[ { "step": 0, "train_loss": 10.51517790555954, "val_loss": 10.160391330718994, "lr": 0.0, "tokens_per_sec": 2158.630169134002, "mfu": 0.6606935809350966, "total_flops": 1161476270194688.0 }, { "step": 50, "train_loss": 6.364298492670059, "val_loss": 6.596065902709961, "lr": 0.01, "tokens_per_sec": 70222.7689057813, "mfu": 21.49313639499045, "total_flops": 5.923528977992909e+16 }, { "step": 100, "train_loss": 5.122354328632355, "val_loss": 5.2908248662948605, "lr": 0.02, "tokens_per_sec": 70112.78707176603, "mfu": 21.459474171237453, "total_flops": 1.1730910328966349e+17 }, { "step": 150, "train_loss": 4.422913044691086, "val_loss": 4.639911162853241, "lr": 0.02, "tokens_per_sec": 70142.59368173171, "mfu": 21.46859710306494, "total_flops": 1.753829167993979e+17 }, { "step": 200, "train_loss": 3.8527721911668777, "val_loss": 4.023697471618652, "lr": 0.02, "tokens_per_sec": 70024.24338660217, "mfu": 21.432373538041187, "total_flops": 2.334567303091323e+17 }, { "step": 250, "train_loss": 3.522311434149742, "val_loss": 3.7592015504837035, "lr": 0.02, "tokens_per_sec": 69977.49930725315, "mfu": 21.41806654205168, "total_flops": 2.915305438188667e+17 }, { "step": 300, "train_loss": 3.396108627319336, "val_loss": 3.6023362636566163, "lr": 0.02, "tokens_per_sec": 69978.57065144836, "mfu": 21.418394449185968, "total_flops": 3.496043573286011e+17 }, { "step": 350, "train_loss": 3.4941859245300293, "val_loss": 3.5021621108055117, "lr": 0.02, "tokens_per_sec": 70027.76531890646, "mfu": 21.433451498545583, "total_flops": 4.076781708383355e+17 }, { "step": 400, "train_loss": 3.309720605611801, "val_loss": 3.464106559753418, "lr": 0.02, "tokens_per_sec": 70089.33947406986, "mfu": 21.452297547141427, "total_flops": 4.657519843480699e+17 }, { "step": 450, "train_loss": 3.178159549832344, "val_loss": 3.4234487891197203, "lr": 0.02, "tokens_per_sec": 69999.7253376068, "mfu": 21.424869280100232, "total_flops": 5.238257978578043e+17 }, { "step": 500, "train_loss": 3.371193677186966, "val_loss": 3.3580533504486083, "lr": 0.02, "tokens_per_sec": 69956.24699933741, "mfu": 21.411561831971763, "total_flops": 5.818996113675387e+17 }, { "step": 550, "train_loss": 3.199240431189537, "val_loss": 3.290653383731842, "lr": 0.02, "tokens_per_sec": 70012.44780624029, "mfu": 21.42876325577054, "total_flops": 6.399734248772731e+17 }, { "step": 600, "train_loss": 3.006466805934906, "val_loss": 3.247062921524048, "lr": 0.02, "tokens_per_sec": 70007.14506416151, "mfu": 21.42714024146133, "total_flops": 6.980472383870075e+17 }, { "step": 650, "train_loss": 3.0513384342193604, "val_loss": 3.206310844421387, "lr": 0.02, "tokens_per_sec": 70062.60969625264, "mfu": 21.444116343674445, "total_flops": 7.561210518967419e+17 }, { "step": 700, "train_loss": 3.1359871327877045, "val_loss": 3.163962686061859, "lr": 0.02, "tokens_per_sec": 70051.95890783441, "mfu": 21.44085644874643, "total_flops": 8.141948654064763e+17 }, { "step": 750, "train_loss": 3.0369881242513657, "val_loss": 3.144151282310486, "lr": 0.02, "tokens_per_sec": 69949.8472232128, "mfu": 21.409603047616102, "total_flops": 8.722686789162107e+17 }, { "step": 800, "train_loss": 3.091953292489052, "val_loss": 3.1556016325950624, "lr": 0.02, "tokens_per_sec": 69977.76396697006, "mfu": 21.41814754665293, "total_flops": 9.303424924259451e+17 }, { "step": 850, "train_loss": 3.090632662177086, "val_loss": 3.136023700237274, "lr": 0.02, "tokens_per_sec": 70045.18964450127, "mfu": 21.43878457515927, "total_flops": 9.884163059356795e+17 }, { "step": 900, "train_loss": 3.0360343009233475, "val_loss": 3.1322978258132936, "lr": 0.02, "tokens_per_sec": 70022.76171204478, "mfu": 21.431920040780238, "total_flops": 1.0464901194454139e+18 }, { "step": 950, "train_loss": 2.9815044552087784, "val_loss": 3.1037992238998413, "lr": 0.02, "tokens_per_sec": 70034.69091604456, "mfu": 21.435571221339565, "total_flops": 1.1045639329551483e+18 }, { "step": 1000, "train_loss": 3.0359542667865753, "val_loss": 3.1057387828826903, "lr": 0.02, "tokens_per_sec": 70007.97170503181, "mfu": 21.427393252062473, "total_flops": 1.1626377464648827e+18 }, { "step": 1050, "train_loss": 2.9832030683755875, "val_loss": 3.0936940908432007, "lr": 0.02, "tokens_per_sec": 69997.46563636261, "mfu": 21.424177651618333, "total_flops": 1.220711559974617e+18 }, { "step": 1100, "train_loss": 2.9214714765548706, "val_loss": 3.0613293051719666, "lr": 0.02, "tokens_per_sec": 70055.72019986459, "mfu": 21.442007670264545, "total_flops": 1.2787853734843515e+18 }, { "step": 1150, "train_loss": 2.9381714314222336, "val_loss": 3.0642426133155825, "lr": 0.02, "tokens_per_sec": 69989.99293893101, "mfu": 21.421890477420586, "total_flops": 1.336859186994086e+18 }, { "step": 1200, "train_loss": 3.0071898102760315, "val_loss": 3.063199543952942, "lr": 0.02, "tokens_per_sec": 69989.67483553122, "mfu": 21.421793115270585, "total_flops": 1.3949330005038203e+18 }, { "step": 1250, "train_loss": 3.078396648168564, "val_loss": 3.0594762563705444, "lr": 0.02, "tokens_per_sec": 70000.17087793698, "mfu": 21.425005646968636, "total_flops": 1.4530068140135547e+18 }, { "step": 1300, "train_loss": 3.049118548631668, "val_loss": 3.040031540393829, "lr": 0.02, "tokens_per_sec": 69999.04751218345, "mfu": 21.42466181755635, "total_flops": 1.511080627523289e+18 }, { "step": 1350, "train_loss": 3.1371348053216934, "val_loss": 3.052434301376343, "lr": 0.02, "tokens_per_sec": 70121.85376955378, "mfu": 21.46224922233898, "total_flops": 1.5691544410330235e+18 }, { "step": 1400, "train_loss": 2.9777035415172577, "val_loss": 3.0453516125679014, "lr": 0.02, "tokens_per_sec": 70000.47080475416, "mfu": 21.42509744579814, "total_flops": 1.627228254542758e+18 }, { "step": 1450, "train_loss": 2.9373037219047546, "val_loss": 3.0371802687644958, "lr": 0.02, "tokens_per_sec": 69993.52379124102, "mfu": 21.422971168077932, "total_flops": 1.6853020680524923e+18 }, { "step": 1500, "train_loss": 2.8891674876213074, "val_loss": 3.0314683616161346, "lr": 0.02, "tokens_per_sec": 70034.81448436435, "mfu": 21.43560904198937, "total_flops": 1.7433758815622267e+18 }, { "step": 1550, "train_loss": 2.8753224909305573, "val_loss": 3.0194230318069457, "lr": 0.02, "tokens_per_sec": 70032.54236910073, "mfu": 21.43491361393908, "total_flops": 1.801449695071961e+18 }, { "step": 1600, "train_loss": 2.873562440276146, "val_loss": 3.0389770805835723, "lr": 0.02, "tokens_per_sec": 70093.54067527145, "mfu": 21.453583411994888, "total_flops": 1.8595235085816955e+18 }, { "step": 1650, "train_loss": 2.8681159019470215, "val_loss": 3.0404816269874573, "lr": 0.02, "tokens_per_sec": 70092.78506979671, "mfu": 21.4533521432514, "total_flops": 1.91759732209143e+18 }, { "step": 1700, "train_loss": 2.8477375507354736, "val_loss": 3.03014994263649, "lr": 0.02, "tokens_per_sec": 70043.1170883897, "mfu": 21.438150226330734, "total_flops": 1.9756711356011643e+18 }, { "step": 1750, "train_loss": 2.924525633454323, "val_loss": 3.0115107893943787, "lr": 0.02, "tokens_per_sec": 70089.97079499098, "mfu": 21.45249077601688, "total_flops": 2.0337449491108987e+18 }, { "step": 1800, "train_loss": 2.845635160803795, "val_loss": 3.021808397769928, "lr": 0.02, "tokens_per_sec": 70073.76133559164, "mfu": 21.447529534425474, "total_flops": 2.091818762620633e+18 }, { "step": 1850, "train_loss": 2.7774328142404556, "val_loss": 3.0103733837604523, "lr": 0.02, "tokens_per_sec": 70157.2173610318, "mfu": 21.47307298373286, "total_flops": 2.1498925761303675e+18 }, { "step": 1900, "train_loss": 2.7959036976099014, "val_loss": 3.010297340154648, "lr": 0.02, "tokens_per_sec": 70144.7278424267, "mfu": 21.469250307255273, "total_flops": 2.207966389640102e+18 }, { "step": 1950, "train_loss": 2.8740807473659515, "val_loss": 3.0071927487850187, "lr": 0.02, "tokens_per_sec": 70076.6894273708, "mfu": 21.448425737707787, "total_flops": 2.2660402031498363e+18 }, { "step": 2000, "train_loss": 2.964038208127022, "val_loss": 3.0082029044628142, "lr": 0.02, "tokens_per_sec": 69929.21289313334, "mfu": 21.403287482483492, "total_flops": 2.3241140166595707e+18 }, { "step": 2050, "train_loss": 2.812971353530884, "val_loss": 3.002509093284607, "lr": 0.02, "tokens_per_sec": 70071.29176833555, "mfu": 21.446773672093368, "total_flops": 2.382187830169305e+18 }, { "step": 2100, "train_loss": 2.9027664810419083, "val_loss": 2.985026216506958, "lr": 0.02, "tokens_per_sec": 70077.55031615285, "mfu": 21.44868923059338, "total_flops": 2.4402616436790395e+18 }, { "step": 2150, "train_loss": 2.7932936400175095, "val_loss": 2.9898294627666475, "lr": 0.02, "tokens_per_sec": 70081.02769582007, "mfu": 21.449753554838203, "total_flops": 2.498335457188774e+18 }, { "step": 2200, "train_loss": 2.8083430528640747, "val_loss": 2.979103463888168, "lr": 0.02, "tokens_per_sec": 70060.1254019016, "mfu": 21.443355973809172, "total_flops": 2.5564092706985083e+18 }, { "step": 2250, "train_loss": 2.8303453028202057, "val_loss": 2.9749884247779845, "lr": 0.02, "tokens_per_sec": 69815.61494395022, "mfu": 21.368518471605263, "total_flops": 2.6144830842082427e+18 }, { "step": 2300, "train_loss": 2.7061598151922226, "val_loss": 2.9586526334285734, "lr": 0.02, "tokens_per_sec": 70126.97363792777, "mfu": 21.463816264354005, "total_flops": 2.672556897717977e+18 }, { "step": 2350, "train_loss": 2.920316368341446, "val_loss": 2.9535837829113007, "lr": 0.02, "tokens_per_sec": 70137.85909871584, "mfu": 21.467147985633094, "total_flops": 2.7306307112277115e+18 }, { "step": 2400, "train_loss": 2.9253035485744476, "val_loss": 2.944886934757233, "lr": 0.02, "tokens_per_sec": 70066.0209632928, "mfu": 21.445160432777115, "total_flops": 2.788704524737446e+18 }, { "step": 2450, "train_loss": 2.8674067705869675, "val_loss": 2.940703195333481, "lr": 0.02, "tokens_per_sec": 70068.62992160743, "mfu": 21.44595895863673, "total_flops": 2.8467783382471803e+18 }, { "step": 2500, "train_loss": 2.8450111150741577, "val_loss": 2.9333001136779786, "lr": 0.02, "tokens_per_sec": 69838.27776156714, "mfu": 21.375454897464362, "total_flops": 2.9048521517569147e+18 }, { "step": 2550, "train_loss": 2.759008914232254, "val_loss": 2.9429529547691344, "lr": 0.02, "tokens_per_sec": 70104.37553002524, "mfu": 21.45689964424651, "total_flops": 2.962925965266649e+18 }, { "step": 2600, "train_loss": 2.565167874097824, "val_loss": 2.940195286273956, "lr": 0.02, "tokens_per_sec": 70102.36405519309, "mfu": 21.45628399061171, "total_flops": 3.0209997787763835e+18 }, { "step": 2650, "train_loss": 2.830553784966469, "val_loss": 2.9388512551784514, "lr": 0.02, "tokens_per_sec": 70080.79229376302, "mfu": 21.4496815051512, "total_flops": 3.079073592286118e+18 }, { "step": 2700, "train_loss": 2.769469901919365, "val_loss": 2.9488451719284057, "lr": 0.02, "tokens_per_sec": 70073.8045434036, "mfu": 21.44754275907341, "total_flops": 3.1371474057958523e+18 }, { "step": 2750, "train_loss": 2.7408286333084106, "val_loss": 2.943127530813217, "lr": 0.02, "tokens_per_sec": 69843.7630299741, "mfu": 21.37713377774594, "total_flops": 3.1952212193055867e+18 }, { "step": 2550, "train_loss": 2.690155401825905, "val_loss": 2.753552335500717, "lr": 0.02, "tokens_per_sec": 70145.36765726679, "mfu": 21.499149298759562, "total_flops": 3.253295032815321e+18 }, { "step": 2600, "train_loss": 2.7558372765779495, "val_loss": 2.7880488693714143, "lr": 0.02, "tokens_per_sec": 70354.6993534577, "mfu": 21.563308252368028, "total_flops": 3.3113688463250555e+18 }, { "step": 2650, "train_loss": 2.748625800013542, "val_loss": 2.8070174753665924, "lr": 0.02, "tokens_per_sec": 70307.24311361065, "mfu": 21.548763189455038, "total_flops": 3.36944265983479e+18 }, { "step": 2700, "train_loss": 2.657764434814453, "val_loss": 2.6618025243282317, "lr": 0.02, "tokens_per_sec": 70264.88750243967, "mfu": 21.53578143971706, "total_flops": 3.4275164733445243e+18 }, { "step": 2750, "train_loss": 2.597847282886505, "val_loss": 2.733213084936142, "lr": 0.02, "tokens_per_sec": 70309.20634173382, "mfu": 21.54936490751479, "total_flops": 3.4855902868542587e+18 }, { "step": 2800, "train_loss": 2.6393312364816666, "val_loss": 2.76564337015152, "lr": 0.02, "tokens_per_sec": 70335.78134187244, "mfu": 21.557509991284064, "total_flops": 3.543664100363993e+18 }, { "step": 2850, "train_loss": 2.7604094594717026, "val_loss": 2.7764769434928893, "lr": 0.02, "tokens_per_sec": 70224.05195676716, "mfu": 21.52326558125867, "total_flops": 3.6017379138737275e+18 }, { "step": 2900, "train_loss": 2.8098009526729584, "val_loss": 2.7969186305999756, "lr": 0.02, "tokens_per_sec": 70314.45950237487, "mfu": 21.55097497085446, "total_flops": 3.659811727383462e+18 }, { "step": 2950, "train_loss": 2.8605630844831467, "val_loss": 2.800990068912506, "lr": 0.02, "tokens_per_sec": 70267.142723665, "mfu": 21.536472651973043, "total_flops": 3.7178855408931963e+18 }, { "step": 3000, "train_loss": 2.826046869158745, "val_loss": 2.7932212591171264, "lr": 0.02, "tokens_per_sec": 70312.75564909614, "mfu": 21.550452749683682, "total_flops": 3.7759593544029307e+18 }, { "step": 3050, "train_loss": 2.7675381749868393, "val_loss": 2.809084439277649, "lr": 0.02, "tokens_per_sec": 70366.1560541514, "mfu": 21.566819664837748, "total_flops": 3.834033167912665e+18 }, { "step": 3100, "train_loss": 2.6846775263547897, "val_loss": 2.8106513619422913, "lr": 0.02, "tokens_per_sec": 70376.63593211089, "mfu": 21.570031686791708, "total_flops": 3.8921069814223995e+18 }, { "step": 3150, "train_loss": 2.8498146533966064, "val_loss": 2.8211086273193358, "lr": 0.02, "tokens_per_sec": 70318.95745570271, "mfu": 21.55235356752254, "total_flops": 3.950180794932134e+18 }, { "step": 3200, "train_loss": 2.882045492529869, "val_loss": 2.8087339460849763, "lr": 0.02, "tokens_per_sec": 70306.22248511235, "mfu": 21.54845037272606, "total_flops": 4.0082546084418683e+18 }, { "step": 3250, "train_loss": 2.7344588190317154, "val_loss": 2.8052581429481505, "lr": 0.02, "tokens_per_sec": 70309.31255946922, "mfu": 21.549397462636076, "total_flops": 4.0663284219516027e+18 }, { "step": 3300, "train_loss": 2.6739814579486847, "val_loss": 2.8148413836956023, "lr": 0.02, "tokens_per_sec": 70365.00773946103, "mfu": 21.56646771302974, "total_flops": 4.124402235461337e+18 }, { "step": 3350, "train_loss": 2.687133699655533, "val_loss": 2.7819845259189604, "lr": 0.02, "tokens_per_sec": 70311.11920932027, "mfu": 21.549951190361448, "total_flops": 4.1824760489710715e+18 }, { "step": 3400, "train_loss": 2.738806650042534, "val_loss": 2.777050405740738, "lr": 0.02, "tokens_per_sec": 70313.55276381515, "mfu": 21.550697060732645, "total_flops": 4.240549862480806e+18 }, { "step": 3450, "train_loss": 2.8085870444774628, "val_loss": 2.7993926048278808, "lr": 0.02, "tokens_per_sec": 70308.41404623268, "mfu": 21.549122073528217, "total_flops": 4.2986236759905403e+18 }, { "step": 3500, "train_loss": 2.7735735923051834, "val_loss": 2.822412884235382, "lr": 0.02, "tokens_per_sec": 70330.90176386677, "mfu": 21.556014428860447, "total_flops": 4.3566974895002747e+18 }, { "step": 3550, "train_loss": 2.7676771581172943, "val_loss": 2.816472613811493, "lr": 0.02, "tokens_per_sec": 70321.96971640864, "mfu": 21.553276807998806, "total_flops": 4.414771303010009e+18 }, { "step": 3600, "train_loss": 2.7102452218532562, "val_loss": 2.818382143974304, "lr": 0.02, "tokens_per_sec": 70311.3801038398, "mfu": 21.550031153022044, "total_flops": 4.4728451165197435e+18 }, { "step": 3650, "train_loss": 2.7262541949748993, "val_loss": 2.8222990155220034, "lr": 0.02, "tokens_per_sec": 70288.59055530882, "mfu": 21.543046288266435, "total_flops": 4.530918930029478e+18 }, { "step": 3700, "train_loss": 2.655830502510071, "val_loss": 2.8140450358390807, "lr": 0.02, "tokens_per_sec": 70323.33015866883, "mfu": 21.553693775679513, "total_flops": 4.5889927435392123e+18 }, { "step": 3750, "train_loss": 2.7190513610839844, "val_loss": 2.8193438947200775, "lr": 0.02, "tokens_per_sec": 70344.47533367388, "mfu": 21.560174649464408, "total_flops": 4.647066557048947e+18 }, { "step": 3800, "train_loss": 2.7415073215961456, "val_loss": 2.8139825463294983, "lr": 0.02, "tokens_per_sec": 70199.24064193484, "mfu": 21.51566105683042, "total_flops": 4.705140370558681e+18 }, { "step": 3850, "train_loss": 2.731362007558346, "val_loss": 2.8108940064907073, "lr": 0.02, "tokens_per_sec": 70298.58135692739, "mfu": 21.546108411151888, "total_flops": 4.763214184068415e+18 }, { "step": 3900, "train_loss": 2.796496495604515, "val_loss": 2.810216724872589, "lr": 0.02, "tokens_per_sec": 70265.2313500534, "mfu": 21.535886827021027, "total_flops": 4.82128799757815e+18 }, { "step": 3950, "train_loss": 2.636091724038124, "val_loss": 2.803800421953201, "lr": 0.02, "tokens_per_sec": 70308.34795684878, "mfu": 21.549101817514472, "total_flops": 4.879361811087884e+18 }, { "step": 4000, "train_loss": 2.8011103123426437, "val_loss": 2.829415625333786, "lr": 0.02, "tokens_per_sec": 70348.76681625859, "mfu": 21.56148996404445, "total_flops": 4.937435624597619e+18 }, { "step": 4050, "train_loss": 2.7537961602211, "val_loss": 2.828208714723587, "lr": 0.02, "tokens_per_sec": 70338.02334258889, "mfu": 21.55819715153057, "total_flops": 4.995509438107353e+18 }, { "step": 4100, "train_loss": 2.7435563504695892, "val_loss": 2.8136156499385834, "lr": 0.02, "tokens_per_sec": 70324.44382873985, "mfu": 21.554035109112153, "total_flops": 5.053583251617087e+18 }, { "step": 4150, "train_loss": 2.708776831626892, "val_loss": 2.8181759417057037, "lr": 0.02, "tokens_per_sec": 70153.27003630337, "mfu": 21.50157133220805, "total_flops": 5.111657065126822e+18 }, { "step": 4200, "train_loss": 2.6885371059179306, "val_loss": 2.8155935287475584, "lr": 0.02, "tokens_per_sec": 70219.75953179927, "mfu": 21.521949977843647, "total_flops": 5.169730878636556e+18 }, { "step": 4250, "train_loss": 2.6093505173921585, "val_loss": 2.84355810880661, "lr": 0.02, "tokens_per_sec": 70308.2900725415, "mfu": 21.54908407630923, "total_flops": 5.227804692146291e+18 }, { "step": 4300, "train_loss": 2.75183042883873, "val_loss": 2.8319630563259124, "lr": 0.02, "tokens_per_sec": 70308.23432386186, "mfu": 21.549066989661508, "total_flops": 5.285878505656025e+18 }, { "step": 4350, "train_loss": 2.743694692850113, "val_loss": 2.8407095193862917, "lr": 0.02, "tokens_per_sec": 70338.29039891162, "mfu": 21.558279002748197, "total_flops": 5.34395231916576e+18 }, { "step": 4400, "train_loss": 2.6878776401281357, "val_loss": 2.8353062510490417, "lr": 0.02, "tokens_per_sec": 70268.21217729135, "mfu": 21.536800433321247, "total_flops": 5.402026132675494e+18 }, { "step": 4450, "train_loss": 2.7711153477430344, "val_loss": 2.8343011856079103, "lr": 0.02, "tokens_per_sec": 70268.73984328873, "mfu": 21.536962159896163, "total_flops": 5.460099946185228e+18 }, { "step": 4500, "train_loss": 2.770058751106262, "val_loss": 2.8305183827877043, "lr": 0.02, "tokens_per_sec": 70330.05274618218, "mfu": 21.555754209852864, "total_flops": 5.518173759694963e+18 }, { "step": 4550, "train_loss": 2.7573233991861343, "val_loss": 2.8480226695537567, "lr": 0.02, "tokens_per_sec": 70306.38769881551, "mfu": 21.54850100977009, "total_flops": 5.576247573204697e+18 }, { "step": 4600, "train_loss": 2.683898538351059, "val_loss": 2.842072516679764, "lr": 0.02, "tokens_per_sec": 70336.00878633632, "mfu": 21.557579701696163, "total_flops": 5.634321386714431e+18 }, { "step": 4650, "train_loss": 2.4759643375873566, "val_loss": 2.2887446224689483, "lr": 0.02, "tokens_per_sec": 70300.01955342875, "mfu": 21.54654921005759, "total_flops": 5.692395200224166e+18 }, { "step": 4700, "train_loss": 2.796401672065258, "val_loss": 2.297064536809921, "lr": 0.02, "tokens_per_sec": 70308.00739617708, "mfu": 21.548997437639215, "total_flops": 5.7504690137339e+18 }, { "step": 4750, "train_loss": 2.596481144428253, "val_loss": 2.4246182203292848, "lr": 0.02, "tokens_per_sec": 70007.82326929359, "mfu": 21.456992739730733, "total_flops": 5.808542827243635e+18 }, { "step": 4800, "train_loss": 2.595515549182892, "val_loss": 2.3040037989616393, "lr": 0.02, "tokens_per_sec": 70316.84112334541, "mfu": 21.55170492390095, "total_flops": 5.866616640753369e+18 }, { "step": 4850, "train_loss": 2.62885881960392, "val_loss": 2.126067227125168, "lr": 0.02, "tokens_per_sec": 70333.82472782901, "mfu": 21.556910300401476, "total_flops": 5.924690454263103e+18 }, { "step": 4900, "train_loss": 2.7220041751861572, "val_loss": 2.335285782814026, "lr": 0.02, "tokens_per_sec": 70141.45779781595, "mfu": 21.497950949461835, "total_flops": 5.982764267772838e+18 }, { "step": 4950, "train_loss": 2.6748989075422287, "val_loss": 2.4292496860027315, "lr": 0.02, "tokens_per_sec": 70252.59415868955, "mfu": 21.532013600991682, "total_flops": 6.040838081282572e+18 }, { "step": 5000, "train_loss": 2.4526360854506493, "val_loss": 2.277197653055191, "lr": 0.02, "tokens_per_sec": 70309.10619387723, "mfu": 21.549334212776245, "total_flops": 6.098911894792307e+18 }, { "step": 5050, "train_loss": 2.559838131070137, "val_loss": 2.317918872833252, "lr": 0.02, "tokens_per_sec": 70267.75883994007, "mfu": 21.632433190049806, "total_flops": 6.156985708302041e+18 }, { "step": 5100, "train_loss": 2.591684401035309, "val_loss": 2.399251621961594, "lr": 0.02, "tokens_per_sec": 70248.37588972368, "mfu": 21.626466009901975, "total_flops": 6.215059521811775e+18 }, { "step": 5150, "train_loss": 2.69405797123909, "val_loss": 2.473741942644119, "lr": 0.02, "tokens_per_sec": 70290.93863159792, "mfu": 21.639569255048546, "total_flops": 6.27313333532151e+18 }, { "step": 5200, "train_loss": 2.7576390206813812, "val_loss": 2.5225872576236723, "lr": 0.02, "tokens_per_sec": 70268.60523157175, "mfu": 21.632693757779972, "total_flops": 6.331207148831244e+18 }, { "step": 5250, "train_loss": 2.6569458171725273, "val_loss": 2.5516360282897947, "lr": 0.02, "tokens_per_sec": 70275.72239653025, "mfu": 21.634884828023413, "total_flops": 6.389280962340979e+18 }, { "step": 5300, "train_loss": 2.737651005387306, "val_loss": 2.5737712025642394, "lr": 0.02, "tokens_per_sec": 70077.68129369298, "mfu": 21.573916455091958, "total_flops": 6.447354775850713e+18 }, { "step": 5350, "train_loss": 2.610097751021385, "val_loss": 2.5789273977279663, "lr": 0.02, "tokens_per_sec": 70291.56662424024, "mfu": 21.639762587084505, "total_flops": 6.505428589360447e+18 }, { "step": 5400, "train_loss": 2.6154174357652664, "val_loss": 2.39971661567688, "lr": 0.02, "tokens_per_sec": 70280.81837590337, "mfu": 21.636453661797916, "total_flops": 6.563502402870182e+18 }, { "step": 5450, "train_loss": 2.560743674635887, "val_loss": 2.471098029613495, "lr": 0.02, "tokens_per_sec": 70284.76569050449, "mfu": 21.637668870889517, "total_flops": 6.621576216379916e+18 }, { "step": 5500, "train_loss": 2.5454889237880707, "val_loss": 2.506662356853485, "lr": 0.02, "tokens_per_sec": 70268.84998053168, "mfu": 21.63276910550148, "total_flops": 6.679650029889651e+18 }, { "step": 5550, "train_loss": 2.656815826892853, "val_loss": 2.5483492255210876, "lr": 0.02, "tokens_per_sec": 70284.1575875986, "mfu": 21.637481662051478, "total_flops": 6.737723843399385e+18 }, { "step": 5600, "train_loss": 2.691612422466278, "val_loss": 2.5813338339328764, "lr": 0.02, "tokens_per_sec": 70268.87950770035, "mfu": 21.632778195651987, "total_flops": 6.79579765690912e+18 }, { "step": 5650, "train_loss": 2.6659550815820694, "val_loss": 2.594164401292801, "lr": 0.02, "tokens_per_sec": 70075.86473585301, "mfu": 21.57335721474108, "total_flops": 6.853871470418854e+18 }, { "step": 5700, "train_loss": 2.722677171230316, "val_loss": 2.6006647884845733, "lr": 0.02, "tokens_per_sec": 70276.25938345322, "mfu": 21.635050143296404, "total_flops": 6.911945283928588e+18 }, { "step": 5750, "train_loss": 2.780638873577118, "val_loss": 2.627731019258499, "lr": 0.02, "tokens_per_sec": 70283.08787318766, "mfu": 21.637152342803084, "total_flops": 6.970019097438323e+18 }, { "step": 5800, "train_loss": 2.6606620252132416, "val_loss": 2.639297807216644, "lr": 0.02, "tokens_per_sec": 70288.14931152055, "mfu": 21.638710542870793, "total_flops": 7.028092910948057e+18 }, { "step": 5850, "train_loss": 2.961484670639038, "val_loss": 2.6399263024330137, "lr": 0.02, "tokens_per_sec": 70300.75165277069, "mfu": 21.6425902639496, "total_flops": 7.086166724457791e+18 }, { "step": 5900, "train_loss": 2.790245294570923, "val_loss": 2.6511899530887604, "lr": 0.02, "tokens_per_sec": 70263.6710969525, "mfu": 21.63117474907294, "total_flops": 7.144240537967526e+18 }, { "step": 5950, "train_loss": 2.71671961247921, "val_loss": 2.6397358894348146, "lr": 0.02, "tokens_per_sec": 70225.60293793224, "mfu": 21.61945519347206, "total_flops": 7.20231435147726e+18 }, { "step": 6000, "train_loss": 2.6487374305725098, "val_loss": 2.6443570375442507, "lr": 0.02, "tokens_per_sec": 70068.37071382166, "mfu": 21.57105012634671, "total_flops": 7.260388164986995e+18 }, { "step": 6050, "train_loss": 2.54835969209671, "val_loss": 2.635010451078415, "lr": 0.02, "tokens_per_sec": 70263.49811395227, "mfu": 21.631121495016494, "total_flops": 7.318461978496729e+18 }, { "step": 6100, "train_loss": 2.5841805785894394, "val_loss": 2.635376179218292, "lr": 0.02, "tokens_per_sec": 70280.29288693903, "mfu": 21.6362918862539, "total_flops": 7.376535792006463e+18 }, { "step": 6150, "train_loss": 2.6273926347494125, "val_loss": 2.6592205107212066, "lr": 0.02, "tokens_per_sec": 70289.42059841492, "mfu": 21.639101917652937, "total_flops": 7.434609605516198e+18 }, { "step": 6200, "train_loss": 2.7345776110887527, "val_loss": 2.68223152756691, "lr": 0.02, "tokens_per_sec": 70245.30586681265, "mfu": 21.625520881344993, "total_flops": 7.492683419025932e+18 }, { "step": 6250, "train_loss": 2.643840879201889, "val_loss": 2.6727379858493805, "lr": 0.02, "tokens_per_sec": 70289.36971010863, "mfu": 21.639086251323114, "total_flops": 7.550757232535667e+18 }, { "step": 6300, "train_loss": 2.755268767476082, "val_loss": 2.6929265022277833, "lr": 0.02, "tokens_per_sec": 70267.55979098973, "mfu": 21.632371911403798, "total_flops": 7.608831046045401e+18 }, { "step": 6350, "train_loss": 2.768963858485222, "val_loss": 2.694007235765457, "lr": 0.02, "tokens_per_sec": 70111.05374796665, "mfu": 21.584190404331004, "total_flops": 7.666904859555135e+18 }, { "step": 6400, "train_loss": 2.6468516141176224, "val_loss": 2.6841990649700165, "lr": 0.02, "tokens_per_sec": 70281.87847469341, "mfu": 21.63678002080849, "total_flops": 7.72497867306487e+18 }, { "step": 6450, "train_loss": 2.712806150317192, "val_loss": 2.684369647502899, "lr": 0.02, "tokens_per_sec": 70267.07985526483, "mfu": 21.632224159751033, "total_flops": 7.783052486574604e+18 }, { "step": 6500, "train_loss": 2.5837599337100983, "val_loss": 2.6865688979625704, "lr": 0.02, "tokens_per_sec": 70286.26060807702, "mfu": 21.63812909197877, "total_flops": 7.841126300084339e+18 }, { "step": 6550, "train_loss": 2.7502972185611725, "val_loss": 2.696410584449768, "lr": 0.02, "tokens_per_sec": 70281.19023232057, "mfu": 21.63656814046116, "total_flops": 7.899200113594073e+18 }, { "step": 6600, "train_loss": 2.7288148403167725, "val_loss": 2.686836302280426, "lr": 0.02, "tokens_per_sec": 70292.68692806634, "mfu": 21.64010748064694, "total_flops": 7.957273927103807e+18 }, { "step": 6650, "train_loss": 2.747691661119461, "val_loss": 2.6874350547790526, "lr": 0.02, "tokens_per_sec": 70281.46943607075, "mfu": 21.636654095336798, "total_flops": 8.015347740613542e+18 }, { "step": 6700, "train_loss": 2.7368868738412857, "val_loss": 2.711217999458313, "lr": 0.02, "tokens_per_sec": 70153.36962880565, "mfu": 21.597217651538543, "total_flops": 8.073421554123276e+18 }, { "step": 6750, "train_loss": 2.686244487762451, "val_loss": 2.7106106102466585, "lr": 0.02, "tokens_per_sec": 70274.80138537045, "mfu": 21.63460128813669, "total_flops": 8.131495367633011e+18 }, { "step": 6800, "train_loss": 2.6627790927886963, "val_loss": 2.703348255157471, "lr": 0.02, "tokens_per_sec": 70283.19165337789, "mfu": 21.637184292278473, "total_flops": 8.189569181142745e+18 }, { "step": 6850, "train_loss": 2.628435805439949, "val_loss": 2.7075068950653076, "lr": 0.02, "tokens_per_sec": 70271.51706881971, "mfu": 21.63359018774681, "total_flops": 8.24764299465248e+18 }, { "step": 6900, "train_loss": 2.694452792406082, "val_loss": 2.7081696808338167, "lr": 0.02, "tokens_per_sec": 70299.27567071345, "mfu": 21.642135872294453, "total_flops": 8.305716808162214e+18 }, { "step": 6950, "train_loss": 2.6705697178840637, "val_loss": 2.7355406820774077, "lr": 0.02, "tokens_per_sec": 70283.07765242794, "mfu": 21.63714919626895, "total_flops": 8.363790621671948e+18 }, { "step": 7000, "train_loss": 2.6670849472284317, "val_loss": 2.714768886566162, "lr": 0.02, "tokens_per_sec": 70282.50900099597, "mfu": 21.636974132849918, "total_flops": 8.421864435181683e+18 }, { "step": 7050, "train_loss": 2.727104902267456, "val_loss": 2.735373377799988, "lr": 0.02, "tokens_per_sec": 70280.4880772634, "mfu": 21.63635197699425, "total_flops": 8.479938248691417e+18 }, { "step": 7100, "train_loss": 2.7122743129730225, "val_loss": 2.729275506734848, "lr": 0.02, "tokens_per_sec": 70293.89287865201, "mfu": 21.640478741178168, "total_flops": 8.538012062201151e+18 }, { "step": 7150, "train_loss": 2.6882054656744003, "val_loss": 2.7395458936691286, "lr": 0.02, "tokens_per_sec": 70289.38835791209, "mfu": 21.63909199218306, "total_flops": 8.596085875710886e+18 }, { "step": 7200, "train_loss": 2.770738333463669, "val_loss": 2.7353258550167086, "lr": 0.02, "tokens_per_sec": 70304.98790253063, "mfu": 21.64389442380073, "total_flops": 8.65415968922062e+18 }, { "step": 7250, "train_loss": 2.681736469268799, "val_loss": 2.7411834478378294, "lr": 0.02, "tokens_per_sec": 70285.81164313164, "mfu": 21.63799087490246, "total_flops": 8.712233502730355e+18 }, { "step": 7300, "train_loss": 2.6936238557100296, "val_loss": 2.7522403955459596, "lr": 0.02, "tokens_per_sec": 70295.82648556051, "mfu": 21.64107401592372, "total_flops": 8.770307316240089e+18 }, { "step": 7350, "train_loss": 2.725588798522949, "val_loss": 2.7442231357097624, "lr": 0.02, "tokens_per_sec": 70298.27954427102, "mfu": 21.641829207629595, "total_flops": 8.828381129749823e+18 }, { "step": 7400, "train_loss": 2.646110162138939, "val_loss": 2.7503407299518585, "lr": 0.02, "tokens_per_sec": 70284.60641909202, "mfu": 21.63761983804364, "total_flops": 8.886454943259558e+18 }, { "step": 7450, "train_loss": 2.7466748505830765, "val_loss": 2.7323903262615206, "lr": 0.02, "tokens_per_sec": 70288.42486357277, "mfu": 21.63879537354504, "total_flops": 8.944528756769292e+18 }, { "step": 7500, "train_loss": 2.725738510489464, "val_loss": 2.746686214208603, "lr": 0.02, "tokens_per_sec": 70294.84079339539, "mfu": 21.64077056352592, "total_flops": 9.002602570279027e+18 }, { "step": 7550, "train_loss": 2.6565753519535065, "val_loss": 2.7430707275867463, "lr": 0.02, "tokens_per_sec": 70298.68934041717, "mfu": 21.641955366310324, "total_flops": 9.060676383788761e+18 }, { "step": 7600, "train_loss": 2.7076889276504517, "val_loss": 2.7611879467964173, "lr": 0.02, "tokens_per_sec": 70256.34739863472, "mfu": 21.628920096054557, "total_flops": 9.118750197298495e+18 }, { "step": 7650, "train_loss": 2.7396631836891174, "val_loss": 2.7371096432209017, "lr": 0.02, "tokens_per_sec": 70277.87960201211, "mfu": 21.635548939192372, "total_flops": 9.17682401080823e+18 }, { "step": 7700, "train_loss": 2.6412112563848495, "val_loss": 2.7318264484405517, "lr": 0.02, "tokens_per_sec": 70300.82907752205, "mfu": 21.642614099714233, "total_flops": 9.234897824317964e+18 }, { "step": 7750, "train_loss": 2.54968099296093, "val_loss": 2.738618016242981, "lr": 0.02, "tokens_per_sec": 70302.56022262946, "mfu": 21.64314704514321, "total_flops": 9.292971637827699e+18 }, { "step": 7800, "train_loss": 2.649760976433754, "val_loss": 2.7415043830871584, "lr": 0.02, "tokens_per_sec": 70278.13295039837, "mfu": 21.635626934308913, "total_flops": 9.351045451337433e+18 }, { "step": 7550, "train_loss": 2.5435707718133926, "val_loss": 2.573974919319153, "lr": 0.02, "tokens_per_sec": 69895.20805720806, "mfu": 21.52970835915071, "total_flops": 9.409119264847167e+18 }, { "step": 7600, "train_loss": 2.6072546392679214, "val_loss": 2.596907156705856, "lr": 0.02, "tokens_per_sec": 70064.23548152526, "mfu": 21.5817736044143, "total_flops": 9.467193078356902e+18 }, { "step": 7650, "train_loss": 2.609854079782963, "val_loss": 2.6193448185920714, "lr": 0.02, "tokens_per_sec": 70083.55278385278, "mfu": 21.587723881936274, "total_flops": 9.525266891866636e+18 }, { "step": 7700, "train_loss": 2.6742303520441055, "val_loss": 2.633003169298172, "lr": 0.02, "tokens_per_sec": 70080.20580557312, "mfu": 21.586692917607696, "total_flops": 9.58334070537637e+18 }, { "step": 7750, "train_loss": 2.5816686004400253, "val_loss": 2.6430088937282563, "lr": 0.02, "tokens_per_sec": 70082.89633950195, "mfu": 21.58752167844021, "total_flops": 9.641414518886105e+18 }, { "step": 7800, "train_loss": 2.672006607055664, "val_loss": 2.6543741524219513, "lr": 0.02, "tokens_per_sec": 70083.62760901512, "mfu": 21.587746930210518, "total_flops": 9.69948833239584e+18 }, { "step": 7850, "train_loss": 2.549602121114731, "val_loss": 2.6550877690315247, "lr": 0.02, "tokens_per_sec": 69909.72719958337, "mfu": 21.53418066719091, "total_flops": 9.757562145905574e+18 }, { "step": 7900, "train_loss": 2.7753456830978394, "val_loss": 2.6652530431747437, "lr": 0.02, "tokens_per_sec": 70079.33580661588, "mfu": 21.58642493323019, "total_flops": 9.815635959415308e+18 }, { "step": 7950, "train_loss": 2.656064510345459, "val_loss": 2.684378969669342, "lr": 0.02, "tokens_per_sec": 70077.42447544802, "mfu": 21.585836188969093, "total_flops": 9.873709772925043e+18 }, { "step": 8000, "train_loss": 2.650539129972458, "val_loss": 2.679304671287537, "lr": 0.02, "tokens_per_sec": 70074.83105163992, "mfu": 21.585037340810658, "total_flops": 9.931783586434777e+18 }, { "step": 8050, "train_loss": 2.6973265260457993, "val_loss": 2.67431520819664, "lr": 0.02, "tokens_per_sec": 70097.39491881663, "mfu": 21.591987652473964, "total_flops": 9.989857399944511e+18 }, { "step": 8100, "train_loss": 2.7974444180727005, "val_loss": 2.6660505414009092, "lr": 0.02, "tokens_per_sec": 70063.2442190119, "mfu": 21.581468267418913, "total_flops": 1.0047931213454246e+19 }, { "step": 8150, "train_loss": 2.6419722586870193, "val_loss": 2.6528017103672026, "lr": 0.02, "tokens_per_sec": 70082.9707165531, "mfu": 21.58754458868348, "total_flops": 1.010600502696398e+19 }, { "step": 8200, "train_loss": 2.627345636487007, "val_loss": 2.649439036846161, "lr": 0.02, "tokens_per_sec": 70074.56811350217, "mfu": 21.584956348399597, "total_flops": 1.0164078840473715e+19 }, { "step": 8250, "train_loss": 2.7267237305641174, "val_loss": 2.6591224551200865, "lr": 0.02, "tokens_per_sec": 70057.47656743563, "mfu": 21.579691666993703, "total_flops": 1.022215265398345e+19 }, { "step": 8300, "train_loss": 2.440946727991104, "val_loss": 2.672474372386932, "lr": 0.02, "tokens_per_sec": 70054.52809839167, "mfu": 21.57878345482313, "total_flops": 1.0280226467493183e+19 }, { "step": 8350, "train_loss": 2.662572145462036, "val_loss": 2.6874975621700288, "lr": 0.02, "tokens_per_sec": 70080.70251085235, "mfu": 21.586845916934827, "total_flops": 1.0338300281002918e+19 }, { "step": 8400, "train_loss": 2.7091101929545403, "val_loss": 2.686986118555069, "lr": 0.02, "tokens_per_sec": 70085.49862505833, "mfu": 21.588323256268698, "total_flops": 1.0396374094512652e+19 }, { "step": 8450, "train_loss": 2.6898171603679657, "val_loss": 2.699655681848526, "lr": 0.02, "tokens_per_sec": 69901.0587853534, "mfu": 21.531510549517677, "total_flops": 1.0454447908022387e+19 }, { "step": 8500, "train_loss": 2.6142636835575104, "val_loss": 2.7020849883556366, "lr": 0.02, "tokens_per_sec": 70093.13326739804, "mfu": 21.590674942851646, "total_flops": 1.0512521721532121e+19 }, { "step": 8550, "train_loss": 2.5702943354845047, "val_loss": 2.6926130056381226, "lr": 0.02, "tokens_per_sec": 70079.20247791831, "mfu": 21.586383864206226, "total_flops": 1.0570595535041855e+19 }, { "step": 8600, "train_loss": 2.730316862463951, "val_loss": 2.6773874521255494, "lr": 0.02, "tokens_per_sec": 70071.69010267203, "mfu": 21.584069839359135, "total_flops": 1.062866934855159e+19 }, { "step": 8650, "train_loss": 2.7265526801347733, "val_loss": 2.6893074452877044, "lr": 0.02, "tokens_per_sec": 70073.10473971444, "mfu": 21.584505587728785, "total_flops": 1.0686743162061324e+19 }, { "step": 8700, "train_loss": 2.5364869236946106, "val_loss": 2.690558123588562, "lr": 0.02, "tokens_per_sec": 70063.07769055446, "mfu": 21.58141697192639, "total_flops": 1.0744816975571059e+19 }, { "step": 8750, "train_loss": 2.694303020834923, "val_loss": 2.6988609194755555, "lr": 0.02, "tokens_per_sec": 70084.14993777053, "mfu": 21.58790782229574, "total_flops": 1.0802890789080793e+19 }, { "step": 8800, "train_loss": 2.6549243479967117, "val_loss": 2.688143104314804, "lr": 0.02, "tokens_per_sec": 69908.60717930882, "mfu": 21.53383566914973, "total_flops": 1.0860964602590527e+19 }, { "step": 8850, "train_loss": 2.6796169131994247, "val_loss": 2.703415608406067, "lr": 0.02, "tokens_per_sec": 70085.21784901444, "mfu": 21.58823676927614, "total_flops": 1.0919038416100262e+19 }, { "step": 8900, "train_loss": 2.7315694838762283, "val_loss": 2.706169080734253, "lr": 0.02, "tokens_per_sec": 70079.59185720404, "mfu": 21.586503804080444, "total_flops": 1.0977112229609996e+19 }, { "step": 8950, "train_loss": 2.6746401339769363, "val_loss": 2.696428966522217, "lr": 0.02, "tokens_per_sec": 70065.1631482729, "mfu": 21.5820593521084, "total_flops": 1.103518604311973e+19 }, { "step": 9000, "train_loss": 2.6321981251239777, "val_loss": 2.696945106983185, "lr": 0.02, "tokens_per_sec": 70080.09558887687, "mfu": 21.586658967736277, "total_flops": 1.1093259856629465e+19 }, { "step": 9050, "train_loss": 2.574547991156578, "val_loss": 2.6979831516742707, "lr": 0.02, "tokens_per_sec": 70080.95175949985, "mfu": 21.586922692594257, "total_flops": 1.11513336701392e+19 }, { "step": 9100, "train_loss": 2.6053851395845413, "val_loss": 2.729407620429993, "lr": 0.02, "tokens_per_sec": 70049.16029463499, "mfu": 21.577130018877764, "total_flops": 1.1209407483648934e+19 }, { "step": 9150, "train_loss": 2.6026515066623688, "val_loss": 2.7150147020816804, "lr": 0.02, "tokens_per_sec": 70077.26122935086, "mfu": 21.58578590453672, "total_flops": 1.1267481297158668e+19 }, { "step": 9200, "train_loss": 2.638631984591484, "val_loss": 2.724787563085556, "lr": 0.02, "tokens_per_sec": 70069.194224259, "mfu": 21.583301037951628, "total_flops": 1.1325555110668403e+19 }, { "step": 9250, "train_loss": 2.687884122133255, "val_loss": 2.7205813467502593, "lr": 0.02, "tokens_per_sec": 70071.53235369503, "mfu": 21.584021248195835, "total_flops": 1.1383628924178137e+19 }, { "step": 9300, "train_loss": 2.6440293043851852, "val_loss": 2.7422130703926086, "lr": 0.02, "tokens_per_sec": 70093.24296655692, "mfu": 21.590708733306865, "total_flops": 1.1441702737687871e+19 }, { "step": 9350, "train_loss": 2.6162674874067307, "val_loss": 2.7284474194049837, "lr": 0.02, "tokens_per_sec": 70091.7371465895, "mfu": 21.59024489800769, "total_flops": 1.1499776551197606e+19 }, { "step": 9400, "train_loss": 2.61313958466053, "val_loss": 2.727823382616043, "lr": 0.02, "tokens_per_sec": 69939.90379685092, "mfu": 21.543475915842432, "total_flops": 1.155785036470734e+19 }, { "step": 9450, "train_loss": 2.645725265145302, "val_loss": 2.740615874528885, "lr": 0.02, "tokens_per_sec": 70059.97998144061, "mfu": 21.580462789577464, "total_flops": 1.1615924178217075e+19 }, { "step": 9500, "train_loss": 2.6479314118623734, "val_loss": 2.734837234020233, "lr": 0.02, "tokens_per_sec": 70060.40698137914, "mfu": 21.58059431768077, "total_flops": 1.167399799172681e+19 }, { "step": 9550, "train_loss": 2.748686507344246, "val_loss": 2.728809928894043, "lr": 0.02, "tokens_per_sec": 70073.24083636962, "mfu": 21.58454750936227, "total_flops": 1.1732071805236543e+19 }, { "step": 9600, "train_loss": 2.582173064351082, "val_loss": 2.719840705394745, "lr": 0.02, "tokens_per_sec": 70062.49574008441, "mfu": 21.581237714660436, "total_flops": 1.1790145618746278e+19 }, { "step": 9650, "train_loss": 2.7906376272439957, "val_loss": 2.7412230372428894, "lr": 0.02, "tokens_per_sec": 70085.55469118859, "mfu": 21.588340526228496, "total_flops": 1.1848219432256012e+19 }, { "step": 9700, "train_loss": 2.62669138610363, "val_loss": 2.728373169898987, "lr": 0.02, "tokens_per_sec": 70065.22197239265, "mfu": 21.58207747160726, "total_flops": 1.1906293245765747e+19 }, { "step": 9750, "train_loss": 2.624654397368431, "val_loss": 2.7269852578639986, "lr": 0.02, "tokens_per_sec": 69933.18341549518, "mfu": 21.5414058475978, "total_flops": 1.1964367059275481e+19 }, { "step": 9800, "train_loss": 2.648192882537842, "val_loss": 2.7176962196826935, "lr": 0.02, "tokens_per_sec": 70090.91467002199, "mfu": 21.589991551875347, "total_flops": 1.2022440872785215e+19 }, { "step": 9850, "train_loss": 2.6563764810562134, "val_loss": 2.724771249294281, "lr": 0.02, "tokens_per_sec": 70081.3885072608, "mfu": 21.587057223303603, "total_flops": 1.208051468629495e+19 }, { "step": 9900, "train_loss": 2.681455746293068, "val_loss": 2.7229500114917755, "lr": 0.02, "tokens_per_sec": 70072.09994073665, "mfu": 21.584196081118066, "total_flops": 1.2138588499804684e+19 }, { "step": 9950, "train_loss": 2.62647445499897, "val_loss": 2.7153716802597048, "lr": 0.02, "tokens_per_sec": 70062.32050948143, "mfu": 21.58118373866004, "total_flops": 1.2196662313314419e+19 }, { "step": 10000, "train_loss": 2.5326335802674294, "val_loss": 2.717075729370117, "lr": 0.02, "tokens_per_sec": 70096.4264025015, "mfu": 21.591689321953382, "total_flops": 1.2254736126824153e+19 }, { "step": 10050, "train_loss": 2.5265747159719467, "val_loss": 2.550719881057739, "lr": 0.02, "tokens_per_sec": 70016.48820081342, "mfu": 21.406339952207, "total_flops": 1.2312809940333887e+19 }, { "step": 10100, "train_loss": 2.5934030413627625, "val_loss": 2.584617280960083, "lr": 0.02, "tokens_per_sec": 70021.87329482104, "mfu": 21.407986352304516, "total_flops": 1.2370883753843622e+19 }, { "step": 10150, "train_loss": 2.5912967920303345, "val_loss": 2.610349267721176, "lr": 0.02, "tokens_per_sec": 70013.94999834869, "mfu": 21.405563940352884, "total_flops": 1.2428957567353356e+19 }, { "step": 10200, "train_loss": 2.6515150517225266, "val_loss": 2.622386109828949, "lr": 0.02, "tokens_per_sec": 70030.94537464414, "mfu": 21.410759985055275, "total_flops": 1.248703138086309e+19 }, { "step": 10050, "train_loss": 2.526685908436775, "val_loss": 2.552409863471985, "lr": 0.02, "tokens_per_sec": 70125.69793352022, "mfu": 21.60070578450605, "total_flops": 1.2545105194372825e+19 }, { "step": 10100, "train_loss": 2.593015134334564, "val_loss": 2.583539479970932, "lr": 0.02, "tokens_per_sec": 70113.93309309517, "mfu": 21.5970818796593, "total_flops": 1.260317900788256e+19 }, { "step": 10150, "train_loss": 2.5912905409932137, "val_loss": 2.61044899225235, "lr": 0.02, "tokens_per_sec": 70132.29410189865, "mfu": 21.60273759162735, "total_flops": 1.2661252821392294e+19 }, { "step": 10200, "train_loss": 2.6520476937294006, "val_loss": 2.6223703742027284, "lr": 0.02, "tokens_per_sec": 70178.88626107713, "mfu": 21.617089299374108, "total_flops": 1.2719326634902028e+19 }, { "step": 10050, "train_loss": 2.52651284635067, "val_loss": 2.5529006838798525, "lr": 0.02, "tokens_per_sec": 70204.14387890541, "mfu": 21.487435818820614, "total_flops": 1.2777400448411763e+19 }, { "step": 10100, "train_loss": 2.5928536653518677, "val_loss": 2.582450878620148, "lr": 0.02, "tokens_per_sec": 70192.73068313688, "mfu": 21.483942573293874, "total_flops": 1.2835474261921497e+19 }, { "step": 10150, "train_loss": 2.5910468697547913, "val_loss": 2.6105341851711272, "lr": 0.02, "tokens_per_sec": 70218.74469285099, "mfu": 21.49190470107206, "total_flops": 1.2893548075431231e+19 }, { "step": 10200, "train_loss": 2.6514677554368973, "val_loss": 2.624534994363785, "lr": 0.02, "tokens_per_sec": 70219.55436424627, "mfu": 21.492152517813057, "total_flops": 1.2951621888940966e+19 }, { "step": 10250, "train_loss": 2.564076505601406, "val_loss": 2.6290538609027863, "lr": 0.02, "tokens_per_sec": 70220.83740988419, "mfu": 21.492545220569248, "total_flops": 1.30096957024507e+19 }, { "step": 10300, "train_loss": 2.652680352330208, "val_loss": 2.6442084670066834, "lr": 0.02, "tokens_per_sec": 70235.22701111587, "mfu": 21.496949456784385, "total_flops": 1.3067769515960435e+19 }, { "step": 10350, "train_loss": 2.5286182314157486, "val_loss": 2.6446537137031556, "lr": 0.02, "tokens_per_sec": 70221.15829073807, "mfu": 21.492643432816735, "total_flops": 1.312584332947017e+19 }, { "step": 10400, "train_loss": 2.7530941367149353, "val_loss": 2.656357008218765, "lr": 0.02, "tokens_per_sec": 70236.16616565398, "mfu": 21.497236904529586, "total_flops": 1.3183917142979903e+19 }, { "step": 10450, "train_loss": 2.6332227289676666, "val_loss": 2.673243999481201, "lr": 0.02, "tokens_per_sec": 70229.65653406357, "mfu": 21.495244496627787, "total_flops": 1.3241990956489638e+19 }, { "step": 10500, "train_loss": 2.6284239292144775, "val_loss": 2.662519484758377, "lr": 0.02, "tokens_per_sec": 70214.50167879055, "mfu": 21.49060603852503, "total_flops": 1.3300064769999372e+19 }, { "step": 10550, "train_loss": 2.6753184646368027, "val_loss": 2.6621902704238893, "lr": 0.02, "tokens_per_sec": 70230.3363657566, "mfu": 21.495452573232164, "total_flops": 1.3358138583509107e+19 }, { "step": 10600, "train_loss": 2.775514140725136, "val_loss": 2.6512930035591125, "lr": 0.02, "tokens_per_sec": 70222.59029373217, "mfu": 21.49308172706445, "total_flops": 1.3416212397018841e+19 }, { "step": 10650, "train_loss": 2.621285066008568, "val_loss": 2.647852545976639, "lr": 0.02, "tokens_per_sec": 70225.17930426326, "mfu": 21.49387414749076, "total_flops": 1.3474286210528575e+19 }, { "step": 10700, "train_loss": 2.6071013510227203, "val_loss": 2.6399790406227113, "lr": 0.02, "tokens_per_sec": 70208.71694363162, "mfu": 21.488835500226454, "total_flops": 1.353236002403831e+19 }, { "step": 10750, "train_loss": 2.705596625804901, "val_loss": 2.649714803695679, "lr": 0.02, "tokens_per_sec": 70220.28075246631, "mfu": 21.492374844009134, "total_flops": 1.3590433837548044e+19 }, { "step": 10800, "train_loss": 2.4201852083206177, "val_loss": 2.6591944634914397, "lr": 0.02, "tokens_per_sec": 70213.37512246182, "mfu": 21.49026123257096, "total_flops": 1.3648507651057779e+19 }, { "step": 10850, "train_loss": 2.64433017373085, "val_loss": 2.6790621519088744, "lr": 0.02, "tokens_per_sec": 70210.31600380773, "mfu": 21.48932492579318, "total_flops": 1.3706581464567513e+19 }, { "step": 10900, "train_loss": 2.6901479735970497, "val_loss": 2.6743451356887817, "lr": 0.02, "tokens_per_sec": 70049.41735203927, "mfu": 21.44007855282732, "total_flops": 1.3764655278077247e+19 }, { "step": 10950, "train_loss": 2.6701857894659042, "val_loss": 2.683942812681198, "lr": 0.02, "tokens_per_sec": 70224.3843041066, "mfu": 21.49363082118695, "total_flops": 1.3822729091586982e+19 }, { "step": 11000, "train_loss": 2.5939954668283463, "val_loss": 2.6858055651187898, "lr": 0.02, "tokens_per_sec": 70231.74048079467, "mfu": 21.495882331791933, "total_flops": 1.3880802905096716e+19 }, { "step": 11050, "train_loss": 2.549482077360153, "val_loss": 2.6854387283325196, "lr": 0.02, "tokens_per_sec": 70214.96890201059, "mfu": 21.490749041892048, "total_flops": 1.393887671860645e+19 }, { "step": 11100, "train_loss": 2.711370125412941, "val_loss": 2.6654779374599458, "lr": 0.02, "tokens_per_sec": 70197.87995457259, "mfu": 21.485518614726402, "total_flops": 1.3996950532116185e+19 }, { "step": 11150, "train_loss": 2.707639902830124, "val_loss": 2.6756514489650725, "lr": 0.02, "tokens_per_sec": 70224.52323160897, "mfu": 21.493673342833475, "total_flops": 1.405502434562592e+19 }, { "step": 11200, "train_loss": 2.520912617444992, "val_loss": 2.6752108156681063, "lr": 0.02, "tokens_per_sec": 70210.97573402476, "mfu": 21.489526849923475, "total_flops": 1.4113098159135654e+19 }, { "step": 11250, "train_loss": 2.674424484372139, "val_loss": 2.6870215952396395, "lr": 0.02, "tokens_per_sec": 70065.63273939754, "mfu": 21.445041608793634, "total_flops": 1.4171171972645388e+19 }, { "step": 11300, "train_loss": 2.635969787836075, "val_loss": 2.6746863961219787, "lr": 0.02, "tokens_per_sec": 70204.44947845214, "mfu": 21.48752935390678, "total_flops": 1.4229245786155123e+19 }, { "step": 11350, "train_loss": 2.6615608781576157, "val_loss": 2.6862066984176636, "lr": 0.02, "tokens_per_sec": 70219.71524689811, "mfu": 21.49220175928891, "total_flops": 1.4287319599664857e+19 }, { "step": 11400, "train_loss": 2.712399512529373, "val_loss": 2.690833550691605, "lr": 0.02, "tokens_per_sec": 70220.12457604909, "mfu": 21.492327042974093, "total_flops": 1.4345393413174591e+19 }, { "step": 11450, "train_loss": 2.6577775329351425, "val_loss": 2.6866107404232027, "lr": 0.02, "tokens_per_sec": 70229.6213204413, "mfu": 21.49523371876759, "total_flops": 1.4403467226684326e+19 }, { "step": 11500, "train_loss": 2.614283174276352, "val_loss": 2.688479536771774, "lr": 0.02, "tokens_per_sec": 70236.05882282139, "mfu": 21.497204050027012, "total_flops": 1.446154104019406e+19 }, { "step": 11550, "train_loss": 2.5547793954610825, "val_loss": 2.6836730897426606, "lr": 0.02, "tokens_per_sec": 70199.13173122169, "mfu": 21.48590174695936, "total_flops": 1.4519614853703795e+19 }, { "step": 11600, "train_loss": 2.5856354534626007, "val_loss": 2.7118352591991424, "lr": 0.02, "tokens_per_sec": 70029.10581562246, "mfu": 21.433861785397028, "total_flops": 1.457768866721353e+19 }, { "step": 11650, "train_loss": 2.584961473941803, "val_loss": 2.697824990749359, "lr": 0.02, "tokens_per_sec": 70228.01487665919, "mfu": 21.49474203329494, "total_flops": 1.4635762480723263e+19 }, { "step": 11700, "train_loss": 2.6195385307073593, "val_loss": 2.7136584401130674, "lr": 0.02, "tokens_per_sec": 70223.3484761831, "mfu": 21.493313784545016, "total_flops": 1.4693836294232998e+19 }, { "step": 11750, "train_loss": 2.670626640319824, "val_loss": 2.7112324714660643, "lr": 0.02, "tokens_per_sec": 70213.6927964773, "mfu": 21.49035846329897, "total_flops": 1.4751910107742732e+19 }, { "step": 11800, "train_loss": 2.627764016389847, "val_loss": 2.73016876578331, "lr": 0.02, "tokens_per_sec": 70190.65073107202, "mfu": 21.483305960780275, "total_flops": 1.4809983921252467e+19 }, { "step": 11850, "train_loss": 2.5967312157154083, "val_loss": 2.7091954708099366, "lr": 0.02, "tokens_per_sec": 70217.17753509365, "mfu": 21.49142503990296, "total_flops": 1.4868057734762201e+19 }, { "step": 11900, "train_loss": 2.596388280391693, "val_loss": 2.7094963908195497, "lr": 0.02, "tokens_per_sec": 70215.31170073652, "mfu": 21.49085396255934, "total_flops": 1.4926131548271935e+19 }, { "step": 11950, "train_loss": 2.624119371175766, "val_loss": 2.728128719329834, "lr": 0.02, "tokens_per_sec": 70048.5275888958, "mfu": 21.43980622234392, "total_flops": 1.498420536178167e+19 }, { "step": 12000, "train_loss": 2.626932516694069, "val_loss": 2.712033325433731, "lr": 0.02, "tokens_per_sec": 70215.11474239366, "mfu": 21.49079367936924, "total_flops": 1.5042279175291404e+19 }, { "step": 12050, "train_loss": 2.731254890561104, "val_loss": 2.712426573038101, "lr": 0.02, "tokens_per_sec": 70177.50172730487, "mfu": 21.47928143517655, "total_flops": 1.5100352988801139e+19 }, { "step": 12100, "train_loss": 2.562843546271324, "val_loss": 2.712849426269531, "lr": 0.02, "tokens_per_sec": 70220.87609033482, "mfu": 21.492557059524252, "total_flops": 1.5158426802310873e+19 }, { "step": 12150, "train_loss": 2.764897644519806, "val_loss": 2.728255307674408, "lr": 0.02, "tokens_per_sec": 70222.98788337511, "mfu": 21.49320341762947, "total_flops": 1.5216500615820607e+19 }, { "step": 12200, "train_loss": 2.6027358323335648, "val_loss": 2.718504363298416, "lr": 0.02, "tokens_per_sec": 70226.1347859551, "mfu": 21.494166592500346, "total_flops": 1.5274574429330342e+19 }, { "step": 12250, "train_loss": 2.600238487124443, "val_loss": 2.7240344643592835, "lr": 0.02, "tokens_per_sec": 70217.72225994278, "mfu": 21.49159176425266, "total_flops": 1.5332648242840076e+19 }, { "step": 12300, "train_loss": 2.624945491552353, "val_loss": 2.7019182324409483, "lr": 0.02, "tokens_per_sec": 70037.50231473647, "mfu": 21.43643170827973, "total_flops": 1.539072205634981e+19 }, { "step": 12350, "train_loss": 2.6347887068986893, "val_loss": 2.713118499517441, "lr": 0.02, "tokens_per_sec": 70236.29818544973, "mfu": 21.497277311929082, "total_flops": 1.5448795869859545e+19 }, { "step": 12400, "train_loss": 2.6599233150482178, "val_loss": 2.7061276376247405, "lr": 0.02, "tokens_per_sec": 70235.88967720557, "mfu": 21.497152279499474, "total_flops": 1.550686968336928e+19 }, { "step": 12450, "train_loss": 2.60282726585865, "val_loss": 2.705427312850952, "lr": 0.02, "tokens_per_sec": 70221.31604141547, "mfu": 21.49269171568679, "total_flops": 1.5564943496879014e+19 }, { "step": 12500, "train_loss": 2.509654685854912, "val_loss": 2.7021617352962495, "lr": 0.02, "tokens_per_sec": 70214.04645600512, "mfu": 21.490466708140172, "total_flops": 1.5623017310388748e+19 }, { "step": 12550, "train_loss": 2.6215831637382507, "val_loss": 2.7008169412612917, "lr": 0.02, "tokens_per_sec": 70214.42803116795, "mfu": 21.490583497141053, "total_flops": 1.5681091123898483e+19 }, { "step": 12600, "train_loss": 2.5994905531406403, "val_loss": 2.7037415862083436, "lr": 0.02, "tokens_per_sec": 70192.67590175563, "mfu": 21.483925806314765, "total_flops": 1.5739164937408217e+19 }, { "step": 12650, "train_loss": 2.525320991873741, "val_loss": 2.7214488089084625, "lr": 0.02, "tokens_per_sec": 70028.52955854105, "mfu": 21.43368540995296, "total_flops": 1.5797238750917951e+19 }, { "step": 12700, "train_loss": 2.614622578024864, "val_loss": 2.7287656903266906, "lr": 0.02, "tokens_per_sec": 70190.07763727287, "mfu": 21.483130553524422, "total_flops": 1.5855312564427686e+19 }, { "step": 12750, "train_loss": 2.609872132539749, "val_loss": 2.728152310848236, "lr": 0.02, "tokens_per_sec": 70161.95478187368, "mfu": 21.474522969740267, "total_flops": 1.591338637793742e+19 }, { "step": 12800, "train_loss": 2.667870581150055, "val_loss": 2.7394967794418337, "lr": 0.02, "tokens_per_sec": 70205.6370498768, "mfu": 21.48789283479776, "total_flops": 1.5971460191447155e+19 }, { "step": 12550, "train_loss": 2.511006385087967, "val_loss": 2.526371055841446, "lr": 0.02, "tokens_per_sec": 70175.22683880641, "mfu": 21.56199713951752, "total_flops": 1.602953400495689e+19 }, { "step": 12600, "train_loss": 2.5791891515254974, "val_loss": 2.5669545888900758, "lr": 0.02, "tokens_per_sec": 70163.2790416696, "mfu": 21.558326066700335, "total_flops": 1.6087607818466623e+19 }, { "step": 12650, "train_loss": 2.576032876968384, "val_loss": 2.592360574007034, "lr": 0.02, "tokens_per_sec": 70180.9795146999, "mfu": 21.563764703182752, "total_flops": 1.6145681631976358e+19 }, { "step": 12700, "train_loss": 2.639482229948044, "val_loss": 2.603729647397995, "lr": 0.02, "tokens_per_sec": 70206.94939833201, "mfu": 21.57174419369264, "total_flops": 1.6203755445486092e+19 }, { "step": 12750, "train_loss": 2.5492171943187714, "val_loss": 2.6102789878845214, "lr": 0.02, "tokens_per_sec": 70194.73188656349, "mfu": 21.567990248523213, "total_flops": 1.6261829258995827e+19 }, { "step": 12800, "train_loss": 2.636518955230713, "val_loss": 2.6257937788963317, "lr": 0.02, "tokens_per_sec": 70190.43946045496, "mfu": 21.566671360310906, "total_flops": 1.6319903072505561e+19 }, { "step": 12850, "train_loss": 2.5141112208366394, "val_loss": 2.627405750751495, "lr": 0.02, "tokens_per_sec": 70171.50622025452, "mfu": 21.56085394454421, "total_flops": 1.6377976886015295e+19 }, { "step": 12900, "train_loss": 2.7366262525320053, "val_loss": 2.6409005224704742, "lr": 0.02, "tokens_per_sec": 70196.86765263397, "mfu": 21.568646482695538, "total_flops": 1.643605069952503e+19 }, { "step": 12950, "train_loss": 2.619333431124687, "val_loss": 2.658438378572464, "lr": 0.02, "tokens_per_sec": 70251.94520442004, "mfu": 21.58556957746217, "total_flops": 1.6494124513034764e+19 }, { "step": 13000, "train_loss": 2.611935794353485, "val_loss": 2.658729982376099, "lr": 0.02, "tokens_per_sec": 70170.87432532028, "mfu": 21.56065978889444, "total_flops": 1.6552198326544499e+19 }, { "step": 13050, "train_loss": 2.6581060141324997, "val_loss": 2.649619275331497, "lr": 0.02, "tokens_per_sec": 70227.76581436433, "mfu": 21.57814023290994, "total_flops": 1.6610272140054233e+19 }, { "step": 13100, "train_loss": 2.7615827918052673, "val_loss": 2.641291284561157, "lr": 0.02, "tokens_per_sec": 70210.33651511092, "mfu": 21.57278491711582, "total_flops": 1.6668345953563967e+19 }, { "step": 13150, "train_loss": 2.6068419814109802, "val_loss": 2.628552460670471, "lr": 0.02, "tokens_per_sec": 70227.09287280244, "mfu": 21.577933465298543, "total_flops": 1.6726419767073702e+19 }, { "step": 13200, "train_loss": 2.5915939286351204, "val_loss": 2.6240718185901644, "lr": 0.02, "tokens_per_sec": 70242.12482435642, "mfu": 21.58255217350953, "total_flops": 1.6784493580583436e+19 }, { "step": 13250, "train_loss": 2.692144051194191, "val_loss": 2.6368056356906893, "lr": 0.02, "tokens_per_sec": 70234.45645879045, "mfu": 21.58019599621098, "total_flops": 1.684256739409317e+19 }, { "step": 13300, "train_loss": 2.404242806136608, "val_loss": 2.6440458834171294, "lr": 0.02, "tokens_per_sec": 70212.24254943595, "mfu": 21.573370564053, "total_flops": 1.6900641207602905e+19 }, { "step": 13350, "train_loss": 2.6278859823942184, "val_loss": 2.6582672238349914, "lr": 0.02, "tokens_per_sec": 70219.35244915294, "mfu": 21.575555147477463, "total_flops": 1.695871502111264e+19 }, { "step": 13400, "train_loss": 2.6749037951231003, "val_loss": 2.660851800441742, "lr": 0.02, "tokens_per_sec": 70055.26580612546, "mfu": 21.52513798622313, "total_flops": 1.7016788834622374e+19 }, { "step": 13450, "train_loss": 2.6571804583072662, "val_loss": 2.6723348915576937, "lr": 0.02, "tokens_per_sec": 70243.49485641379, "mfu": 21.582973128718834, "total_flops": 1.7074862648132108e+19 }, { "step": 13500, "train_loss": 2.5777985006570816, "val_loss": 2.6738951981067656, "lr": 0.02, "tokens_per_sec": 70221.8123944745, "mfu": 21.576310988767563, "total_flops": 1.7132936461641843e+19 }, { "step": 13550, "train_loss": 2.5361034870147705, "val_loss": 2.6723001420497896, "lr": 0.02, "tokens_per_sec": 70233.72170186242, "mfu": 21.579970235248112, "total_flops": 1.7191010275151577e+19 }, { "step": 13600, "train_loss": 2.6962577551603317, "val_loss": 2.6532412350177763, "lr": 0.02, "tokens_per_sec": 70226.71687890886, "mfu": 21.57781793764551, "total_flops": 1.7249084088661311e+19 }, { "step": 13650, "train_loss": 2.6913505494594574, "val_loss": 2.66121444106102, "lr": 0.02, "tokens_per_sec": 70210.21893955192, "mfu": 21.572748790921707, "total_flops": 1.7307157902171046e+19 }, { "step": 13700, "train_loss": 2.5043933764100075, "val_loss": 2.6647586822509766, "lr": 0.02, "tokens_per_sec": 70221.3220958266, "mfu": 21.57616033990585, "total_flops": 1.736523171568078e+19 }, { "step": 13750, "train_loss": 2.6589005440473557, "val_loss": 2.674451744556427, "lr": 0.02, "tokens_per_sec": 70083.35522385615, "mfu": 21.533768723479614, "total_flops": 1.7423305529190515e+19 }, { "step": 13800, "train_loss": 2.62132728099823, "val_loss": 2.6628175556659697, "lr": 0.02, "tokens_per_sec": 70216.07364959198, "mfu": 21.574547705536208, "total_flops": 1.748137934270025e+19 }, { "step": 13850, "train_loss": 2.6479734033346176, "val_loss": 2.6767729818820953, "lr": 0.02, "tokens_per_sec": 70236.02270545306, "mfu": 21.5806772401995, "total_flops": 1.7539453156209983e+19 }, { "step": 13900, "train_loss": 2.7008413672447205, "val_loss": 2.6886323153972627, "lr": 0.02, "tokens_per_sec": 70175.82813137904, "mfu": 21.562181892304434, "total_flops": 1.7597526969719718e+19 }, { "step": 13950, "train_loss": 2.6451959162950516, "val_loss": 2.679285502433777, "lr": 0.02, "tokens_per_sec": 70229.14493070339, "mfu": 21.578563979350065, "total_flops": 1.7655600783229452e+19 }, { "step": 14000, "train_loss": 2.6014012545347214, "val_loss": 2.680087924003601, "lr": 0.02, "tokens_per_sec": 70238.65353790967, "mfu": 21.581485588165858, "total_flops": 1.7713674596739187e+19 }, { "step": 14050, "train_loss": 2.5424724519252777, "val_loss": 2.6780451416969298, "lr": 0.02, "tokens_per_sec": 70229.17812525907, "mfu": 21.578574178688875, "total_flops": 1.7771748410248921e+19 }, { "step": 14100, "train_loss": 2.5748010873794556, "val_loss": 2.6970475554466247, "lr": 0.02, "tokens_per_sec": 70071.97021212461, "mfu": 21.53027057176073, "total_flops": 1.7829822223758655e+19 }, { "step": 14150, "train_loss": 2.5738344490528107, "val_loss": 2.6904390811920167, "lr": 0.02, "tokens_per_sec": 70275.71666961232, "mfu": 21.592873583271164, "total_flops": 1.788789603726839e+19 }, { "step": 14200, "train_loss": 2.607429787516594, "val_loss": 2.699114978313446, "lr": 0.02, "tokens_per_sec": 70223.35666132403, "mfu": 21.576785479252198, "total_flops": 1.7945969850778124e+19 }, { "step": 14250, "train_loss": 2.658610910177231, "val_loss": 2.7006686329841614, "lr": 0.02, "tokens_per_sec": 70218.85175854548, "mfu": 21.575401305588635, "total_flops": 1.8004043664287859e+19 }, { "step": 14300, "train_loss": 2.613561823964119, "val_loss": 2.7155104994773867, "lr": 0.02, "tokens_per_sec": 70206.51545496765, "mfu": 21.57161086052077, "total_flops": 1.8062117477797593e+19 }, { "step": 14350, "train_loss": 2.5862353146076202, "val_loss": 2.706381583213806, "lr": 0.02, "tokens_per_sec": 70232.6663078057, "mfu": 21.579645955517833, "total_flops": 1.8120191291307327e+19 }, { "step": 14400, "train_loss": 2.584561437368393, "val_loss": 2.7018906712532043, "lr": 0.02, "tokens_per_sec": 70251.8311925962, "mfu": 21.585534546259137, "total_flops": 1.8178265104817062e+19 }, { "step": 14450, "train_loss": 2.610737219452858, "val_loss": 2.7197303116321563, "lr": 0.02, "tokens_per_sec": 70059.37408795145, "mfu": 21.526400294376174, "total_flops": 1.8236338918326796e+19 }, { "step": 14500, "train_loss": 2.6162658035755157, "val_loss": 2.71315957903862, "lr": 0.02, "tokens_per_sec": 70209.70044612157, "mfu": 21.57258947894844, "total_flops": 1.829441273183653e+19 }, { "step": 14550, "train_loss": 2.719397410750389, "val_loss": 2.7064169228076933, "lr": 0.02, "tokens_per_sec": 70227.90610092366, "mfu": 21.578183337272076, "total_flops": 1.8352486545346265e+19 }, { "step": 14600, "train_loss": 2.550754353404045, "val_loss": 2.704353654384613, "lr": 0.02, "tokens_per_sec": 70209.50901122653, "mfu": 21.57253065878005, "total_flops": 1.8410560358856e+19 }, { "step": 14650, "train_loss": 2.751055210828781, "val_loss": 2.7144021570682524, "lr": 0.02, "tokens_per_sec": 70246.63136750608, "mfu": 21.583936851192515, "total_flops": 1.8468634172365734e+19 }, { "step": 14700, "train_loss": 2.588581830263138, "val_loss": 2.703622704744339, "lr": 0.02, "tokens_per_sec": 70220.90636202156, "mfu": 21.576032601792008, "total_flops": 1.852670798587547e+19 }, { "step": 14750, "train_loss": 2.5866953879594803, "val_loss": 2.7090446412563325, "lr": 0.02, "tokens_per_sec": 70219.54752534849, "mfu": 21.575615086469607, "total_flops": 1.8584781799385203e+19 }, { "step": 14800, "train_loss": 2.6120127886533737, "val_loss": 2.6936957955360414, "lr": 0.02, "tokens_per_sec": 70056.30594351621, "mfu": 21.525457578199592, "total_flops": 1.8642855612894937e+19 }, { "step": 14850, "train_loss": 2.6211347430944443, "val_loss": 2.7043174862861634, "lr": 0.02, "tokens_per_sec": 70201.50397715726, "mfu": 21.570071036923753, "total_flops": 1.870092942640467e+19 }, { "step": 14900, "train_loss": 2.646672412753105, "val_loss": 2.7002520680427553, "lr": 0.02, "tokens_per_sec": 70214.55380391362, "mfu": 21.57408071868564, "total_flops": 1.8759003239914406e+19 }, { "step": 14950, "train_loss": 2.5893217772245407, "val_loss": 2.68537123799324, "lr": 0.02, "tokens_per_sec": 70209.82900398185, "mfu": 21.572628979557706, "total_flops": 1.881707705342414e+19 }, { "step": 15000, "train_loss": 2.496798686683178, "val_loss": 2.6873112499713896, "lr": 0.02, "tokens_per_sec": 70225.07861072318, "mfu": 21.57731456436801, "total_flops": 1.8875150866933875e+19 }, { "step": 15050, "train_loss": 2.6120809465646744, "val_loss": 2.6986561715602875, "lr": 0.02, "tokens_per_sec": 70235.34377313984, "mfu": 21.580468631873526, "total_flops": 1.893322468044361e+19 }, { "step": 15100, "train_loss": 2.589638665318489, "val_loss": 2.6991723120212554, "lr": 0.02, "tokens_per_sec": 70226.08421351104, "mfu": 21.577623545263574, "total_flops": 1.8991298493953343e+19 }, { "step": 15150, "train_loss": 2.5153046399354935, "val_loss": 2.7059680104255674, "lr": 0.02, "tokens_per_sec": 70237.30029581553, "mfu": 21.58106979183035, "total_flops": 1.9049372307463078e+19 }, { "step": 15200, "train_loss": 2.603958636522293, "val_loss": 2.7175380647182465, "lr": 0.02, "tokens_per_sec": 70217.51508637612, "mfu": 21.57499060051805, "total_flops": 1.9107446120972812e+19 }, { "step": 15250, "train_loss": 2.599534645676613, "val_loss": 2.717556869983673, "lr": 0.02, "tokens_per_sec": 70212.6102057161, "mfu": 21.57348352989605, "total_flops": 1.9165519934482547e+19 }, { "step": 15300, "train_loss": 2.6571564823389053, "val_loss": 2.720191490650177, "lr": 0.02, "tokens_per_sec": 70215.26484315922, "mfu": 21.574299192167956, "total_flops": 1.922359374799228e+19 }, { "step": 15050, "train_loss": 2.5011045336723328, "val_loss": 2.5259799718856812, "lr": 0.02, "tokens_per_sec": 70822.64640515842, "mfu": 21.658787127552376, "total_flops": 1.9281667561502015e+19 }, { "step": 15100, "train_loss": 2.5692661106586456, "val_loss": 2.5606980443000795, "lr": 0.02, "tokens_per_sec": 70767.62368058978, "mfu": 21.64196022910816, "total_flops": 1.933974137501175e+19 }, { "step": 15150, "train_loss": 2.5663330778479576, "val_loss": 2.5890700042247774, "lr": 0.02, "tokens_per_sec": 70799.76122967034, "mfu": 21.65178844606514, "total_flops": 1.9397815188521484e+19 }, { "step": 15200, "train_loss": 2.6242035180330276, "val_loss": 2.595632380247116, "lr": 0.02, "tokens_per_sec": 70832.7004053005, "mfu": 21.66186181425092, "total_flops": 1.945588900203122e+19 }, { "step": 15250, "train_loss": 2.5373122468590736, "val_loss": 2.5982221364974976, "lr": 0.02, "tokens_per_sec": 70710.97738859533, "mfu": 21.62463681573479, "total_flops": 1.9513962815540953e+19 }, { "step": 15300, "train_loss": 2.626233994960785, "val_loss": 2.6194027781486513, "lr": 0.02, "tokens_per_sec": 70690.91638907796, "mfu": 21.618501815983056, "total_flops": 1.9572036629050687e+19 }, { "step": 15350, "train_loss": 2.5037445425987244, "val_loss": 2.6230907142162323, "lr": 0.02, "tokens_per_sec": 70841.31925088949, "mfu": 21.664497605927345, "total_flops": 1.963011044256042e+19 }, { "step": 15400, "train_loss": 2.725143328309059, "val_loss": 2.6347573161125184, "lr": 0.02, "tokens_per_sec": 70836.69318627186, "mfu": 21.663082875556803, "total_flops": 1.9688184256070156e+19 }, { "step": 15450, "train_loss": 2.607233941555023, "val_loss": 2.6485621273517608, "lr": 0.02, "tokens_per_sec": 70812.53403285159, "mfu": 21.655694589638795, "total_flops": 1.974625806957989e+19 }, { "step": 15500, "train_loss": 2.6106958985328674, "val_loss": 2.6626220107078553, "lr": 0.02, "tokens_per_sec": 70740.52877982834, "mfu": 21.63367414100469, "total_flops": 1.9804331883089625e+19 }, { "step": 15550, "train_loss": 2.654520571231842, "val_loss": 2.6427935302257537, "lr": 0.02, "tokens_per_sec": 70855.78450661922, "mfu": 21.66892133068911, "total_flops": 1.986240569659936e+19 }, { "step": 15600, "train_loss": 2.7503251135349274, "val_loss": 2.63273109793663, "lr": 0.02, "tokens_per_sec": 70792.46802091923, "mfu": 21.649558057004086, "total_flops": 1.9920479510109094e+19 }, { "step": 15650, "train_loss": 2.598275765776634, "val_loss": 2.621151161193848, "lr": 0.02, "tokens_per_sec": 70701.00678300665, "mfu": 21.621587632529426, "total_flops": 1.997855332361883e+19 }, { "step": 15700, "train_loss": 2.5822859704494476, "val_loss": 2.617991364002228, "lr": 0.02, "tokens_per_sec": 70797.39009105547, "mfu": 21.651063310969004, "total_flops": 2.0036627137128563e+19 }, { "step": 15750, "train_loss": 2.6830928325653076, "val_loss": 2.6261918783187865, "lr": 0.02, "tokens_per_sec": 70786.01949477864, "mfu": 21.647585986458083, "total_flops": 2.0094700950638297e+19 }, { "step": 15800, "train_loss": 2.395069792866707, "val_loss": 2.6370294749736787, "lr": 0.02, "tokens_per_sec": 70823.31358734316, "mfu": 21.658991163374928, "total_flops": 2.015277476414803e+19 }, { "step": 15850, "train_loss": 2.6195893734693527, "val_loss": 2.6514434933662416, "lr": 0.02, "tokens_per_sec": 70822.91954916564, "mfu": 21.658870659702163, "total_flops": 2.0210848577657766e+19 }, { "step": 15900, "train_loss": 2.6659241691231728, "val_loss": 2.6493006229400633, "lr": 0.02, "tokens_per_sec": 70844.27349820061, "mfu": 21.66540106572279, "total_flops": 2.02689223911675e+19 }, { "step": 15950, "train_loss": 2.6470499634742737, "val_loss": 2.6604777455329893, "lr": 0.02, "tokens_per_sec": 70789.01379554643, "mfu": 21.648501695291586, "total_flops": 2.0326996204677235e+19 }, { "step": 16000, "train_loss": 2.567369759082794, "val_loss": 2.66320680975914, "lr": 0.02, "tokens_per_sec": 70693.97673794771, "mfu": 21.619437723465627, "total_flops": 2.038507001818697e+19 }, { "step": 16050, "train_loss": 2.5259591192007065, "val_loss": 2.6549350917339325, "lr": 0.01943255903154212, "tokens_per_sec": 70815.6263577521, "mfu": 21.656640275943236, "total_flops": 2.0443143831696703e+19 }, { "step": 16100, "train_loss": 2.6835711896419525, "val_loss": 2.635953629016876, "lr": 0.018881217525718468, "tokens_per_sec": 70834.84756150097, "mfu": 21.66251845166051, "total_flops": 2.0501217645206438e+19 }, { "step": 16150, "train_loss": 2.676758661866188, "val_loss": 2.644348645210266, "lr": 0.01834551870779559, "tokens_per_sec": 70833.7125334854, "mfu": 21.662171340511456, "total_flops": 2.0559291458716172e+19 }, { "step": 16200, "train_loss": 2.491854965686798, "val_loss": 2.6453607082366943, "lr": 0.01782501876267491, "tokens_per_sec": 70804.74951601207, "mfu": 21.653313950088084, "total_flops": 2.0617365272225907e+19 }, { "step": 16250, "train_loss": 2.638686791062355, "val_loss": 2.6507443487644196, "lr": 0.01731928646720131, "tokens_per_sec": 70825.65635173608, "mfu": 21.65970762114447, "total_flops": 2.067543908573564e+19 }, { "step": 16300, "train_loss": 2.600520685315132, "val_loss": 2.6338987708091737, "lr": 0.016827902832903904, "tokens_per_sec": 70849.17663872243, "mfu": 21.66690053068504, "total_flops": 2.0733512899245375e+19 }, { "step": 16350, "train_loss": 2.6237353086471558, "val_loss": 2.6439998149871826, "lr": 0.016350460758873, "tokens_per_sec": 70703.86261251083, "mfu": 21.622460994461562, "total_flops": 2.079158671275511e+19 }, { "step": 16400, "train_loss": 2.673539698123932, "val_loss": 2.6453349232673644, "lr": 0.01588656469448563, "tokens_per_sec": 70845.07186851281, "mfu": 21.665645221138018, "total_flops": 2.0849660526264844e+19 }, { "step": 16450, "train_loss": 2.6117686182260513, "val_loss": 2.6378572285175323, "lr": 0.01543583031170025, "tokens_per_sec": 70741.21808301542, "mfu": 21.6338849418114, "total_flops": 2.090773433977458e+19 }, { "step": 16500, "train_loss": 2.569175586104393, "val_loss": 2.630873453617096, "lr": 0.014997884186649118, "tokens_per_sec": 70848.69488819782, "mfu": 21.666753203063287, "total_flops": 2.0965808153284313e+19 }, { "step": 16550, "train_loss": 2.5066262930631638, "val_loss": 2.6248888432979585, "lr": 0.014572363490264556, "tokens_per_sec": 70831.78663993345, "mfu": 21.661582369037077, "total_flops": 2.1023881966794047e+19 }, { "step": 16600, "train_loss": 2.533449575304985, "val_loss": 2.643479061126709, "lr": 0.014158915687682759, "tokens_per_sec": 70832.01285455696, "mfu": 21.661651549371683, "total_flops": 2.108195578030378e+19 }, { "step": 16650, "train_loss": 2.5322001576423645, "val_loss": 2.6342837035655977, "lr": 0.013757198246176152, "tokens_per_sec": 70853.31213328421, "mfu": 21.668165236269022, "total_flops": 2.1140029593813516e+19 }, { "step": 16700, "train_loss": 2.5647076219320297, "val_loss": 2.636631351709366, "lr": 0.013366878351372293, "tokens_per_sec": 70662.51164198358, "mfu": 21.60981515427305, "total_flops": 2.119810340732325e+19 }, { "step": 16750, "train_loss": 2.6134786158800125, "val_loss": 2.632091587781906, "lr": 0.012987632631524226, "tokens_per_sec": 70829.18487637353, "mfu": 21.660786704853933, "total_flops": 2.1256177220832985e+19 }, { "step": 16800, "train_loss": 2.563320219516754, "val_loss": 2.6496681392192842, "lr": 0.012619146889603866, "tokens_per_sec": 70828.32515462447, "mfu": 21.66052378711085, "total_flops": 2.131425103434272e+19 }, { "step": 16850, "train_loss": 2.5337288677692413, "val_loss": 2.62899107336998, "lr": 0.012261115842996415, "tokens_per_sec": 70843.56255750875, "mfu": 21.66518364779382, "total_flops": 2.1372324847852454e+19 }, { "step": 16900, "train_loss": 2.531308576464653, "val_loss": 2.624918353557587, "lr": 0.01191324287058021, "tokens_per_sec": 70843.99254368903, "mfu": 21.66531514498602, "total_flops": 2.143039866136219e+19 }, { "step": 16950, "train_loss": 2.552768364548683, "val_loss": 2.6276935577392577, "lr": 0.011575239766982413, "tokens_per_sec": 70791.12024610583, "mfu": 21.649145884213866, "total_flops": 2.1488472474871923e+19 }, { "step": 17000, "train_loss": 2.5594132989645004, "val_loss": 2.6281732857227325, "lr": 0.011246826503806982, "tokens_per_sec": 70846.84612722936, "mfu": 21.66618782006367, "total_flops": 2.1546546288381657e+19 }, { "step": 17050, "train_loss": 2.6567023247480392, "val_loss": 2.620693105459213, "lr": 0.010927730997637084, "tokens_per_sec": 70635.0011758644, "mfu": 21.601401979113696, "total_flops": 2.160462010189139e+19 }, { "step": 17100, "train_loss": 2.491407111287117, "val_loss": 2.6200718581676483, "lr": 0.010617688884619766, "tokens_per_sec": 70835.64195443952, "mfu": 21.66276139072626, "total_flops": 2.1662693915401126e+19 }, { "step": 17150, "train_loss": 2.6877047419548035, "val_loss": 2.6259263932704924, "lr": 0.010316443301446114, "tokens_per_sec": 70749.92382525704, "mfu": 21.636547307983363, "total_flops": 2.172076772891086e+19 }, { "step": 17200, "train_loss": 2.525592103600502, "val_loss": 2.611518979072571, "lr": 0.010023744672545448, "tokens_per_sec": 70787.86815875485, "mfu": 21.648151339795795, "total_flops": 2.1778841542420595e+19 }, { "step": 17250, "train_loss": 2.5193743258714676, "val_loss": 2.610617792606354, "lr": 0.009739350503317262, "tokens_per_sec": 70782.25900458822, "mfu": 21.64643596368057, "total_flops": 2.183691535593033e+19 }, { "step": 17300, "train_loss": 2.5447549670934677, "val_loss": 2.6041156351566315, "lr": 0.00946302517922961, "tokens_per_sec": 70798.21395966616, "mfu": 21.65131526420382, "total_flops": 2.1894989169440063e+19 }, { "step": 17050, "train_loss": 2.4430361688137054, "val_loss": 2.492651104927063, "lr": 0.010927730997637084, "tokens_per_sec": 70194.53280299473, "mfu": 21.56194795535554, "total_flops": 2.1953062982949798e+19 }, { "step": 17100, "train_loss": 2.5078289210796356, "val_loss": 2.5014576375484467, "lr": 0.010617688884619766, "tokens_per_sec": 70187.43364078536, "mfu": 21.559767276035185, "total_flops": 2.2011136796459532e+19 }, { "step": 17150, "train_loss": 2.500446632504463, "val_loss": 2.515454649925232, "lr": 0.010316443301446114, "tokens_per_sec": 70188.95186454445, "mfu": 21.560233635171304, "total_flops": 2.2069210609969267e+19 }, { "step": 17200, "train_loss": 2.5545513927936554, "val_loss": 2.526023209095001, "lr": 0.010023744672545448, "tokens_per_sec": 70195.58066519728, "mfu": 21.56226983156688, "total_flops": 2.2127284423479e+19 }, { "step": 17250, "train_loss": 2.4661522805690765, "val_loss": 2.52562854886055, "lr": 0.009739350503317262, "tokens_per_sec": 70194.09453183571, "mfu": 21.561813329772264, "total_flops": 2.2185358236988735e+19 }, { "step": 17300, "train_loss": 2.5579015016555786, "val_loss": 2.5291117131710052, "lr": 0.00946302517922961, "tokens_per_sec": 70178.53306763705, "mfu": 21.557033249789363, "total_flops": 2.224343205049847e+19 }, { "step": 17350, "train_loss": 2.43189400434494, "val_loss": 2.528818500041962, "lr": 0.009194539770617445, "tokens_per_sec": 70201.79005590732, "mfu": 21.564177196057784, "total_flops": 2.2301505864008204e+19 }, { "step": 17400, "train_loss": 2.6497219800949097, "val_loss": 2.5407229542732237, "lr": 0.008933671843019263, "tokens_per_sec": 70196.11978921283, "mfu": 21.562435436543513, "total_flops": 2.235957967751794e+19 }, { "step": 17450, "train_loss": 2.535208210349083, "val_loss": 2.543413257598877, "lr": 0.008680205272894878, "tokens_per_sec": 70208.10410382063, "mfu": 21.566116708539145, "total_flops": 2.2417653491027673e+19 }, { "step": 17050, "train_loss": 2.4430273175239563, "val_loss": 2.492583030462265, "lr": 0.010927730997637084, "tokens_per_sec": 69976.98224120555, "mfu": 21.590922027044734, "total_flops": 2.2475727304537407e+19 }, { "step": 17100, "train_loss": 2.5079168379306793, "val_loss": 2.5017378509044645, "lr": 0.010617688884619766, "tokens_per_sec": 70199.67471947223, "mfu": 21.65963227690506, "total_flops": 2.253380111804714e+19 }, { "step": 17150, "train_loss": 2.5002063512802124, "val_loss": 2.5166682302951813, "lr": 0.010316443301446114, "tokens_per_sec": 70221.68480180511, "mfu": 21.666423338140316, "total_flops": 2.2591874931556876e+19 }, { "step": 17200, "train_loss": 2.5549302101135254, "val_loss": 2.526432579755783, "lr": 0.010023744672545448, "tokens_per_sec": 70243.03028256173, "mfu": 21.6730093410785, "total_flops": 2.264994874506661e+19 }, { "step": 17250, "train_loss": 2.466187208890915, "val_loss": 2.5247345089912416, "lr": 0.009739350503317262, "tokens_per_sec": 70228.35040146303, "mfu": 21.668479963588688, "total_flops": 2.2708022558576345e+19 }, { "step": 17300, "train_loss": 2.557755470275879, "val_loss": 2.528937244415283, "lr": 0.00946302517922961, "tokens_per_sec": 70228.95406261715, "mfu": 21.668666219132895, "total_flops": 2.276609637208608e+19 }, { "step": 17350, "train_loss": 2.432028129696846, "val_loss": 2.527354818582535, "lr": 0.009194539770617445, "tokens_per_sec": 70061.32081891438, "mfu": 21.616944121694466, "total_flops": 2.2824170185595814e+19 }, { "step": 17400, "train_loss": 2.649830251932144, "val_loss": 2.5389151513576507, "lr": 0.008933671843019263, "tokens_per_sec": 70234.34250372445, "mfu": 21.670328785995153, "total_flops": 2.288224399910555e+19 }, { "step": 17450, "train_loss": 2.534702032804489, "val_loss": 2.542128974199295, "lr": 0.008680205272894878, "tokens_per_sec": 70238.1502163951, "mfu": 21.671503629847237, "total_flops": 2.2940317812615283e+19 }, { "step": 17500, "train_loss": 2.5225761979818344, "val_loss": 2.539433038234711, "lr": 0.008433930068571646, "tokens_per_sec": 70230.80548197155, "mfu": 21.669237461985045, "total_flops": 2.2998391626125017e+19 }, { "step": 17550, "train_loss": 2.5620259642601013, "val_loss": 2.5314638674259187, "lr": 0.008194642196270831, "tokens_per_sec": 70189.97693235113, "mfu": 21.6566400906338, "total_flops": 2.305646543963475e+19 }, { "step": 17600, "train_loss": 2.6607564091682434, "val_loss": 2.5253191769123076, "lr": 0.007962143411069945, "tokens_per_sec": 70225.33539103418, "mfu": 21.66754970262294, "total_flops": 2.3114539253144486e+19 }, { "step": 17650, "train_loss": 2.5130698829889297, "val_loss": 2.5219079494476317, "lr": 0.007736241092661044, "tokens_per_sec": 70207.26667798325, "mfu": 21.66197472407832, "total_flops": 2.317261306665422e+19 }, { "step": 17700, "train_loss": 2.4957679137587547, "val_loss": 2.5156410336494446, "lr": 0.007516748085768884, "tokens_per_sec": 70217.80555572028, "mfu": 21.66522642313971, "total_flops": 2.3230686880163955e+19 }, { "step": 17750, "train_loss": 2.5936799198389053, "val_loss": 2.5133773267269133, "lr": 0.007303482545096754, "tokens_per_sec": 70207.28292869608, "mfu": 21.66197973812526, "total_flops": 2.328876069367369e+19 }, { "step": 17800, "train_loss": 2.3071534037590027, "val_loss": 2.523169732093811, "lr": 0.00709626778467151, "tokens_per_sec": 70210.6295030122, "mfu": 21.663012300873792, "total_flops": 2.3346834507183423e+19 }, { "step": 17850, "train_loss": 2.52487476170063, "val_loss": 2.5232641220092775, "lr": 0.0068949321314629875, "tokens_per_sec": 70216.21063847265, "mfu": 21.66473432226839, "total_flops": 2.3404908320693158e+19 }, { "step": 17900, "train_loss": 2.57292353361845, "val_loss": 2.525147634744644, "lr": 0.006699308783156554, "tokens_per_sec": 70208.71739194557, "mfu": 21.662422332006543, "total_flops": 2.3462982134202892e+19 }, { "step": 17950, "train_loss": 2.5538694113492966, "val_loss": 2.5261887848377227, "lr": 0.006509235669960918, "tokens_per_sec": 70191.6308148483, "mfu": 21.6571503848319, "total_flops": 2.3521055947712627e+19 }, { "step": 18000, "train_loss": 2.475142151117325, "val_loss": 2.5296940565109254, "lr": 0.006324555320336759, "tokens_per_sec": 70216.04248525678, "mfu": 21.66468243973708, "total_flops": 2.357912976122236e+19 }, { "step": 18050, "train_loss": 2.4305569678545, "val_loss": 2.5247761785984038, "lr": 0.006145114730534894, "tokens_per_sec": 70213.43430780503, "mfu": 21.663877704319948, "total_flops": 2.3637203574732095e+19 }, { "step": 18100, "train_loss": 2.5882591605186462, "val_loss": 2.519233852624893, "lr": 0.005970765237835919, "tokens_per_sec": 70215.29241962059, "mfu": 21.664451012085483, "total_flops": 2.369527738824183e+19 }, { "step": 18150, "train_loss": 2.579355038702488, "val_loss": 2.520945113897324, "lr": 0.005801362397386308, "tokens_per_sec": 70217.91418806078, "mfu": 21.665259940909674, "total_flops": 2.3753351201751564e+19 }, { "step": 18200, "train_loss": 2.3963413536548615, "val_loss": 2.521712827682495, "lr": 0.005636765862528908, "tokens_per_sec": 70216.60131755349, "mfu": 21.66485486364217, "total_flops": 2.38114250152613e+19 }, { "step": 18250, "train_loss": 2.548183634877205, "val_loss": 2.524004542827606, "lr": 0.0054768392685287225, "tokens_per_sec": 70199.54698313973, "mfu": 21.659592864728282, "total_flops": 2.3869498828771033e+19 }, { "step": 18300, "train_loss": 2.514558032155037, "val_loss": 2.5144589722156523, "lr": 0.005321450119597621, "tokens_per_sec": 70043.88400575175, "mfu": 21.61156411156346, "total_flops": 2.3927572642280767e+19 }, { "step": 18350, "train_loss": 2.5371487140655518, "val_loss": 2.517710816860199, "lr": 0.005170469679124381, "tokens_per_sec": 70222.98205292481, "mfu": 21.66682359615199, "total_flops": 2.39856464557905e+19 }, { "step": 18400, "train_loss": 2.5916066020727158, "val_loss": 2.5215661764144897, "lr": 0.005023772863019161, "tokens_per_sec": 70194.03168204919, "mfu": 21.657891156080836, "total_flops": 2.4043720269300236e+19 }, { "step": 18450, "train_loss": 2.526241347193718, "val_loss": 2.5169094502925873, "lr": 0.0048812381360839605, "tokens_per_sec": 70216.15335418667, "mfu": 21.664716647591487, "total_flops": 2.410179408280997e+19 }, { "step": 18500, "train_loss": 2.487909287214279, "val_loss": 2.511305093765259, "lr": 0.004742747411323311, "tokens_per_sec": 70188.60697085253, "mfu": 21.656217398329147, "total_flops": 2.4159867896319705e+19 }, { "step": 18550, "train_loss": 2.426807537674904, "val_loss": 2.5122386038303377, "lr": 0.004608185952111692, "tokens_per_sec": 70229.14605214057, "mfu": 21.668725456194814, "total_flops": 2.421794170982944e+19 }, { "step": 18600, "train_loss": 2.4519720375537872, "val_loss": 2.5156042873859406, "lr": 0.004477442277136679, "tokens_per_sec": 70199.08175790055, "mfu": 21.659449322644264, "total_flops": 2.4276015523339174e+19 }, { "step": 18650, "train_loss": 2.456761196255684, "val_loss": 2.5095840215682985, "lr": 0.004350408068039046, "tokens_per_sec": 70222.81487638665, "mfu": 21.666772014967943, "total_flops": 2.433408933684891e+19 }, { "step": 18700, "train_loss": 2.487348422408104, "val_loss": 2.514208322763443, "lr": 0.004226978079673293, "tokens_per_sec": 70193.76919298085, "mfu": 21.65781016686373, "total_flops": 2.4392163150358643e+19 }, { "step": 18750, "train_loss": 2.5389831364154816, "val_loss": 2.5142862021923067, "lr": 0.004107050052914292, "tokens_per_sec": 70220.9965045322, "mfu": 21.66621096869718, "total_flops": 2.4450236963868377e+19 }, { "step": 18800, "train_loss": 2.487064689397812, "val_loss": 2.52103568315506, "lr": 0.0039905246299377604, "tokens_per_sec": 70206.9877272651, "mfu": 21.6618886557311, "total_flops": 2.450831077737811e+19 }, { "step": 18850, "train_loss": 2.458393931388855, "val_loss": 2.5108425080776215, "lr": 0.0038773052719044145, "tokens_per_sec": 70196.25076219505, "mfu": 21.658575838170098, "total_flops": 2.4566384590887846e+19 }, { "step": 18900, "train_loss": 2.461804822087288, "val_loss": 2.5080023407936096, "lr": 0.003767298178979601, "tokens_per_sec": 70043.49959678273, "mfu": 21.61144550479005, "total_flops": 2.462445840439758e+19 }, { "step": 18950, "train_loss": 2.4762733578681946, "val_loss": 2.5128548741340637, "lr": 0.003660412212622112, "tokens_per_sec": 70186.46487495826, "mfu": 21.65555646920722, "total_flops": 2.4682532217907315e+19 }, { "step": 19000, "train_loss": 2.492058739066124, "val_loss": 2.5074475705623627, "lr": 0.003556558820077846, "tokens_per_sec": 70216.60468066345, "mfu": 21.664855901306883, "total_flops": 2.474060603141705e+19 }, { "step": 19050, "train_loss": 2.6086084991693497, "val_loss": 2.509049689769745, "lr": 0.003455651961015727, "tokens_per_sec": 70218.33986415529, "mfu": 21.66539128037404, "total_flops": 2.4798679844926783e+19 }, { "step": 19100, "train_loss": 2.4477206766605377, "val_loss": 2.511517381668091, "lr": 0.0033576080362451204, "tokens_per_sec": 70222.40125522732, "mfu": 21.666644394971854, "total_flops": 2.4856753658436518e+19 }, { "step": 19150, "train_loss": 2.6432832330465317, "val_loss": 2.511579215526581, "lr": 0.003262345818455677, "tokens_per_sec": 70208.37241605691, "mfu": 21.66231589204193, "total_flops": 2.4914827471946252e+19 }, { "step": 19200, "train_loss": 2.485611543059349, "val_loss": 2.5106746435165403, "lr": 0.003169786384922227, "tokens_per_sec": 70229.11240904075, "mfu": 21.668715075845142, "total_flops": 2.4972901285455987e+19 }, { "step": 19250, "train_loss": 2.477419838309288, "val_loss": 2.507012498378754, "lr": 0.0030798530521189845, "tokens_per_sec": 70049.52457141867, "mfu": 21.613304469744147, "total_flops": 2.503097509896572e+19 }, { "step": 19300, "train_loss": 2.505843862891197, "val_loss": 2.4998787224292753, "lr": 0.0029924713121888677, "tokens_per_sec": 70222.45843758582, "mfu": 21.66666203819973, "total_flops": 2.5089048912475455e+19 }, { "step": 19350, "train_loss": 2.514888495206833, "val_loss": 2.5030120730400087, "lr": 0.0029075687712153235, "tokens_per_sec": 70215.38904950574, "mfu": 21.66448082657946, "total_flops": 2.514712272598519e+19 }, { "step": 19400, "train_loss": 2.5361353158950806, "val_loss": 2.502032619714737, "lr": 0.0028250750892455092, "tokens_per_sec": 70220.36315738855, "mfu": 21.666015553742763, "total_flops": 2.5205196539494924e+19 }, { "step": 19450, "train_loss": 2.4763139337301254, "val_loss": 2.5012336313724517, "lr": 0.002744921922015124, "tokens_per_sec": 70211.37520972393, "mfu": 21.663242383608893, "total_flops": 2.526327035300466e+19 }, { "step": 19500, "train_loss": 2.384517587721348, "val_loss": 2.497239035367966, "lr": 0.002667042864326648, "tokens_per_sec": 70220.23478517876, "mfu": 21.66597594537037, "total_flops": 2.5321344166514393e+19 }, { "step": 19550, "train_loss": 2.4997721165418625, "val_loss": 2.4959007024765016, "lr": 0.0025913733950340394, "tokens_per_sec": 70220.8249648787, "mfu": 21.666158041303564, "total_flops": 2.5379417980024127e+19 }, { "step": 19600, "train_loss": 2.4810939878225327, "val_loss": 2.4974619328975676, "lr": 0.0025178508235883345, "tokens_per_sec": 70220.9263189256, "mfu": 21.66618931340553, "total_flops": 2.543749179353386e+19 }, { "step": 19650, "train_loss": 2.4052472189068794, "val_loss": 2.494679808616638, "lr": 0.0024464142380998632, "tokens_per_sec": 70213.12111983016, "mfu": 21.663781072299983, "total_flops": 2.5495565607043596e+19 }, { "step": 19700, "train_loss": 2.4900813549757004, "val_loss": 2.499013751745224, "lr": 0.0023770044548740367, "tokens_per_sec": 70217.72730476891, "mfu": 21.66520227934077, "total_flops": 2.555363942055333e+19 }, { "step": 19750, "train_loss": 2.484506592154503, "val_loss": 2.4978590846061706, "lr": 0.0023095639693789167, "tokens_per_sec": 70216.77720863638, "mfu": 21.664909133639735, "total_flops": 2.5611713234063065e+19 }, { "step": 19800, "train_loss": 2.538615956902504, "val_loss": 2.4995242774486544, "lr": 0.0022440369086039274, "tokens_per_sec": 70221.13945488365, "mfu": 21.66625507505606, "total_flops": 2.56697870475728e+19 }, { "step": 19550, "train_loss": 2.369577407836914, "val_loss": 2.434698057174683, "lr": 0.0025913733950340394, "tokens_per_sec": 70228.78181122785, "mfu": 21.61442648762866, "total_flops": 2.5727860861082534e+19 }, { "step": 19600, "train_loss": 2.436405763030052, "val_loss": 2.435771381855011, "lr": 0.0025178508235883345, "tokens_per_sec": 70236.42684073243, "mfu": 21.616779410802906, "total_flops": 2.578593467459227e+19 }, { "step": 19650, "train_loss": 2.4230239391326904, "val_loss": 2.4384620606899263, "lr": 0.0024464142380998632, "tokens_per_sec": 70220.75051917222, "mfu": 21.61195468380037, "total_flops": 2.5844008488102003e+19 }, { "step": 19700, "train_loss": 2.4815607368946075, "val_loss": 2.44344682097435, "lr": 0.0023770044548740367, "tokens_per_sec": 70243.49205167516, "mfu": 21.61895388227461, "total_flops": 2.5902082301611737e+19 }, { "step": 19750, "train_loss": 2.3950591534376144, "val_loss": 2.4454896986484527, "lr": 0.0023095639693789167, "tokens_per_sec": 70258.77042879947, "mfu": 21.623656133270384, "total_flops": 2.596015611512147e+19 }, { "step": 19800, "train_loss": 2.4920027405023575, "val_loss": 2.448630851507187, "lr": 0.0022440369086039274, "tokens_per_sec": 70099.8157121793, "mfu": 21.57473438140968, "total_flops": 2.6018229928631206e+19 }, { "step": 19850, "train_loss": 2.3662567362189293, "val_loss": 2.4463677644729613, "lr": 0.0021803689847702556, "tokens_per_sec": 70249.89910754852, "mfu": 21.620925792289356, "total_flops": 2.607630374214094e+19 }, { "step": 19900, "train_loss": 2.584068328142166, "val_loss": 2.4525881469249726, "lr": 0.0021185074503545782, "tokens_per_sec": 70253.34300150532, "mfu": 21.621985725137975, "total_flops": 2.6134377555650675e+19 }, { "step": 19950, "train_loss": 2.47366926074028, "val_loss": 2.4520198345184325, "lr": 0.0020584010543888563, "tokens_per_sec": 70260.01483742267, "mfu": 21.624039126937838, "total_flops": 2.619245136916041e+19 }, { "step": 19999, "train_loss": 2.5081333369016647, "val_loss": 2.45241938829422, "lr": 0.0020011516239787216, "tokens_per_sec": 70258.21878164218, "mfu": 21.62348635192688, "total_flops": 2.624936370639995e+19 } ]