, "learning_rate": 1.9812979452836117e-05, "loss": 0.0286, "step": 990 }, { "epoch": 0.04537488159991832, "grad_norm": 0.2646290957927704, "learning_rate": 1.9807278938803853e-05, "loss": 0.0281, "step": 1000 }, { "epoch": 0.04582863041591751, "grad_norm": 0.16532529890537262, "learning_rate": 1.980149369128634e-05, "loss": 0.0234, "step": 1010 }, { "epoch": 0.04628237923191669, "grad_norm": 0.2719618082046509, "learning_rate": 1.9795623760267294e-05, "loss": 0.0243, "step": 1020 }, { "epoch": 0.046736128047915874, "grad_norm": 0.16915248334407806, "learning_rate": 1.9789669196462072e-05, "loss": 0.0261, "step": 1030 }, { "epoch": 0.047189876863915056, "grad_norm": 0.26485320925712585, "learning_rate": 1.978363005131725e-05, "loss": 0.0265, "step": 1040 }, { "epoch": 0.047643625679914245, "grad_norm": 0.17664441466331482, "learning_rate": 1.9777506377010182e-05, "loss": 0.0241, "step": 1050 }, { "epoch": 0.04809737449591343, "grad_norm": 0.19183993339538574, "learning_rate": 1.9771298226448535e-05, "loss": 0.0222, "step": 1060 }, { "epoch": 0.04855112331191261, "grad_norm": 0.22114388644695282, "learning_rate": 1.9765005653269842e-05, "loss": 0.0203, "step": 1070 }, { "epoch": 0.04900487212791179, "grad_norm": 0.3064383566379547, "learning_rate": 1.9758628711841035e-05, "loss": 0.0283, "step": 1080 }, { "epoch": 0.04945862094391097, "grad_norm": 0.18376608192920685, "learning_rate": 1.975216745725797e-05, "loss": 0.0267, "step": 1090 }, { "epoch": 0.04991236975991016, "grad_norm": 0.279770165681839, "learning_rate": 1.974562194534496e-05, "loss": 0.0242, "step": 1100 }, { "epoch": 0.05036611857590934, "grad_norm": 0.37529680132865906, "learning_rate": 1.9738992232654296e-05, "loss": 0.0243, "step": 1110 }, { "epoch": 0.050819867391908524, "grad_norm": 0.20544101297855377, "learning_rate": 1.9732278376465746e-05, "loss": 0.0267, "step": 1120 }, { "epoch": 0.051273616207907706, "grad_norm": 0.31520286202430725, "learning_rate": 1.9725480434786065e-05, "loss": 0.0201, "step": 1130 }, { "epoch": 0.05172736502390689, "grad_norm": 0.2995913028717041, "learning_rate": 1.971859846634849e-05, "loss": 0.0264, "step": 1140 }, { "epoch": 0.052181113839906076, "grad_norm": 0.18620726466178894, "learning_rate": 1.9711632530612247e-05, "loss": 0.0268, "step": 1150 }, { "epoch": 0.05263486265590526, "grad_norm": 0.26744094491004944, "learning_rate": 1.970458268776202e-05, "loss": 0.0238, "step": 1160 }, { "epoch": 0.05308861147190444, "grad_norm": 0.16825206577777863, "learning_rate": 1.9697448998707448e-05, "loss": 0.023, "step": 1170 }, { "epoch": 0.05354236028790362, "grad_norm": 0.16264687478542328, "learning_rate": 1.9690231525082576e-05, "loss": 0.0244, "step": 1180 }, { "epoch": 0.05399610910390281, "grad_norm": 0.22812557220458984, "learning_rate": 1.968293032924535e-05, "loss": 0.0209, "step": 1190 }, { "epoch": 0.05444985791990199, "grad_norm": 0.22756661474704742, "learning_rate": 1.9675545474277045e-05, "loss": 0.0253, "step": 1200 }, { "epoch": 0.054903606735901174, "grad_norm": 0.24875099956989288, "learning_rate": 1.966807702398176e-05, "loss": 0.0213, "step": 1210 }, { "epoch": 0.055357355551900356, "grad_norm": 0.265860915184021, "learning_rate": 1.9660525042885828e-05, "loss": 0.0225, "step": 1220 }, { "epoch": 0.05581110436789954, "grad_norm": 0.17177896201610565, "learning_rate": 1.965288959623729e-05, "loss": 0.0246, "step": 1230 }, { "epoch": 0.056264853183898726, "grad_norm": 0.20631885528564453, "learning_rate": 1.964517075000531e-05, "loss": 0.0262, "step": 1240 }, { "epoch": 0.05671860199989791, "grad_norm": 0.1978541612625122, "learning_rate": 1.9637368570879612e-05, "loss": 0.0228, "step": 1250 }, { "epoch": 0.05717235081589709, "grad_norm": 0.30454206466674805, "learning_rate": 1.9629483126269904e-05, "loss": 0.0294, "step": 1260 }, { "epoch": 0.05762609963189627, "grad_norm": 0.21276606619358063, "learning_rate": 1.9621514484305308e-05, "loss": 0.0256, "step": 1270 }, { "epoch": 0.05807984844789545, "grad_norm": 0.2528305947780609, "learning_rate": 1.9613462713833734e-05, "loss": 0.0251, "step": 1280 }, { "epoch": 0.05853359726389464, "grad_norm": 0.16171447932720184, "learning_rate": 1.9605327884421338e-05, "loss": 0.0249, "step": 1290 }, { "epoch": 0.058987346079893824, "grad_norm": 0.2728146016597748, "learning_rate": 1.9597110066351875e-05, "loss": 0.024, "step": 1300 }, { "epoch": 0.059441094895893005, "grad_norm": 0.20263759791851044, "learning_rate": 1.958880933062612e-05, "loss": 0.0217, "step": 1310 }, { "epoch": 0.05989484371189219, "grad_norm": 0.18950790166854858, "learning_rate": 1.958042574896124e-05, "loss": 0.0211, "step": 1320 }, { "epoch": 0.060348592527891376, "grad_norm": 0.18953697383403778, "learning_rate": 1.9571959393790174e-05, "loss": 0.0222, "step": 1330 }, { "epoch": 0.06080234134389056, "grad_norm": 0.18523451685905457, "learning_rate": 1.9563410338261022e-05, "loss": 0.0238, "step": 1340 }, { "epoch": 0.06125609015988974, "grad_norm": 0.18328619003295898, "learning_rate": 1.9554778656236402e-05, "loss": 0.0238, "step": 1350 }, { "epoch": 0.06170983897588892, "grad_norm": 0.18588705360889435, "learning_rate": 1.9546064422292806e-05, "loss": 0.0217, "step": 1360 }, { "epoch": 0.0621635877918881, "grad_norm": 0.3565521836280823, "learning_rate": 1.9537267711719966e-05, "loss": 0.0282, "step": 1370 }, { "epoch": 0.06261733660788729, "grad_norm": 0.2516770362854004, "learning_rate": 1.9528388600520208e-05, "loss": 0.0224, "step": 1380 }, { "epoch": 0.06307108542388647, "grad_norm": 0.20207948982715607, "learning_rate": 1.9519427165407773e-05, "loss": 0.0212, "step": 1390 }, { "epoch": 0.06352483423988566, "grad_norm": 0.19543805718421936, "learning_rate": 1.9510383483808183e-05, "loss": 0.0233, "step": 1400 }, { "epoch": 0.06397858305588484, "grad_norm": 0.16665951907634735, "learning_rate": 1.950125763385755e-05, "loss": 0.0182, "step": 1410 }, { "epoch": 0.06443233187188402, "grad_norm": 0.13467726111412048, "learning_rate": 1.949204969440191e-05, "loss": 0.024, "step": 1420 }, { "epoch": 0.06488608068788321, "grad_norm": 0.18848440051078796, "learning_rate": 1.9482759744996537e-05, "loss": 0.0257, "step": 1430 }, { "epoch": 0.06533982950388238, "grad_norm": 0.13075998425483704, "learning_rate": 1.9473387865905268e-05, "loss": 0.0261, "step": 1440 }, { "epoch": 0.06579357831988157, "grad_norm": 0.24777375161647797, "learning_rate": 1.9463934138099796e-05, "loss": 0.0242, "step": 1450 }, { "epoch": 0.06624732713588076, "grad_norm": 0.13372217118740082, "learning_rate": 1.945439864325897e-05, "loss": 0.0216, "step": 1460 }, { "epoch": 0.06670107595187993, "grad_norm": 0.2603487968444824, "learning_rate": 1.944478146376811e-05, "loss": 0.0243, "step": 1470 }, { "epoch": 0.06715482476787912, "grad_norm": 0.20325487852096558, "learning_rate": 1.943508268271826e-05, "loss": 0.0262, "step": 1480 }, { "epoch": 0.0676085735838783, "grad_norm": 0.23849761486053467, "learning_rate": 1.9425302383905497e-05, "loss": 0.0247, "step": 1490 }, { "epoch": 0.06806232239987749, "grad_norm": 0.17411549389362335, "learning_rate": 1.941544065183021e-05, "loss": 0.0213, "step": 1500 }, { "epoch": 0.06851607121587668, "grad_norm": 0.2012050598859787, "learning_rate": 1.9405497571696347e-05, "loss": 0.0215, "step": 1510 }, { "epoch": 0.06896982003187585, "grad_norm": 0.19042851030826569, "learning_rate": 1.93954732294107e-05, "loss": 0.0228, "step": 1520 }, { "epoch": 0.06942356884787504, "grad_norm": 0.20282061398029327, "learning_rate": 1.9385367711582142e-05, "loss": 0.0199, "step": 1530 }, { "epoch": 0.06987731766387423, "grad_norm": 0.22637954354286194, "learning_rate": 1.9375181105520907e-05, "loss": 0.0188, "step": 1540 }, { "epoch": 0.0703310664798734, "grad_norm": 0.17371779680252075, "learning_rate": 1.9364913499237814e-05, "loss": 0.0169, "step": 1550 }, { "epoch": 0.07078481529587259, "grad_norm": 0.20600782334804535, "learning_rate": 1.93545649814435e-05, "loss": 0.0194, "step": 1560 }, { "epoch": 0.07123856411187177, "grad_norm": 0.19030754268169403, "learning_rate": 1.934413564154769e-05, "loss": 0.0187, "step": 1570 }, { "epoch": 0.07169231292787095, "grad_norm": 0.3098970651626587, "learning_rate": 1.9333625569658377e-05, "loss": 0.0299, "step": 1580 }, { "epoch": 0.07214606174387014, "grad_norm": 0.16969068348407745, "learning_rate": 1.9323034856581083e-05, "loss": 0.0212, "step": 1590 }, { "epoch": 0.07259981055986932, "grad_norm": 0.17913642525672913, "learning_rate": 1.9312363593818045e-05, "loss": 0.0238, "step": 1600 }, { "epoch": 0.07305355937586851, "grad_norm": 0.27264851331710815, "learning_rate": 1.930161187356745e-05, "loss": 0.0202, "step": 1610 }, { "epoch": 0.07350730819186768, "grad_norm": 0.23685918748378754, "learning_rate": 1.929077978872262e-05, "loss": 0.0225, "step": 1620 }, { "epoch": 0.07396105700786687, "grad_norm": 0.23093511164188385, "learning_rate": 1.9279867432871215e-05, "loss": 0.0206, "step": 1630 }, { "epoch": 0.07441480582386606, "grad_norm": 0.19286802411079407, "learning_rate": 1.9268874900294426e-05, "loss": 0.0242, "step": 1640 }, { "epoch": 0.07486855463986523, "grad_norm": 0.15499348938465118, "learning_rate": 1.9257802285966166e-05, "loss": 0.0211, "step": 1650 }, { "epoch": 0.07532230345586442, "grad_norm": 0.25941014289855957, "learning_rate": 1.924664968555223e-05, "loss": 0.0234, "step": 1660 }, { "epoch": 0.0757760522718636, "grad_norm": 0.2614118456840515, "learning_rate": 1.9235417195409487e-05, "loss": 0.0203, "step": 1670 }, { "epoch": 0.07622980108786279, "grad_norm": 0.24374106526374817, "learning_rate": 1.922410491258505e-05, "loss": 0.0225, "step": 1680 }, { "epoch": 0.07668354990386198, "grad_norm": 0.2553516626358032, "learning_rate": 1.9212712934815413e-05, "loss": 0.0253, "step": 1690 }, { "epoch": 0.07713729871986115, "grad_norm": 0.18607620894908905, "learning_rate": 1.9201241360525643e-05, "loss": 0.0205, "step": 1700 }, { "epoch": 0.07759104753586034, "grad_norm": 0.2262287437915802, "learning_rate": 1.9189690288828487e-05, "loss": 0.0245, "step": 1710 }, { "epoch": 0.07804479635185951, "grad_norm": 0.24479566514492035, "learning_rate": 1.9178059819523563e-05, "loss": 0.0208, "step": 1720 }, { "epoch": 0.0784985451678587, "grad_norm": 0.2556163966655731, "learning_rate": 1.9166350053096453e-05, "loss": 0.0215, "step": 1730 }, { "epoch": 0.07895229398385789, "grad_norm": 0.8043221831321716, "learning_rate": 1.9154561090717857e-05, "loss": 0.021, "step": 1740 }, { "epoch": 0.07940604279985707, "grad_norm": 0.2874302864074707, "learning_rate": 1.9142693034242726e-05, "loss": 0.0194, "step": 1750 }, { "epoch": 0.07985979161585625, "grad_norm": 0.2852942943572998, "learning_rate": 1.913074598620937e-05, "loss": 0.0228, "step": 1760 }, { "epoch": 0.08031354043185543, "grad_norm": 0.18303954601287842, "learning_rate": 1.9118720049838567e-05, "loss": 0.0212, "step": 1770 }, { "epoch": 0.08076728924785462, "grad_norm": 0.14635030925273895, "learning_rate": 1.9106615329032695e-05, "loss": 0.0226, "step": 1780 }, { "epoch": 0.0812210380638538, "grad_norm": 0.17153240740299225, "learning_rate": 1.9094431928374798e-05, "loss": 0.0195, "step": 1790 }, { "epoch": 0.08167478687985298, "grad_norm": 0.23255468904972076, "learning_rate": 1.9082169953127714e-05, "loss": 0.0206, "step": 1800 }, { "epoch": 0.08212853569585217, "grad_norm": 0.17257371544837952, "learning_rate": 1.9069829509233156e-05, "loss": 0.0191, "step": 1810 }, { "epoch": 0.08258228451185136, "grad_norm": 0.15475516021251678, "learning_rate": 1.9057410703310788e-05, "loss": 0.0193, "step": 1820 }, { "epoch": 0.08303603332785053, "grad_norm": 0.17117761075496674, "learning_rate": 1.9044913642657318e-05, "loss": 0.0204, "step": 1830 }, { "epoch": 0.08348978214384972, "grad_norm": 0.20223674178123474, "learning_rate": 1.9032338435245557e-05, "loss": 0.0204, "step": 1840 }, { "epoch": 0.0839435309598489, "grad_norm": 0.24889586865901947, "learning_rate": 1.9019685189723497e-05, "loss": 0.0244, "step": 1850 }, { "epoch": 0.08439727977584809, "grad_norm": 0.2555040419101715, "learning_rate": 1.900695401541337e-05, "loss": 0.0203, "step": 1860 }, { "epoch": 0.08485102859184727, "grad_norm": 0.33612123131752014, "learning_rate": 1.8994145022310693e-05, "loss": 0.0187, "step": 1870 }, { "epoch": 0.08530477740784645, "grad_norm": 0.17665745317935944, "learning_rate": 1.8981258321083335e-05, "loss": 0.0221, "step": 1880 }, { "epoch": 0.08575852622384564, "grad_norm": 0.22196491062641144, "learning_rate": 1.8968294023070548e-05, "loss": 0.0203, "step": 1890 }, { "epoch": 0.08621227503984481, "grad_norm": 0.16377434134483337, "learning_rate": 1.895525224028201e-05, "loss": 0.0213, "step": 1900 }, { "epoch": 0.086666023855844, "grad_norm": 0.1716269701719284, "learning_rate": 1.8942133085396855e-05, "loss": 0.0175, "step": 1910 }, { "epoch": 0.08711977267184319, "grad_norm": 0.15549753606319427, "learning_rate": 1.8928936671762704e-05, "loss": 0.0211, "step": 1920 }, { "epoch": 0.08757352148784237, "grad_norm": 0.20391225814819336, "learning_rate": 1.8915663113394677e-05, "loss": 0.0243, "step": 1930 }, { "epoch": 0.08802727030384155, "grad_norm": 0.20363971590995789, "learning_rate": 1.890231252497442e-05, "loss": 0.0191, "step": 1940 }, { "epoch": 0.08848101911984073, "grad_norm": 0.16531780362129211, "learning_rate": 1.8888885021849103e-05, "loss": 0.0202, "step": 1950 }, { "epoch": 0.08893476793583992, "grad_norm": 0.19257767498493195, "learning_rate": 1.8875380720030434e-05, "loss": 0.0224, "step": 1960 }, { "epoch": 0.0893885167518391, "grad_norm": 0.21032428741455078, "learning_rate": 1.886179973619364e-05, "loss": 0.0221, "step": 1970 }, { "epoch": 0.08984226556783828, "grad_norm": 0.258282870054245, "learning_rate": 1.8848142187676485e-05, "loss": 0.0203, "step": 1980 }, { "epoch": 0.09029601438383747, "grad_norm": 0.15617872774600983, "learning_rate": 1.883440819247822e-05, "loss": 0.0225, "step": 1990 }, { "epoch": 0.09074976319983664, "grad_norm": 0.10580911487340927, "learning_rate": 1.8820597869258606e-05, "loss": 0.0208, "step": 2000 }, { "epoch": 0.09120351201583583, "grad_norm": 0.15046417713165283, "learning_rate": 1.8806711337336852e-05, "loss": 0.0167, "step": 2010 }, { "epoch": 0.09165726083183502, "grad_norm": 0.19403544068336487, "learning_rate": 1.8792748716690608e-05, "loss": 0.0233, "step": 2020 }, { "epoch": 0.0921110096478342, "grad_norm": 0.18249863386154175, "learning_rate": 1.8778710127954912e-05, "loss": 0.0233, "step": 2030 }, { "epoch": 0.09256475846383339, "grad_norm": 0.19264620542526245, "learning_rate": 1.8764595692421163e-05, "loss": 0.0199, "step": 2040 }, { "epoch": 0.09301850727983256, "grad_norm": 0.21285632252693176, "learning_rate": 1.8750405532036064e-05, "loss": 0.0207, "step": 2050 }, { "epoch": 0.09347225609583175, "grad_norm": 0.18511101603507996, "learning_rate": 1.8736139769400567e-05, "loss": 0.0211, "step": 2060 }, { "epoch": 0.09392600491183094, "grad_norm": 0.15465500950813293, "learning_rate": 1.8721798527768813e-05, "loss": 0.0199, "step": 2070 }, { "epoch": 0.09437975372783011, "grad_norm": 0.155814990401268, "learning_rate": 1.870738193104708e-05, "loss": 0.0193, "step": 2080 }, { "epoch": 0.0948335025438293, "grad_norm": 0.11972299963235855, "learning_rate": 1.86928901037927e-05, "loss": 0.0184, "step": 2090 }, { "epoch": 0.09528725135982849, "grad_norm": 0.19595398008823395, "learning_rate": 1.8678323171212982e-05, "loss": 0.0257, "step": 2100 }, { "epoch": 0.09574100017582766, "grad_norm": 0.24550503492355347, "learning_rate": 1.866368125916414e-05, "loss": 0.0189, "step": 2110 }, { "epoch": 0.09619474899182685, "grad_norm": 0.1504066437482834, "learning_rate": 1.864896449415019e-05, "loss": 0.0172, "step": 2120 }, { "epoch": 0.09664849780782603, "grad_norm": 0.2157190591096878, "learning_rate": 1.863417300332188e-05, "loss": 0.0193, "step": 2130 }, { "epoch": 0.09710224662382522, "grad_norm": 0.2792034149169922, "learning_rate": 1.8619306914475573e-05, "loss": 0.0199, "step": 2140 }, { "epoch": 0.0975559954398244, "grad_norm": 0.14018727838993073, "learning_rate": 1.860436635605214e-05, "loss": 0.0216, "step": 2150 }, { "epoch": 0.09800974425582358, "grad_norm": 0.17746350169181824, "learning_rate": 1.8589351457135873e-05, "loss": 0.0171, "step": 2160 }, { "epoch": 0.09846349307182277, "grad_norm": 0.1806003302335739, "learning_rate": 1.8574262347453344e-05, "loss": 0.0238, "step": 2170 }, { "epoch": 0.09891724188782194, "grad_norm": 0.15680097043514252, "learning_rate": 1.85590991573723e-05, "loss": 0.0205, "step": 2180 }, { "epoch": 0.09937099070382113, "grad_norm": 0.14457744359970093, "learning_rate": 1.854386201790053e-05, "loss": 0.0173, "step": 2190 }, { "epoch": 0.09982473951982032, "grad_norm": 0.1748875379562378, "learning_rate": 1.8528551060684744e-05, "loss": 0.0204, "step": 2200 }, { "epoch": 0.1002784883358195, "grad_norm": 0.2859458923339844, "learning_rate": 1.851316641800941e-05, "loss": 0.0204, "step": 2210 }, { "epoch": 0.10073223715181868, "grad_norm": 0.16153216361999512, "learning_rate": 1.8497708222795638e-05, "loss": 0.019, "step": 2220 }, { "epoch": 0.10118598596781786, "grad_norm": 0.1441313922405243, "learning_rate": 1.8482176608600025e-05, "loss": 0.0202, "step": 2230 }, { "epoch": 0.10163973478381705, "grad_norm": 0.19473570585250854, "learning_rate": 1.846657170961349e-05, "loss": 0.0237, "step": 2240 }, { "epoch": 0.10209348359981624, "grad_norm": 1.6714526414871216, "learning_rate": 1.8450893660660126e-05, "loss": 0.0226, "step": 2250 }, { "epoch": 0.10254723241581541, "grad_norm": 0.1397905796766281, "learning_rate": 1.8435142597196033e-05, "loss": 0.02, "step": 2260 }, { "epoch": 0.1030009812318146, "grad_norm": 0.31626781821250916, "learning_rate": 1.8419318655308135e-05, "loss": 0.021, "step": 2270 }, { "epoch": 0.10345473004781378, "grad_norm": 0.2210594266653061, "learning_rate": 1.8403421971713034e-05, "loss": 0.0249, "step": 2280 }, { "epoch": 0.10390847886381296, "grad_norm": 0.26183414459228516, "learning_rate": 1.838745268375579e-05, "loss": 0.0182, "step": 2290 }, { "epoch": 0.10436222767981215, "grad_norm": 0.21002666652202606, "learning_rate": 1.8371410929408767e-05, "loss": 0.0202, "step": 2300 }, { "epoch": 0.10481597649581133, "grad_norm": 0.206701397895813, "learning_rate": 1.835529684727043e-05, "loss": 0.0173, "step": 2310 }, { "epoch": 0.10526972531181052, "grad_norm": 0.16457580029964447, "learning_rate": 1.8339110576564132e-05, "loss": 0.0147, "step": 2320 }, { "epoch": 0.10572347412780969, "grad_norm": 0.1727597862482071, "learning_rate": 1.8322852257136935e-05, "loss": 0.0174, "step": 2330 }, { "epoch": 0.10617722294380888, "grad_norm": 0.18292547762393951, "learning_rate": 1.8306522029458395e-05, "loss": 0.0171, "step": 2340 }, { "epoch": 0.10663097175980807, "grad_norm": 0.22897396981716156, "learning_rate": 1.8290120034619335e-05, "loss": 0.0197, "step": 2350 }, { "epoch": 0.10708472057580724, "grad_norm": 0.18409202992916107, "learning_rate": 1.8273646414330645e-05, "loss": 0.0207, "step": 2360 }, { "epoch": 0.10753846939180643, "grad_norm": 0.22689014673233032, "learning_rate": 1.8257101310922042e-05, "loss": 0.0237, "step": 2370 }, { "epoch": 0.10799221820780562, "grad_norm": 0.19369293749332428, "learning_rate": 1.8240484867340852e-05, "loss": 0.0159, "step": 2380 }, { "epoch": 0.1084459670238048, "grad_norm": 0.13972799479961395, "learning_rate": 1.8223797227150762e-05, "loss": 0.02, "step": 2390 }, { "epoch": 0.10889971583980398, "grad_norm": 0.16729070246219635, "learning_rate": 1.8207038534530598e-05, "loss": 0.0183, "step": 2400 }, { "epoch": 0.10935346465580316, "grad_norm": 0.21379055082798004, "learning_rate": 1.819020893427306e-05, "loss": 0.0184, "step": 2410 }, { "epoch": 0.10980721347180235, "grad_norm": 0.24638915061950684, "learning_rate": 1.817330857178349e-05, "loss": 0.0235, "step": 2420 }, { "epoch": 0.11026096228780154, "grad_norm": 0.1810813695192337, "learning_rate": 1.8156337593078594e-05, "loss": 0.0179, "step": 2430 }, { "epoch": 0.11071471110380071, "grad_norm": 0.21317435801029205, "learning_rate": 1.81392961447852e-05, "loss": 0.0189, "step": 2440 }, { "epoch": 0.1111684599197999, "grad_norm": 0.17048487067222595, "learning_rate": 1.8122184374138973e-05, "loss": 0.0183, "step": 2450 }, { "epoch": 0.11162220873579907, "grad_norm": 0.1683121770620346, "learning_rate": 1.810500242898317e-05, "loss": 0.0195, "step": 2460 }, { "epoch": 0.11207595755179826, "grad_norm": 0.15048103034496307, "learning_rate": 1.808775045776733e-05, "loss": 0.0205, "step": 2470 }, { "epoch": 0.11252970636779745, "grad_norm": 0.21140700578689575, "learning_rate": 1.8070428609546012e-05, "loss": 0.0214, "step": 2480 }, { "epoch": 0.11298345518379663, "grad_norm": 0.18310706317424774, "learning_rate": 1.8053037033977513e-05, "loss": 0.0158, "step": 2490 }, { "epoch": 0.11343720399979582, "grad_norm": 0.20921604335308075, "learning_rate": 1.803557588132254e-05, "loss": 0.0171, "step": 2500 }, { "epoch": 0.11389095281579499, "grad_norm": 0.20439156889915466, "learning_rate": 1.8018045302442966e-05, "loss": 0.0221, "step": 2510 }, { "epoch": 0.11434470163179418, "grad_norm": 0.19221094250679016, "learning_rate": 1.8000445448800473e-05, "loss": 0.017, "step": 2520 }, { "epoch": 0.11479845044779337, "grad_norm": 0.18272683024406433, "learning_rate": 1.7982776472455274e-05, "loss": 0.0187, "step": 2530 }, { "epoch": 0.11525219926379254, "grad_norm": 0.20446264743804932, "learning_rate": 1.7965038526064796e-05, "loss": 0.0253, "step": 2540 }, { "epoch": 0.11570594807979173, "grad_norm": 0.1428219974040985, "learning_rate": 1.794723176288236e-05, "loss": 0.0221, "step": 2550 }, { "epoch": 0.1161596968957909, "grad_norm": 0.3095003068447113, "learning_rate": 1.7929356336755842e-05, "loss": 0.019, "step": 2560 }, { "epoch": 0.1166134457117901, "grad_norm": 0.2172461450099945, "learning_rate": 1.7911412402126366e-05, "loss": 0.0209, "step": 2570 }, { "epoch": 0.11706719452778928, "grad_norm": 0.15558798611164093, "learning_rate": 1.789340011402696e-05, "loss": 0.0203, "step": 2580 }, { "epoch": 0.11752094334378846, "grad_norm": 0.20151139795780182, "learning_rate": 1.7875319628081205e-05, "loss": 0.0199, "step": 2590 }, { "epoch": 0.11797469215978765, "grad_norm": 0.2480895221233368, "learning_rate": 1.785717110050192e-05, "loss": 0.0196, "step": 2600 }, { "epoch": 0.11842844097578682, "grad_norm": 0.24924373626708984, "learning_rate": 1.7838954688089777e-05, "loss": 0.0195, "step": 2610 }, { "epoch": 0.11888218979178601, "grad_norm": 0.20633499324321747, "learning_rate": 1.782067054823197e-05, "loss": 0.0198, "step": 2620 }, { "epoch": 0.1193359386077852, "grad_norm": 0.2311122715473175, "learning_rate": 1.7802318838900855e-05, "loss": 0.0193, "step": 2630 }, { "epoch": 0.11978968742378437, "grad_norm": 0.2317935824394226, "learning_rate": 1.7783899718652563e-05, "loss": 0.0184, "step": 2640 }, { "epoch": 0.12024343623978356, "grad_norm": 0.1745319962501526, "learning_rate": 1.776541334662566e-05, "loss": 0.0209, "step": 2650 }, { "epoch": 0.12069718505578275, "grad_norm": 0.19152095913887024, "learning_rate": 1.7746859882539747e-05, "loss": 0.0208, "step": 2660 }, { "epoch": 0.12115093387178193, "grad_norm": 0.21103863418102264, "learning_rate": 1.7728239486694104e-05, "loss": 0.0216, "step": 2670 }, { "epoch": 0.12160468268778112, "grad_norm": 0.11558584868907928, "learning_rate": 1.7709552319966275e-05, "loss": 0.0195, "step": 2680 }, { "epoch": 0.12205843150378029, "grad_norm": 0.19106024503707886, "learning_rate": 1.7690798543810715e-05, "loss": 0.0206, "step": 2690 }, { "epoch": 0.12251218031977948, "grad_norm": 0.22221724689006805, "learning_rate": 1.7671978320257356e-05, "loss": 0.0193, "step": 2700 }, { "epoch": 0.12296592913577867, "grad_norm": 0.2026868611574173, "learning_rate": 1.7653091811910236e-05, "loss": 0.0173, "step": 2710 }, { "epoch": 0.12341967795177784, "grad_norm": 0.16690614819526672, "learning_rate": 1.763413918194608e-05, "loss": 0.0201, "step": 2720 }, { "epoch": 0.12387342676777703, "grad_norm": 0.13100586831569672, "learning_rate": 1.7615120594112895e-05, "loss": 0.0132, "step": 2730 }, { "epoch": 0.1243271755837762, "grad_norm": 0.20827825367450714, "learning_rate": 1.7596036212728558e-05, "loss": 0.0149, "step": 2740 }, { "epoch": 0.1247809243997754, "grad_norm": 0.11902619153261185, "learning_rate": 1.757688620267939e-05, "loss": 0.0184, "step": 2750 }, { "epoch": 0.12523467321577458, "grad_norm": 0.18187430500984192, "learning_rate": 1.755767072941874e-05, "loss": 0.0187, "step": 2760 }, { "epoch": 0.12568842203177377, "grad_norm": 0.1461418867111206, "learning_rate": 1.7538389958965537e-05, "loss": 0.0207, "step": 2770 }, { "epoch": 0.12614217084777293, "grad_norm": 0.1332821249961853, "learning_rate": 1.7519044057902877e-05, "loss": 0.0201, "step": 2780 }, { "epoch": 0.12659591966377212, "grad_norm": 0.16073133051395416, "learning_rate": 1.749963319337658e-05, "loss": 0.0219, "step": 2790 }, { "epoch": 0.1270496684797713, "grad_norm": 0.1910991072654724, "learning_rate": 1.748015753309373e-05, "loss": 0.0148, "step": 2800 }, { "epoch": 0.1275034172957705, "grad_norm": 0.1660872995853424, "learning_rate": 1.746061724532124e-05, "loss": 0.0192, "step": 2810 }, { "epoch": 0.1279571661117697, "grad_norm": 0.1372659057378769, "learning_rate": 1.7441012498884402e-05, "loss": 0.0158, "step": 2820 }, { "epoch": 0.12841091492776885, "grad_norm": 0.14543624222278595, "learning_rate": 1.7421343463165415e-05, "loss": 0.0192, "step": 2830 }, { "epoch": 0.12886466374376804, "grad_norm": 0.2286180555820465, "learning_rate": 1.7401610308101933e-05, "loss": 0.0205, "step": 2840 }, { "epoch": 0.12931841255976723, "grad_norm": 0.11448660492897034, "learning_rate": 1.7381813204185585e-05, "loss": 0.0139, "step": 2850 }, { "epoch": 0.12977216137576641, "grad_norm": 0.10879474133253098, "learning_rate": 1.7361952322460513e-05, "loss": 0.0174, "step": 2860 }, { "epoch": 0.1302259101917656, "grad_norm": 0.17434784770011902, "learning_rate": 1.7342027834521896e-05, "loss": 0.0173, "step": 2870 }, { "epoch": 0.13067965900776476, "grad_norm": 0.21135075390338898, "learning_rate": 1.7322039912514453e-05, "loss": 0.0148, "step": 2880 }, { "epoch": 0.13113340782376395, "grad_norm": 0.1227855533361435, "learning_rate": 1.7301988729130964e-05, "loss": 0.0178, "step": 2890 }, { "epoch": 0.13158715663976314, "grad_norm": 0.1610281616449356, "learning_rate": 1.7281874457610787e-05, "loss": 0.0166, "step": 2900 }, { "epoch": 0.13204090545576233, "grad_norm": 0.14950548112392426, "learning_rate": 1.7261697271738337e-05, "loss": 0.0203, "step": 2910 }, { "epoch": 0.13249465427176152, "grad_norm": 0.19927935302257538, "learning_rate": 1.724145734584162e-05, "loss": 0.016, "step": 2920 }, { "epoch": 0.13294840308776068, "grad_norm": 0.12208227068185806, "learning_rate": 1.7221154854790696e-05, "loss": 0.0164, "step": 2930 }, { "epoch": 0.13340215190375987, "grad_norm": 0.19058668613433838, "learning_rate": 1.7200789973996172e-05, "loss": 0.0176, "step": 2940 }, { "epoch": 0.13385590071975906, "grad_norm": 0.1380406767129898, "learning_rate": 1.7180362879407707e-05, "loss": 0.0173, "step": 2950 }, { "epoch": 0.13430964953575825, "grad_norm": 0.1319362074136734, "learning_rate": 1.7159873747512472e-05, "loss": 0.0181, "step": 2960 }, { "epoch": 0.13476339835175744, "grad_norm": 0.16067825257778168, "learning_rate": 1.713932275533363e-05, "loss": 0.0186, "step": 2970 }, { "epoch": 0.1352171471677566, "grad_norm": 0.1704990565776825, "learning_rate": 1.7118710080428807e-05, "loss": 0.0187, "step": 2980 }, { "epoch": 0.13567089598375578, "grad_norm": 0.22068627178668976, "learning_rate": 1.7098035900888566e-05, "loss": 0.0173, "step": 2990 }, { "epoch": 0.13612464479975497, "grad_norm": 0.15751402080059052, "learning_rate": 1.7077300395334857e-05, "loss": 0.0216, "step": 3000 }, { "epoch": 0.13657839361575416, "grad_norm": 0.26060059666633606, "learning_rate": 1.7056503742919476e-05, "loss": 0.0183, "step": 3010 }, { "epoch": 0.13703214243175335, "grad_norm": 0.25333723425865173, "learning_rate": 1.703564612332252e-05, "loss": 0.0169, "step": 3020 }, { "epoch": 0.1374858912477525, "grad_norm": 0.11502993106842041, "learning_rate": 1.7014727716750842e-05, "loss": 0.0177, "step": 3030 }, { "epoch": 0.1379396400637517, "grad_norm": 0.15380710363388062, "learning_rate": 1.699374870393647e-05, "loss": 0.0179, "step": 3040 }, { "epoch": 0.1383933888797509, "grad_norm": 0.12410028278827667, "learning_rate": 1.697270926613507e-05, "loss": 0.0169, "step": 3050 }, { "epoch": 0.13884713769575008, "grad_norm": 0.22475175559520721, "learning_rate": 1.6951609585124377e-05, "loss": 0.0201, "step": 3060 }, { "epoch": 0.13930088651174927, "grad_norm": 0.1766625940799713, "learning_rate": 1.6930449843202607e-05, "loss": 0.0179, "step": 3070 }, { "epoch": 0.13975463532774846, "grad_norm": 0.1324724704027176, "learning_rate": 1.69092302231869e-05, "loss": 0.0198, "step": 3080 }, { "epoch": 0.14020838414374762, "grad_norm": 0.13342659175395966, "learning_rate": 1.688795090841173e-05, "loss": 0.0162, "step": 3090 }, { "epoch": 0.1406621329597468, "grad_norm": 0.17313902080059052, "learning_rate": 1.686661208272734e-05, "loss": 0.0165, "step": 3100 }, { "epoch": 0.141115881775746, "grad_norm": 0.1365835964679718, "learning_rate": 1.6845213930498122e-05, "loss": 0.0189, "step": 3110 }, { "epoch": 0.14156963059174518, "grad_norm": 0.11911448836326599, "learning_rate": 1.682375663660104e-05, "loss": 0.0155, "step": 3120 }, { "epoch": 0.14202337940774437, "grad_norm": 0.15301299095153809, "learning_rate": 1.680224038642405e-05, "loss": 0.0151, "step": 3130 }, { "epoch": 0.14247712822374353, "grad_norm": 0.18948419392108917, "learning_rate": 1.6780665365864465e-05, "loss": 0.0197, "step": 3140 }, { "epoch": 0.14293087703974272, "grad_norm": 0.17763479053974152, "learning_rate": 1.675903176132737e-05, "loss": 0.0163, "step": 3150 }, { "epoch": 0.1433846258557419, "grad_norm": 0.14466649293899536, "learning_rate": 1.6737339759724016e-05, "loss": 0.0177, "step": 3160 }, { "epoch": 0.1438383746717411, "grad_norm": 0.1917739361524582, "learning_rate": 1.6715589548470187e-05, "loss": 0.0165, "step": 3170 }, { "epoch": 0.1442921234877403, "grad_norm": 0.15840497612953186, "learning_rate": 1.669378131548459e-05, "loss": 0.0162, "step": 3180 }, { "epoch": 0.14474587230373945, "grad_norm": 0.13481085002422333, "learning_rate": 1.6671915249187237e-05, "loss": 0.0145, "step": 3190 }, { "epoch": 0.14519962111973864, "grad_norm": 0.1597055047750473, "learning_rate": 1.6649991538497808e-05, "loss": 0.0176, "step": 3200 }, { "epoch": 0.14565336993573783, "grad_norm": 0.19727839529514313, "learning_rate": 1.6628010372834028e-05, "loss": 0.0196, "step": 3210 }, { "epoch": 0.14610711875173701, "grad_norm": 0.15541918575763702, "learning_rate": 1.660597194211001e-05, "loss": 0.0177, "step": 3220 }, { "epoch": 0.1465608675677362, "grad_norm": 0.15050934255123138, "learning_rate": 1.6583876436734646e-05, "loss": 0.0176, "step": 3230 }, { "epoch": 0.14701461638373536, "grad_norm": 0.18570442497730255, "learning_rate": 1.6561724047609936e-05, "loss": 0.0164, "step": 3240 }, { "epoch": 0.14746836519973455, "grad_norm": 0.22303296625614166, "learning_rate": 1.653951496612935e-05, "loss": 0.0153, "step": 3250 }, { "epoch": 0.14792211401573374, "grad_norm": 0.14683152735233307, "learning_rate": 1.6517249384176163e-05, "loss": 0.0139, "step": 3260 }, { "epoch": 0.14837586283173293, "grad_norm": 0.32259002327919006, "learning_rate": 1.6494927494121827e-05, "loss": 0.015, "step": 3270 }, { "epoch": 0.14882961164773212, "grad_norm": 0.1250266432762146, "learning_rate": 1.647254948882426e-05, "loss": 0.0156, "step": 3280 }, { "epoch": 0.14928336046373128, "grad_norm": 0.14093117415905, "learning_rate": 1.6450115561626237e-05, "loss": 0.0201, "step": 3290 }, { "epoch": 0.14973710927973047, "grad_norm": 0.13918519020080566, "learning_rate": 1.6427625906353667e-05, "loss": 0.0173, "step": 3300 }, { "epoch": 0.15019085809572966, "grad_norm": 0.1556759774684906, "learning_rate": 1.640508071731395e-05, "loss": 0.0194, "step": 3310 }, { "epoch": 0.15064460691172885, "grad_norm": 0.15330232679843903, "learning_rate": 1.6382480189294293e-05, "loss": 0.0127, "step": 3320 }, { "epoch": 0.15109835572772803, "grad_norm": 0.22334079444408417, "learning_rate": 1.635982451756002e-05, "loss": 0.0185, "step": 3330 }, { "epoch": 0.1515521045437272, "grad_norm": 0.14111435413360596, "learning_rate": 1.6337113897852887e-05, "loss": 0.0145, "step": 3340 }, { "epoch": 0.15200585335972638, "grad_norm": 0.19717109203338623, "learning_rate": 1.6314348526389396e-05, "loss": 0.0155, "step": 3350 }, { "epoch": 0.15245960217572557, "grad_norm": 0.18821868300437927, "learning_rate": 1.6291528599859102e-05, "loss": 0.0156, "step": 3360 }, { "epoch": 0.15291335099172476, "grad_norm": 0.21783433854579926, "learning_rate": 1.6268654315422892e-05, "loss": 0.0192, "step": 3370 }, { "epoch": 0.15336709980772395, "grad_norm": 0.18887248635292053, "learning_rate": 1.6245725870711314e-05, "loss": 0.0167, "step": 3380 }, { "epoch": 0.1538208486237231, "grad_norm": 0.13798531889915466, "learning_rate": 1.6222743463822842e-05, "loss": 0.0161, "step": 3390 }, { "epoch": 0.1542745974397223, "grad_norm": 0.11533325910568237, "learning_rate": 1.6199707293322183e-05, "loss": 0.0166, "step": 3400 }, { "epoch": 0.1547283462557215, "grad_norm": 0.15065830945968628, "learning_rate": 1.6176617558238548e-05, "loss": 0.0203, "step": 3410 }, { "epoch": 0.15518209507172068, "grad_norm": 0.15858496725559235, "learning_rate": 1.615347445806394e-05, "loss": 0.0197, "step": 3420 }, { "epoch": 0.15563584388771987, "grad_norm": 0.14009806513786316, "learning_rate": 1.613027819275143e-05, "loss": 0.0174, "step": 3430 }, { "epoch": 0.15608959270371903, "grad_norm": 0.10621248185634613, "learning_rate": 1.6107028962713433e-05, "loss": 0.0167, "step": 3440 }, { "epoch": 0.15654334151971822, "grad_norm": 0.1479751616716385, "learning_rate": 1.608372696881996e-05, "loss": 0.0163, "step": 3450 }, { "epoch": 0.1569970903357174, "grad_norm": 0.1862780600786209, "learning_rate": 1.60603724123969e-05, "loss": 0.0161, "step": 3460 }, { "epoch": 0.1574508391517166, "grad_norm": 0.12772873044013977, "learning_rate": 1.603696549522428e-05, "loss": 0.0167, "step": 3470 }, { "epoch": 0.15790458796771578, "grad_norm": 0.1471777707338333, "learning_rate": 1.6013506419534505e-05, "loss": 0.02, "step": 3480 }, { "epoch": 0.15835833678371494, "grad_norm": 0.1465907096862793, "learning_rate": 1.598999538801064e-05, "loss": 0.0154, "step": 3490 }, { "epoch": 0.15881208559971413, "grad_norm": 0.1613445281982422, "learning_rate": 1.5966432603784615e-05, "loss": 0.0144, "step": 3500 }, { "epoch": 0.15926583441571332, "grad_norm": 0.12628807127475739, "learning_rate": 1.594281827043552e-05, "loss": 0.015, "step": 3510 }, { "epoch": 0.1597195832317125, "grad_norm": 0.22475102543830872, "learning_rate": 1.5919152591987814e-05, "loss": 0.0157, "step": 3520 }, { "epoch": 0.1601733320477117, "grad_norm": 0.1369459331035614, "learning_rate": 1.5895435772909564e-05, "loss": 0.0161, "step": 3530 }, { "epoch": 0.16062708086371086, "grad_norm": 0.14013464748859406, "learning_rate": 1.5871668018110694e-05, "loss": 0.0197, "step": 3540 }, { "epoch": 0.16108082967971005, "grad_norm": 0.1429699957370758, "learning_rate": 1.5847849532941196e-05, "loss": 0.0167, "step": 3550 }, { "epoch": 0.16153457849570924, "grad_norm": 0.1818908452987671, "learning_rate": 1.5823980523189373e-05, "loss": 0.0134, "step": 3560 }, { "epoch": 0.16198832731170842, "grad_norm": 0.20686101913452148, "learning_rate": 1.580006119508005e-05, "loss": 0.0162, "step": 3570 }, { "epoch": 0.1624420761277076, "grad_norm": 0.210786372423172, "learning_rate": 1.5776091755272792e-05, "loss": 0.0141, "step": 3580 }, { "epoch": 0.16289582494370677, "grad_norm": 0.1615171581506729, "learning_rate": 1.5752072410860132e-05, "loss": 0.0201, "step": 3590 }, { "epoch": 0.16334957375970596, "grad_norm": 0.20601245760917664, "learning_rate": 1.5728003369365763e-05, "loss": 0.0166, "step": 3600 }, { "epoch": 0.16380332257570515, "grad_norm": 0.1881730854511261, "learning_rate": 1.5703884838742755e-05, "loss": 0.0166, "step": 3610 }, { "epoch": 0.16425707139170434, "grad_norm": 0.19188909232616425, "learning_rate": 1.5679717027371756e-05, "loss": 0.0184, "step": 3620 }, { "epoch": 0.16471082020770353, "grad_norm": 0.15665607154369354, "learning_rate": 1.5655500144059202e-05, "loss": 0.0145, "step": 3630 }, { "epoch": 0.16516456902370272, "grad_norm": 0.12721318006515503, "learning_rate": 1.5631234398035483e-05, "loss": 0.0161, "step": 3640 }, { "epoch": 0.16561831783970188, "grad_norm": 0.1763559728860855, "learning_rate": 1.5606919998953182e-05, "loss": 0.0172, "step": 3650 }, { "epoch": 0.16607206665570107, "grad_norm": 0.21826671063899994, "learning_rate": 1.5582557156885218e-05, "loss": 0.0167, "step": 3660 }, { "epoch": 0.16652581547170026, "grad_norm": 0.1324024349451065, "learning_rate": 1.5558146082323056e-05, "loss": 0.0178, "step": 3670 }, { "epoch": 0.16697956428769944, "grad_norm": 0.18612372875213623, "learning_rate": 1.5533686986174885e-05, "loss": 0.0152, "step": 3680 }, { "epoch": 0.16743331310369863, "grad_norm": 0.15666766464710236, "learning_rate": 1.5509180079763794e-05, "loss": 0.0198, "step": 3690 }, { "epoch": 0.1678870619196978, "grad_norm": 0.14181271195411682, "learning_rate": 1.548462557482594e-05, "loss": 0.0165, "step": 3700 }, { "epoch": 0.16834081073569698, "grad_norm": 0.12797847390174866, "learning_rate": 1.546002368350873e-05, "loss": 0.0141, "step": 3710 }, { "epoch": 0.16879455955169617, "grad_norm": 0.18827445805072784, "learning_rate": 1.5435374618368987e-05, "loss": 0.0171, "step": 3720 }, { "epoch": 0.16924830836769536, "grad_norm": 0.24247965216636658, "learning_rate": 1.5410678592371097e-05, "loss": 0.0186, "step": 3730 }, { "epoch": 0.16970205718369455, "grad_norm": 0.15781782567501068, "learning_rate": 1.5385935818885185e-05, "loss": 0.0147, "step": 3740 }, { "epoch": 0.1701558059996937, "grad_norm": 0.1117536649107933, "learning_rate": 1.5361146511685275e-05, "loss": 0.0139, "step": 3750 }, { "epoch": 0.1706095548156929, "grad_norm": 0.12188609689474106, "learning_rate": 1.5336310884947424e-05, "loss": 0.0141, "step": 3760 }, { "epoch": 0.1710633036316921, "grad_norm": 0.25685781240463257, "learning_rate": 1.5311429153247898e-05, "loss": 0.0133, "step": 3770 }, { "epoch": 0.17151705244769128, "grad_norm": 0.19103960692882538, "learning_rate": 1.5286501531561292e-05, "loss": 0.0151, "step": 3780 }, { "epoch": 0.17197080126369046, "grad_norm": 0.16486330330371857, "learning_rate": 1.526152823525868e-05, "loss": 0.0164, "step": 3790 }, { "epoch": 0.17242455007968963, "grad_norm": 0.18085534870624542, "learning_rate": 1.5236509480105781e-05, "loss": 0.0188, "step": 3800 }, { "epoch": 0.17287829889568881, "grad_norm": 0.1550951898097992, "learning_rate": 1.5211445482261039e-05, "loss": 0.0194, "step": 3810 }, { "epoch": 0.173332047711688, "grad_norm": 0.137322798371315, "learning_rate": 1.5186336458273809e-05, "loss": 0.0174, "step": 3820 }, { "epoch": 0.1737857965276872, "grad_norm": 0.17041611671447754, "learning_rate": 1.5161182625082469e-05, "loss": 0.0161, "step": 3830 }, { "epoch": 0.17423954534368638, "grad_norm": 0.1689881682395935, "learning_rate": 1.5135984200012526e-05, "loss": 0.0188, "step": 3840 }, { "epoch": 0.17469329415968554, "grad_norm": 0.1273307353258133, "learning_rate": 1.511074140077477e-05, "loss": 0.0154, "step": 3850 }, { "epoch": 0.17514704297568473, "grad_norm": 0.17848558723926544, "learning_rate": 1.5085454445463367e-05, "loss": 0.0177, "step": 3860 }, { "epoch": 0.17560079179168392, "grad_norm": 0.12270861119031906, "learning_rate": 1.506012355255399e-05, "loss": 0.0138, "step": 3870 }, { "epoch": 0.1760545406076831, "grad_norm": 0.19836512207984924, "learning_rate": 1.503474894090193e-05, "loss": 0.015, "step": 3880 }, { "epoch": 0.1765082894236823, "grad_norm": 0.1289934664964676, "learning_rate": 1.5009330829740183e-05, "loss": 0.0143, "step": 3890 }, { "epoch": 0.17696203823968146, "grad_norm": 0.09760569036006927, "learning_rate": 1.4983869438677605e-05, "loss": 0.0167, "step": 3900 }, { "epoch": 0.17741578705568065, "grad_norm": 0.19151850044727325, "learning_rate": 1.4958364987696956e-05, "loss": 0.0128, "step": 3910 }, { "epoch": 0.17786953587167983, "grad_norm": 0.1450025737285614, "learning_rate": 1.4932817697153046e-05, "loss": 0.0147, "step": 3920 }, { "epoch": 0.17832328468767902, "grad_norm": 0.17141938209533691, "learning_rate": 1.4907227787770805e-05, "loss": 0.0173, "step": 3930 }, { "epoch": 0.1787770335036782, "grad_norm": 0.15546472370624542, "learning_rate": 1.4881595480643379e-05, "loss": 0.0163, "step": 3940 }, { "epoch": 0.17923078231967737, "grad_norm": 0.11263284087181091, "learning_rate": 1.4855920997230238e-05, "loss": 0.016, "step": 3950 }, { "epoch": 0.17968453113567656, "grad_norm": 0.1784234344959259, "learning_rate": 1.4830204559355234e-05, "loss": 0.0165, "step": 3960 }, { "epoch": 0.18013827995167575, "grad_norm": 0.126837357878685, "learning_rate": 1.4804446389204715e-05, "loss": 0.0144, "step": 3970 }, { "epoch": 0.18059202876767494, "grad_norm": 0.18918831646442413, "learning_rate": 1.4778646709325573e-05, "loss": 0.0155, "step": 3980 }, { "epoch": 0.18104577758367413, "grad_norm": 0.13399192690849304, "learning_rate": 1.4752805742623349e-05, "loss": 0.0161, "step": 3990 }, { "epoch": 0.1814995263996733, "grad_norm": 0.22408077120780945, "learning_rate": 1.47269237123603e-05, "loss": 0.0184, "step": 4000 }, { "epoch": 0.18195327521567248, "grad_norm": 0.07992484420537949, "learning_rate": 1.470100084215345e-05, "loss": 0.0148, "step": 4010 }, { "epoch": 0.18240702403167167, "grad_norm": 0.25986817479133606, "learning_rate": 1.4675037355972693e-05, "loss": 0.0201, "step": 4020 }, { "epoch": 0.18286077284767085, "grad_norm": 0.16225387156009674, "learning_rate": 1.4649033478138825e-05, "loss": 0.0155, "step": 4030 }, { "epoch": 0.18331452166367004, "grad_norm": 0.13349366188049316, "learning_rate": 1.4622989433321627e-05, "loss": 0.0218, "step": 4040 }, { "epoch": 0.1837682704796692, "grad_norm": 0.1226244643330574, "learning_rate": 1.459690544653792e-05, "loss": 0.0174, "step": 4050 }, { "epoch": 0.1842220192956684, "grad_norm": 0.1379251480102539, "learning_rate": 1.457078174314961e-05, "loss": 0.0154, "step": 4060 }, { "epoch": 0.18467576811166758, "grad_norm": 0.131155326962471, "learning_rate": 1.4544618548861753e-05, "loss": 0.0153, "step": 4070 }, { "epoch": 0.18512951692766677, "grad_norm": 0.12272419035434723, "learning_rate": 1.45184160897206e-05, "loss": 0.0131, "step": 4080 }, { "epoch": 0.18558326574366596, "grad_norm": 0.13305765390396118, "learning_rate": 1.4492174592111642e-05, "loss": 0.0182, "step": 4090 }, { "epoch": 0.18603701455966512, "grad_norm": 0.17420318722724915, "learning_rate": 1.4465894282757662e-05, "loss": 0.0155, "step": 4100 }, { "epoch": 0.1864907633756643, "grad_norm": 0.15594936907291412, "learning_rate": 1.4439575388716768e-05, "loss": 0.0161, "step": 4110 }, { "epoch": 0.1869445121916635, "grad_norm": 0.14801789820194244, "learning_rate": 1.441321813738044e-05, "loss": 0.0151, "step": 4120 }, { "epoch": 0.1873982610076627, "grad_norm": 0.19852624833583832, "learning_rate": 1.4386822756471545e-05, "loss": 0.0165, "step": 4130 }, { "epoch": 0.18785200982366188, "grad_norm": 0.08428862690925598, "learning_rate": 1.43603894740424e-05, "loss": 0.0154, "step": 4140 }, { "epoch": 0.18830575863966104, "grad_norm": 0.18242783844470978, "learning_rate": 1.4333918518472773e-05, "loss": 0.0163, "step": 4150 }, { "epoch": 0.18875950745566022, "grad_norm": 0.14235183596611023, "learning_rate": 1.4307410118467932e-05, "loss": 0.0127, "step": 4160 }, { "epoch": 0.1892132562716594, "grad_norm": 0.10563075542449951, "learning_rate": 1.428086450305666e-05, "loss": 0.0158, "step": 4170 }, { "epoch": 0.1896670050876586, "grad_norm": 0.16389837861061096, "learning_rate": 1.4254281901589263e-05, "loss": 0.0155, "step": 4180 }, { "epoch": 0.1901207539036578, "grad_norm": 0.11111581325531006, "learning_rate": 1.4227662543735618e-05, "loss": 0.0149, "step": 4190 }, { "epoch": 0.19057450271965698, "grad_norm": 0.15358184278011322, "learning_rate": 1.4201006659483156e-05, "loss": 0.0147, "step": 4200 }, { "epoch": 0.19102825153565614, "grad_norm": 0.18707069754600525, "learning_rate": 1.4174314479134909e-05, "loss": 0.014, "step": 4210 }, { "epoch": 0.19148200035165533, "grad_norm": 0.1715817004442215, "learning_rate": 1.4147586233307485e-05, "loss": 0.0172, "step": 4220 }, { "epoch": 0.19193574916765452, "grad_norm": 0.17705568671226501, "learning_rate": 1.4120822152929099e-05, "loss": 0.0162, "step": 4230 }, { "epoch": 0.1923894979836537, "grad_norm": 0.19288791716098785, "learning_rate": 1.4094022469237577e-05, "loss": 0.0175, "step": 4240 }, { "epoch": 0.1928432467996529, "grad_norm": 0.0979590117931366, "learning_rate": 1.4067187413778338e-05, "loss": 0.0129, "step": 4250 }, { "epoch": 0.19329699561565206, "grad_norm": 0.15775372087955475, "learning_rate": 1.4040317218402426e-05, "loss": 0.0143, "step": 4260 }, { "epoch": 0.19375074443165125, "grad_norm": 0.17242823541164398, "learning_rate": 1.4013412115264477e-05, "loss": 0.0184, "step": 4270 }, { "epoch": 0.19420449324765043, "grad_norm": 0.15347068011760712, "learning_rate": 1.398647233682073e-05, "loss": 0.0157, "step": 4280 }, { "epoch": 0.19465824206364962, "grad_norm": 0.14956066012382507, "learning_rate": 1.3959498115827007e-05, "loss": 0.0122, "step": 4290 }, { "epoch": 0.1951119908796488, "grad_norm": 0.1466713845729828, "learning_rate": 1.3932489685336722e-05, "loss": 0.0155, "step": 4300 }, { "epoch": 0.19556573969564797, "grad_norm": 0.15853281319141388, "learning_rate": 1.3905447278698838e-05, "loss": 0.0162, "step": 4310 }, { "epoch": 0.19601948851164716, "grad_norm": 0.16414372622966766, "learning_rate": 1.3878371129555874e-05, "loss": 0.0137, "step": 4320 }, { "epoch": 0.19647323732764635, "grad_norm": 0.13074234127998352, "learning_rate": 1.3851261471841891e-05, "loss": 0.0139, "step": 4330 }, { "epoch": 0.19692698614364554, "grad_norm": 0.1296066790819168, "learning_rate": 1.382411853978044e-05, "loss": 0.0141, "step": 4340 }, { "epoch": 0.19738073495964473, "grad_norm": 0.11560890078544617, "learning_rate": 1.3796942567882565e-05, "loss": 0.0131, "step": 4350 }, { "epoch": 0.1978344837756439, "grad_norm": 0.14548613131046295, "learning_rate": 1.3769733790944777e-05, "loss": 0.0143, "step": 4360 }, { "epoch": 0.19828823259164308, "grad_norm": 0.1252041906118393, "learning_rate": 1.3742492444047e-05, "loss": 0.0147, "step": 4370 }, { "epoch": 0.19874198140764227, "grad_norm": 0.18141184747219086, "learning_rate": 1.3715218762550584e-05, "loss": 0.0138, "step": 4380 }, { "epoch": 0.19919573022364145, "grad_norm": 0.14405906200408936, "learning_rate": 1.368791298209622e-05, "loss": 0.0152, "step": 4390 }, { "epoch": 0.19964947903964064, "grad_norm": 0.14750778675079346, "learning_rate": 1.3660575338601945e-05, "loss": 0.0135, "step": 4400 }, { "epoch": 0.2001032278556398, "grad_norm": 0.14470447599887848, "learning_rate": 1.363320606826108e-05, "loss": 0.0149, "step": 4410 }, { "epoch": 0.200556976671639, "grad_norm": 0.19202418625354767, "learning_rate": 1.36058054075402e-05, "loss": 0.0134, "step": 4420 }, { "epoch": 0.20101072548763818, "grad_norm": 0.14392217993736267, "learning_rate": 1.3578373593177091e-05, "loss": 0.0132, "step": 4430 }, { "epoch": 0.20146447430363737, "grad_norm": 0.19024603068828583, "learning_rate": 1.35509108621787e-05, "loss": 0.0158, "step": 4440 }, { "epoch": 0.20191822311963656, "grad_norm": 0.19015447795391083, "learning_rate": 1.3523417451819087e-05, "loss": 0.0144, "step": 4450 }, { "epoch": 0.20237197193563572, "grad_norm": 0.1269548237323761, "learning_rate": 1.3495893599637385e-05, "loss": 0.0143, "step": 4460 }, { "epoch": 0.2028257207516349, "grad_norm": 0.5116326808929443, "learning_rate": 1.3468339543435725e-05, "loss": 0.0146, "step": 4470 }, { "epoch": 0.2032794695676341, "grad_norm": 0.1450648307800293, "learning_rate": 1.3440755521277209e-05, "loss": 0.0174, "step": 4480 }, { "epoch": 0.20373321838363329, "grad_norm": 0.10610708594322205, "learning_rate": 1.3413141771483842e-05, "loss": 0.0147, "step": 4490 }, { "epoch": 0.20418696719963247, "grad_norm": 0.15620329976081848, "learning_rate": 1.3385498532634465e-05, "loss": 0.0149, "step": 4500 }, { "epoch": 0.20464071601563164, "grad_norm": 0.19468973577022552, "learning_rate": 1.3357826043562698e-05, "loss": 0.0173, "step": 4510 }, { "epoch": 0.20509446483163082, "grad_norm": 0.1022220253944397, "learning_rate": 1.3330124543354888e-05, "loss": 0.0155, "step": 4520 }, { "epoch": 0.20554821364763, "grad_norm": 0.21672423183918, "learning_rate": 1.3302394271348026e-05, "loss": 0.0167, "step": 4530 }, { "epoch": 0.2060019624636292, "grad_norm": 0.2700170576572418, "learning_rate": 1.3274635467127688e-05, "loss": 0.0126, "step": 4540 }, { "epoch": 0.2064557112796284, "grad_norm": 0.1277543008327484, "learning_rate": 1.3246848370525973e-05, "loss": 0.0158, "step": 4550 }, { "epoch": 0.20690946009562755, "grad_norm": 0.1226678118109703, "learning_rate": 1.3219033221619408e-05, "loss": 0.0142, "step": 4560 }, { "epoch": 0.20736320891162674, "grad_norm": 0.13949912786483765, "learning_rate": 1.3191190260726903e-05, "loss": 0.0142, "step": 4570 }, { "epoch": 0.20781695772762593, "grad_norm": 0.19324158132076263, "learning_rate": 1.3163319728407645e-05, "loss": 0.0157, "step": 4580 }, { "epoch": 0.20827070654362512, "grad_norm": 0.2552565336227417, "learning_rate": 1.3135421865459042e-05, "loss": 0.0175, "step": 4590 }, { "epoch": 0.2087244553596243, "grad_norm": 0.12216931581497192, "learning_rate": 1.3107496912914636e-05, "loss": 0.0125, "step": 4600 }, { "epoch": 0.20917820417562347, "grad_norm": 0.1292363703250885, "learning_rate": 1.307954511204202e-05, "loss": 0.0164, "step": 4610 }, { "epoch": 0.20963195299162266, "grad_norm": 0.13501308858394623, "learning_rate": 1.3051566704340746e-05, "loss": 0.0133, "step": 4620 }, { "epoch": 0.21008570180762184, "grad_norm": 0.14220742881298065, "learning_rate": 1.3023561931540247e-05, "loss": 0.0145, "step": 4630 }, { "epoch": 0.21053945062362103, "grad_norm": 0.14298029243946075, "learning_rate": 1.2995531035597753e-05, "loss": 0.0138, "step": 4640 }, { "epoch": 0.21099319943962022, "grad_norm": 0.17490971088409424, "learning_rate": 1.2967474258696186e-05, "loss": 0.0136, "step": 4650 }, { "epoch": 0.21144694825561938, "grad_norm": 0.1275247186422348, "learning_rate": 1.2939391843242082e-05, "loss": 0.0176, "step": 4660 }, { "epoch": 0.21190069707161857, "grad_norm": 0.08051750063896179, "learning_rate": 1.291128403186349e-05, "loss": 0.0119, "step": 4670 }, { "epoch": 0.21235444588761776, "grad_norm": 0.21460536122322083, "learning_rate": 1.2883151067407866e-05, "loss": 0.0146, "step": 4680 }, { "epoch": 0.21280819470361695, "grad_norm": 0.09482090175151825, "learning_rate": 1.2854993192940005e-05, "loss": 0.0148, "step": 4690 }, { "epoch": 0.21326194351961614, "grad_norm": 0.12067610770463943, "learning_rate": 1.2826810651739899e-05, "loss": 0.0166, "step": 4700 }, { "epoch": 0.2137156923356153, "grad_norm": 0.12952324748039246, "learning_rate": 1.279860368730067e-05, "loss": 0.0142, "step": 4710 }, { "epoch": 0.2141694411516145, "grad_norm": 0.11326833069324493, "learning_rate": 1.2770372543326454e-05, "loss": 0.0173, "step": 4720 }, { "epoch": 0.21462318996761368, "grad_norm": 0.13386495411396027, "learning_rate": 1.2742117463730289e-05, "loss": 0.0107, "step": 4730 }, { "epoch": 0.21507693878361286, "grad_norm": 0.12700419127941132, "learning_rate": 1.2713838692632015e-05, "loss": 0.0155, "step": 4740 }, { "epoch": 0.21553068759961205, "grad_norm": 0.13237282633781433, "learning_rate": 1.2685536474356161e-05, "loss": 0.0157, "step": 4750 }, { "epoch": 0.21598443641561124, "grad_norm": 0.12272775173187256, "learning_rate": 1.2657211053429844e-05, "loss": 0.0142, "step": 4760 }, { "epoch": 0.2164381852316104, "grad_norm": 0.2121793031692505, "learning_rate": 1.2628862674580642e-05, "loss": 0.0154, "step": 4770 }, { "epoch": 0.2168919340476096, "grad_norm": 0.14809030294418335, "learning_rate": 1.2600491582734484e-05, "loss": 0.0151, "step": 4780 }, { "epoch": 0.21734568286360878, "grad_norm": 0.16274145245552063, "learning_rate": 1.2572098023013544e-05, "loss": 0.0164, "step": 4790 }, { "epoch": 0.21779943167960797, "grad_norm": 0.12132669985294342, "learning_rate": 1.254368224073411e-05, "loss": 0.0151, "step": 4800 }, { "epoch": 0.21825318049560716, "grad_norm": 0.13608211278915405, "learning_rate": 1.251524448140447e-05, "loss": 0.0171, "step": 4810 }, { "epoch": 0.21870692931160632, "grad_norm": 0.1499515175819397, "learning_rate": 1.2486784990722791e-05, "loss": 0.0169, "step": 4820 }, { "epoch": 0.2191606781276055, "grad_norm": 0.08290209621191025, "learning_rate": 1.2458304014574996e-05, "loss": 0.0139, "step": 4830 }, { "epoch": 0.2196144269436047, "grad_norm": 0.19991664588451385, "learning_rate": 1.242980179903264e-05, "loss": 0.0172, "step": 4840 }, { "epoch": 0.22006817575960388, "grad_norm": 0.15450483560562134, "learning_rate": 1.2401278590350782e-05, "loss": 0.0173, "step": 4850 }, { "epoch": 0.22052192457560307, "grad_norm": 0.14158503711223602, "learning_rate": 1.2372734634965861e-05, "loss": 0.0159, "step": 4860 }, { "epoch": 0.22097567339160223, "grad_norm": 0.16519369184970856, "learning_rate": 1.234417017949356e-05, "loss": 0.0151, "step": 4870 }, { "epoch": 0.22142942220760142, "grad_norm": 0.1639871746301651, "learning_rate": 1.2315585470726685e-05, "loss": 0.0167, "step": 4880 }, { "epoch": 0.2218831710236006, "grad_norm": 0.1480032354593277, "learning_rate": 1.2286980755633027e-05, "loss": 0.0149, "step": 4890 }, { "epoch": 0.2223369198395998, "grad_norm": 0.1687663197517395, "learning_rate": 1.225835628135322e-05, "loss": 0.0143, "step": 4900 }, { "epoch": 0.222790668655599, "grad_norm": 0.19108609855175018, "learning_rate": 1.2229712295198633e-05, "loss": 0.0145, "step": 4910 }, { "epoch": 0.22324441747159815, "grad_norm": 0.14848002791404724, "learning_rate": 1.2201049044649192e-05, "loss": 0.0165, "step": 4920 }, { "epoch": 0.22369816628759734, "grad_norm": 0.16095581650733948, "learning_rate": 1.217236677735128e-05, "loss": 0.0129, "step": 4930 }, { "epoch": 0.22415191510359653, "grad_norm": 0.11495261639356613, "learning_rate": 1.2143665741115581e-05, "loss": 0.0123, "step": 4940 }, { "epoch": 0.22460566391959572, "grad_norm": 0.1297575980424881, "learning_rate": 1.2114946183914935e-05, "loss": 0.0129, "step": 4950 }, { "epoch": 0.2250594127355949, "grad_norm": 0.2002328485250473, "learning_rate": 1.2086208353882203e-05, "loss": 0.0181, "step": 4960 }, { "epoch": 0.22551316155159407, "grad_norm": 0.1517852246761322, "learning_rate": 1.2057452499308117e-05, "loss": 0.016, "step": 4970 }, { "epoch": 0.22596691036759325, "grad_norm": 0.1382087916135788, "learning_rate": 1.2028678868639147e-05, "loss": 0.013, "step": 4980 }, { "epoch": 0.22642065918359244, "grad_norm": 0.16167710721492767, "learning_rate": 1.1999887710475337e-05, "loss": 0.0146, "step": 4990 }, { "epoch": 0.22687440799959163, "grad_norm": 0.16487346589565277, "learning_rate": 1.197107927356817e-05, "loss": 0.0133, "step": 5000 }, { "epoch": 0.22732815681559082, "grad_norm": 0.16545337438583374, "learning_rate": 1.1942253806818414e-05, "loss": 0.016, "step": 5010 }, { "epoch": 0.22778190563158998, "grad_norm": 0.13121789693832397, "learning_rate": 1.1913411559273973e-05, "loss": 0.0127, "step": 5020 }, { "epoch": 0.22823565444758917, "grad_norm": 0.13568609952926636, "learning_rate": 1.1884552780127736e-05, "loss": 0.0153, "step": 5030 }, { "epoch": 0.22868940326358836, "grad_norm": 0.12294431775808334, "learning_rate": 1.1855677718715417e-05, "loss": 0.0127, "step": 5040 }, { "epoch": 0.22914315207958755, "grad_norm": 0.15326954424381256, "learning_rate": 1.1826786624513416e-05, "loss": 0.0156, "step": 5050 }, { "epoch": 0.22959690089558674, "grad_norm": 0.22886556386947632, "learning_rate": 1.1797879747136645e-05, "loss": 0.012, "step": 5060 }, { "epoch": 0.2300506497115859, "grad_norm": 0.14157430827617645, "learning_rate": 1.1768957336336384e-05, "loss": 0.0124, "step": 5070 }, { "epoch": 0.23050439852758509, "grad_norm": 0.09879475831985474, "learning_rate": 1.1740019641998124e-05, "loss": 0.0152, "step": 5080 }, { "epoch": 0.23095814734358427, "grad_norm": 0.1359705924987793, "learning_rate": 1.171106691413939e-05, "loss": 0.0134, "step": 5090 }, { "epoch": 0.23141189615958346, "grad_norm": 0.11448939144611359, "learning_rate": 1.1682099402907612e-05, "loss": 0.0155, "step": 5100 }, { "epoch": 0.23186564497558265, "grad_norm": 0.15960171818733215, "learning_rate": 1.1653117358577937e-05, "loss": 0.0141, "step": 5110 }, { "epoch": 0.2323193937915818, "grad_norm": 0.18288302421569824, "learning_rate": 1.1624121031551073e-05, "loss": 0.0146, "step": 5120 }, { "epoch": 0.232773142607581, "grad_norm": 0.13190463185310364, "learning_rate": 1.1595110672351132e-05, "loss": 0.0102, "step": 5130 }, { "epoch": 0.2332268914235802, "grad_norm": 0.1572716385126114, "learning_rate": 1.1566086531623464e-05, "loss": 0.0149, "step": 5140 }, { "epoch": 0.23368064023957938, "grad_norm": 0.15305542945861816, "learning_rate": 1.1537048860132487e-05, "loss": 0.0141, "step": 5150 }, { "epoch": 0.23413438905557857, "grad_norm": 0.1395915448665619, "learning_rate": 1.1507997908759525e-05, "loss": 0.014, "step": 5160 }, { "epoch": 0.23458813787157773, "grad_norm": 0.15027165412902832, "learning_rate": 1.1478933928500635e-05, "loss": 0.0123, "step": 5170 }, { "epoch": 0.23504188668757692, "grad_norm": 0.12792900204658508, "learning_rate": 1.1449857170464445e-05, "loss": 0.013, "step": 5180 }, { "epoch": 0.2354956355035761, "grad_norm": 0.18118540942668915, "learning_rate": 1.1420767885869974e-05, "loss": 0.0168, "step": 5190 }, { "epoch": 0.2359493843195753, "grad_norm": 0.10059325397014618, "learning_rate": 1.1391666326044484e-05, "loss": 0.0126, "step": 5200 }, { "epoch": 0.23640313313557448, "grad_norm": 0.08281472325325012, "learning_rate": 1.1362552742421269e-05, "loss": 0.0147, "step": 5210 }, { "epoch": 0.23685688195157364, "grad_norm": 0.16372814774513245, "learning_rate": 1.1333427386537537e-05, "loss": 0.0155, "step": 5220 }, { "epoch": 0.23731063076757283, "grad_norm": 0.15756796300411224, "learning_rate": 1.1304290510032184e-05, "loss": 0.0125, "step": 5230 }, { "epoch": 0.23776437958357202, "grad_norm": 0.10431189090013504, "learning_rate": 1.1275142364643645e-05, "loss": 0.0132, "step": 5240 }, { "epoch": 0.2382181283995712, "grad_norm": 0.17831729352474213, "learning_rate": 1.1245983202207729e-05, "loss": 0.0146, "step": 5250 }, { "epoch": 0.2386718772155704, "grad_norm": 0.11174460500478745, "learning_rate": 1.1216813274655417e-05, "loss": 0.0111, "step": 5260 }, { "epoch": 0.23912562603156956, "grad_norm": 0.18887224793434143, "learning_rate": 1.1187632834010707e-05, "loss": 0.0119, "step": 5270 }, { "epoch": 0.23957937484756875, "grad_norm": 0.2523764371871948, "learning_rate": 1.1158442132388427e-05, "loss": 0.0147, "step": 5280 }, { "epoch": 0.24003312366356794, "grad_norm": 0.22768224775791168, "learning_rate": 1.1129241421992059e-05, "loss": 0.0164, "step": 5290 }, { "epoch": 0.24048687247956713, "grad_norm": 0.13157208263874054, "learning_rate": 1.1100030955111554e-05, "loss": 0.0119, "step": 5300 }, { "epoch": 0.24094062129556632, "grad_norm": 0.13875404000282288, "learning_rate": 1.1070810984121164e-05, "loss": 0.0132, "step": 5310 }, { "epoch": 0.2413943701115655, "grad_norm": 0.18443532288074493, "learning_rate": 1.1041581761477252e-05, "loss": 0.0166, "step": 5320 }, { "epoch": 0.24184811892756466, "grad_norm": 0.16607621312141418, "learning_rate": 1.1012343539716115e-05, "loss": 0.0185, "step": 5330 }, { "epoch": 0.24230186774356385, "grad_norm": 0.17037932574748993, "learning_rate": 1.0983096571451805e-05, "loss": 0.0151, "step": 5340 }, { "epoch": 0.24275561655956304, "grad_norm": 0.15466848015785217, "learning_rate": 1.0953841109373935e-05, "loss": 0.0167, "step": 5350 }, { "epoch": 0.24320936537556223, "grad_norm": 0.11539296060800552, "learning_rate": 1.0924577406245507e-05, "loss": 0.0116, "step": 5360 }, { "epoch": 0.24366311419156142, "grad_norm": 0.11359529942274094, "learning_rate": 1.0895305714900721e-05, "loss": 0.0106, "step": 5370 }, { "epoch": 0.24411686300756058, "grad_norm": 0.12212841212749481, "learning_rate": 1.0866026288242803e-05, "loss": 0.0119, "step": 5380 }, { "epoch": 0.24457061182355977, "grad_norm": 0.1847812980413437, "learning_rate": 1.0836739379241805e-05, "loss": 0.0148, "step": 5390 }, { "epoch": 0.24502436063955896, "grad_norm": 0.19463911652565002, "learning_rate": 1.0807445240932422e-05, "loss": 0.0191, "step": 5400 }, { "epoch": 0.24547810945555815, "grad_norm": 0.08880347013473511, "learning_rate": 1.0778144126411815e-05, "loss": 0.0123, "step": 5410 }, { "epoch": 0.24593185827155734, "grad_norm": 0.1543634831905365, "learning_rate": 1.0748836288837418e-05, "loss": 0.0153, "step": 5420 }, { "epoch": 0.2463856070875565, "grad_norm": 0.16077403724193573, "learning_rate": 1.0719521981424745e-05, "loss": 0.0137, "step": 5430 }, { "epoch": 0.24683935590355569, "grad_norm": 0.16406534612178802, "learning_rate": 1.0690201457445218e-05, "loss": 0.0121, "step": 5440 }, { "epoch": 0.24729310471955487, "grad_norm": 0.15214091539382935, "learning_rate": 1.0660874970223963e-05, "loss": 0.0139, "step": 5450 }, { "epoch": 0.24774685353555406, "grad_norm": 0.16863113641738892, "learning_rate": 1.0631542773137627e-05, "loss": 0.0115, "step": 5460 }, { "epoch": 0.24820060235155325, "grad_norm": 0.1798219382762909, "learning_rate": 1.060220511961219e-05, "loss": 0.0128, "step": 5470 }, { "epoch": 0.2486543511675524, "grad_norm": 0.13751454651355743, "learning_rate": 1.0572862263120784e-05, "loss": 0.0139, "step": 5480 }, { "epoch": 0.2491080999835516, "grad_norm": 0.07154980301856995, "learning_rate": 1.0543514457181476e-05, "loss": 0.0132, "step": 5490 }, { "epoch": 0.2495618487995508, "grad_norm": 0.07695973664522171, "learning_rate": 1.051416195535511e-05, "loss": 0.0134, "step": 5500 }, { "epoch": 0.25001559761555, "grad_norm": 0.111163429915905, "learning_rate": 1.0484805011243102e-05, "loss": 0.012, "step": 5510 }, { "epoch": 0.25046934643154917, "grad_norm": 0.1088341623544693, "learning_rate": 1.0455443878485238e-05, "loss": 0.0132, "step": 5520 }, { "epoch": 0.25092309524754836, "grad_norm": 0.1207728236913681, "learning_rate": 1.0426078810757502e-05, "loss": 0.0139, "step": 5530 }, { "epoch": 0.25137684406354754, "grad_norm": 0.1364191770553589, "learning_rate": 1.039671006176987e-05, "loss": 0.0135, "step": 5540 }, { "epoch": 0.2518305928795467, "grad_norm": 0.14863966405391693, "learning_rate": 1.0367337885264128e-05, "loss": 0.0139, "step": 5550 }, { "epoch": 0.25228434169554587, "grad_norm": 0.10519534349441528, "learning_rate": 1.0337962535011679e-05, "loss": 0.0124, "step": 5560 }, { "epoch": 0.25273809051154505, "grad_norm": 0.07501281052827835, "learning_rate": 1.0308584264811332e-05, "loss": 0.0108, "step": 5570 }, { "epoch": 0.25319183932754424, "grad_norm": 0.1170443519949913, "learning_rate": 1.0279203328487142e-05, "loss": 0.0103, "step": 5580 }, { "epoch": 0.25364558814354343, "grad_norm": 0.1034766435623169, "learning_rate": 1.0249819979886184e-05, "loss": 0.0103, "step": 5590 }, { "epoch": 0.2540993369595426, "grad_norm": 0.12563982605934143, "learning_rate": 1.0220434472876384e-05, "loss": 0.0087, "step": 5600 }, { "epoch": 0.2545530857755418, "grad_norm": 0.11183736473321915, "learning_rate": 1.0191047061344315e-05, "loss": 0.0125, "step": 5610 }, { "epoch": 0.255006834591541, "grad_norm": 0.1348152458667755, "learning_rate": 1.0161657999192998e-05, "loss": 0.0127, "step": 5620 }, { "epoch": 0.2554605834075402, "grad_norm": 0.09866070747375488, "learning_rate": 1.0132267540339726e-05, "loss": 0.0107, "step": 5630 }, { "epoch": 0.2559143322235394, "grad_norm": 0.12439010292291641, "learning_rate": 1.010287593871385e-05, "loss": 0.0118, "step": 5640 }, { "epoch": 0.25636808103953856, "grad_norm": 0.09671192616224289, "learning_rate": 1.0073483448254599e-05, "loss": 0.0121, "step": 5650 }, { "epoch": 0.2568218298555377, "grad_norm": 0.14748752117156982, "learning_rate": 1.0044090322908884e-05, "loss": 0.0108, "step": 5660 }, { "epoch": 0.2572755786715369, "grad_norm": 0.14265309274196625, "learning_rate": 1.0014696816629093e-05, "loss": 0.0127, "step": 5670 }, { "epoch": 0.2577293274875361, "grad_norm": 0.14851878583431244, "learning_rate": 9.985303183370909e-06, "loss": 0.0156, "step": 5680 }, { "epoch": 0.25818307630353526, "grad_norm": 0.10572591423988342, "learning_rate": 9.95590967709112e-06, "loss": 0.0115, "step": 5690 }, { "epoch": 0.25863682511953445, "grad_norm": 0.13734707236289978, "learning_rate": 9.926516551745401e-06, "loss": 0.0115, "step": 5700 }, { "epoch": 0.25909057393553364, "grad_norm": 0.13167889416217804, "learning_rate": 9.897124061286152e-06, "loss": 0.01, "step": 5710 }, { "epoch": 0.25954432275153283, "grad_norm": 0.1438244879245758, "learning_rate": 9.867732459660277e-06, "loss": 0.0113, "step": 5720 }, { "epoch": 0.259998071567532, "grad_norm": 0.16616496443748474, "learning_rate": 9.838342000807006e-06, "loss": 0.013, "step": 5730 }, { "epoch": 0.2604518203835312, "grad_norm": 0.12727349996566772, "learning_rate": 9.808952938655689e-06, "loss": 0.0146, "step": 5740 }, { "epoch": 0.2609055691995304, "grad_norm": 0.159709170460701, "learning_rate": 9.77956552712362e-06, "loss": 0.0156, "step": 5750 }, { "epoch": 0.26135931801552953, "grad_norm": 0.09865106642246246, "learning_rate": 9.75018002011382e-06, "loss": 0.0095, "step": 5760 }, { "epoch": 0.2618130668315287, "grad_norm": 0.17693890631198883, "learning_rate": 9.720796671512863e-06, "loss": 0.014, "step": 5770 }, { "epoch": 0.2622668156475279, "grad_norm": 0.12724372744560242, "learning_rate": 9.69141573518867e-06, "loss": 0.0134, "step": 5780 }, { "epoch": 0.2627205644635271, "grad_norm": 0.11850858479738235, "learning_rate": 9.662037464988323e-06, "loss": 0.0115, "step": 5790 }, { "epoch": 0.2631743132795263, "grad_norm": 0.13083550333976746, "learning_rate": 9.63266211473587e-06, "loss": 0.0128, "step": 5800 }, { "epoch": 0.2636280620955255, "grad_norm": 0.16072624921798706, "learning_rate": 9.603289938230132e-06, "loss": 0.0129, "step": 5810 }, { "epoch": 0.26408181091152466, "grad_norm": 0.1012595146894455, "learning_rate": 9.573921189242501e-06, "loss": 0.0126, "step": 5820 }, { "epoch": 0.26453555972752385, "grad_norm": 0.17362189292907715, "learning_rate": 9.544556121514765e-06, "loss": 0.014, "step": 5830 }, { "epoch": 0.26498930854352304, "grad_norm": 0.15318715572357178, "learning_rate": 9.5151949887569e-06, "loss": 0.0133, "step": 5840 }, { "epoch": 0.2654430573595222, "grad_norm": 0.1297263503074646, "learning_rate": 9.485838044644891e-06, "loss": 0.0131, "step": 5850 }, { "epoch": 0.26589680617552136, "grad_norm": 0.13620354235172272, "learning_rate": 9.456485542818527e-06, "loss": 0.0142, "step": 5860 }, { "epoch": 0.26635055499152055, "grad_norm": 0.14983710646629333, "learning_rate": 9.427137736879222e-06, "loss": 0.0113, "step": 5870 }, { "epoch": 0.26680430380751974, "grad_norm": 0.13920733332633972, "learning_rate": 9.397794880387812e-06, "loss": 0.0131, "step": 5880 }, { "epoch": 0.2672580526235189, "grad_norm": 0.13266463577747345, "learning_rate": 9.368457226862378e-06, "loss": 0.0117, "step": 5890 }, { "epoch": 0.2677118014395181, "grad_norm": 0.15641669929027557, "learning_rate": 9.339125029776039e-06, "loss": 0.015, "step": 5900 }, { "epoch": 0.2681655502555173, "grad_norm": 0.14706820249557495, "learning_rate": 9.309798542554782e-06, "loss": 0.011, "step": 5910 }, { "epoch": 0.2686192990715165, "grad_norm": 0.09607023000717163, "learning_rate": 9.280478018575257e-06, "loss": 0.012, "step": 5920 }, { "epoch": 0.2690730478875157, "grad_norm": 0.13049288094043732, "learning_rate": 9.251163711162584e-06, "loss": 0.0126, "step": 5930 }, { "epoch": 0.26952679670351487, "grad_norm": 0.10754423588514328, "learning_rate": 9.221855873588187e-06, "loss": 0.0125, "step": 5940 }, { "epoch": 0.26998054551951406, "grad_norm": 0.14173737168312073, "learning_rate": 9.192554759067581e-06, "loss": 0.015, "step": 5950 }, { "epoch": 0.2704342943355132, "grad_norm": 0.08642645925283432, "learning_rate": 9.163260620758197e-06, "loss": 0.0102, "step": 5960 }, { "epoch": 0.2708880431515124, "grad_norm": 0.09053780883550644, "learning_rate": 9.133973711757198e-06, "loss": 0.0116, "step": 5970 }, { "epoch": 0.27134179196751157, "grad_norm": 0.17731283605098724, "learning_rate": 9.10469428509928e-06, "loss": 0.0134, "step": 5980 }, { "epoch": 0.27179554078351076, "grad_norm": 0.14436522126197815, "learning_rate": 9.075422593754498e-06, "loss": 0.0122, "step": 5990 }, { "epoch": 0.27224928959950995, "grad_norm": 0.21791435778141022, "learning_rate": 9.046158890626069e-06, "loss": 0.0163, "step": 6000 }, { "epoch": 0.27270303841550914, "grad_norm": 0.1288122683763504, "learning_rate": 9.016903428548195e-06, "loss": 0.0137, "step": 6010 }, { "epoch": 0.2731567872315083, "grad_norm": 0.179439514875412, "learning_rate": 8.987656460283885e-06, "loss": 0.0119, "step": 6020 }, { "epoch": 0.2736105360475075, "grad_norm": 0.1351165622472763, "learning_rate": 8.958418238522748e-06, "loss": 0.0141, "step": 6030 }, { "epoch": 0.2740642848635067, "grad_norm": 0.12908992171287537, "learning_rate": 8.929189015878838e-06, "loss": 0.0127, "step": 6040 }, { "epoch": 0.2745180336795059, "grad_norm": 0.19940853118896484, "learning_rate": 8.899969044888448e-06, "loss": 0.012, "step": 6050 }, { "epoch": 0.274971782495505, "grad_norm": 0.16033363342285156, "learning_rate": 8.870758578007944e-06, "loss": 0.0122, "step": 6060 }, { "epoch": 0.2754255313115042, "grad_norm": 0.13968591392040253, "learning_rate": 8.841557867611576e-06, "loss": 0.013, "step": 6070 }, { "epoch": 0.2758792801275034, "grad_norm": 0.1749831885099411, "learning_rate": 8.812367165989295e-06, "loss": 0.0134, "step": 6080 }, { "epoch": 0.2763330289435026, "grad_norm": 0.11019530147314072, "learning_rate": 8.783186725344588e-06, "loss": 0.0118, "step": 6090 }, { "epoch": 0.2767867777595018, "grad_norm": 0.1606026291847229, "learning_rate": 8.754016797792276e-06, "loss": 0.0144, "step": 6100 }, { "epoch": 0.27724052657550097, "grad_norm": 0.13256487250328064, "learning_rate": 8.72485763535636e-06, "loss": 0.0119, "step": 6110 }, { "epoch": 0.27769427539150016, "grad_norm": 0.1421864628791809, "learning_rate": 8.695709489967821e-06, "loss": 0.0126, "step": 6120 }, { "epoch": 0.27814802420749934, "grad_norm": 0.14893385767936707, "learning_rate": 8.666572613462465e-06, "loss": 0.0143, "step": 6130 }, { "epoch": 0.27860177302349853, "grad_norm": 0.14165258407592773, "learning_rate": 8.63744725757873e-06, "loss": 0.0114, "step": 6140 }, { "epoch": 0.2790555218394977, "grad_norm": 0.0763816237449646, "learning_rate": 8.60833367395552e-06, "loss": 0.0115, "step": 6150 }, { "epoch": 0.2795092706554969, "grad_norm": 0.12333854287862778, "learning_rate": 8.579232114130027e-06, "loss": 0.0143, "step": 6160 }, { "epoch": 0.27996301947149604, "grad_norm": 0.11713657528162003, "learning_rate": 8.550142829535559e-06, "loss": 0.0087, "step": 6170 }, { "epoch": 0.28041676828749523, "grad_norm": 0.21120575070381165, "learning_rate": 8.521066071499368e-06, "loss": 0.0119, "step": 6180 }, { "epoch": 0.2808705171034944, "grad_norm": 0.146609827876091, "learning_rate": 8.492002091240478e-06, "loss": 0.0127, "step": 6190 }, { "epoch": 0.2813242659194936, "grad_norm": 0.10008296370506287, "learning_rate": 8.462951139867514e-06, "loss": 0.0129, "step": 6200 }, { "epoch": 0.2817780147354928, "grad_norm": 0.12781381607055664, "learning_rate": 8.43391346837654e-06, "loss": 0.0121, "step": 6210 }, { "epoch": 0.282231763551492, "grad_norm": 0.14611364901065826, "learning_rate": 8.404889327648873e-06, "loss": 0.0122, "step": 6220 }, { "epoch": 0.2826855123674912, "grad_norm": 0.17796698212623596, "learning_rate": 8.375878968448934e-06, "loss": 0.0126, "step": 6230 }, { "epoch": 0.28313926118349036, "grad_norm": 0.0997767299413681, "learning_rate": 8.346882641422066e-06, "loss": 0.0115, "step": 6240 }, { "epoch": 0.28359300999948955, "grad_norm": 0.17882458865642548, "learning_rate": 8.317900597092388e-06, "loss": 0.0136, "step": 6250 }, { "epoch": 0.28404675881548874, "grad_norm": 0.12098672240972519, "learning_rate": 8.288933085860611e-06, "loss": 0.0117, "step": 6260 }, { "epoch": 0.2845005076314879, "grad_norm": 0.09727644175291061, "learning_rate": 8.25998035800188e-06, "loss": 0.0128, "step": 6270 }, { "epoch": 0.28495425644748706, "grad_norm": 0.07772582769393921, "learning_rate": 8.231042663663619e-06, "loss": 0.0138, "step": 6280 }, { "epoch": 0.28540800526348625, "grad_norm": 0.14961278438568115, "learning_rate": 8.202120252863359e-06, "loss": 0.0113, "step": 6290 }, { "epoch": 0.28586175407948544, "grad_norm": 0.0879221111536026, "learning_rate": 8.173213375486589e-06, "loss": 0.0135, "step": 6300 }, { "epoch": 0.28631550289548463, "grad_norm": 0.1098606288433075, "learning_rate": 8.144322281284586e-06, "loss": 0.0147, "step": 6310 }, { "epoch": 0.2867692517114838, "grad_norm": 0.18294364213943481, "learning_rate": 8.11544721987227e-06, "loss": 0.0131, "step": 6320 }, { "epoch": 0.287223000527483, "grad_norm": 0.15368086099624634, "learning_rate": 8.086588440726034e-06, "loss": 0.0122, "step": 6330 }, { "epoch": 0.2876767493434822, "grad_norm": 0.16978329420089722, "learning_rate": 8.057746193181591e-06, "loss": 0.0118, "step": 6340 }, { "epoch": 0.2881304981594814, "grad_norm": 0.20677420496940613, "learning_rate": 8.028920726431832e-06, "loss": 0.0122, "step": 6350 }, { "epoch": 0.2885842469754806, "grad_norm": 0.11161544919013977, "learning_rate": 8.000112289524666e-06, "loss": 0.0114, "step": 6360 }, { "epoch": 0.2890379957914797, "grad_norm": 0.14191238582134247, "learning_rate": 7.971321131360855e-06, "loss": 0.0119, "step": 6370 }, { "epoch": 0.2894917446074789, "grad_norm": 0.10466880351305008, "learning_rate": 7.942547500691884e-06, "loss": 0.0136, "step": 6380 }, { "epoch": 0.2899454934234781, "grad_norm": 0.17010283470153809, "learning_rate": 7.913791646117798e-06, "loss": 0.0126, "step": 6390 }, { "epoch": 0.2903992422394773, "grad_norm": 0.11536948382854462, "learning_rate": 7.885053816085067e-06, "loss": 0.0151, "step": 6400 }, { "epoch": 0.29085299105547646, "grad_norm": 0.1053222119808197, "learning_rate": 7.85633425888442e-06, "loss": 0.0117, "step": 6410 }, { "epoch": 0.29130673987147565, "grad_norm": 0.11969737708568573, "learning_rate": 7.827633222648722e-06, "loss": 0.0126, "step": 6420 }, { "epoch": 0.29176048868747484, "grad_norm": 0.13598807156085968, "learning_rate": 7.798950955350812e-06, "loss": 0.0132, "step": 6430 }, { "epoch": 0.29221423750347403, "grad_norm": 0.12094290554523468, "learning_rate": 7.770287704801374e-06, "loss": 0.0128, "step": 6440 }, { "epoch": 0.2926679863194732, "grad_norm": 0.12500420212745667, "learning_rate": 7.741643718646783e-06, "loss": 0.0123, "step": 6450 }, { "epoch": 0.2931217351354724, "grad_norm": 0.088765949010849, "learning_rate": 7.713019244366977e-06, "loss": 0.0123, "step": 6460 }, { "epoch": 0.29357548395147154, "grad_norm": 0.13747364282608032, "learning_rate": 7.684414529273315e-06, "loss": 0.0116, "step": 6470 }, { "epoch": 0.2940292327674707, "grad_norm": 0.10825503617525101, "learning_rate": 7.655829820506442e-06, "loss": 0.0111, "step": 6480 }, { "epoch": 0.2944829815834699, "grad_norm": 0.1315292865037918, "learning_rate": 7.627265365034141e-06, "loss": 0.0092, "step": 6490 }, { "epoch": 0.2949367303994691, "grad_norm": 0.09187211841344833, "learning_rate": 7.59872140964922e-06, "loss": 0.0132, "step": 6500 }, { "epoch": 0.2953904792154683, "grad_norm": 0.13788200914859772, "learning_rate": 7.570198200967363e-06, "loss": 0.0101, "step": 6510 }, { "epoch": 0.2958442280314675, "grad_norm": 0.08541585505008698, "learning_rate": 7.5416959854250076e-06, "loss": 0.0103, "step": 6520 }, { "epoch": 0.29629797684746667, "grad_norm": 0.5758150219917297, "learning_rate": 7.513215009277212e-06, "loss": 0.0135, "step": 6530 }, { "epoch": 0.29675172566346586, "grad_norm": 0.2074153572320938, "learning_rate": 7.484755518595534e-06, "loss": 0.0146, "step": 6540 }, { "epoch": 0.29720547447946505, "grad_norm": 0.13672949373722076, "learning_rate": 7.456317759265893e-06, "loss": 0.011, "step": 6550 }, { "epoch": 0.29765922329546424, "grad_norm": 0.15952400863170624, "learning_rate": 7.4279019769864605e-06, "loss": 0.0132, "step": 6560 }, { "epoch": 0.29811297211146337, "grad_norm": 0.1452173888683319, "learning_rate": 7.399508417265517e-06, "loss": 0.0133, "step": 6570 }, { "epoch": 0.29856672092746256, "grad_norm": 0.12948384881019592, "learning_rate": 7.3711373254193595e-06, "loss": 0.0118, "step": 6580 }, { "epoch": 0.29902046974346175, "grad_norm": 0.11394894123077393, "learning_rate": 7.342788946570159e-06, "loss": 0.0121, "step": 6590 }, { "epoch": 0.29947421855946094, "grad_norm": 0.14150279760360718, "learning_rate": 7.314463525643842e-06, "loss": 0.0123, "step": 6600 }, { "epoch": 0.2999279673754601, "grad_norm": 0.12348120659589767, "learning_rate": 7.286161307367989e-06, "loss": 0.0128, "step": 6610 }, { "epoch": 0.3003817161914593, "grad_norm": 0.12587220966815948, "learning_rate": 7.257882536269716e-06, "loss": 0.0101, "step": 6620 }, { "epoch": 0.3008354650074585, "grad_norm": 0.17185339331626892, "learning_rate": 7.2296274566735494e-06, "loss": 0.0117, "step": 6630 }, { "epoch": 0.3012892138234577, "grad_norm": 0.1300371140241623, "learning_rate": 7.201396312699334e-06, "loss": 0.0089, "step": 6640 }, { "epoch": 0.3017429626394569, "grad_norm": 0.1046392098069191, "learning_rate": 7.173189348260105e-06, "loss": 0.0105, "step": 6650 }, { "epoch": 0.30219671145545607, "grad_norm": 0.1329888254404068, "learning_rate": 7.145006807060002e-06, "loss": 0.0112, "step": 6660 }, { "epoch": 0.3026504602714552, "grad_norm": 0.12515147030353546, "learning_rate": 7.116848932592136e-06, "loss": 0.0132, "step": 6670 }, { "epoch": 0.3031042090874544, "grad_norm": 0.10097947716712952, "learning_rate": 7.088715968136513e-06, "loss": 0.0139, "step": 6680 }, { "epoch": 0.3035579579034536, "grad_norm": 0.1693345010280609, "learning_rate": 7.06060815675792e-06, "loss": 0.0116, "step": 6690 }, { "epoch": 0.30401170671945277, "grad_norm": 0.12788210809230804, "learning_rate": 7.032525741303815e-06, "loss": 0.0136, "step": 6700 }, { "epoch": 0.30446545553545196, "grad_norm": 0.10201315581798553, "learning_rate": 7.00446896440225e-06, "loss": 0.0129, "step": 6710 }, { "epoch": 0.30491920435145115, "grad_norm": 0.13046912848949432, "learning_rate": 6.976438068459756e-06, "loss": 0.0102, "step": 6720 }, { "epoch": 0.30537295316745033, "grad_norm": 0.13688482344150543, "learning_rate": 6.948433295659258e-06, "loss": 0.0126, "step": 6730 }, { "epoch": 0.3058267019834495, "grad_norm": 0.14753024280071259, "learning_rate": 6.920454887957984e-06, "loss": 0.0145, "step": 6740 }, { "epoch": 0.3062804507994487, "grad_norm": 0.2111503630876541, "learning_rate": 6.892503087085365e-06, "loss": 0.0131, "step": 6750 }, { "epoch": 0.3067341996154479, "grad_norm": 0.09545427560806274, "learning_rate": 6.864578134540961e-06, "loss": 0.0124, "step": 6760 }, { "epoch": 0.3071879484314471, "grad_norm": 0.11243736743927002, "learning_rate": 6.83668027159236e-06, "loss": 0.0118, "step": 6770 }, { "epoch": 0.3076416972474462, "grad_norm": 0.1419551521539688, "learning_rate": 6.8088097392731035e-06, "loss": 0.0136, "step": 6780 }, { "epoch": 0.3080954460634454, "grad_norm": 0.21375605463981628, "learning_rate": 6.7809667783805934e-06, "loss": 0.0129, "step": 6790 }, { "epoch": 0.3085491948794446, "grad_norm": 0.11559443175792694, "learning_rate": 6.753151629474028e-06, "loss": 0.0113, "step": 6800 }, { "epoch": 0.3090029436954438, "grad_norm": 0.26248887181282043, "learning_rate": 6.725364532872312e-06, "loss": 0.0124, "step": 6810 }, { "epoch": 0.309456692511443, "grad_norm": 0.1434881091117859, "learning_rate": 6.697605728651977e-06, "loss": 0.0125, "step": 6820 }, { "epoch": 0.30991044132744217, "grad_norm": 0.14054977893829346, "learning_rate": 6.669875456645115e-06, "loss": 0.0128, "step": 6830 }, { "epoch": 0.31036419014344135, "grad_norm": 0.14137424528598785, "learning_rate": 6.642173956437306e-06, "loss": 0.0101, "step": 6840 }, { "epoch": 0.31081793895944054, "grad_norm": 0.11850173771381378, "learning_rate": 6.614501467365539e-06, "loss": 0.0113, "step": 6850 }, { "epoch": 0.31127168777543973, "grad_norm": 0.09213166683912277, "learning_rate": 6.586858228516162e-06, "loss": 0.0113, "step": 6860 }, { "epoch": 0.3117254365914389, "grad_norm": 0.18658258020877838, "learning_rate": 6.559244478722792e-06, "loss": 0.0113, "step": 6870 }, { "epoch": 0.31217918540743805, "grad_norm": 0.14191250503063202, "learning_rate": 6.531660456564282e-06, "loss": 0.0171, "step": 6880 }, { "epoch": 0.31263293422343724, "grad_norm": 0.14040282368659973, "learning_rate": 6.504106400362621e-06, "loss": 0.0107, "step": 6890 }, { "epoch": 0.31308668303943643, "grad_norm": 0.10161775350570679, "learning_rate": 6.476582548180912e-06, "loss": 0.0108, "step": 6900 }, { "epoch": 0.3135404318554356, "grad_norm": 0.09201598167419434, "learning_rate": 6.449089137821301e-06, "loss": 0.0142, "step": 6910 }, { "epoch": 0.3139941806714348, "grad_norm": 0.15806888043880463, "learning_rate": 6.421626406822909e-06, "loss": 0.0103, "step": 6920 }, { "epoch": 0.314447929487434, "grad_norm": 0.15518872439861298, "learning_rate": 6.394194592459801e-06, "loss": 0.0145, "step": 6930 }, { "epoch": 0.3149016783034332, "grad_norm": 0.16833215951919556, "learning_rate": 6.366793931738922e-06, "loss": 0.0136, "step": 6940 }, { "epoch": 0.3153554271194324, "grad_norm": 0.08252312988042831, "learning_rate": 6.339424661398058e-06, "loss": 0.0114, "step": 6950 }, { "epoch": 0.31580917593543156, "grad_norm": 0.13390517234802246, "learning_rate": 6.312087017903783e-06, "loss": 0.0147, "step": 6960 }, { "epoch": 0.31626292475143075, "grad_norm": 0.11417072266340256, "learning_rate": 6.284781237449419e-06, "loss": 0.0125, "step": 6970 }, { "epoch": 0.3167166735674299, "grad_norm": 0.11521535366773605, "learning_rate": 6.257507555953002e-06, "loss": 0.0113, "step": 6980 }, { "epoch": 0.3171704223834291, "grad_norm": 0.133391872048378, "learning_rate": 6.230266209055229e-06, "loss": 0.0098, "step": 6990 }, { "epoch": 0.31762417119942826, "grad_norm": 0.14856669306755066, "learning_rate": 6.20305743211744e-06, "loss": 0.0124, "step": 7000 }, { "epoch": 0.31807792001542745, "grad_norm": 0.14302918314933777, "learning_rate": 6.175881460219565e-06, "loss": 0.0098, "step": 7010 }, { "epoch": 0.31853166883142664, "grad_norm": 0.11710617691278458, "learning_rate": 6.148738528158109e-06, "loss": 0.0106, "step": 7020 }, { "epoch": 0.31898541764742583, "grad_norm": 0.10232780873775482, "learning_rate": 6.1216288704441255e-06, "loss": 0.0098, "step": 7030 }, { "epoch": 0.319439166463425, "grad_norm": 0.08459633588790894, "learning_rate": 6.094552721301164e-06, "loss": 0.0103, "step": 7040 }, { "epoch": 0.3198929152794242, "grad_norm": 0.12328217923641205, "learning_rate": 6.067510314663283e-06, "loss": 0.0092, "step": 7050 }, { "epoch": 0.3203466640954234, "grad_norm": 0.11582467705011368, "learning_rate": 6.0405018841729934e-06, "loss": 0.0102, "step": 7060 }, { "epoch": 0.3208004129114226, "grad_norm": 0.12689736485481262, "learning_rate": 6.013527663179275e-06, "loss": 0.012, "step": 7070 }, { "epoch": 0.3212541617274217, "grad_norm": 0.17244750261306763, "learning_rate": 5.986587884735526e-06, "loss": 0.0148, "step": 7080 }, { "epoch": 0.3217079105434209, "grad_norm": 0.1367264688014984, "learning_rate": 5.9596827815975775e-06, "loss": 0.0126, "step": 7090 }, { "epoch": 0.3221616593594201, "grad_norm": 0.15937544405460358, "learning_rate": 5.9328125862216676e-06, "loss": 0.0122, "step": 7100 }, { "epoch": 0.3226154081754193, "grad_norm": 0.1453155279159546, "learning_rate": 5.90597753076243e-06, "loss": 0.0117, "step": 7110 }, { "epoch": 0.32306915699141847, "grad_norm": 0.1392073631286621, "learning_rate": 5.879177847070906e-06, "loss": 0.0115, "step": 7120 }, { "epoch": 0.32352290580741766, "grad_norm": 0.12530963122844696, "learning_rate": 5.8524137666925174e-06, "loss": 0.0102, "step": 7130 }, { "epoch": 0.32397665462341685, "grad_norm": 0.11018750071525574, "learning_rate": 5.825685520865092e-06, "loss": 0.0114, "step": 7140 }, { "epoch": 0.32443040343941604, "grad_norm": 0.1296696662902832, "learning_rate": 5.798993340516843e-06, "loss": 0.0115, "step": 7150 }, { "epoch": 0.3248841522554152, "grad_norm": 0.16327860951423645, "learning_rate": 5.772337456264386e-06, "loss": 0.0127, "step": 7160 }, { "epoch": 0.3253379010714144, "grad_norm": 0.14514560997486115, "learning_rate": 5.745718098410737e-06, "loss": 0.0129, "step": 7170 }, { "epoch": 0.32579164988741355, "grad_norm": 0.11294388025999069, "learning_rate": 5.719135496943343e-06, "loss": 0.0099, "step": 7180 }, { "epoch": 0.32624539870341274, "grad_norm": 0.1433974653482437, "learning_rate": 5.69258988153207e-06, "loss": 0.0115, "step": 7190 }, { "epoch": 0.3266991475194119, "grad_norm": 0.1322980523109436, "learning_rate": 5.666081481527232e-06, "loss": 0.0118, "step": 7200 }, { "epoch": 0.3271528963354111, "grad_norm": 0.14572355151176453, "learning_rate": 5.639610525957604e-06, "loss": 0.0114, "step": 7210 }, { "epoch": 0.3276066451514103, "grad_norm": 0.07759486883878708, "learning_rate": 5.613177243528458e-06, "loss": 0.0116, "step": 7220 }, { "epoch": 0.3280603939674095, "grad_norm": 0.10589128732681274, "learning_rate": 5.586781862619566e-06, "loss": 0.0106, "step": 7230 }, { "epoch": 0.3285141427834087, "grad_norm": 0.1211918294429779, "learning_rate": 5.560424611283231e-06, "loss": 0.011, "step": 7240 }, { "epoch": 0.32896789159940787, "grad_norm": 0.12552636861801147, "learning_rate": 5.53410571724234e-06, "loss": 0.0111, "step": 7250 }, { "epoch": 0.32942164041540706, "grad_norm": 0.15319830179214478, "learning_rate": 5.507825407888362e-06, "loss": 0.0125, "step": 7260 }, { "epoch": 0.32987538923140625, "grad_norm": 0.1404561698436737, "learning_rate": 5.481583910279402e-06, "loss": 0.0131, "step": 7270 }, { "epoch": 0.33032913804740544, "grad_norm": 0.13395948708057404, "learning_rate": 5.4553814511382485e-06, "loss": 0.0105, "step": 7280 }, { "epoch": 0.33078288686340457, "grad_norm": 0.13249263167381287, "learning_rate": 5.429218256850393e-06, "loss": 0.011, "step": 7290 }, { "epoch": 0.33123663567940376, "grad_norm": 0.1319146752357483, "learning_rate": 5.403094553462083e-06, "loss": 0.0121, "step": 7300 }, { "epoch": 0.33169038449540295, "grad_norm": 0.07734081894159317, "learning_rate": 5.377010566678371e-06, "loss": 0.013, "step": 7310 }, { "epoch": 0.33214413331140213, "grad_norm": 0.10048568248748779, "learning_rate": 5.350966521861178e-06, "loss": 0.0099, "step": 7320 }, { "epoch": 0.3325978821274013, "grad_norm": 0.1091848835349083, "learning_rate": 5.324962644027312e-06, "loss": 0.0146, "step": 7330 }, { "epoch": 0.3330516309434005, "grad_norm": 0.08323021978139877, "learning_rate": 5.298999157846555e-06, "loss": 0.0114, "step": 7340 }, { "epoch": 0.3335053797593997, "grad_norm": 0.11683864891529083, "learning_rate": 5.273076287639704e-06, "loss": 0.0122, "step": 7350 }, { "epoch": 0.3339591285753989, "grad_norm": 0.10861805081367493, "learning_rate": 5.247194257376653e-06, "loss": 0.0119, "step": 7360 }, { "epoch": 0.3344128773913981, "grad_norm": 0.11935005336999893, "learning_rate": 5.221353290674429e-06, "loss": 0.011, "step": 7370 }, { "epoch": 0.33486662620739727, "grad_norm": 0.09110080450773239, "learning_rate": 5.1955536107952885e-06, "loss": 0.0101, "step": 7380 }, { "epoch": 0.3353203750233964, "grad_norm": 0.10009435564279556, "learning_rate": 5.169795440644767e-06, "loss": 0.0109, "step": 7390 }, { "epoch": 0.3357741238393956, "grad_norm": 0.13338936865329742, "learning_rate": 5.144079002769766e-06, "loss": 0.0104, "step": 7400 }, { "epoch": 0.3362278726553948, "grad_norm": 0.10619942098855972, "learning_rate": 5.118404519356621e-06, "loss": 0.0105, "step": 7410 }, { "epoch": 0.33668162147139397, "grad_norm": 0.19180886447429657, "learning_rate": 5.0927722122292e-06, "loss": 0.0111, "step": 7420 }, { "epoch": 0.33713537028739315, "grad_norm": 0.13941583037376404, "learning_rate": 5.067182302846958e-06, "loss": 0.0111, "step": 7430 }, { "epoch": 0.33758911910339234, "grad_norm": 0.14227934181690216, "learning_rate": 5.041635012303048e-06, "loss": 0.0121, "step": 7440 }, { "epoch": 0.33804286791939153, "grad_norm": 0.11831536144018173, "learning_rate": 5.016130561322399e-06, "loss": 0.009, "step": 7450 }, { "epoch": 0.3384966167353907, "grad_norm": 0.10274643450975418, "learning_rate": 4.990669170259816e-06, "loss": 0.0119, "step": 7460 }, { "epoch": 0.3389503655513899, "grad_norm": 0.22540533542633057, "learning_rate": 4.965251059098074e-06, "loss": 0.0094, "step": 7470 }, { "epoch": 0.3394041143673891, "grad_norm": 0.1153121292591095, "learning_rate": 4.93987644744601e-06, "loss": 0.0118, "step": 7480 }, { "epoch": 0.33985786318338823, "grad_norm": 0.13815748691558838, "learning_rate": 4.9145455545366335e-06, "loss": 0.0094, "step": 7490 }, { "epoch": 0.3403116119993874, "grad_norm": 0.1963374763727188, "learning_rate": 4.889258599225233e-06, "loss": 0.0136, "step": 7500 }, { "epoch": 0.3407653608153866, "grad_norm": 0.10098478198051453, "learning_rate": 4.864015799987474e-06, "loss": 0.0105, "step": 7510 }, { "epoch": 0.3412191096313858, "grad_norm": 0.13396380841732025, "learning_rate": 4.838817374917534e-06, "loss": 0.0147, "step": 7520 }, { "epoch": 0.341672858447385, "grad_norm": 0.14336052536964417, "learning_rate": 4.8136635417261935e-06, "loss": 0.0118, "step": 7530 }, { "epoch": 0.3421266072633842, "grad_norm": 0.10059235990047455, "learning_rate": 4.788554517738967e-06, "loss": 0.0109, "step": 7540 }, { "epoch": 0.34258035607938336, "grad_norm": 0.1468759924173355, "learning_rate": 4.763490519894223e-06, "loss": 0.0097, "step": 7550 }, { "epoch": 0.34303410489538255, "grad_norm": 0.12398961931467056, "learning_rate": 4.738471764741319e-06, "loss": 0.0107, "step": 7560 }, { "epoch": 0.34348785371138174, "grad_norm": 0.09385305643081665, "learning_rate": 4.713498468438709e-06, "loss": 0.0128, "step": 7570 }, { "epoch": 0.34394160252738093, "grad_norm": 0.21803854405879974, "learning_rate": 4.6885708467521015e-06, "loss": 0.0147, "step": 7580 }, { "epoch": 0.34439535134338006, "grad_norm": 0.1330396980047226, "learning_rate": 4.6636891150525765e-06, "loss": 0.0107, "step": 7590 }, { "epoch": 0.34484910015937925, "grad_norm": 0.12276133894920349, "learning_rate": 4.638853488314727e-06, "loss": 0.0104, "step": 7600 }, { "epoch": 0.34530284897537844, "grad_norm": 0.14330561459064484, "learning_rate": 4.614064181114817e-06, "loss": 0.0097, "step": 7610 }, { "epoch": 0.34575659779137763, "grad_norm": 0.1479530930519104, "learning_rate": 4.589321407628907e-06, "loss": 0.0096, "step": 7620 }, { "epoch": 0.3462103466073768, "grad_norm": 0.10039757937192917, "learning_rate": 4.5646253816310175e-06, "loss": 0.0106, "step": 7630 }, { "epoch": 0.346664095423376, "grad_norm": 0.14793506264686584, "learning_rate": 4.539976316491272e-06, "loss": 0.0117, "step": 7640 }, { "epoch": 0.3471178442393752, "grad_norm": 0.11778666079044342, "learning_rate": 4.515374425174062e-06, "loss": 0.01, "step": 7650 }, { "epoch": 0.3475715930553744, "grad_norm": 0.143499955534935, "learning_rate": 4.49081992023621e-06, "loss": 0.0085, "step": 7660 }, { "epoch": 0.3480253418713736, "grad_norm": 0.1290217489004135, "learning_rate": 4.466313013825119e-06, "loss": 0.0099, "step": 7670 }, { "epoch": 0.34847909068737276, "grad_norm": 0.1522817611694336, "learning_rate": 4.4418539176769456e-06, "loss": 0.0112, "step": 7680 }, { "epoch": 0.3489328395033719, "grad_norm": 0.11977915465831757, "learning_rate": 4.417442843114786e-06, "loss": 0.0104, "step": 7690 }, { "epoch": 0.3493865883193711, "grad_norm": 0.10978726297616959, "learning_rate": 4.393080001046818e-06, "loss": 0.0111, "step": 7700 }, { "epoch": 0.34984033713537027, "grad_norm": 0.14366991817951202, "learning_rate": 4.368765601964516e-06, "loss": 0.0133, "step": 7710 }, { "epoch": 0.35029408595136946, "grad_norm": 0.1092318445444107, "learning_rate": 4.3444998559408025e-06, "loss": 0.0093, "step": 7720 }, { "epoch": 0.35074783476736865, "grad_norm": 0.11551111936569214, "learning_rate": 4.320282972628246e-06, "loss": 0.0113, "step": 7730 }, { "epoch": 0.35120158358336784, "grad_norm": 0.15743663907051086, "learning_rate": 4.2961151612572495e-06, "loss": 0.0123, "step": 7740 }, { "epoch": 0.351655332399367, "grad_norm": 0.10656075179576874, "learning_rate": 4.2719966306342386e-06, "loss": 0.0122, "step": 7750 }, { "epoch": 0.3521090812153662, "grad_norm": 0.1292315423488617, "learning_rate": 4.247927589139869e-06, "loss": 0.0102, "step": 7760 }, { "epoch": 0.3525628300313654, "grad_norm": 0.061548516154289246, "learning_rate": 4.223908244727211e-06, "loss": 0.0101, "step": 7770 }, { "epoch": 0.3530165788473646, "grad_norm": 0.13080838322639465, "learning_rate": 4.199938804919957e-06, "loss": 0.0093, "step": 7780 }, { "epoch": 0.3534703276633637, "grad_norm": 0.18990670144557953, "learning_rate": 4.176019476810631e-06, "loss": 0.014, "step": 7790 }, { "epoch": 0.3539240764793629, "grad_norm": 0.1360110193490982, "learning_rate": 4.152150467058805e-06, "loss": 0.0132, "step": 7800 }, { "epoch": 0.3543778252953621, "grad_norm": 0.10822978615760803, "learning_rate": 4.128331981889309e-06, "loss": 0.0106, "step": 7810 }, { "epoch": 0.3548315741113613, "grad_norm": 0.10335934162139893, "learning_rate": 4.104564227090437e-06, "loss": 0.0126, "step": 7820 }, { "epoch": 0.3552853229273605, "grad_norm": 0.12353721261024475, "learning_rate": 4.080847408012189e-06, "loss": 0.011, "step": 7830 }, { "epoch": 0.35573907174335967, "grad_norm": 0.12535031139850616, "learning_rate": 4.057181729564478e-06, "loss": 0.0092, "step": 7840 }, { "epoch": 0.35619282055935886, "grad_norm": 0.13195911049842834, "learning_rate": 4.033567396215387e-06, "loss": 0.009, "step": 7850 }, { "epoch": 0.35664656937535805, "grad_norm": 0.15567561984062195, "learning_rate": 4.0100046119893654e-06, "loss": 0.0142, "step": 7860 }, { "epoch": 0.35710031819135724, "grad_norm": 0.12166192382574081, "learning_rate": 3.986493580465498e-06, "loss": 0.0099, "step": 7870 }, { "epoch": 0.3575540670073564, "grad_norm": 0.15031491219997406, "learning_rate": 3.963034504775727e-06, "loss": 0.0114, "step": 7880 }, { "epoch": 0.3580078158233556, "grad_norm": 0.1494155079126358, "learning_rate": 3.939627587603103e-06, "loss": 0.0141, "step": 7890 }, { "epoch": 0.35846156463935475, "grad_norm": 0.16597557067871094, "learning_rate": 3.9162730311800455e-06, "loss": 0.012, "step": 7900 }, { "epoch": 0.35891531345535393, "grad_norm": 0.08207192271947861, "learning_rate": 3.8929710372865696e-06, "loss": 0.0104, "step": 7910 }, { "epoch": 0.3593690622713531, "grad_norm": 0.19074343144893646, "learning_rate": 3.869721807248571e-06, "loss": 0.011, "step": 7920 }, { "epoch": 0.3598228110873523, "grad_norm": 0.12037303298711777, "learning_rate": 3.8465255419360635e-06, "loss": 0.0112, "step": 7930 }, { "epoch": 0.3602765599033515, "grad_norm": 0.10831968486309052, "learning_rate": 3.823382441761454e-06, "loss": 0.0095, "step": 7940 }, { "epoch": 0.3607303087193507, "grad_norm": 0.08468834310770035, "learning_rate": 3.8002927066778193e-06, "loss": 0.0125, "step": 7950 }, { "epoch": 0.3611840575353499, "grad_norm": 0.10347823798656464, "learning_rate": 3.7772565361771596e-06, "loss": 0.0131, "step": 7960 }, { "epoch": 0.36163780635134907, "grad_norm": 0.1401902288198471, "learning_rate": 3.75427412928869e-06, "loss": 0.014, "step": 7970 }, { "epoch": 0.36209155516734826, "grad_norm": 0.1368122547864914, "learning_rate": 3.731345684577109e-06, "loss": 0.0122, "step": 7980 }, { "epoch": 0.36254530398334744, "grad_norm": 0.1546686738729477, "learning_rate": 3.7084714001409016e-06, "loss": 0.0127, "step": 7990 }, { "epoch": 0.3629990527993466, "grad_norm": 0.08418429642915726, "learning_rate": 3.6856514736106063e-06, "loss": 0.0102, "step": 8000 }, { "epoch": 0.36345280161534577, "grad_norm": 0.09853670001029968, "learning_rate": 3.6628861021471185e-06, "loss": 0.0121, "step": 8010 }, { "epoch": 0.36390655043134496, "grad_norm": 0.09226000308990479, "learning_rate": 3.6401754824399837e-06, "loss": 0.011, "step": 8020 }, { "epoch": 0.36436029924734414, "grad_norm": 0.09666698426008224, "learning_rate": 3.6175198107057107e-06, "loss": 0.0106, "step": 8030 }, { "epoch": 0.36481404806334333, "grad_norm": 0.15817151963710785, "learning_rate": 3.5949192826860513e-06, "loss": 0.0086, "step": 8040 }, { "epoch": 0.3652677968793425, "grad_norm": 0.10350827127695084, "learning_rate": 3.572374093646336e-06, "loss": 0.0107, "step": 8050 }, { "epoch": 0.3657215456953417, "grad_norm": 0.15225833654403687, "learning_rate": 3.5498844383737653e-06, "loss": 0.0109, "step": 8060 }, { "epoch": 0.3661752945113409, "grad_norm": 0.1303558200597763, "learning_rate": 3.5274505111757405e-06, "loss": 0.0102, "step": 8070 }, { "epoch": 0.3666290433273401, "grad_norm": 0.145473450422287, "learning_rate": 3.5050725058781765e-06, "loss": 0.013, "step": 8080 }, { "epoch": 0.3670827921433393, "grad_norm": 0.1219491958618164, "learning_rate": 3.482750615823838e-06, "loss": 0.0097, "step": 8090 }, { "epoch": 0.3675365409593384, "grad_norm": 0.09301994740962982, "learning_rate": 3.4604850338706554e-06, "loss": 0.0107, "step": 8100 }, { "epoch": 0.3679902897753376, "grad_norm": 0.09377678483724594, "learning_rate": 3.4382759523900678e-06, "loss": 0.01, "step": 8110 }, { "epoch": 0.3684440385913368, "grad_norm": 0.13796672224998474, "learning_rate": 3.4161235632653587e-06, "loss": 0.0105, "step": 8120 }, { "epoch": 0.368897787407336, "grad_norm": 0.11785976588726044, "learning_rate": 3.394028057889992e-06, "loss": 0.0101, "step": 8130 }, { "epoch": 0.36935153622333516, "grad_norm": 0.14343872666358948, "learning_rate": 3.3719896271659734e-06, "loss": 0.0109, "step": 8140 }, { "epoch": 0.36980528503933435, "grad_norm": 0.16964061558246613, "learning_rate": 3.3500084615021912e-06, "loss": 0.0103, "step": 8150 }, { "epoch": 0.37025903385533354, "grad_norm": 0.10996939241886139, "learning_rate": 3.3280847508127644e-06, "loss": 0.0089, "step": 8160 }, { "epoch": 0.37071278267133273, "grad_norm": 0.13238005340099335, "learning_rate": 3.306218684515413e-06, "loss": 0.0087, "step": 8170 }, { "epoch": 0.3711665314873319, "grad_norm": 0.12384450435638428, "learning_rate": 3.284410451529816e-06, "loss": 0.0087, "step": 8180 }, { "epoch": 0.3716202803033311, "grad_norm": 0.1461249142885208, "learning_rate": 3.2626602402759865e-06, "loss": 0.0094, "step": 8190 }, { "epoch": 0.37207402911933024, "grad_norm": 0.14945949614048004, "learning_rate": 3.240968238672633e-06, "loss": 0.0094, "step": 8200 }, { "epoch": 0.37252777793532943, "grad_norm": 0.11225463449954987, "learning_rate": 3.2193346341355413e-06, "loss": 0.0116, "step": 8210 }, { "epoch": 0.3729815267513286, "grad_norm": 0.13144910335540771, "learning_rate": 3.1977596135759524e-06, "loss": 0.0089, "step": 8220 }, { "epoch": 0.3734352755673278, "grad_norm": 0.14924506843090057, "learning_rate": 3.176243363398961e-06, "loss": 0.0126, "step": 8230 }, { "epoch": 0.373889024383327, "grad_norm": 0.257318913936615, "learning_rate": 3.1547860695018793e-06, "loss": 0.0125, "step": 8240 }, { "epoch": 0.3743427731993262, "grad_norm": 0.0802023783326149, "learning_rate": 3.13338791727266e-06, "loss": 0.0116, "step": 8250 }, { "epoch": 0.3747965220153254, "grad_norm": 0.13596881926059723, "learning_rate": 3.1120490915882694e-06, "loss": 0.0126, "step": 8260 }, { "epoch": 0.37525027083132456, "grad_norm": 0.1355019360780716, "learning_rate": 3.090769776813106e-06, "loss": 0.0091, "step": 8270 }, { "epoch": 0.37570401964732375, "grad_norm": 0.09479621052742004, "learning_rate": 3.0695501567973983e-06, "loss": 0.0084, "step": 8280 }, { "epoch": 0.37615776846332294, "grad_norm": 0.17646194994449615, "learning_rate": 3.0483904148756284e-06, "loss": 0.0126, "step": 8290 }, { "epoch": 0.37661151727932207, "grad_norm": 0.13487325608730316, "learning_rate": 3.0272907338649337e-06, "loss": 0.0128, "step": 8300 }, { "epoch": 0.37706526609532126, "grad_norm": 0.13014310598373413, "learning_rate": 3.006251296063536e-06, "loss": 0.0104, "step": 8310 }, { "epoch": 0.37751901491132045, "grad_norm": 0.10160695761442184, "learning_rate": 2.985272283249161e-06, "loss": 0.0114, "step": 8320 }, { "epoch": 0.37797276372731964, "grad_norm": 0.11651884764432907, "learning_rate": 2.9643538766774793e-06, "loss": 0.0101, "step": 8330 }, { "epoch": 0.3784265125433188, "grad_norm": 0.08997322618961334, "learning_rate": 2.943496257080527e-06, "loss": 0.0095, "step": 8340 }, { "epoch": 0.378880261359318, "grad_norm": 0.10125599801540375, "learning_rate": 2.9226996046651435e-06, "loss": 0.0125, "step": 8350 }, { "epoch": 0.3793340101753172, "grad_norm": 0.10461420565843582, "learning_rate": 2.901964099111435e-06, "loss": 0.0105, "step": 8360 }, { "epoch": 0.3797877589913164, "grad_norm": 0.1022501215338707, "learning_rate": 2.881289919571193e-06, "loss": 0.0121, "step": 8370 }, { "epoch": 0.3802415078073156, "grad_norm": 0.11119214445352554, "learning_rate": 2.860677244666373e-06, "loss": 0.0096, "step": 8380 }, { "epoch": 0.38069525662331477, "grad_norm": 0.10773328691720963, "learning_rate": 2.840126252487532e-06, "loss": 0.0128, "step": 8390 }, { "epoch": 0.38114900543931396, "grad_norm": 0.13215845823287964, "learning_rate": 2.8196371205922955e-06, "loss": 0.0096, "step": 8400 }, { "epoch": 0.3816027542553131, "grad_norm": 0.0992102399468422, "learning_rate": 2.799210026003831e-06, "loss": 0.0098, "step": 8410 }, { "epoch": 0.3820565030713123, "grad_norm": 0.16861070692539215, "learning_rate": 2.7788451452093067e-06, "loss": 0.0132, "step": 8420 }, { "epoch": 0.38251025188731147, "grad_norm": 0.11336692422628403, "learning_rate": 2.75854265415838e-06, "loss": 0.0089, "step": 8430 }, { "epoch": 0.38296400070331066, "grad_norm": 0.21186785399913788, "learning_rate": 2.738302728261665e-06, "loss": 0.0124, "step": 8440 }, { "epoch": 0.38341774951930985, "grad_norm": 0.09690193086862564, "learning_rate": 2.7181255423892192e-06, "loss": 0.0101, "step": 8450 }, { "epoch": 0.38387149833530904, "grad_norm": 0.1311059594154358, "learning_rate": 2.6980112708690374e-06, "loss": 0.0107, "step": 8460 }, { "epoch": 0.3843252471513082, "grad_norm": 0.13914784789085388, "learning_rate": 2.677960087485547e-06, "loss": 0.0096, "step": 8470 }, { "epoch": 0.3847789959673074, "grad_norm": 0.11388586461544037, "learning_rate": 2.657972165478103e-06, "loss": 0.0089, "step": 8480 }, { "epoch": 0.3852327447833066, "grad_norm": 0.18526536226272583, "learning_rate": 2.638047677539487e-06, "loss": 0.0107, "step": 8490 }, { "epoch": 0.3856864935993058, "grad_norm": 0.08570456504821777, "learning_rate": 2.618186795814418e-06, "loss": 0.0084, "step": 8500 }, { "epoch": 0.3861402424153049, "grad_norm": 0.09870341420173645, "learning_rate": 2.598389691898072e-06, "loss": 0.0131, "step": 8510 }, { "epoch": 0.3865939912313041, "grad_norm": 0.13254578411579132, "learning_rate": 2.578656536834586e-06, "loss": 0.013, "step": 8520 }, { "epoch": 0.3870477400473033, "grad_norm": 0.22662994265556335, "learning_rate": 2.5589875011156008e-06, "loss": 0.0134, "step": 8530 }, { "epoch": 0.3875014888633025, "grad_norm": 0.10000614821910858, "learning_rate": 2.539382754678764e-06, "loss": 0.0122, "step": 8540 }, { "epoch": 0.3879552376793017, "grad_norm": 0.14206960797309875, "learning_rate": 2.519842466906276e-06, "loss": 0.0084, "step": 8550 }, { "epoch": 0.38840898649530087, "grad_norm": 0.14792758226394653, "learning_rate": 2.5003668066234233e-06, "loss": 0.0091, "step": 8560 }, { "epoch": 0.38886273531130006, "grad_norm": 0.08291668444871902, "learning_rate": 2.480955942097121e-06, "loss": 0.0108, "step": 8570 }, { "epoch": 0.38931648412729924, "grad_norm": 0.11204250901937485, "learning_rate": 2.4616100410344634e-06, "loss": 0.0106, "step": 8580 }, { "epoch": 0.38977023294329843, "grad_norm": 0.16575933992862701, "learning_rate": 2.442329270581262e-06, "loss": 0.013, "step": 8590 }, { "epoch": 0.3902239817592976, "grad_norm": 0.10479316860437393, "learning_rate": 2.4231137973206097e-06, "loss": 0.0138, "step": 8600 }, { "epoch": 0.39067773057529676, "grad_norm": 0.1315084546804428, "learning_rate": 2.4039637872714417e-06, "loss": 0.0093, "step": 8610 }, { "epoch": 0.39113147939129594, "grad_norm": 0.14845795929431915, "learning_rate": 2.3848794058871073e-06, "loss": 0.0092, "step": 8620 }, { "epoch": 0.39158522820729513, "grad_norm": 0.09940434992313385, "learning_rate": 2.3658608180539243e-06, "loss": 0.0115, "step": 8630 }, { "epoch": 0.3920389770232943, "grad_norm": 0.13721604645252228, "learning_rate": 2.3469081880897694e-06, "loss": 0.0107, "step": 8640 }, { "epoch": 0.3924927258392935, "grad_norm": 0.11317367106676102, "learning_rate": 2.328021679742648e-06, "loss": 0.0087, "step": 8650 }, { "epoch": 0.3929464746552927, "grad_norm": 0.10845492035150528, "learning_rate": 2.309201456189286e-06, "loss": 0.0102, "step": 8660 }, { "epoch": 0.3934002234712919, "grad_norm": 0.11880891025066376, "learning_rate": 2.290447680033725e-06, "loss": 0.011, "step": 8670 }, { "epoch": 0.3938539722872911, "grad_norm": 0.11626913398504257, "learning_rate": 2.2717605133059007e-06, "loss": 0.0113, "step": 8680 }, { "epoch": 0.39430772110329027, "grad_norm": 0.10748831182718277, "learning_rate": 2.253140117460255e-06, "loss": 0.0096, "step": 8690 }, { "epoch": 0.39476146991928945, "grad_norm": 0.12304140627384186, "learning_rate": 2.2345866533743453e-06, "loss": 0.0097, "step": 8700 }, { "epoch": 0.3952152187352886, "grad_norm": 0.10171318054199219, "learning_rate": 2.2161002813474397e-06, "loss": 0.0091, "step": 8710 }, { "epoch": 0.3956689675512878, "grad_norm": 0.13390186429023743, "learning_rate": 2.197681161099149e-06, "loss": 0.0073, "step": 8720 }, { "epoch": 0.39612271636728696, "grad_norm": 0.1202036514878273, "learning_rate": 2.179329451768031e-06, "loss": 0.0096, "step": 8730 }, { "epoch": 0.39657646518328615, "grad_norm": 0.11887665092945099, "learning_rate": 2.161045311910227e-06, "loss": 0.0116, "step": 8740 }, { "epoch": 0.39703021399928534, "grad_norm": 0.12759540975093842, "learning_rate": 2.1428288994980816e-06, "loss": 0.0105, "step": 8750 }, { "epoch": 0.39748396281528453, "grad_norm": 0.1575375497341156, "learning_rate": 2.124680371918796e-06, "loss": 0.0121, "step": 8760 }, { "epoch": 0.3979377116312837, "grad_norm": 0.10857710242271423, "learning_rate": 2.106599885973044e-06, "loss": 0.0113, "step": 8770 }, { "epoch": 0.3983914604472829, "grad_norm": 0.13608422875404358, "learning_rate": 2.088587597873637e-06, "loss": 0.0094, "step": 8780 }, { "epoch": 0.3988452092632821, "grad_norm": 0.18532246351242065, "learning_rate": 2.070643663244163e-06, "loss": 0.0111, "step": 8790 }, { "epoch": 0.3992989580792813, "grad_norm": 0.14446675777435303, "learning_rate": 2.052768237117644e-06, "loss": 0.0123, "step": 8800 }, { "epoch": 0.3997527068952804, "grad_norm": 0.16273485124111176, "learning_rate": 2.034961473935203e-06, "loss": 0.0084, "step": 8810 }, { "epoch": 0.4002064557112796, "grad_norm": 0.0890413299202919, "learning_rate": 2.0172235275447284e-06, "loss": 0.009, "step": 8820 }, { "epoch": 0.4006602045272788, "grad_norm": 0.1420058310031891, "learning_rate": 1.9995545511995316e-06, "loss": 0.0094, "step": 8830 }, { "epoch": 0.401113953343278, "grad_norm": 0.10661931335926056, "learning_rate": 1.9819546975570382e-06, "loss": 0.0091, "step": 8840 }, { "epoch": 0.4015677021592772, "grad_norm": 0.15149742364883423, "learning_rate": 1.9644241186774593e-06, "loss": 0.0107, "step": 8850 }, { "epoch": 0.40202145097527636, "grad_norm": 0.09179948270320892, "learning_rate": 1.9469629660224907e-06, "loss": 0.0107, "step": 8860 }, { "epoch": 0.40247519979127555, "grad_norm": 0.12367530912160873, "learning_rate": 1.9295713904539892e-06, "loss": 0.0071, "step": 8870 }, { "epoch": 0.40292894860727474, "grad_norm": 0.08769435435533524, "learning_rate": 1.912249542232675e-06, "loss": 0.0113, "step": 8880 }, { "epoch": 0.40338269742327393, "grad_norm": 0.12802784144878387, "learning_rate": 1.8949975710168357e-06, "loss": 0.0112, "step": 8890 }, { "epoch": 0.4038364462392731, "grad_norm": 0.102344810962677, "learning_rate": 1.8778156258610292e-06, "loss": 0.0127, "step": 8900 }, { "epoch": 0.40429019505527225, "grad_norm": 0.1247081309556961, "learning_rate": 1.8607038552148039e-06, "loss": 0.0101, "step": 8910 }, { "epoch": 0.40474394387127144, "grad_norm": 0.08801490813493729, "learning_rate": 1.8436624069214071e-06, "loss": 0.0078, "step": 8920 }, { "epoch": 0.4051976926872706, "grad_norm": 0.10792089253664017, "learning_rate": 1.8266914282165116e-06, "loss": 0.0097, "step": 8930 }, { "epoch": 0.4056514415032698, "grad_norm": 0.073944590985775, "learning_rate": 1.80979106572694e-06, "loss": 0.009, "step": 8940 }, { "epoch": 0.406105190319269, "grad_norm": 0.12981021404266357, "learning_rate": 1.792961465469404e-06, "loss": 0.0104, "step": 8950 }, { "epoch": 0.4065589391352682, "grad_norm": 0.19840888679027557, "learning_rate": 1.7762027728492405e-06, "loss": 0.0128, "step": 8960 }, { "epoch": 0.4070126879512674, "grad_norm": 0.14734525978565216, "learning_rate": 1.759515132659153e-06, "loss": 0.0108, "step": 8970 }, { "epoch": 0.40746643676726657, "grad_norm": 0.1168968454003334, "learning_rate": 1.742898689077961e-06, "loss": 0.0109, "step": 8980 }, { "epoch": 0.40792018558326576, "grad_norm": 0.07594756782054901, "learning_rate": 1.726353585669356e-06, "loss": 0.0087, "step": 8990 }, { "epoch": 0.40837393439926495, "grad_norm": 0.1320488303899765, "learning_rate": 1.7098799653806663e-06, "loss": 0.0091, "step": 9000 }, { "epoch": 0.40882768321526414, "grad_norm": 0.12867127358913422, "learning_rate": 1.6934779705416082e-06, "loss": 0.0093, "step": 9010 }, { "epoch": 0.40928143203126327, "grad_norm": 0.13493511080741882, "learning_rate": 1.6771477428630656e-06, "loss": 0.0112, "step": 9020 }, { "epoch": 0.40973518084726246, "grad_norm": 0.13557837903499603, "learning_rate": 1.6608894234358708e-06, "loss": 0.0103, "step": 9030 }, { "epoch": 0.41018892966326165, "grad_norm": 0.1060461550951004, "learning_rate": 1.6447031527295744e-06, "loss": 0.0119, "step": 9040 }, { "epoch": 0.41064267847926084, "grad_norm": 0.14427106082439423, "learning_rate": 1.628589070591232e-06, "loss": 0.0113, "step": 9050 }, { "epoch": 0.41109642729526, "grad_norm": 0.11962931603193283, "learning_rate": 1.6125473162442107e-06, "loss": 0.0099, "step": 9060 }, { "epoch": 0.4115501761112592, "grad_norm": 0.17841389775276184, "learning_rate": 1.5965780282869693e-06, "loss": 0.0098, "step": 9070 }, { "epoch": 0.4120039249272584, "grad_norm": 0.06603584438562393, "learning_rate": 1.5806813446918657e-06, "loss": 0.0081, "step": 9080 }, { "epoch": 0.4124576737432576, "grad_norm": 0.10497824102640152, "learning_rate": 1.56485740280397e-06, "loss": 0.0104, "step": 9090 }, { "epoch": 0.4129114225592568, "grad_norm": 0.10903108865022659, "learning_rate": 1.5491063393398742e-06, "loss": 0.0112, "step": 9100 }, { "epoch": 0.41336517137525597, "grad_norm": 0.15178728103637695, "learning_rate": 1.5334282903865116e-06, "loss": 0.0109, "step": 9110 }, { "epoch": 0.4138189201912551, "grad_norm": 0.07373558729887009, "learning_rate": 1.5178233913999784e-06, "loss": 0.0097, "step": 9120 }, { "epoch": 0.4142726690072543, "grad_norm": 0.13341090083122253, "learning_rate": 1.5022917772043633e-06, "loss": 0.0097, "step": 9130 }, { "epoch": 0.4147264178232535, "grad_norm": 0.08545683324337006, "learning_rate": 1.4868335819905922e-06, "loss": 0.0093, "step": 9140 }, { "epoch": 0.41518016663925267, "grad_norm": 0.11698795109987259, "learning_rate": 1.4714489393152586e-06, "loss": 0.0099, "step": 9150 }, { "epoch": 0.41563391545525186, "grad_norm": 0.10725052654743195, "learning_rate": 1.4561379820994692e-06, "loss": 0.0092, "step": 9160 }, { "epoch": 0.41608766427125105, "grad_norm": 0.14751501381397247, "learning_rate": 1.4409008426277028e-06, "loss": 0.0098, "step": 9170 }, { "epoch": 0.41654141308725023, "grad_norm": 0.11136484146118164, "learning_rate": 1.4257376525466594e-06, "loss": 0.0113, "step": 9180 }, { "epoch": 0.4169951619032494, "grad_norm": 0.0880153626203537, "learning_rate": 1.4106485428641292e-06, "loss": 0.0087, "step": 9190 }, { "epoch": 0.4174489107192486, "grad_norm": 0.1130552664399147, "learning_rate": 1.3956336439478612e-06, "loss": 0.0091, "step": 9200 }, { "epoch": 0.4179026595352478, "grad_norm": 0.0887952521443367, "learning_rate": 1.3806930855244315e-06, "loss": 0.0094, "step": 9210 }, { "epoch": 0.41835640835124693, "grad_norm": 0.13212358951568604, "learning_rate": 1.3658269966781223e-06, "loss": 0.0099, "step": 9220 }, { "epoch": 0.4188101571672461, "grad_norm": 0.10839957743883133, "learning_rate": 1.3510355058498114e-06, "loss": 0.0083, "step": 9230 }, { "epoch": 0.4192639059832453, "grad_norm": 0.08262749761343002, "learning_rate": 1.3363187408358612e-06, "loss": 0.0095, "step": 9240 }, { "epoch": 0.4197176547992445, "grad_norm": 0.11758922785520554, "learning_rate": 1.3216768287870185e-06, "loss": 0.0095, "step": 9250 }, { "epoch": 0.4201714036152437, "grad_norm": 0.13314858078956604, "learning_rate": 1.3071098962073004e-06, "loss": 0.0076, "step": 9260 }, { "epoch": 0.4206251524312429, "grad_norm": 0.08712945133447647, "learning_rate": 1.292618068952921e-06, "loss": 0.0105, "step": 9270 }, { "epoch": 0.42107890124724207, "grad_norm": 0.12162909656763077, "learning_rate": 1.2782014722311897e-06, "loss": 0.0086, "step": 9280 }, { "epoch": 0.42153265006324125, "grad_norm": 0.0974271148443222, "learning_rate": 1.2638602305994364e-06, "loss": 0.0104, "step": 9290 }, { "epoch": 0.42198639887924044, "grad_norm": 0.08132126927375793, "learning_rate": 1.2495944679639383e-06, "loss": 0.0108, "step": 9300 }, { "epoch": 0.42244014769523963, "grad_norm": 0.09410033375024796, "learning_rate": 1.2354043075788391e-06, "loss": 0.0103, "step": 9310 }, { "epoch": 0.42289389651123876, "grad_norm": 0.11205838620662689, "learning_rate": 1.2212898720450915e-06, "loss": 0.0109, "step": 9320 }, { "epoch": 0.42334764532723795, "grad_norm": 0.08991268277168274, "learning_rate": 1.2072512833093964e-06, "loss": 0.0103, "step": 9330 }, { "epoch": 0.42380139414323714, "grad_norm": 0.15082821249961853, "learning_rate": 1.1932886626631512e-06, "loss": 0.0117, "step": 9340 }, { "epoch": 0.42425514295923633, "grad_norm": 0.16661033034324646, "learning_rate": 1.179402130741396e-06, "loss": 0.0082, "step": 9350 }, { "epoch": 0.4247088917752355, "grad_norm": 0.12957607209682465, "learning_rate": 1.165591807521781e-06, "loss": 0.0117, "step": 9360 }, { "epoch": 0.4251626405912347, "grad_norm": 0.12426209449768066, "learning_rate": 1.1518578123235191e-06, "loss": 0.0084, "step": 9370 }, { "epoch": 0.4256163894072339, "grad_norm": 0.1846769154071808, "learning_rate": 1.1382002638063584e-06, "loss": 0.0103, "step": 9380 }, { "epoch": 0.4260701382232331, "grad_norm": 0.13810695707798004, "learning_rate": 1.1246192799695666e-06, "loss": 0.01, "step": 9390 }, { "epoch": 0.4265238870392323, "grad_norm": 0.16302891075611115, "learning_rate": 1.1111149781508968e-06, "loss": 0.0106, "step": 9400 }, { "epoch": 0.42697763585523146, "grad_norm": 0.12007147818803787, "learning_rate": 1.0976874750255828e-06, "loss": 0.0091, "step": 9410 }, { "epoch": 0.4274313846712306, "grad_norm": 0.10976436734199524, "learning_rate": 1.0843368866053271e-06, "loss": 0.01, "step": 9420 }, { "epoch": 0.4278851334872298, "grad_norm": 0.12193258106708527, "learning_rate": 1.0710633282372996e-06, "loss": 0.0107, "step": 9430 }, { "epoch": 0.428338882303229, "grad_norm": 0.12088105827569962, "learning_rate": 1.0578669146031484e-06, "loss": 0.0104, "step": 9440 }, { "epoch": 0.42879263111922816, "grad_norm": 0.1749945878982544, "learning_rate": 1.0447477597179945e-06, "loss": 0.0104, "step": 9450 }, { "epoch": 0.42924637993522735, "grad_norm": 0.09087590128183365, "learning_rate": 1.0317059769294557e-06, "loss": 0.01, "step": 9460 }, { "epoch": 0.42970012875122654, "grad_norm": 0.1413136124610901, "learning_rate": 1.0187416789166672e-06, "loss": 0.0096, "step": 9470 }, { "epoch": 0.43015387756722573, "grad_norm": 0.1012006476521492, "learning_rate": 1.0058549776893068e-06, "loss": 0.012, "step": 9480 }, { "epoch": 0.4306076263832249, "grad_norm": 0.15488676726818085, "learning_rate": 9.930459845866313e-07, "loss": 0.0093, "step": 9490 }, { "epoch": 0.4310613751992241, "grad_norm": 0.12633702158927917, "learning_rate": 9.803148102765026e-07, "loss": 0.0089, "step": 9500 }, { "epoch": 0.4315151240152233, "grad_norm": 0.13584880530834198, "learning_rate": 9.676615647544452e-07, "loss": 0.0086, "step": 9510 }, { "epoch": 0.4319688728312225, "grad_norm": 0.17178885638713837, "learning_rate": 9.550863573426838e-07, "loss": 0.0106, "step": 9520 }, { "epoch": 0.4324226216472216, "grad_norm": 0.13634297251701355, "learning_rate": 9.425892966892136e-07, "loss": 0.0112, "step": 9530 }, { "epoch": 0.4328763704632208, "grad_norm": 0.1509203016757965, "learning_rate": 9.301704907668474e-07, "loss": 0.01, "step": 9540 }, { "epoch": 0.43333011927922, "grad_norm": 0.10559886693954468, "learning_rate": 9.178300468722901e-07, "loss": 0.0097, "step": 9550 }, { "epoch": 0.4337838680952192, "grad_norm": 0.1276906430721283, "learning_rate": 9.055680716252068e-07, "loss": 0.0097, "step": 9560 }, { "epoch": 0.43423761691121837, "grad_norm": 0.12274769693613052, "learning_rate": 8.933846709673078e-07, "loss": 0.0098, "step": 9570 }, { "epoch": 0.43469136572721756, "grad_norm": 0.12317559868097305, "learning_rate": 8.812799501614311e-07, "loss": 0.0087, "step": 9580 }, { "epoch": 0.43514511454321675, "grad_norm": 0.08936764299869537, "learning_rate": 8.692540137906314e-07, "loss": 0.0095, "step": 9590 }, { "epoch": 0.43559886335921594, "grad_norm": 0.13898974657058716, "learning_rate": 8.573069657572752e-07, "loss": 0.0097, "step": 9600 }, { "epoch": 0.4360526121752151, "grad_norm": 0.10988219082355499, "learning_rate": 8.454389092821458e-07, "loss": 0.0102, "step": 9610 }, { "epoch": 0.4365063609912143, "grad_norm": 0.13083656132221222, "learning_rate": 8.336499469035509e-07, "loss": 0.0114, "step": 9620 }, { "epoch": 0.43696010980721345, "grad_norm": 0.08637557923793793, "learning_rate": 8.219401804764382e-07, "loss": 0.0105, "step": 9630 }, { "epoch": 0.43741385862321264, "grad_norm": 0.12404067814350128, "learning_rate": 8.10309711171512e-07, "loss": 0.011, "step": 9640 }, { "epoch": 0.4378676074392118, "grad_norm": 0.1374865174293518, "learning_rate": 7.987586394743608e-07, "loss": 0.0099, "step": 9650 }, { "epoch": 0.438321356255211, "grad_norm": 0.13313066959381104, "learning_rate": 7.872870651845888e-07, "loss": 0.0111, "step": 9660 }, { "epoch": 0.4387751050712102, "grad_norm": 0.1515800505876541, "learning_rate": 7.758950874149541e-07, "loss": 0.0118, "step": 9670 }, { "epoch": 0.4392288538872094, "grad_norm": 0.19143113493919373, "learning_rate": 7.645828045905157e-07, "loss": 0.0114, "step": 9680 }, { "epoch": 0.4396826027032086, "grad_norm": 0.125772625207901, "learning_rate": 7.533503144477738e-07, "loss": 0.0098, "step": 9690 }, { "epoch": 0.44013635151920777, "grad_norm": 0.13882988691329956, "learning_rate": 7.421977140338376e-07, "loss": 0.0109, "step": 9700 }, { "epoch": 0.44059010033520696, "grad_norm": 0.12801378965377808, "learning_rate": 7.311250997055752e-07, "loss": 0.0105, "step": 9710 }, { "epoch": 0.44104384915120615, "grad_norm": 0.12278799712657928, "learning_rate": 7.201325671287862e-07, "loss": 0.0106, "step": 9720 }, { "epoch": 0.4414975979672053, "grad_norm": 0.10260124504566193, "learning_rate": 7.092202112773817e-07, "loss": 0.01, "step": 9730 }, { "epoch": 0.44195134678320447, "grad_norm": 0.07633895426988602, "learning_rate": 6.983881264325521e-07, "loss": 0.0074, "step": 9740 }, { "epoch": 0.44240509559920366, "grad_norm": 0.1047191321849823, "learning_rate": 6.876364061819574e-07, "loss": 0.01, "step": 9750 }, { "epoch": 0.44285884441520285, "grad_norm": 0.1236443743109703, "learning_rate": 6.769651434189195e-07, "loss": 0.0084, "step": 9760 }, { "epoch": 0.44331259323120203, "grad_norm": 0.1673034131526947, "learning_rate": 6.663744303416231e-07, "loss": 0.0092, "step": 9770 }, { "epoch": 0.4437663420472012, "grad_norm": 0.11086335778236389, "learning_rate": 6.558643584523117e-07, "loss": 0.0094, "step": 9780 }, { "epoch": 0.4442200908632004, "grad_norm": 0.12456338852643967, "learning_rate": 6.454350185564994e-07, "loss": 0.0089, "step": 9790 }, { "epoch": 0.4446738396791996, "grad_norm": 0.0809222161769867, "learning_rate": 6.350865007621887e-07, "loss": 0.0105, "step": 9800 }, { "epoch": 0.4451275884951988, "grad_norm": 0.12166428565979004, "learning_rate": 6.248188944790933e-07, "loss": 0.0072, "step": 9810 }, { "epoch": 0.445581337311198, "grad_norm": 0.051547128707170486, "learning_rate": 6.146322884178591e-07, "loss": 0.0086, "step": 9820 }, { "epoch": 0.4460350861271971, "grad_norm": 0.15253229439258575, "learning_rate": 6.045267705893043e-07, "loss": 0.0094, "step": 9830 }, { "epoch": 0.4464888349431963, "grad_norm": 0.094617560505867, "learning_rate": 5.945024283036549e-07, "loss": 0.0083, "step": 9840 }, { "epoch": 0.4469425837591955, "grad_norm": 0.10362132638692856, "learning_rate": 5.845593481697931e-07, "loss": 0.0091, "step": 9850 }, { "epoch": 0.4473963325751947, "grad_norm": 0.11719254404306412, "learning_rate": 5.746976160945051e-07, "loss": 0.0094, "step": 9860 }, { "epoch": 0.44785008139119387, "grad_norm": 0.18373169004917145, "learning_rate": 5.649173172817457e-07, "loss": 0.0115, "step": 9870 }, { "epoch": 0.44830383020719305, "grad_norm": 0.15360820293426514, "learning_rate": 5.55218536231894e-07, "loss": 0.0125, "step": 9880 }, { "epoch": 0.44875757902319224, "grad_norm": 0.09183119237422943, "learning_rate": 5.456013567410312e-07, "loss": 0.009, "step": 9890 }, { "epoch": 0.44921132783919143, "grad_norm": 0.11628349125385284, "learning_rate": 5.360658619002068e-07, "loss": 0.0118, "step": 9900 }, { "epoch": 0.4496650766551906, "grad_norm": 0.10773396492004395, "learning_rate": 5.266121340947327e-07, "loss": 0.0093, "step": 9910 }, { "epoch": 0.4501188254711898, "grad_norm": 0.10770761966705322, "learning_rate": 5.172402550034639e-07, "loss": 0.0109, "step": 9920 }, { "epoch": 0.45057257428718894, "grad_norm": 0.06062379479408264, "learning_rate": 5.079503055980939e-07, "loss": 0.0083, "step": 9930 }, { "epoch": 0.45102632310318813, "grad_norm": 0.09935268759727478, "learning_rate": 4.987423661424517e-07, "loss": 0.0102, "step": 9940 }, { "epoch": 0.4514800719191873, "grad_norm": 0.10356450080871582, "learning_rate": 4.896165161918176e-07, "loss": 0.0099, "step": 9950 }, { "epoch": 0.4519338207351865, "grad_norm": 0.16462033987045288, "learning_rate": 4.805728345922267e-07, "loss": 0.009, "step": 9960 }, { "epoch": 0.4523875695511857, "grad_norm": 0.13190323114395142, "learning_rate": 4.716113994797944e-07, "loss": 0.0133, "step": 9970 }, { "epoch": 0.4528413183671849, "grad_norm": 0.10049797594547272, "learning_rate": 4.627322882800345e-07, "loss": 0.0117, "step": 9980 }, { "epoch": 0.4532950671831841, "grad_norm": 0.15959760546684265, "learning_rate": 4.5393557770719744e-07, "loss": 0.0109, "step": 9990 }, { "epoch": 0.45374881599918326, "grad_norm": 0.18049407005310059, "learning_rate": 4.4522134376359995e-07, "loss": 0.0103, "step": 10000 }, { "epoch": 0.45420256481518245, "grad_norm": 0.1254950761795044, "learning_rate": 4.3658966173897866e-07, "loss": 0.0099, "step": 10010 }, { "epoch": 0.45465631363118164, "grad_norm": 0.1041630282998085, "learning_rate": 4.2804060620982747e-07, "loss": 0.0081, "step": 10020 }, { "epoch": 0.4551100624471808, "grad_norm": 0.13439397513866425, "learning_rate": 4.1957425103876235e-07, "loss": 0.0104, "step": 10030 }, { "epoch": 0.45556381126317996, "grad_norm": 0.09225888550281525, "learning_rate": 4.111906693738799e-07, "loss": 0.0076, "step": 10040 }, { "epoch": 0.45601756007917915, "grad_norm": 0.11776017397642136, "learning_rate": 4.02889933648124e-07, "loss": 0.0098, "step": 10050 }, { "epoch": 0.45647130889517834, "grad_norm": 0.08760102838277817, "learning_rate": 3.946721155786615e-07, "loss": 0.0069, "step": 10060 }, { "epoch": 0.45692505771117753, "grad_norm": 0.11535688489675522, "learning_rate": 3.865372861662664e-07, "loss": 0.0094, "step": 10070 }, { "epoch": 0.4573788065271767, "grad_norm": 0.13672105967998505, "learning_rate": 3.784855156946965e-07, "loss": 0.0092, "step": 10080 }, { "epoch": 0.4578325553431759, "grad_norm": 0.12640318274497986, "learning_rate": 3.705168737300968e-07, "loss": 0.0133, "step": 10090 }, { "epoch": 0.4582863041591751, "grad_norm": 0.12096633017063141, "learning_rate": 3.626314291203914e-07, "loss": 0.0108, "step": 10100 }, { "epoch": 0.4587400529751743, "grad_norm": 0.15392784774303436, "learning_rate": 3.548292499946937e-07, "loss": 0.0102, "step": 10110 }, { "epoch": 0.4591938017911735, "grad_norm": 0.123865507543087, "learning_rate": 3.4711040376271264e-07, "loss": 0.0094, "step": 10120 }, { "epoch": 0.45964755060717266, "grad_norm": 0.10113903135061264, "learning_rate": 3.394749571141731e-07, "loss": 0.01, "step": 10130 }, { "epoch": 0.4601012994231718, "grad_norm": 0.1781150996685028, "learning_rate": 3.319229760182441e-07, "loss": 0.0113, "step": 10140 }, { "epoch": 0.460555048239171, "grad_norm": 0.14278154075145721, "learning_rate": 3.244545257229559e-07, "loss": 0.0126, "step": 10150 }, { "epoch": 0.46100879705517017, "grad_norm": 0.14660927653312683, "learning_rate": 3.170696707546539e-07, "loss": 0.0089, "step": 10160 }, { "epoch": 0.46146254587116936, "grad_norm": 0.15213704109191895, "learning_rate": 3.0976847491742347e-07, "loss": 0.011, "step": 10170 }, { "epoch": 0.46191629468716855, "grad_norm": 0.10338576883077621, "learning_rate": 3.0255100129255364e-07, "loss": 0.0094, "step": 10180 }, { "epoch": 0.46237004350316774, "grad_norm": 0.13408440351486206, "learning_rate": 2.9541731223797997e-07, "loss": 0.0101, "step": 10190 }, { "epoch": 0.4628237923191669, "grad_norm": 0.10805374383926392, "learning_rate": 2.883674693877558e-07, "loss": 0.0095, "step": 10200 }, { "epoch": 0.4632775411351661, "grad_norm": 0.06863702088594437, "learning_rate": 2.8140153365151304e-07, "loss": 0.0121, "step": 10210 }, { "epoch": 0.4637312899511653, "grad_norm": 0.05823841691017151, "learning_rate": 2.7451956521393983e-07, "loss": 0.0097, "step": 10220 }, { "epoch": 0.4641850387671645, "grad_norm": 0.12672090530395508, "learning_rate": 2.677216235342561e-07, "loss": 0.0093, "step": 10230 }, { "epoch": 0.4646387875831636, "grad_norm": 0.13820886611938477, "learning_rate": 2.6100776734570345e-07, "loss": 0.0114, "step": 10240 }, { "epoch": 0.4650925363991628, "grad_norm": 0.11121530085802078, "learning_rate": 2.543780546550401e-07, "loss": 0.0101, "step": 10250 }, { "epoch": 0.465546285215162, "grad_norm": 0.17037838697433472, "learning_rate": 2.478325427420336e-07, "loss": 0.0105, "step": 10260 }, { "epoch": 0.4660000340311612, "grad_norm": 0.0819673016667366, "learning_rate": 2.4137128815896803e-07, "loss": 0.0108, "step": 10270 }, { "epoch": 0.4664537828471604, "grad_norm": 0.10760702937841415, "learning_rate": 2.3499434673015852e-07, "loss": 0.01, "step": 10280 }, { "epoch": 0.46690753166315957, "grad_norm": 0.0882890373468399, "learning_rate": 2.2870177355146406e-07, "loss": 0.0099, "step": 10290 }, { "epoch": 0.46736128047915876, "grad_norm": 0.135042205452919, "learning_rate": 2.2249362298981892e-07, "loss": 0.0124, "step": 10300 }, { "epoch": 0.46781502929515795, "grad_norm": 0.11944553256034851, "learning_rate": 2.1636994868275085e-07, "loss": 0.0079, "step": 10310 }, { "epoch": 0.46826877811115714, "grad_norm": 0.09973999112844467, "learning_rate": 2.1033080353793144e-07, "loss": 0.0091, "step": 10320 }, { "epoch": 0.4687225269271563, "grad_norm": 0.10816528648138046, "learning_rate": 2.043762397327087e-07, "loss": 0.0096, "step": 10330 }, { "epoch": 0.46917627574315546, "grad_norm": 0.1579977422952652, "learning_rate": 1.985063087136596e-07, "loss": 0.0086, "step": 10340 }, { "epoch": 0.46963002455915465, "grad_norm": 0.11629801988601685, "learning_rate": 1.927210611961494e-07, "loss": 0.0097, "step": 10350 }, { "epoch": 0.47008377337515383, "grad_norm": 0.12378907203674316, "learning_rate": 1.870205471638864e-07, "loss": 0.0088, "step": 10360 }, { "epoch": 0.470537522191153, "grad_norm": 0.11535657197237015, "learning_rate": 1.814048158684978e-07, "loss": 0.0091, "step": 10370 }, { "epoch": 0.4709912710071522, "grad_norm": 0.16682352125644684, "learning_rate": 1.7587391582909452e-07, "loss": 0.0098, "step": 10380 }, { "epoch": 0.4714450198231514, "grad_norm": 0.11436964571475983, "learning_rate": 1.7042789483186273e-07, "loss": 0.0103, "step": 10390 }, { "epoch": 0.4718987686391506, "grad_norm": 0.11008341610431671, "learning_rate": 1.6506679992964292e-07, "loss": 0.0089, "step": 10400 }, { "epoch": 0.4723525174551498, "grad_norm": 0.1441035121679306, "learning_rate": 1.597906774415281e-07, "loss": 0.0092, "step": 10410 }, { "epoch": 0.47280626627114897, "grad_norm": 0.11642053723335266, "learning_rate": 1.5459957295245965e-07, "loss": 0.0098, "step": 10420 }, { "epoch": 0.47326001508714816, "grad_norm": 0.12789706885814667, "learning_rate": 1.494935313128376e-07, "loss": 0.0092, "step": 10430 }, { "epoch": 0.4737137639031473, "grad_norm": 0.127280130982399, "learning_rate": 1.4447259663812886e-07, "loss": 0.0096, "step": 10440 }, { "epoch": 0.4741675127191465, "grad_norm": 0.13570982217788696, "learning_rate": 1.395368123084917e-07, "loss": 0.0076, "step": 10450 }, { "epoch": 0.47462126153514567, "grad_norm": 0.15512555837631226, "learning_rate": 1.3468622096839524e-07, "loss": 0.0111, "step": 10460 }, { "epoch": 0.47507501035114486, "grad_norm": 0.09604503214359283, "learning_rate": 1.2992086452625175e-07, "loss": 0.0083, "step": 10470 }, { "epoch": 0.47552875916714404, "grad_norm": 0.1286637783050537, "learning_rate": 1.252407841540626e-07, "loss": 0.0101, "step": 10480 }, { "epoch": 0.47598250798314323, "grad_norm": 0.12484487891197205, "learning_rate": 1.2064602028704742e-07, "loss": 0.0094, "step": 10490 }, { "epoch": 0.4764362567991424, "grad_norm": 0.10206886380910873, "learning_rate": 1.1613661262331099e-07, "loss": 0.0094, "step": 10500 }, { "epoch": 0.4768900056151416, "grad_norm": 0.12927402555942535, "learning_rate": 1.1171260012348805e-07, "loss": 0.0104, "step": 10510 }, { "epoch": 0.4773437544311408, "grad_norm": 0.09732365608215332, "learning_rate": 1.0737402101041349e-07, "loss": 0.0091, "step": 10520 }, { "epoch": 0.47779750324714, "grad_norm": 0.10109642148017883, "learning_rate": 1.0312091276878821e-07, "loss": 0.0099, "step": 10530 }, { "epoch": 0.4782512520631391, "grad_norm": 0.09376849234104156, "learning_rate": 9.895331214485937e-08, "loss": 0.0094, "step": 10540 }, { "epoch": 0.4787050008791383, "grad_norm": 0.11916600167751312, "learning_rate": 9.487125514610063e-08, "loss": 0.0092, "step": 10550 }, { "epoch": 0.4791587496951375, "grad_norm": 0.1303083449602127, "learning_rate": 9.087477704089686e-08, "loss": 0.0093, "step": 10560 }, { "epoch": 0.4796124985111367, "grad_norm": 0.11214817315340042, "learning_rate": 8.696391235824886e-08, "loss": 0.0101, "step": 10570 }, { "epoch": 0.4800662473271359, "grad_norm": 0.14808695018291473, "learning_rate": 8.313869488746574e-08, "loss": 0.0085, "step": 10580 }, { "epoch": 0.48051999614313506, "grad_norm": 0.15055687725543976, "learning_rate": 7.939915767787853e-08, "loss": 0.0073, "step": 10590 }, { "epoch": 0.48097374495913425, "grad_norm": 0.12333395332098007, "learning_rate": 7.574533303855491e-08, "loss": 0.0106, "step": 10600 }, { "epoch": 0.48142749377513344, "grad_norm": 0.10209726542234421, "learning_rate": 7.217725253801488e-08, "loss": 0.0101, "step": 10610 }, { "epoch": 0.48188124259113263, "grad_norm": 0.13615788519382477, "learning_rate": 6.869494700396328e-08, "loss": 0.0096, "step": 10620 }, { "epoch": 0.4823349914071318, "grad_norm": 0.12133138626813889, "learning_rate": 6.529844652301997e-08, "loss": 0.0113, "step": 10630 }, { "epoch": 0.482788740223131, "grad_norm": 0.11862960457801819, "learning_rate": 6.19877804404645e-08, "loss": 0.0073, "step": 10640 }, { "epoch": 0.48324248903913014, "grad_norm": 0.11843565106391907, "learning_rate": 5.876297735997738e-08, "loss": 0.01, "step": 10650 }, { "epoch": 0.48369623785512933, "grad_norm": 0.10824906080961227, "learning_rate": 5.562406514339369e-08, "loss": 0.0091, "step": 10660 }, { "epoch": 0.4841499866711285, "grad_norm": 0.10267908871173859, "learning_rate": 5.257107091046654e-08, "loss": 0.0086, "step": 10670 }, { "epoch": 0.4846037354871277, "grad_norm": 0.1007552221417427, "learning_rate": 4.9604021038628384e-08, "loss": 0.0088, "step": 10680 }, { "epoch": 0.4850574843031269, "grad_norm": 0.1403747946023941, "learning_rate": 4.6722941162764546e-08, "loss": 0.0091, "step": 10690 }, { "epoch": 0.4855112331191261, "grad_norm": 0.1846962571144104, "learning_rate": 4.392785617499451e-08, "loss": 0.0096, "step": 10700 }, { "epoch": 0.4859649819351253, "grad_norm": 0.11575271934270859, "learning_rate": 4.1218790224450965e-08, "loss": 0.0112, "step": 10710 }, { "epoch": 0.48641873075112446, "grad_norm": 0.1052660271525383, "learning_rate": 3.859576671707554e-08, "loss": 0.0092, "step": 10720 }, { "epoch": 0.48687247956712365, "grad_norm": 0.12023202329874039, "learning_rate": 3.605880831541564e-08, "loss": 0.0122, "step": 10730 }, { "epoch": 0.48732622838312284, "grad_norm": 0.13231338560581207, "learning_rate": 3.36079369384279e-08, "loss": 0.0118, "step": 10740 }, { "epoch": 0.487779977199122, "grad_norm": 0.09625431150197983, "learning_rate": 3.124317376129171e-08, "loss": 0.0068, "step": 10750 }, { "epoch": 0.48823372601512116, "grad_norm": 0.1553715318441391, "learning_rate": 2.8964539215220468e-08, "loss": 0.0097, "step": 10760 }, { "epoch": 0.48868747483112035, "grad_norm": 0.10737349838018417, "learning_rate": 2.6772052987290575e-08, "loss": 0.009, "step": 10770 }, { "epoch": 0.48914122364711954, "grad_norm": 0.0638694316148758, "learning_rate": 2.4665734020270503e-08, "loss": 0.0086, "step": 10780 }, { "epoch": 0.4895949724631187, "grad_norm": 0.06933131068944931, "learning_rate": 2.2645600512452016e-08, "loss": 0.0104, "step": 10790 }, { "epoch": 0.4900487212791179, "grad_norm": 0.1229269951581955, "learning_rate": 2.0711669917501398e-08, "loss": 0.0078, "step": 10800 }, { "epoch": 0.4905024700951171, "grad_norm": 0.14476285874843597, "learning_rate": 1.8863958944300708e-08, "loss": 0.0101, "step": 10810 }, { "epoch": 0.4909562189111163, "grad_norm": 0.1204502135515213, "learning_rate": 1.710248355680788e-08, "loss": 0.011, "step": 10820 }, { "epoch": 0.4914099677271155, "grad_norm": 0.09992097318172455, "learning_rate": 1.5427258973919058e-08, "loss": 0.0088, "step": 10830 }, { "epoch": 0.49186371654311467, "grad_norm": 0.1434265375137329, "learning_rate": 1.3838299669334255e-08, "loss": 0.009, "step": 10840 }, { "epoch": 0.4923174653591138, "grad_norm": 0.08076232671737671, "learning_rate": 1.2335619371434126e-08, "loss": 0.0082, "step": 10850 }, { "epoch": 0.492771214175113, "grad_norm": 0.10127487778663635, "learning_rate": 1.0919231063161173e-08, "loss": 0.0095, "step": 10860 }, { "epoch": 0.4932249629911122, "grad_norm": 0.11844775080680847, "learning_rate": 9.589146981907604e-09, "loss": 0.0087, "step": 10870 }, { "epoch": 0.49367871180711137, "grad_norm": 0.10561437904834747, "learning_rate": 8.345378619408762e-09, "loss": 0.0071, "step": 10880 }, { "epoch": 0.49413246062311056, "grad_norm": 0.14991527795791626, "learning_rate": 7.187936721646527e-09, "loss": 0.0102, "step": 10890 }, { "epoch": 0.49458620943910975, "grad_norm": 0.12201883643865585, "learning_rate": 6.116831288751624e-09, "loss": 0.0098, "step": 10900 }, { "epoch": 0.49503995825510894, "grad_norm": 0.09622867405414581, "learning_rate": 5.13207157492257e-09, "loss": 0.0099, "step": 10910 }, { "epoch": 0.4954937070711081, "grad_norm": 0.1273748129606247, "learning_rate": 4.233666088341304e-09, "loss": 0.01, "step": 10920 }, { "epoch": 0.4959474558871073, "grad_norm": 0.13358333706855774, "learning_rate": 3.4216225911032354e-09, "loss": 0.0098, "step": 10930 }, { "epoch": 0.4964012047031065, "grad_norm": 0.11978624016046524, "learning_rate": 2.6959480991484157e-09, "loss": 0.0081, "step": 10940 }, { "epoch": 0.49685495351910564, "grad_norm": 0.15657968819141388, "learning_rate": 2.0566488821993635e-09, "loss": 0.012, "step": 10950 }, { "epoch": 0.4973087023351048, "grad_norm": 0.13145264983177185, "learning_rate": 1.503730463709996e-09, "loss": 0.0079, "step": 10960 }, { "epoch": 0.497762451151104, "grad_norm": 0.19312238693237305, "learning_rate": 1.0371976208167766e-09, "loss": 0.0119, "step": 10970 }, { "epoch": 0.4982161999671032, "grad_norm": 0.09484487026929855, "learning_rate": 6.570543842965293e-10, "loss": 0.0104, "step": 10980 }, { "epoch": 0.4986699487831024, "grad_norm": 0.06194043532013893, "learning_rate": 3.6330403853201966e-10, "loss": 0.0074, "step": 10990 }, { "epoch": 0.4991236975991016, "grad_norm": 0.09161436557769775, "learning_rate": 1.5594912148420017e-10, "loss": 0.0103, "step": 11000 }, { "epoch": 0.49957744641510077, "grad_norm": 0.12720157206058502, "learning_rate": 3.49914246700056e-11, "loss": 0.0129, "step": 11010 } ], "logging_steps": 10, "max_steps": 11019, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.677505360785572e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }