llara1.0-100M-base / checkpoint-19303 /trainer_state.json
helloadhavan's picture
Upload folder using huggingface_hub
ee32672 verified
Raw
History Blame Contribute Delete
72.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 19303,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002590270942340569,
"grad_norm": 2.053372621536255,
"learning_rate": 2.94e-05,
"loss": 10.122694702148438,
"step": 50
},
{
"epoch": 0.005180541884681138,
"grad_norm": 1.547727346420288,
"learning_rate": 5.94e-05,
"loss": 8.92919921875,
"step": 100
},
{
"epoch": 0.0077708128270217065,
"grad_norm": 1.2461143732070923,
"learning_rate": 8.939999999999999e-05,
"loss": 7.86175048828125,
"step": 150
},
{
"epoch": 0.010361083769362275,
"grad_norm": 1.2778165340423584,
"learning_rate": 0.0001194,
"loss": 7.493834228515625,
"step": 200
},
{
"epoch": 0.012951354711702844,
"grad_norm": 1.2245980501174927,
"learning_rate": 0.0001494,
"loss": 7.367841796875,
"step": 250
},
{
"epoch": 0.015541625654043413,
"grad_norm": 1.2784472703933716,
"learning_rate": 0.00017939999999999997,
"loss": 7.193695068359375,
"step": 300
},
{
"epoch": 0.018131896596383982,
"grad_norm": 1.0248782634735107,
"learning_rate": 0.00020939999999999997,
"loss": 7.10666748046875,
"step": 350
},
{
"epoch": 0.02072216753872455,
"grad_norm": 1.1631972789764404,
"learning_rate": 0.0002394,
"loss": 6.988087158203125,
"step": 400
},
{
"epoch": 0.02331243848106512,
"grad_norm": 0.9435396790504456,
"learning_rate": 0.0002694,
"loss": 6.923257446289062,
"step": 450
},
{
"epoch": 0.02590270942340569,
"grad_norm": 1.0248340368270874,
"learning_rate": 0.00029939999999999996,
"loss": 6.855230712890625,
"step": 500
},
{
"epoch": 0.028492980365746257,
"grad_norm": 1.0538331270217896,
"learning_rate": 0.0002992182098601287,
"loss": 6.823525390625,
"step": 550
},
{
"epoch": 0.031083251308086826,
"grad_norm": 1.077528715133667,
"learning_rate": 0.0002984204648194437,
"loss": 6.768828125,
"step": 600
},
{
"epoch": 0.03367352225042739,
"grad_norm": 0.9541271328926086,
"learning_rate": 0.00029762271977875867,
"loss": 6.701234130859375,
"step": 650
},
{
"epoch": 0.036263793192767964,
"grad_norm": 1.0689588785171509,
"learning_rate": 0.0002968249747380737,
"loss": 6.623989868164062,
"step": 700
},
{
"epoch": 0.03885406413510853,
"grad_norm": 0.9765296578407288,
"learning_rate": 0.0002960272296973887,
"loss": 6.646817016601562,
"step": 750
},
{
"epoch": 0.0414443350774491,
"grad_norm": 0.985795259475708,
"learning_rate": 0.0002952294846567037,
"loss": 6.574093627929687,
"step": 800
},
{
"epoch": 0.04403460601978967,
"grad_norm": 0.9846292734146118,
"learning_rate": 0.0002944317396160187,
"loss": 6.57322998046875,
"step": 850
},
{
"epoch": 0.04662487696213024,
"grad_norm": 0.9685710072517395,
"learning_rate": 0.00029363399457533367,
"loss": 6.578934326171875,
"step": 900
},
{
"epoch": 0.049215147904470805,
"grad_norm": 0.9839648008346558,
"learning_rate": 0.00029283624953464874,
"loss": 6.5168524169921875,
"step": 950
},
{
"epoch": 0.05180541884681138,
"grad_norm": 0.9077258706092834,
"learning_rate": 0.0002920385044939637,
"loss": 6.452473754882813,
"step": 1000
},
{
"epoch": 0.05439568978915194,
"grad_norm": 0.9350703358650208,
"learning_rate": 0.0002912407594532787,
"loss": 6.431409912109375,
"step": 1050
},
{
"epoch": 0.056985960731492515,
"grad_norm": 0.9615451097488403,
"learning_rate": 0.0002904430144125937,
"loss": 6.449520874023437,
"step": 1100
},
{
"epoch": 0.05957623167383308,
"grad_norm": 1.1748050451278687,
"learning_rate": 0.0002896452693719087,
"loss": 6.4049951171875,
"step": 1150
},
{
"epoch": 0.06216650261617365,
"grad_norm": 0.9100262522697449,
"learning_rate": 0.00028884752433122374,
"loss": 6.3247802734375,
"step": 1200
},
{
"epoch": 0.06475677355851422,
"grad_norm": 0.9959622025489807,
"learning_rate": 0.0002880497792905387,
"loss": 6.380640258789063,
"step": 1250
},
{
"epoch": 0.06734704450085478,
"grad_norm": 0.9422142505645752,
"learning_rate": 0.0002872520342498537,
"loss": 6.371143188476562,
"step": 1300
},
{
"epoch": 0.06993731544319536,
"grad_norm": 1.0342239141464233,
"learning_rate": 0.0002864542892091687,
"loss": 6.372911376953125,
"step": 1350
},
{
"epoch": 0.07252758638553593,
"grad_norm": 0.9143936634063721,
"learning_rate": 0.00028565654416848373,
"loss": 6.302949829101562,
"step": 1400
},
{
"epoch": 0.0751178573278765,
"grad_norm": 0.9583525657653809,
"learning_rate": 0.00028485879912779874,
"loss": 6.344402465820313,
"step": 1450
},
{
"epoch": 0.07770812827021706,
"grad_norm": 0.9014634490013123,
"learning_rate": 0.00028406105408711375,
"loss": 6.285582275390625,
"step": 1500
},
{
"epoch": 0.08029839921255763,
"grad_norm": 0.94912189245224,
"learning_rate": 0.0002832633090464287,
"loss": 6.273238525390625,
"step": 1550
},
{
"epoch": 0.0828886701548982,
"grad_norm": 0.9462608098983765,
"learning_rate": 0.0002824655640057438,
"loss": 6.25449951171875,
"step": 1600
},
{
"epoch": 0.08547894109723878,
"grad_norm": 0.9752557277679443,
"learning_rate": 0.00028166781896505873,
"loss": 6.239361572265625,
"step": 1650
},
{
"epoch": 0.08806921203957933,
"grad_norm": 1.0141621828079224,
"learning_rate": 0.00028087007392437374,
"loss": 6.25331787109375,
"step": 1700
},
{
"epoch": 0.0906594829819199,
"grad_norm": 0.9530112147331238,
"learning_rate": 0.00028007232888368875,
"loss": 6.2260675048828125,
"step": 1750
},
{
"epoch": 0.09324975392426048,
"grad_norm": 1.020228624343872,
"learning_rate": 0.00027927458384300377,
"loss": 6.190052490234375,
"step": 1800
},
{
"epoch": 0.09584002486660105,
"grad_norm": 0.9993957877159119,
"learning_rate": 0.0002784768388023188,
"loss": 6.179526977539062,
"step": 1850
},
{
"epoch": 0.09843029580894161,
"grad_norm": 0.972730278968811,
"learning_rate": 0.00027767909376163373,
"loss": 6.153101196289063,
"step": 1900
},
{
"epoch": 0.10102056675128218,
"grad_norm": 0.962273359298706,
"learning_rate": 0.00027688134872094874,
"loss": 6.1879443359375,
"step": 1950
},
{
"epoch": 0.10361083769362275,
"grad_norm": 1.0591261386871338,
"learning_rate": 0.00027608360368026376,
"loss": 6.112931518554688,
"step": 2000
},
{
"epoch": 0.10620110863596333,
"grad_norm": 1.0552300214767456,
"learning_rate": 0.00027528585863957877,
"loss": 6.099534301757813,
"step": 2050
},
{
"epoch": 0.10879137957830388,
"grad_norm": 1.008030891418457,
"learning_rate": 0.0002744881135988938,
"loss": 6.118661499023437,
"step": 2100
},
{
"epoch": 0.11138165052064446,
"grad_norm": 1.0966074466705322,
"learning_rate": 0.00027369036855820874,
"loss": 6.07596923828125,
"step": 2150
},
{
"epoch": 0.11397192146298503,
"grad_norm": 1.1636556386947632,
"learning_rate": 0.0002728926235175238,
"loss": 6.071294555664062,
"step": 2200
},
{
"epoch": 0.1165621924053256,
"grad_norm": 0.9836180210113525,
"learning_rate": 0.00027209487847683876,
"loss": 6.070966796875,
"step": 2250
},
{
"epoch": 0.11915246334766616,
"grad_norm": 1.0915582180023193,
"learning_rate": 0.00027129713343615377,
"loss": 6.039877319335938,
"step": 2300
},
{
"epoch": 0.12174273429000673,
"grad_norm": 0.9727494120597839,
"learning_rate": 0.0002704993883954688,
"loss": 6.059029541015625,
"step": 2350
},
{
"epoch": 0.1243330052323473,
"grad_norm": 1.034834623336792,
"learning_rate": 0.0002697016433547838,
"loss": 6.029169921875,
"step": 2400
},
{
"epoch": 0.12692327617468788,
"grad_norm": 1.0125064849853516,
"learning_rate": 0.0002689038983140988,
"loss": 6.04001708984375,
"step": 2450
},
{
"epoch": 0.12951354711702845,
"grad_norm": 1.0140148401260376,
"learning_rate": 0.0002681061532734138,
"loss": 6.013587646484375,
"step": 2500
},
{
"epoch": 0.13210381805936902,
"grad_norm": 0.9864634275436401,
"learning_rate": 0.00026730840823272877,
"loss": 6.027371826171875,
"step": 2550
},
{
"epoch": 0.13469408900170957,
"grad_norm": 1.0383540391921997,
"learning_rate": 0.0002665106631920438,
"loss": 5.929879150390625,
"step": 2600
},
{
"epoch": 0.13728435994405014,
"grad_norm": 1.0422724485397339,
"learning_rate": 0.0002657129181513588,
"loss": 5.992667236328125,
"step": 2650
},
{
"epoch": 0.1398746308863907,
"grad_norm": 1.1296474933624268,
"learning_rate": 0.0002649151731106738,
"loss": 5.917919311523438,
"step": 2700
},
{
"epoch": 0.14246490182873128,
"grad_norm": 0.9819442629814148,
"learning_rate": 0.0002641174280699888,
"loss": 5.95181884765625,
"step": 2750
},
{
"epoch": 0.14505517277107186,
"grad_norm": 0.9991300106048584,
"learning_rate": 0.0002633196830293038,
"loss": 5.92015869140625,
"step": 2800
},
{
"epoch": 0.14764544371341243,
"grad_norm": 0.9796678423881531,
"learning_rate": 0.00026252193798861884,
"loss": 5.9480438232421875,
"step": 2850
},
{
"epoch": 0.150235714655753,
"grad_norm": 0.9904295206069946,
"learning_rate": 0.0002617241929479338,
"loss": 5.948643188476563,
"step": 2900
},
{
"epoch": 0.15282598559809357,
"grad_norm": 0.9728788733482361,
"learning_rate": 0.0002609264479072488,
"loss": 5.874560546875,
"step": 2950
},
{
"epoch": 0.15541625654043412,
"grad_norm": 1.0392069816589355,
"learning_rate": 0.0002601287028665638,
"loss": 5.878630981445313,
"step": 3000
},
{
"epoch": 0.1580065274827747,
"grad_norm": 1.0595394372940063,
"learning_rate": 0.00025933095782587883,
"loss": 5.886410522460937,
"step": 3050
},
{
"epoch": 0.16059679842511526,
"grad_norm": 1.0325700044631958,
"learning_rate": 0.00025853321278519384,
"loss": 5.828760375976563,
"step": 3100
},
{
"epoch": 0.16318706936745583,
"grad_norm": 0.9954292178153992,
"learning_rate": 0.00025773546774450885,
"loss": 5.899686889648438,
"step": 3150
},
{
"epoch": 0.1657773403097964,
"grad_norm": 1.0147758722305298,
"learning_rate": 0.0002569377227038238,
"loss": 5.906227416992188,
"step": 3200
},
{
"epoch": 0.16836761125213698,
"grad_norm": 0.948711633682251,
"learning_rate": 0.0002561399776631389,
"loss": 5.840421752929688,
"step": 3250
},
{
"epoch": 0.17095788219447755,
"grad_norm": 1.0470232963562012,
"learning_rate": 0.00025534223262245383,
"loss": 5.837828369140625,
"step": 3300
},
{
"epoch": 0.17354815313681812,
"grad_norm": 1.0115586519241333,
"learning_rate": 0.00025454448758176885,
"loss": 5.83600830078125,
"step": 3350
},
{
"epoch": 0.17613842407915867,
"grad_norm": 1.035805583000183,
"learning_rate": 0.00025374674254108386,
"loss": 5.834181518554687,
"step": 3400
},
{
"epoch": 0.17872869502149924,
"grad_norm": 1.0050755739212036,
"learning_rate": 0.00025294899750039887,
"loss": 5.810164794921875,
"step": 3450
},
{
"epoch": 0.1813189659638398,
"grad_norm": 1.010113000869751,
"learning_rate": 0.0002521512524597139,
"loss": 5.795087280273438,
"step": 3500
},
{
"epoch": 0.18390923690618038,
"grad_norm": 0.958347737789154,
"learning_rate": 0.00025135350741902884,
"loss": 5.823550415039063,
"step": 3550
},
{
"epoch": 0.18649950784852096,
"grad_norm": 1.0313442945480347,
"learning_rate": 0.00025055576237834385,
"loss": 5.788181762695313,
"step": 3600
},
{
"epoch": 0.18908977879086153,
"grad_norm": 1.0172735452651978,
"learning_rate": 0.00024975801733765886,
"loss": 5.783030395507812,
"step": 3650
},
{
"epoch": 0.1916800497332021,
"grad_norm": 0.9737220406532288,
"learning_rate": 0.00024896027229697387,
"loss": 5.807670288085937,
"step": 3700
},
{
"epoch": 0.19427032067554267,
"grad_norm": 0.9933872222900391,
"learning_rate": 0.0002481625272562889,
"loss": 5.762023315429688,
"step": 3750
},
{
"epoch": 0.19686059161788322,
"grad_norm": 0.9616476893424988,
"learning_rate": 0.00024736478221560384,
"loss": 5.771016845703125,
"step": 3800
},
{
"epoch": 0.1994508625602238,
"grad_norm": 1.0486760139465332,
"learning_rate": 0.00024656703717491885,
"loss": 5.7555908203125,
"step": 3850
},
{
"epoch": 0.20204113350256436,
"grad_norm": 1.0453466176986694,
"learning_rate": 0.00024576929213423386,
"loss": 5.721697998046875,
"step": 3900
},
{
"epoch": 0.20463140444490494,
"grad_norm": 0.9990628361701965,
"learning_rate": 0.00024497154709354887,
"loss": 5.755276489257812,
"step": 3950
},
{
"epoch": 0.2072216753872455,
"grad_norm": 1.0465604066848755,
"learning_rate": 0.0002441738020528639,
"loss": 5.7613458251953125,
"step": 4000
},
{
"epoch": 0.20981194632958608,
"grad_norm": 1.022708773612976,
"learning_rate": 0.00024337605701217887,
"loss": 5.7307891845703125,
"step": 4050
},
{
"epoch": 0.21240221727192665,
"grad_norm": 1.0304657220840454,
"learning_rate": 0.0002425783119714939,
"loss": 5.68947998046875,
"step": 4100
},
{
"epoch": 0.21499248821426722,
"grad_norm": 1.0114036798477173,
"learning_rate": 0.0002417805669308089,
"loss": 5.721485595703125,
"step": 4150
},
{
"epoch": 0.21758275915660777,
"grad_norm": 1.0038365125656128,
"learning_rate": 0.0002409828218901239,
"loss": 5.75361328125,
"step": 4200
},
{
"epoch": 0.22017303009894834,
"grad_norm": 1.030957818031311,
"learning_rate": 0.00024018507684943889,
"loss": 5.660762329101562,
"step": 4250
},
{
"epoch": 0.2227633010412889,
"grad_norm": 0.9766149520874023,
"learning_rate": 0.00023938733180875392,
"loss": 5.695264892578125,
"step": 4300
},
{
"epoch": 0.22535357198362949,
"grad_norm": 0.9994016289710999,
"learning_rate": 0.0002385895867680689,
"loss": 5.690259399414063,
"step": 4350
},
{
"epoch": 0.22794384292597006,
"grad_norm": 1.0340245962142944,
"learning_rate": 0.0002377918417273839,
"loss": 5.654801635742188,
"step": 4400
},
{
"epoch": 0.23053411386831063,
"grad_norm": 1.1064614057540894,
"learning_rate": 0.0002369940966866989,
"loss": 5.675580444335938,
"step": 4450
},
{
"epoch": 0.2331243848106512,
"grad_norm": 1.084747314453125,
"learning_rate": 0.00023619635164601392,
"loss": 5.655352783203125,
"step": 4500
},
{
"epoch": 0.23571465575299178,
"grad_norm": 0.9864429831504822,
"learning_rate": 0.00023539860660532893,
"loss": 5.62256103515625,
"step": 4550
},
{
"epoch": 0.23830492669533232,
"grad_norm": 0.9708291292190552,
"learning_rate": 0.0002346008615646439,
"loss": 5.628120727539063,
"step": 4600
},
{
"epoch": 0.2408951976376729,
"grad_norm": 1.0555731058120728,
"learning_rate": 0.00023380311652395892,
"loss": 5.648033447265625,
"step": 4650
},
{
"epoch": 0.24348546858001346,
"grad_norm": 0.979049801826477,
"learning_rate": 0.00023300537148327393,
"loss": 5.658864135742188,
"step": 4700
},
{
"epoch": 0.24607573952235404,
"grad_norm": 0.9773157835006714,
"learning_rate": 0.00023220762644258895,
"loss": 5.631370239257812,
"step": 4750
},
{
"epoch": 0.2486660104646946,
"grad_norm": 1.0320409536361694,
"learning_rate": 0.00023140988140190393,
"loss": 5.6348028564453125,
"step": 4800
},
{
"epoch": 0.25125628140703515,
"grad_norm": 1.0131503343582153,
"learning_rate": 0.00023061213636121891,
"loss": 5.562122802734375,
"step": 4850
},
{
"epoch": 0.25384655234937575,
"grad_norm": 1.0046422481536865,
"learning_rate": 0.00022981439132053395,
"loss": 5.629116821289062,
"step": 4900
},
{
"epoch": 0.2564368232917163,
"grad_norm": 1.0059541463851929,
"learning_rate": 0.00022901664627984894,
"loss": 5.580728149414062,
"step": 4950
},
{
"epoch": 0.2590270942340569,
"grad_norm": 0.9650059938430786,
"learning_rate": 0.00022821890123916395,
"loss": 5.640592041015625,
"step": 5000
},
{
"epoch": 0.26161736517639744,
"grad_norm": 1.0083752870559692,
"learning_rate": 0.00022742115619847893,
"loss": 5.601475219726563,
"step": 5050
},
{
"epoch": 0.26420763611873804,
"grad_norm": 1.0079939365386963,
"learning_rate": 0.00022662341115779394,
"loss": 5.613602905273438,
"step": 5100
},
{
"epoch": 0.2667979070610786,
"grad_norm": 1.0314289331436157,
"learning_rate": 0.00022582566611710895,
"loss": 5.603458862304688,
"step": 5150
},
{
"epoch": 0.26938817800341913,
"grad_norm": 1.077232837677002,
"learning_rate": 0.00022502792107642397,
"loss": 5.565240478515625,
"step": 5200
},
{
"epoch": 0.27197844894575973,
"grad_norm": 1.0053828954696655,
"learning_rate": 0.00022423017603573895,
"loss": 5.59989013671875,
"step": 5250
},
{
"epoch": 0.2745687198881003,
"grad_norm": 0.9962897300720215,
"learning_rate": 0.00022343243099505393,
"loss": 5.561985473632813,
"step": 5300
},
{
"epoch": 0.2771589908304409,
"grad_norm": 0.9912266135215759,
"learning_rate": 0.00022263468595436897,
"loss": 5.577881469726562,
"step": 5350
},
{
"epoch": 0.2797492617727814,
"grad_norm": 0.9959378242492676,
"learning_rate": 0.00022183694091368396,
"loss": 5.560752563476562,
"step": 5400
},
{
"epoch": 0.282339532715122,
"grad_norm": 0.9649254083633423,
"learning_rate": 0.00022103919587299897,
"loss": 5.5196160888671875,
"step": 5450
},
{
"epoch": 0.28492980365746257,
"grad_norm": 0.9408061504364014,
"learning_rate": 0.00022024145083231395,
"loss": 5.555208129882812,
"step": 5500
},
{
"epoch": 0.28752007459980317,
"grad_norm": 0.9576660990715027,
"learning_rate": 0.000219443705791629,
"loss": 5.521038208007813,
"step": 5550
},
{
"epoch": 0.2901103455421437,
"grad_norm": 1.060819149017334,
"learning_rate": 0.00021864596075094398,
"loss": 5.531424560546875,
"step": 5600
},
{
"epoch": 0.29270061648448426,
"grad_norm": 1.0806543827056885,
"learning_rate": 0.000217848215710259,
"loss": 5.504271850585938,
"step": 5650
},
{
"epoch": 0.29529088742682486,
"grad_norm": 0.9876242876052856,
"learning_rate": 0.00021705047066957397,
"loss": 5.4666552734375,
"step": 5700
},
{
"epoch": 0.2978811583691654,
"grad_norm": 1.0988022089004517,
"learning_rate": 0.000216252725628889,
"loss": 5.555277099609375,
"step": 5750
},
{
"epoch": 0.300471429311506,
"grad_norm": 0.9912010431289673,
"learning_rate": 0.000215454980588204,
"loss": 5.5327294921875,
"step": 5800
},
{
"epoch": 0.30306170025384654,
"grad_norm": 0.9935503602027893,
"learning_rate": 0.00021465723554751898,
"loss": 5.524019775390625,
"step": 5850
},
{
"epoch": 0.30565197119618714,
"grad_norm": 1.0960224866867065,
"learning_rate": 0.000213859490506834,
"loss": 5.49506103515625,
"step": 5900
},
{
"epoch": 0.3082422421385277,
"grad_norm": 1.0867022275924683,
"learning_rate": 0.00021306174546614903,
"loss": 5.5002587890625,
"step": 5950
},
{
"epoch": 0.31083251308086823,
"grad_norm": 1.0140448808670044,
"learning_rate": 0.000212264000425464,
"loss": 5.462314453125,
"step": 6000
},
{
"epoch": 0.31342278402320883,
"grad_norm": 1.0166538953781128,
"learning_rate": 0.000211466255384779,
"loss": 5.486057739257813,
"step": 6050
},
{
"epoch": 0.3160130549655494,
"grad_norm": 0.9597861170768738,
"learning_rate": 0.000210668510344094,
"loss": 5.470111694335937,
"step": 6100
},
{
"epoch": 0.31860332590789,
"grad_norm": 0.9873929619789124,
"learning_rate": 0.00020987076530340902,
"loss": 5.505219116210937,
"step": 6150
},
{
"epoch": 0.3211935968502305,
"grad_norm": 0.9906278848648071,
"learning_rate": 0.00020907302026272403,
"loss": 5.458482055664063,
"step": 6200
},
{
"epoch": 0.3237838677925711,
"grad_norm": 0.9522607922554016,
"learning_rate": 0.00020827527522203901,
"loss": 5.48074951171875,
"step": 6250
},
{
"epoch": 0.32637413873491167,
"grad_norm": 1.0274453163146973,
"learning_rate": 0.00020747753018135403,
"loss": 5.457478637695313,
"step": 6300
},
{
"epoch": 0.32896440967725227,
"grad_norm": 1.0103005170822144,
"learning_rate": 0.000206679785140669,
"loss": 5.503182983398437,
"step": 6350
},
{
"epoch": 0.3315546806195928,
"grad_norm": 0.9858667254447937,
"learning_rate": 0.00020588204009998405,
"loss": 5.424717407226563,
"step": 6400
},
{
"epoch": 0.33414495156193336,
"grad_norm": 1.0011284351348877,
"learning_rate": 0.00020508429505929903,
"loss": 5.419669799804687,
"step": 6450
},
{
"epoch": 0.33673522250427396,
"grad_norm": 0.9662808179855347,
"learning_rate": 0.00020428655001861402,
"loss": 5.426592407226562,
"step": 6500
},
{
"epoch": 0.3393254934466145,
"grad_norm": 0.9978706240653992,
"learning_rate": 0.00020348880497792903,
"loss": 5.404547119140625,
"step": 6550
},
{
"epoch": 0.3419157643889551,
"grad_norm": 0.9442576169967651,
"learning_rate": 0.00020269105993724404,
"loss": 5.452359008789062,
"step": 6600
},
{
"epoch": 0.34450603533129565,
"grad_norm": 1.0593928098678589,
"learning_rate": 0.00020189331489655905,
"loss": 5.419390869140625,
"step": 6650
},
{
"epoch": 0.34709630627363625,
"grad_norm": 1.0044008493423462,
"learning_rate": 0.00020109556985587403,
"loss": 5.416279296875,
"step": 6700
},
{
"epoch": 0.3496865772159768,
"grad_norm": 0.9485145807266235,
"learning_rate": 0.00020029782481518905,
"loss": 5.431090087890625,
"step": 6750
},
{
"epoch": 0.35227684815831734,
"grad_norm": 0.9767470955848694,
"learning_rate": 0.00019950007977450406,
"loss": 5.429151000976563,
"step": 6800
},
{
"epoch": 0.35486711910065794,
"grad_norm": 0.9516136646270752,
"learning_rate": 0.00019870233473381907,
"loss": 5.393023071289062,
"step": 6850
},
{
"epoch": 0.3574573900429985,
"grad_norm": 1.0077961683273315,
"learning_rate": 0.00019790458969313405,
"loss": 5.4060986328125,
"step": 6900
},
{
"epoch": 0.3600476609853391,
"grad_norm": 1.0050827264785767,
"learning_rate": 0.00019710684465244904,
"loss": 5.4089599609375,
"step": 6950
},
{
"epoch": 0.3626379319276796,
"grad_norm": 1.0005428791046143,
"learning_rate": 0.00019630909961176408,
"loss": 5.408168334960937,
"step": 7000
},
{
"epoch": 0.3652282028700202,
"grad_norm": 0.9248517751693726,
"learning_rate": 0.00019551135457107906,
"loss": 5.355136108398438,
"step": 7050
},
{
"epoch": 0.36781847381236077,
"grad_norm": 0.9752355813980103,
"learning_rate": 0.00019471360953039407,
"loss": 5.386387939453125,
"step": 7100
},
{
"epoch": 0.37040874475470137,
"grad_norm": 1.193434238433838,
"learning_rate": 0.00019391586448970906,
"loss": 5.354227905273437,
"step": 7150
},
{
"epoch": 0.3729990156970419,
"grad_norm": 0.967904806137085,
"learning_rate": 0.0001931181194490241,
"loss": 5.36183349609375,
"step": 7200
},
{
"epoch": 0.37558928663938246,
"grad_norm": 0.9852302074432373,
"learning_rate": 0.00019232037440833908,
"loss": 5.32298583984375,
"step": 7250
},
{
"epoch": 0.37817955758172306,
"grad_norm": 1.0207830667495728,
"learning_rate": 0.0001915226293676541,
"loss": 5.336920166015625,
"step": 7300
},
{
"epoch": 0.3807698285240636,
"grad_norm": 0.9589182734489441,
"learning_rate": 0.00019072488432696907,
"loss": 5.341430053710938,
"step": 7350
},
{
"epoch": 0.3833600994664042,
"grad_norm": 1.0570025444030762,
"learning_rate": 0.0001899271392862841,
"loss": 5.38286376953125,
"step": 7400
},
{
"epoch": 0.38595037040874475,
"grad_norm": 0.9938406944274902,
"learning_rate": 0.0001891293942455991,
"loss": 5.3147039794921875,
"step": 7450
},
{
"epoch": 0.38854064135108535,
"grad_norm": 1.020323395729065,
"learning_rate": 0.00018833164920491408,
"loss": 5.34332763671875,
"step": 7500
},
{
"epoch": 0.3911309122934259,
"grad_norm": 1.042059302330017,
"learning_rate": 0.0001875339041642291,
"loss": 5.33760986328125,
"step": 7550
},
{
"epoch": 0.39372118323576644,
"grad_norm": 0.975479245185852,
"learning_rate": 0.00018673615912354408,
"loss": 5.353353271484375,
"step": 7600
},
{
"epoch": 0.39631145417810704,
"grad_norm": 0.9893775582313538,
"learning_rate": 0.00018593841408285911,
"loss": 5.3269775390625,
"step": 7650
},
{
"epoch": 0.3989017251204476,
"grad_norm": 1.1089980602264404,
"learning_rate": 0.0001851406690421741,
"loss": 5.3275830078125,
"step": 7700
},
{
"epoch": 0.4014919960627882,
"grad_norm": 1.0087932348251343,
"learning_rate": 0.0001843429240014891,
"loss": 5.347962036132812,
"step": 7750
},
{
"epoch": 0.4040822670051287,
"grad_norm": 1.0132794380187988,
"learning_rate": 0.0001835451789608041,
"loss": 5.295555419921875,
"step": 7800
},
{
"epoch": 0.4066725379474693,
"grad_norm": 1.0221766233444214,
"learning_rate": 0.00018274743392011913,
"loss": 5.299251098632812,
"step": 7850
},
{
"epoch": 0.40926280888980987,
"grad_norm": 1.0063681602478027,
"learning_rate": 0.00018194968887943412,
"loss": 5.335856323242187,
"step": 7900
},
{
"epoch": 0.41185307983215047,
"grad_norm": 1.0150210857391357,
"learning_rate": 0.0001811519438387491,
"loss": 5.261018676757812,
"step": 7950
},
{
"epoch": 0.414443350774491,
"grad_norm": 0.9695018529891968,
"learning_rate": 0.0001803541987980641,
"loss": 5.286878662109375,
"step": 8000
},
{
"epoch": 0.41703362171683156,
"grad_norm": 0.982979953289032,
"learning_rate": 0.00017955645375737915,
"loss": 5.3470751953125,
"step": 8050
},
{
"epoch": 0.41962389265917216,
"grad_norm": 1.0282922983169556,
"learning_rate": 0.00017875870871669413,
"loss": 5.333184814453125,
"step": 8100
},
{
"epoch": 0.4222141636015127,
"grad_norm": 1.0405290126800537,
"learning_rate": 0.00017796096367600912,
"loss": 5.27518310546875,
"step": 8150
},
{
"epoch": 0.4248044345438533,
"grad_norm": 1.0299566984176636,
"learning_rate": 0.00017716321863532413,
"loss": 5.2979150390625,
"step": 8200
},
{
"epoch": 0.42739470548619385,
"grad_norm": 0.9923998117446899,
"learning_rate": 0.00017636547359463914,
"loss": 5.288524780273438,
"step": 8250
},
{
"epoch": 0.42998497642853445,
"grad_norm": 0.9922242164611816,
"learning_rate": 0.00017556772855395415,
"loss": 5.305535888671875,
"step": 8300
},
{
"epoch": 0.432575247370875,
"grad_norm": 1.057271957397461,
"learning_rate": 0.00017476998351326914,
"loss": 5.241010131835938,
"step": 8350
},
{
"epoch": 0.43516551831321554,
"grad_norm": 1.00055730342865,
"learning_rate": 0.00017397223847258415,
"loss": 5.270580444335938,
"step": 8400
},
{
"epoch": 0.43775578925555614,
"grad_norm": 0.9761698842048645,
"learning_rate": 0.00017317449343189916,
"loss": 5.28654052734375,
"step": 8450
},
{
"epoch": 0.4403460601978967,
"grad_norm": 1.0556055307388306,
"learning_rate": 0.00017237674839121417,
"loss": 5.2256103515625,
"step": 8500
},
{
"epoch": 0.4429363311402373,
"grad_norm": 1.032482624053955,
"learning_rate": 0.00017157900335052916,
"loss": 5.19790283203125,
"step": 8550
},
{
"epoch": 0.4455266020825778,
"grad_norm": 0.9819381833076477,
"learning_rate": 0.00017078125830984414,
"loss": 5.22557373046875,
"step": 8600
},
{
"epoch": 0.4481168730249184,
"grad_norm": 1.0934169292449951,
"learning_rate": 0.00016998351326915918,
"loss": 5.261765747070313,
"step": 8650
},
{
"epoch": 0.45070714396725897,
"grad_norm": 1.0120759010314941,
"learning_rate": 0.00016918576822847416,
"loss": 5.285647583007813,
"step": 8700
},
{
"epoch": 0.4532974149095995,
"grad_norm": 0.9943802356719971,
"learning_rate": 0.00016838802318778917,
"loss": 5.223272094726562,
"step": 8750
},
{
"epoch": 0.4558876858519401,
"grad_norm": 0.992707371711731,
"learning_rate": 0.00016759027814710416,
"loss": 5.220896606445312,
"step": 8800
},
{
"epoch": 0.45847795679428066,
"grad_norm": 1.0634857416152954,
"learning_rate": 0.00016679253310641917,
"loss": 5.202445678710937,
"step": 8850
},
{
"epoch": 0.46106822773662126,
"grad_norm": 1.0071600675582886,
"learning_rate": 0.00016599478806573418,
"loss": 5.221668701171875,
"step": 8900
},
{
"epoch": 0.4636584986789618,
"grad_norm": 1.0665333271026611,
"learning_rate": 0.0001651970430250492,
"loss": 5.246253662109375,
"step": 8950
},
{
"epoch": 0.4662487696213024,
"grad_norm": 0.9957162141799927,
"learning_rate": 0.00016439929798436418,
"loss": 5.2391455078125,
"step": 9000
},
{
"epoch": 0.46883904056364295,
"grad_norm": 1.018820881843567,
"learning_rate": 0.00016360155294367916,
"loss": 5.238643188476562,
"step": 9050
},
{
"epoch": 0.47142931150598355,
"grad_norm": 1.032936692237854,
"learning_rate": 0.0001628038079029942,
"loss": 5.20821044921875,
"step": 9100
},
{
"epoch": 0.4740195824483241,
"grad_norm": 0.9814261794090271,
"learning_rate": 0.00016200606286230918,
"loss": 5.1940673828125,
"step": 9150
},
{
"epoch": 0.47660985339066464,
"grad_norm": 1.0158833265304565,
"learning_rate": 0.0001612083178216242,
"loss": 5.216905517578125,
"step": 9200
},
{
"epoch": 0.47920012433300524,
"grad_norm": 1.1137468814849854,
"learning_rate": 0.00016041057278093918,
"loss": 5.2248681640625,
"step": 9250
},
{
"epoch": 0.4817903952753458,
"grad_norm": 1.0213594436645508,
"learning_rate": 0.00015961282774025422,
"loss": 5.193059692382812,
"step": 9300
},
{
"epoch": 0.4843806662176864,
"grad_norm": 1.0907715559005737,
"learning_rate": 0.0001588150826995692,
"loss": 5.1917724609375,
"step": 9350
},
{
"epoch": 0.48697093716002693,
"grad_norm": 1.042995572090149,
"learning_rate": 0.0001580173376588842,
"loss": 5.22154296875,
"step": 9400
},
{
"epoch": 0.48956120810236753,
"grad_norm": 1.0548311471939087,
"learning_rate": 0.0001572195926181992,
"loss": 5.220418701171875,
"step": 9450
},
{
"epoch": 0.4921514790447081,
"grad_norm": 1.0239475965499878,
"learning_rate": 0.00015642184757751424,
"loss": 5.175424194335937,
"step": 9500
},
{
"epoch": 0.4947417499870486,
"grad_norm": 1.0291051864624023,
"learning_rate": 0.00015562410253682922,
"loss": 5.191710815429688,
"step": 9550
},
{
"epoch": 0.4973320209293892,
"grad_norm": 1.0522440671920776,
"learning_rate": 0.0001548263574961442,
"loss": 5.157244262695312,
"step": 9600
},
{
"epoch": 0.49992229187172976,
"grad_norm": 1.0253952741622925,
"learning_rate": 0.00015402861245545921,
"loss": 5.150928344726562,
"step": 9650
},
{
"epoch": 0.5025125628140703,
"grad_norm": 1.0156198740005493,
"learning_rate": 0.00015323086741477423,
"loss": 5.178720092773437,
"step": 9700
},
{
"epoch": 0.505102833756411,
"grad_norm": 1.0090805292129517,
"learning_rate": 0.00015243312237408924,
"loss": 5.203817749023438,
"step": 9750
},
{
"epoch": 0.5076931046987515,
"grad_norm": 1.0411272048950195,
"learning_rate": 0.00015163537733340422,
"loss": 5.192196044921875,
"step": 9800
},
{
"epoch": 0.510283375641092,
"grad_norm": 0.9986085891723633,
"learning_rate": 0.00015083763229271923,
"loss": 5.123272705078125,
"step": 9850
},
{
"epoch": 0.5128736465834326,
"grad_norm": 1.0344675779342651,
"learning_rate": 0.00015003988725203424,
"loss": 5.1883087158203125,
"step": 9900
},
{
"epoch": 0.5154639175257731,
"grad_norm": 1.1055423021316528,
"learning_rate": 0.00014924214221134926,
"loss": 5.184451904296875,
"step": 9950
},
{
"epoch": 0.5180541884681138,
"grad_norm": 1.0196657180786133,
"learning_rate": 0.00014844439717066424,
"loss": 5.228606567382813,
"step": 10000
},
{
"epoch": 0.5206444594104543,
"grad_norm": 1.0357885360717773,
"learning_rate": 0.00014764665212997925,
"loss": 5.15303466796875,
"step": 10050
},
{
"epoch": 0.5232347303527949,
"grad_norm": 1.053410530090332,
"learning_rate": 0.00014684890708929424,
"loss": 5.157637329101562,
"step": 10100
},
{
"epoch": 0.5258250012951354,
"grad_norm": 1.0483365058898926,
"learning_rate": 0.00014605116204860925,
"loss": 5.137578125,
"step": 10150
},
{
"epoch": 0.5284152722374761,
"grad_norm": 1.03421151638031,
"learning_rate": 0.00014525341700792426,
"loss": 5.17541259765625,
"step": 10200
},
{
"epoch": 0.5310055431798166,
"grad_norm": 1.0358786582946777,
"learning_rate": 0.00014445567196723924,
"loss": 5.094895629882813,
"step": 10250
},
{
"epoch": 0.5335958141221572,
"grad_norm": 1.036773443222046,
"learning_rate": 0.00014365792692655425,
"loss": 5.12354736328125,
"step": 10300
},
{
"epoch": 0.5361860850644977,
"grad_norm": 1.0109663009643555,
"learning_rate": 0.00014286018188586927,
"loss": 5.14419921875,
"step": 10350
},
{
"epoch": 0.5387763560068383,
"grad_norm": 1.0550706386566162,
"learning_rate": 0.00014206243684518428,
"loss": 5.161801147460937,
"step": 10400
},
{
"epoch": 0.5413666269491789,
"grad_norm": 0.992174506187439,
"learning_rate": 0.00014126469180449926,
"loss": 5.097646179199219,
"step": 10450
},
{
"epoch": 0.5439568978915195,
"grad_norm": 1.0084872245788574,
"learning_rate": 0.00014046694676381427,
"loss": 5.099291687011719,
"step": 10500
},
{
"epoch": 0.54654716883386,
"grad_norm": 1.068367838859558,
"learning_rate": 0.00013966920172312928,
"loss": 5.135633544921875,
"step": 10550
},
{
"epoch": 0.5491374397762006,
"grad_norm": 1.1861008405685425,
"learning_rate": 0.0001388714566824443,
"loss": 5.086067810058593,
"step": 10600
},
{
"epoch": 0.5517277107185412,
"grad_norm": 1.0353467464447021,
"learning_rate": 0.00013807371164175928,
"loss": 5.1578912353515625,
"step": 10650
},
{
"epoch": 0.5543179816608818,
"grad_norm": 1.0449336767196655,
"learning_rate": 0.0001372759666010743,
"loss": 5.07172607421875,
"step": 10700
},
{
"epoch": 0.5569082526032223,
"grad_norm": 1.041418194770813,
"learning_rate": 0.00013647822156038927,
"loss": 5.111160888671875,
"step": 10750
},
{
"epoch": 0.5594985235455628,
"grad_norm": 1.0924108028411865,
"learning_rate": 0.00013568047651970429,
"loss": 5.09541015625,
"step": 10800
},
{
"epoch": 0.5620887944879034,
"grad_norm": 1.0294824838638306,
"learning_rate": 0.0001348827314790193,
"loss": 5.103350524902344,
"step": 10850
},
{
"epoch": 0.564679065430244,
"grad_norm": 1.0669180154800415,
"learning_rate": 0.0001340849864383343,
"loss": 5.056949768066406,
"step": 10900
},
{
"epoch": 0.5672693363725846,
"grad_norm": 1.0346755981445312,
"learning_rate": 0.0001332872413976493,
"loss": 5.089794921875,
"step": 10950
},
{
"epoch": 0.5698596073149251,
"grad_norm": 1.0336682796478271,
"learning_rate": 0.0001324894963569643,
"loss": 5.086961059570313,
"step": 11000
},
{
"epoch": 0.5724498782572657,
"grad_norm": 1.112239956855774,
"learning_rate": 0.00013169175131627932,
"loss": 5.12211181640625,
"step": 11050
},
{
"epoch": 0.5750401491996063,
"grad_norm": 1.0269253253936768,
"learning_rate": 0.00013089400627559433,
"loss": 5.078684997558594,
"step": 11100
},
{
"epoch": 0.5776304201419469,
"grad_norm": 1.104164958000183,
"learning_rate": 0.0001300962612349093,
"loss": 5.09882080078125,
"step": 11150
},
{
"epoch": 0.5802206910842874,
"grad_norm": 1.0542353391647339,
"learning_rate": 0.00012929851619422432,
"loss": 5.109794921875,
"step": 11200
},
{
"epoch": 0.582810962026628,
"grad_norm": 1.0559531450271606,
"learning_rate": 0.0001285007711535393,
"loss": 5.06039306640625,
"step": 11250
},
{
"epoch": 0.5854012329689685,
"grad_norm": 1.042899489402771,
"learning_rate": 0.00012770302611285432,
"loss": 5.068010864257812,
"step": 11300
},
{
"epoch": 0.5879915039113092,
"grad_norm": 1.002777099609375,
"learning_rate": 0.00012690528107216933,
"loss": 5.114893798828125,
"step": 11350
},
{
"epoch": 0.5905817748536497,
"grad_norm": 1.0369727611541748,
"learning_rate": 0.0001261075360314843,
"loss": 5.056740112304688,
"step": 11400
},
{
"epoch": 0.5931720457959903,
"grad_norm": 1.09577476978302,
"learning_rate": 0.00012530979099079932,
"loss": 5.068506469726563,
"step": 11450
},
{
"epoch": 0.5957623167383308,
"grad_norm": 1.0519545078277588,
"learning_rate": 0.00012451204595011434,
"loss": 5.057008666992187,
"step": 11500
},
{
"epoch": 0.5983525876806713,
"grad_norm": 1.0662899017333984,
"learning_rate": 0.00012371430090942935,
"loss": 5.0344482421875,
"step": 11550
},
{
"epoch": 0.600942858623012,
"grad_norm": 1.095126748085022,
"learning_rate": 0.00012291655586874433,
"loss": 5.038729553222656,
"step": 11600
},
{
"epoch": 0.6035331295653525,
"grad_norm": 1.0853722095489502,
"learning_rate": 0.00012211881082805934,
"loss": 5.02118408203125,
"step": 11650
},
{
"epoch": 0.6061234005076931,
"grad_norm": 1.063301682472229,
"learning_rate": 0.00012132106578737434,
"loss": 5.050283508300781,
"step": 11700
},
{
"epoch": 0.6087136714500336,
"grad_norm": 1.0829721689224243,
"learning_rate": 0.00012052332074668935,
"loss": 5.066478881835938,
"step": 11750
},
{
"epoch": 0.6113039423923743,
"grad_norm": 1.0675097703933716,
"learning_rate": 0.00011972557570600435,
"loss": 5.023760681152344,
"step": 11800
},
{
"epoch": 0.6138942133347148,
"grad_norm": 1.0228451490402222,
"learning_rate": 0.00011892783066531936,
"loss": 4.974055480957031,
"step": 11850
},
{
"epoch": 0.6164844842770554,
"grad_norm": 1.066720724105835,
"learning_rate": 0.00011813008562463436,
"loss": 5.049208068847657,
"step": 11900
},
{
"epoch": 0.6190747552193959,
"grad_norm": 1.036490559577942,
"learning_rate": 0.00011733234058394937,
"loss": 5.009530029296875,
"step": 11950
},
{
"epoch": 0.6216650261617365,
"grad_norm": 1.1739553213119507,
"learning_rate": 0.00011653459554326435,
"loss": 5.04651123046875,
"step": 12000
},
{
"epoch": 0.6242552971040771,
"grad_norm": 1.0319231748580933,
"learning_rate": 0.00011573685050257937,
"loss": 5.033486328125,
"step": 12050
},
{
"epoch": 0.6268455680464177,
"grad_norm": 1.0789339542388916,
"learning_rate": 0.00011493910546189436,
"loss": 4.994020690917969,
"step": 12100
},
{
"epoch": 0.6294358389887582,
"grad_norm": 1.076440453529358,
"learning_rate": 0.00011414136042120937,
"loss": 5.000180969238281,
"step": 12150
},
{
"epoch": 0.6320261099310988,
"grad_norm": 1.0861104726791382,
"learning_rate": 0.00011334361538052437,
"loss": 4.985296325683594,
"step": 12200
},
{
"epoch": 0.6346163808734394,
"grad_norm": 1.0445137023925781,
"learning_rate": 0.00011254587033983938,
"loss": 5.013554077148438,
"step": 12250
},
{
"epoch": 0.63720665181578,
"grad_norm": 1.0858595371246338,
"learning_rate": 0.00011174812529915438,
"loss": 4.968103637695313,
"step": 12300
},
{
"epoch": 0.6397969227581205,
"grad_norm": 1.1367961168289185,
"learning_rate": 0.00011095038025846939,
"loss": 5.062051391601562,
"step": 12350
},
{
"epoch": 0.642387193700461,
"grad_norm": 1.0913275480270386,
"learning_rate": 0.00011015263521778439,
"loss": 5.018081359863281,
"step": 12400
},
{
"epoch": 0.6449774646428016,
"grad_norm": 1.1007710695266724,
"learning_rate": 0.0001093548901770994,
"loss": 4.987852478027344,
"step": 12450
},
{
"epoch": 0.6475677355851422,
"grad_norm": 1.1163687705993652,
"learning_rate": 0.00010855714513641439,
"loss": 5.0391796875,
"step": 12500
},
{
"epoch": 0.6501580065274828,
"grad_norm": 1.0992286205291748,
"learning_rate": 0.0001077594000957294,
"loss": 4.996618347167969,
"step": 12550
},
{
"epoch": 0.6527482774698233,
"grad_norm": 1.0719584226608276,
"learning_rate": 0.0001069616550550444,
"loss": 5.018509216308594,
"step": 12600
},
{
"epoch": 0.6553385484121639,
"grad_norm": 1.1296252012252808,
"learning_rate": 0.00010616391001435939,
"loss": 5.024125671386718,
"step": 12650
},
{
"epoch": 0.6579288193545045,
"grad_norm": 1.0779918432235718,
"learning_rate": 0.0001053661649736744,
"loss": 4.969640197753907,
"step": 12700
},
{
"epoch": 0.6605190902968451,
"grad_norm": 1.1184015274047852,
"learning_rate": 0.0001045684199329894,
"loss": 5.004659423828125,
"step": 12750
},
{
"epoch": 0.6631093612391856,
"grad_norm": 1.1083693504333496,
"learning_rate": 0.00010377067489230441,
"loss": 4.98875244140625,
"step": 12800
},
{
"epoch": 0.6656996321815262,
"grad_norm": 1.069274663925171,
"learning_rate": 0.00010297292985161941,
"loss": 4.963236083984375,
"step": 12850
},
{
"epoch": 0.6682899031238667,
"grad_norm": 1.0793840885162354,
"learning_rate": 0.00010217518481093442,
"loss": 4.990594177246094,
"step": 12900
},
{
"epoch": 0.6708801740662074,
"grad_norm": 1.0875816345214844,
"learning_rate": 0.00010137743977024942,
"loss": 4.95796875,
"step": 12950
},
{
"epoch": 0.6734704450085479,
"grad_norm": 1.0997555255889893,
"learning_rate": 0.00010057969472956443,
"loss": 4.989603271484375,
"step": 13000
},
{
"epoch": 0.6760607159508885,
"grad_norm": 1.0631086826324463,
"learning_rate": 9.978194968887942e-05,
"loss": 5.020981140136719,
"step": 13050
},
{
"epoch": 0.678650986893229,
"grad_norm": 1.0640684366226196,
"learning_rate": 9.898420464819443e-05,
"loss": 4.959541015625,
"step": 13100
},
{
"epoch": 0.6812412578355695,
"grad_norm": 1.052869439125061,
"learning_rate": 9.818645960750942e-05,
"loss": 4.984897155761718,
"step": 13150
},
{
"epoch": 0.6838315287779102,
"grad_norm": 1.08975088596344,
"learning_rate": 9.738871456682444e-05,
"loss": 4.944871520996093,
"step": 13200
},
{
"epoch": 0.6864217997202507,
"grad_norm": 1.2661519050598145,
"learning_rate": 9.659096952613943e-05,
"loss": 4.930430908203125,
"step": 13250
},
{
"epoch": 0.6890120706625913,
"grad_norm": 1.0473780632019043,
"learning_rate": 9.579322448545445e-05,
"loss": 5.001515808105469,
"step": 13300
},
{
"epoch": 0.6916023416049318,
"grad_norm": 1.0621232986450195,
"learning_rate": 9.499547944476944e-05,
"loss": 4.933506164550781,
"step": 13350
},
{
"epoch": 0.6941926125472725,
"grad_norm": 1.1702978610992432,
"learning_rate": 9.419773440408445e-05,
"loss": 4.983041687011719,
"step": 13400
},
{
"epoch": 0.696782883489613,
"grad_norm": 1.1768927574157715,
"learning_rate": 9.339998936339945e-05,
"loss": 4.953892211914063,
"step": 13450
},
{
"epoch": 0.6993731544319536,
"grad_norm": 1.1313837766647339,
"learning_rate": 9.260224432271446e-05,
"loss": 4.9757080078125,
"step": 13500
},
{
"epoch": 0.7019634253742941,
"grad_norm": 1.091066598892212,
"learning_rate": 9.180449928202945e-05,
"loss": 4.9706729125976565,
"step": 13550
},
{
"epoch": 0.7045536963166347,
"grad_norm": 1.1084681749343872,
"learning_rate": 9.100675424134446e-05,
"loss": 4.945950927734375,
"step": 13600
},
{
"epoch": 0.7071439672589753,
"grad_norm": 1.108149528503418,
"learning_rate": 9.020900920065946e-05,
"loss": 4.929463195800781,
"step": 13650
},
{
"epoch": 0.7097342382013159,
"grad_norm": 1.0872802734375,
"learning_rate": 8.941126415997447e-05,
"loss": 4.941557006835938,
"step": 13700
},
{
"epoch": 0.7123245091436564,
"grad_norm": 1.0591199398040771,
"learning_rate": 8.861351911928947e-05,
"loss": 4.920311279296875,
"step": 13750
},
{
"epoch": 0.714914780085997,
"grad_norm": 1.1315951347351074,
"learning_rate": 8.781577407860448e-05,
"loss": 4.9243032836914065,
"step": 13800
},
{
"epoch": 0.7175050510283376,
"grad_norm": 1.136399745941162,
"learning_rate": 8.701802903791947e-05,
"loss": 4.938468933105469,
"step": 13850
},
{
"epoch": 0.7200953219706782,
"grad_norm": 1.0928754806518555,
"learning_rate": 8.622028399723447e-05,
"loss": 4.978819885253906,
"step": 13900
},
{
"epoch": 0.7226855929130187,
"grad_norm": 1.1162477731704712,
"learning_rate": 8.542253895654948e-05,
"loss": 4.914239807128906,
"step": 13950
},
{
"epoch": 0.7252758638553592,
"grad_norm": 1.143379807472229,
"learning_rate": 8.462479391586447e-05,
"loss": 4.932565612792969,
"step": 14000
},
{
"epoch": 0.7278661347976998,
"grad_norm": 1.1196132898330688,
"learning_rate": 8.382704887517949e-05,
"loss": 4.927554626464843,
"step": 14050
},
{
"epoch": 0.7304564057400404,
"grad_norm": 1.0910526514053345,
"learning_rate": 8.302930383449448e-05,
"loss": 4.92410888671875,
"step": 14100
},
{
"epoch": 0.733046676682381,
"grad_norm": 1.1446021795272827,
"learning_rate": 8.223155879380949e-05,
"loss": 4.9032418823242185,
"step": 14150
},
{
"epoch": 0.7356369476247215,
"grad_norm": 1.1149300336837769,
"learning_rate": 8.143381375312449e-05,
"loss": 4.967186584472656,
"step": 14200
},
{
"epoch": 0.7382272185670621,
"grad_norm": 1.1184216737747192,
"learning_rate": 8.06360687124395e-05,
"loss": 4.884440612792969,
"step": 14250
},
{
"epoch": 0.7408174895094027,
"grad_norm": 1.178585171699524,
"learning_rate": 7.98383236717545e-05,
"loss": 4.921098022460938,
"step": 14300
},
{
"epoch": 0.7434077604517433,
"grad_norm": 1.1591829061508179,
"learning_rate": 7.904057863106951e-05,
"loss": 4.918094177246093,
"step": 14350
},
{
"epoch": 0.7459980313940838,
"grad_norm": 1.1531950235366821,
"learning_rate": 7.82428335903845e-05,
"loss": 4.960716857910156,
"step": 14400
},
{
"epoch": 0.7485883023364244,
"grad_norm": 1.1077449321746826,
"learning_rate": 7.744508854969952e-05,
"loss": 4.948377990722657,
"step": 14450
},
{
"epoch": 0.7511785732787649,
"grad_norm": 1.1129505634307861,
"learning_rate": 7.664734350901451e-05,
"loss": 4.886638793945313,
"step": 14500
},
{
"epoch": 0.7537688442211056,
"grad_norm": 1.104398250579834,
"learning_rate": 7.584959846832953e-05,
"loss": 4.876330871582031,
"step": 14550
},
{
"epoch": 0.7563591151634461,
"grad_norm": 1.0603358745574951,
"learning_rate": 7.505185342764451e-05,
"loss": 4.879733276367188,
"step": 14600
},
{
"epoch": 0.7589493861057867,
"grad_norm": 1.1460893154144287,
"learning_rate": 7.425410838695952e-05,
"loss": 4.906120300292969,
"step": 14650
},
{
"epoch": 0.7615396570481272,
"grad_norm": 1.1097170114517212,
"learning_rate": 7.345636334627452e-05,
"loss": 4.904447021484375,
"step": 14700
},
{
"epoch": 0.7641299279904677,
"grad_norm": 1.1351251602172852,
"learning_rate": 7.265861830558953e-05,
"loss": 4.891495056152344,
"step": 14750
},
{
"epoch": 0.7667201989328084,
"grad_norm": 1.1649805307388306,
"learning_rate": 7.186087326490453e-05,
"loss": 4.919297485351563,
"step": 14800
},
{
"epoch": 0.769310469875149,
"grad_norm": 1.1074897050857544,
"learning_rate": 7.106312822421953e-05,
"loss": 4.88187255859375,
"step": 14850
},
{
"epoch": 0.7719007408174895,
"grad_norm": 1.1506906747817993,
"learning_rate": 7.026538318353454e-05,
"loss": 4.901511840820312,
"step": 14900
},
{
"epoch": 0.77449101175983,
"grad_norm": 1.1153931617736816,
"learning_rate": 6.946763814284953e-05,
"loss": 4.864546203613282,
"step": 14950
},
{
"epoch": 0.7770812827021707,
"grad_norm": 1.2132166624069214,
"learning_rate": 6.866989310216455e-05,
"loss": 4.8552725219726565,
"step": 15000
},
{
"epoch": 0.7796715536445112,
"grad_norm": 1.273779273033142,
"learning_rate": 6.787214806147954e-05,
"loss": 4.8751153564453125,
"step": 15050
},
{
"epoch": 0.7822618245868518,
"grad_norm": 1.1260673999786377,
"learning_rate": 6.707440302079455e-05,
"loss": 4.880099792480468,
"step": 15100
},
{
"epoch": 0.7848520955291923,
"grad_norm": 1.194525122642517,
"learning_rate": 6.627665798010955e-05,
"loss": 4.861859130859375,
"step": 15150
},
{
"epoch": 0.7874423664715329,
"grad_norm": 1.1089609861373901,
"learning_rate": 6.547891293942455e-05,
"loss": 4.869077453613281,
"step": 15200
},
{
"epoch": 0.7900326374138735,
"grad_norm": 1.194669246673584,
"learning_rate": 6.468116789873956e-05,
"loss": 4.892135009765625,
"step": 15250
},
{
"epoch": 0.7926229083562141,
"grad_norm": 1.1765038967132568,
"learning_rate": 6.388342285805456e-05,
"loss": 4.886556701660156,
"step": 15300
},
{
"epoch": 0.7952131792985546,
"grad_norm": 1.1625683307647705,
"learning_rate": 6.308567781736957e-05,
"loss": 4.85548583984375,
"step": 15350
},
{
"epoch": 0.7978034502408952,
"grad_norm": 1.2013742923736572,
"learning_rate": 6.228793277668457e-05,
"loss": 4.857999572753906,
"step": 15400
},
{
"epoch": 0.8003937211832358,
"grad_norm": 1.147049903869629,
"learning_rate": 6.149018773599957e-05,
"loss": 4.8666470336914065,
"step": 15450
},
{
"epoch": 0.8029839921255764,
"grad_norm": 1.1380870342254639,
"learning_rate": 6.069244269531458e-05,
"loss": 4.868302307128906,
"step": 15500
},
{
"epoch": 0.8055742630679169,
"grad_norm": 1.1927944421768188,
"learning_rate": 5.9894697654629575e-05,
"loss": 4.852730407714843,
"step": 15550
},
{
"epoch": 0.8081645340102575,
"grad_norm": 1.1827374696731567,
"learning_rate": 5.909695261394458e-05,
"loss": 4.834335632324219,
"step": 15600
},
{
"epoch": 0.810754804952598,
"grad_norm": 1.1386221647262573,
"learning_rate": 5.8299207573259584e-05,
"loss": 4.80497802734375,
"step": 15650
},
{
"epoch": 0.8133450758949387,
"grad_norm": 1.1220753192901611,
"learning_rate": 5.750146253257459e-05,
"loss": 4.840519409179688,
"step": 15700
},
{
"epoch": 0.8159353468372792,
"grad_norm": 1.178306221961975,
"learning_rate": 5.670371749188958e-05,
"loss": 4.811608581542969,
"step": 15750
},
{
"epoch": 0.8185256177796197,
"grad_norm": 1.108260989189148,
"learning_rate": 5.5905972451204585e-05,
"loss": 4.848948974609375,
"step": 15800
},
{
"epoch": 0.8211158887219603,
"grad_norm": 1.1373419761657715,
"learning_rate": 5.510822741051959e-05,
"loss": 4.879228210449218,
"step": 15850
},
{
"epoch": 0.8237061596643009,
"grad_norm": 1.1401219367980957,
"learning_rate": 5.4310482369834594e-05,
"loss": 4.843096923828125,
"step": 15900
},
{
"epoch": 0.8262964306066415,
"grad_norm": 1.2324658632278442,
"learning_rate": 5.35127373291496e-05,
"loss": 4.8840853881835935,
"step": 15950
},
{
"epoch": 0.828886701548982,
"grad_norm": 1.123178243637085,
"learning_rate": 5.27149922884646e-05,
"loss": 4.8826901245117185,
"step": 16000
},
{
"epoch": 0.8314769724913226,
"grad_norm": 1.1920490264892578,
"learning_rate": 5.19172472477796e-05,
"loss": 4.831457214355469,
"step": 16050
},
{
"epoch": 0.8340672434336631,
"grad_norm": 1.1483428478240967,
"learning_rate": 5.1119502207094605e-05,
"loss": 4.812444152832032,
"step": 16100
},
{
"epoch": 0.8366575143760038,
"grad_norm": 1.110986590385437,
"learning_rate": 5.032175716640961e-05,
"loss": 4.907426452636718,
"step": 16150
},
{
"epoch": 0.8392477853183443,
"grad_norm": 1.1689667701721191,
"learning_rate": 4.9524012125724614e-05,
"loss": 4.833683166503906,
"step": 16200
},
{
"epoch": 0.8418380562606849,
"grad_norm": 1.2110393047332764,
"learning_rate": 4.872626708503962e-05,
"loss": 4.851173706054688,
"step": 16250
},
{
"epoch": 0.8444283272030254,
"grad_norm": 1.1109424829483032,
"learning_rate": 4.7928522044354616e-05,
"loss": 4.8200289916992185,
"step": 16300
},
{
"epoch": 0.847018598145366,
"grad_norm": 1.2129008769989014,
"learning_rate": 4.713077700366962e-05,
"loss": 4.797307434082032,
"step": 16350
},
{
"epoch": 0.8496088690877066,
"grad_norm": 1.1530872583389282,
"learning_rate": 4.6333031962984626e-05,
"loss": 4.856131896972657,
"step": 16400
},
{
"epoch": 0.8521991400300472,
"grad_norm": 1.3170881271362305,
"learning_rate": 4.553528692229963e-05,
"loss": 4.85386474609375,
"step": 16450
},
{
"epoch": 0.8547894109723877,
"grad_norm": 1.1147844791412354,
"learning_rate": 4.4737541881614635e-05,
"loss": 4.820576171875,
"step": 16500
},
{
"epoch": 0.8573796819147282,
"grad_norm": 1.1671563386917114,
"learning_rate": 4.393979684092964e-05,
"loss": 4.826213684082031,
"step": 16550
},
{
"epoch": 0.8599699528570689,
"grad_norm": 1.1505951881408691,
"learning_rate": 4.314205180024464e-05,
"loss": 4.807721557617188,
"step": 16600
},
{
"epoch": 0.8625602237994094,
"grad_norm": 1.2463537454605103,
"learning_rate": 4.234430675955964e-05,
"loss": 4.844184875488281,
"step": 16650
},
{
"epoch": 0.86515049474175,
"grad_norm": 1.2044734954833984,
"learning_rate": 4.1546561718874646e-05,
"loss": 4.8523867797851565,
"step": 16700
},
{
"epoch": 0.8677407656840905,
"grad_norm": 1.2013986110687256,
"learning_rate": 4.074881667818965e-05,
"loss": 4.796150817871093,
"step": 16750
},
{
"epoch": 0.8703310366264311,
"grad_norm": 1.0996414422988892,
"learning_rate": 3.9951071637504655e-05,
"loss": 4.80515625,
"step": 16800
},
{
"epoch": 0.8729213075687717,
"grad_norm": 1.1789201498031616,
"learning_rate": 3.915332659681965e-05,
"loss": 4.838421936035156,
"step": 16850
},
{
"epoch": 0.8755115785111123,
"grad_norm": 1.1779522895812988,
"learning_rate": 3.835558155613466e-05,
"loss": 4.7971868896484375,
"step": 16900
},
{
"epoch": 0.8781018494534528,
"grad_norm": 1.2039135694503784,
"learning_rate": 3.755783651544966e-05,
"loss": 4.779116821289063,
"step": 16950
},
{
"epoch": 0.8806921203957934,
"grad_norm": 1.195914626121521,
"learning_rate": 3.676009147476466e-05,
"loss": 4.842753601074219,
"step": 17000
},
{
"epoch": 0.883282391338134,
"grad_norm": 1.2158962488174438,
"learning_rate": 3.5962346434079664e-05,
"loss": 4.767064514160157,
"step": 17050
},
{
"epoch": 0.8858726622804746,
"grad_norm": 1.1929750442504883,
"learning_rate": 3.516460139339467e-05,
"loss": 4.835277404785156,
"step": 17100
},
{
"epoch": 0.8884629332228151,
"grad_norm": 1.1390390396118164,
"learning_rate": 3.4366856352709673e-05,
"loss": 4.834110107421875,
"step": 17150
},
{
"epoch": 0.8910532041651557,
"grad_norm": 1.2707346677780151,
"learning_rate": 3.356911131202468e-05,
"loss": 4.800746765136719,
"step": 17200
},
{
"epoch": 0.8936434751074962,
"grad_norm": 1.2819527387619019,
"learning_rate": 3.2771366271339676e-05,
"loss": 4.803517456054688,
"step": 17250
},
{
"epoch": 0.8962337460498369,
"grad_norm": 1.130441427230835,
"learning_rate": 3.197362123065468e-05,
"loss": 4.809993591308594,
"step": 17300
},
{
"epoch": 0.8988240169921774,
"grad_norm": 1.2011135816574097,
"learning_rate": 3.117587618996968e-05,
"loss": 4.835908813476562,
"step": 17350
},
{
"epoch": 0.9014142879345179,
"grad_norm": 1.1804677248001099,
"learning_rate": 3.0378131149284686e-05,
"loss": 4.820344543457031,
"step": 17400
},
{
"epoch": 0.9040045588768585,
"grad_norm": 1.1942323446273804,
"learning_rate": 2.9580386108599687e-05,
"loss": 4.803614196777343,
"step": 17450
},
{
"epoch": 0.906594829819199,
"grad_norm": 1.215887188911438,
"learning_rate": 2.8782641067914692e-05,
"loss": 4.784419860839844,
"step": 17500
},
{
"epoch": 0.9091851007615397,
"grad_norm": 1.1484390497207642,
"learning_rate": 2.7984896027229693e-05,
"loss": 4.850234375,
"step": 17550
},
{
"epoch": 0.9117753717038802,
"grad_norm": 1.16694176197052,
"learning_rate": 2.7187150986544697e-05,
"loss": 4.777885437011719,
"step": 17600
},
{
"epoch": 0.9143656426462208,
"grad_norm": 1.17305588722229,
"learning_rate": 2.6389405945859702e-05,
"loss": 4.832542114257812,
"step": 17650
},
{
"epoch": 0.9169559135885613,
"grad_norm": 1.2073004245758057,
"learning_rate": 2.5591660905174703e-05,
"loss": 4.793171081542969,
"step": 17700
},
{
"epoch": 0.919546184530902,
"grad_norm": 1.1849514245986938,
"learning_rate": 2.4793915864489708e-05,
"loss": 4.813887634277344,
"step": 17750
},
{
"epoch": 0.9221364554732425,
"grad_norm": 1.1501691341400146,
"learning_rate": 2.3996170823804712e-05,
"loss": 4.818171081542968,
"step": 17800
},
{
"epoch": 0.9247267264155831,
"grad_norm": 1.1892966032028198,
"learning_rate": 2.3198425783119713e-05,
"loss": 4.815703125,
"step": 17850
},
{
"epoch": 0.9273169973579236,
"grad_norm": 1.225103735923767,
"learning_rate": 2.2400680742434718e-05,
"loss": 4.793248901367187,
"step": 17900
},
{
"epoch": 0.9299072683002642,
"grad_norm": 1.1672247648239136,
"learning_rate": 2.1602935701749716e-05,
"loss": 4.814731750488281,
"step": 17950
},
{
"epoch": 0.9324975392426048,
"grad_norm": 1.1436740159988403,
"learning_rate": 2.080519066106472e-05,
"loss": 4.764231262207031,
"step": 18000
},
{
"epoch": 0.9350878101849454,
"grad_norm": 1.1047722101211548,
"learning_rate": 2.0007445620379725e-05,
"loss": 4.796049499511719,
"step": 18050
},
{
"epoch": 0.9376780811272859,
"grad_norm": 1.1942051649093628,
"learning_rate": 1.9209700579694726e-05,
"loss": 4.743634033203125,
"step": 18100
},
{
"epoch": 0.9402683520696264,
"grad_norm": 1.1561431884765625,
"learning_rate": 1.841195553900973e-05,
"loss": 4.780207214355468,
"step": 18150
},
{
"epoch": 0.9428586230119671,
"grad_norm": 1.1528445482254028,
"learning_rate": 1.7614210498324735e-05,
"loss": 4.7709219360351565,
"step": 18200
},
{
"epoch": 0.9454488939543076,
"grad_norm": 1.1248111724853516,
"learning_rate": 1.6816465457639736e-05,
"loss": 4.76812744140625,
"step": 18250
},
{
"epoch": 0.9480391648966482,
"grad_norm": 1.2417064905166626,
"learning_rate": 1.601872041695474e-05,
"loss": 4.804192199707031,
"step": 18300
},
{
"epoch": 0.9506294358389887,
"grad_norm": 1.1914931535720825,
"learning_rate": 1.5220975376269744e-05,
"loss": 4.7780892944335935,
"step": 18350
},
{
"epoch": 0.9532197067813293,
"grad_norm": 1.1965242624282837,
"learning_rate": 1.4423230335584745e-05,
"loss": 4.782641296386719,
"step": 18400
},
{
"epoch": 0.9558099777236699,
"grad_norm": 1.2170450687408447,
"learning_rate": 1.3625485294899748e-05,
"loss": 4.839132690429688,
"step": 18450
},
{
"epoch": 0.9584002486660105,
"grad_norm": 1.212841272354126,
"learning_rate": 1.2827740254214752e-05,
"loss": 4.804522399902344,
"step": 18500
},
{
"epoch": 0.960990519608351,
"grad_norm": 1.2126749753952026,
"learning_rate": 1.2029995213529755e-05,
"loss": 4.806041870117188,
"step": 18550
},
{
"epoch": 0.9635807905506916,
"grad_norm": 1.1030791997909546,
"learning_rate": 1.1232250172844758e-05,
"loss": 4.794331359863281,
"step": 18600
},
{
"epoch": 0.9661710614930322,
"grad_norm": 1.1634585857391357,
"learning_rate": 1.043450513215976e-05,
"loss": 4.783181762695312,
"step": 18650
},
{
"epoch": 0.9687613324353728,
"grad_norm": 1.1583950519561768,
"learning_rate": 9.636760091474764e-06,
"loss": 4.768502197265625,
"step": 18700
},
{
"epoch": 0.9713516033777133,
"grad_norm": 1.264675498008728,
"learning_rate": 8.839015050789766e-06,
"loss": 4.739063110351562,
"step": 18750
},
{
"epoch": 0.9739418743200539,
"grad_norm": 1.1584491729736328,
"learning_rate": 8.04127001010477e-06,
"loss": 4.728269348144531,
"step": 18800
},
{
"epoch": 0.9765321452623944,
"grad_norm": 1.1771270036697388,
"learning_rate": 7.243524969419773e-06,
"loss": 4.793826293945313,
"step": 18850
},
{
"epoch": 0.9791224162047351,
"grad_norm": 1.230299711227417,
"learning_rate": 6.445779928734776e-06,
"loss": 4.7584912109375,
"step": 18900
},
{
"epoch": 0.9817126871470756,
"grad_norm": 1.1764832735061646,
"learning_rate": 5.648034888049779e-06,
"loss": 4.755926818847656,
"step": 18950
},
{
"epoch": 0.9843029580894161,
"grad_norm": 1.1985116004943848,
"learning_rate": 4.850289847364782e-06,
"loss": 4.724908142089844,
"step": 19000
},
{
"epoch": 0.9868932290317567,
"grad_norm": 1.1627134084701538,
"learning_rate": 4.0525448066797845e-06,
"loss": 4.764407043457031,
"step": 19050
},
{
"epoch": 0.9894834999740972,
"grad_norm": 1.2293567657470703,
"learning_rate": 3.2547997659947877e-06,
"loss": 4.771585693359375,
"step": 19100
},
{
"epoch": 0.9920737709164379,
"grad_norm": 1.1796555519104004,
"learning_rate": 2.4570547253097906e-06,
"loss": 4.763060302734375,
"step": 19150
},
{
"epoch": 0.9946640418587784,
"grad_norm": 1.1778082847595215,
"learning_rate": 1.6593096846247936e-06,
"loss": 4.758525085449219,
"step": 19200
},
{
"epoch": 0.997254312801119,
"grad_norm": 1.1932587623596191,
"learning_rate": 8.615646439397968e-07,
"loss": 4.812500915527344,
"step": 19250
},
{
"epoch": 0.9998445837434595,
"grad_norm": 1.1879589557647705,
"learning_rate": 6.381960325479977e-08,
"loss": 4.800540161132813,
"step": 19300
}
],
"logging_steps": 50,
"max_steps": 19303,
"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": 1.401424037216256e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}