CHIPSAL-SUBTASK-B-FewShot / trainer_state.json
arhamm40182's picture
Few Shot Prompting SFT with Gemma2-9B Model | 0.89 Weighted F1 and 0.75 Macro F1 Score on Dev Set
dcf7a13 verified
raw
history blame
55.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1563,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003198976327575176,
"grad_norm": 2.5142228603363037,
"learning_rate": 0.00019936020473448497,
"loss": 2.2007,
"step": 5
},
{
"epoch": 0.006397952655150352,
"grad_norm": 0.7006776332855225,
"learning_rate": 0.00019872040946896996,
"loss": 0.461,
"step": 10
},
{
"epoch": 0.009596928982725527,
"grad_norm": 0.6147411465644836,
"learning_rate": 0.0001980806142034549,
"loss": 0.3499,
"step": 15
},
{
"epoch": 0.012795905310300703,
"grad_norm": 0.4334537386894226,
"learning_rate": 0.00019744081893793988,
"loss": 0.3154,
"step": 20
},
{
"epoch": 0.01599488163787588,
"grad_norm": 0.5372670292854309,
"learning_rate": 0.00019680102367242483,
"loss": 0.2982,
"step": 25
},
{
"epoch": 0.019193857965451054,
"grad_norm": 0.4113191068172455,
"learning_rate": 0.0001961612284069098,
"loss": 0.2751,
"step": 30
},
{
"epoch": 0.022392834293026232,
"grad_norm": 0.3651261627674103,
"learning_rate": 0.00019552143314139478,
"loss": 0.279,
"step": 35
},
{
"epoch": 0.025591810620601407,
"grad_norm": 0.4230581820011139,
"learning_rate": 0.00019488163787587971,
"loss": 0.248,
"step": 40
},
{
"epoch": 0.028790786948176585,
"grad_norm": 0.39593610167503357,
"learning_rate": 0.0001942418426103647,
"loss": 0.2902,
"step": 45
},
{
"epoch": 0.03198976327575176,
"grad_norm": 0.41984233260154724,
"learning_rate": 0.00019360204734484966,
"loss": 0.2932,
"step": 50
},
{
"epoch": 0.035188739603326934,
"grad_norm": 0.26596859097480774,
"learning_rate": 0.00019296225207933462,
"loss": 0.2375,
"step": 55
},
{
"epoch": 0.03838771593090211,
"grad_norm": 0.44939491152763367,
"learning_rate": 0.0001923224568138196,
"loss": 0.2794,
"step": 60
},
{
"epoch": 0.04158669225847729,
"grad_norm": 0.3508070111274719,
"learning_rate": 0.00019168266154830454,
"loss": 0.2787,
"step": 65
},
{
"epoch": 0.044785668586052464,
"grad_norm": 0.40634146332740784,
"learning_rate": 0.0001910428662827895,
"loss": 0.2767,
"step": 70
},
{
"epoch": 0.04798464491362764,
"grad_norm": 0.6950023770332336,
"learning_rate": 0.00019040307101727449,
"loss": 0.2776,
"step": 75
},
{
"epoch": 0.05118362124120281,
"grad_norm": 0.34798476099967957,
"learning_rate": 0.00018976327575175944,
"loss": 0.2492,
"step": 80
},
{
"epoch": 0.05438259756877799,
"grad_norm": 0.30487826466560364,
"learning_rate": 0.00018912348048624443,
"loss": 0.2344,
"step": 85
},
{
"epoch": 0.05758157389635317,
"grad_norm": 0.2966866195201874,
"learning_rate": 0.00018848368522072936,
"loss": 0.2717,
"step": 90
},
{
"epoch": 0.060780550223928344,
"grad_norm": 0.41749754548072815,
"learning_rate": 0.00018784388995521432,
"loss": 0.2855,
"step": 95
},
{
"epoch": 0.06397952655150352,
"grad_norm": 0.3517999053001404,
"learning_rate": 0.0001872040946896993,
"loss": 0.2537,
"step": 100
},
{
"epoch": 0.0671785028790787,
"grad_norm": 0.3616477847099304,
"learning_rate": 0.00018656429942418427,
"loss": 0.2548,
"step": 105
},
{
"epoch": 0.07037747920665387,
"grad_norm": 0.3890167772769928,
"learning_rate": 0.00018592450415866926,
"loss": 0.2678,
"step": 110
},
{
"epoch": 0.07357645553422905,
"grad_norm": 0.35902726650238037,
"learning_rate": 0.0001852847088931542,
"loss": 0.2751,
"step": 115
},
{
"epoch": 0.07677543186180422,
"grad_norm": 0.3401743471622467,
"learning_rate": 0.00018464491362763915,
"loss": 0.2696,
"step": 120
},
{
"epoch": 0.0799744081893794,
"grad_norm": 0.2922796905040741,
"learning_rate": 0.00018400511836212414,
"loss": 0.2326,
"step": 125
},
{
"epoch": 0.08317338451695458,
"grad_norm": 0.3094709515571594,
"learning_rate": 0.0001833653230966091,
"loss": 0.2393,
"step": 130
},
{
"epoch": 0.08637236084452975,
"grad_norm": 0.2731708884239197,
"learning_rate": 0.00018272552783109406,
"loss": 0.2446,
"step": 135
},
{
"epoch": 0.08957133717210493,
"grad_norm": 0.281548410654068,
"learning_rate": 0.00018208573256557901,
"loss": 0.2739,
"step": 140
},
{
"epoch": 0.0927703134996801,
"grad_norm": 0.31158748269081116,
"learning_rate": 0.00018144593730006397,
"loss": 0.2562,
"step": 145
},
{
"epoch": 0.09596928982725528,
"grad_norm": 0.31344011425971985,
"learning_rate": 0.00018080614203454896,
"loss": 0.2457,
"step": 150
},
{
"epoch": 0.09916826615483046,
"grad_norm": 0.25769639015197754,
"learning_rate": 0.00018016634676903392,
"loss": 0.2413,
"step": 155
},
{
"epoch": 0.10236724248240563,
"grad_norm": 0.3215169310569763,
"learning_rate": 0.00017952655150351888,
"loss": 0.2587,
"step": 160
},
{
"epoch": 0.10556621880998081,
"grad_norm": 0.31406131386756897,
"learning_rate": 0.00017888675623800384,
"loss": 0.252,
"step": 165
},
{
"epoch": 0.10876519513755598,
"grad_norm": 0.2920863926410675,
"learning_rate": 0.0001782469609724888,
"loss": 0.2469,
"step": 170
},
{
"epoch": 0.11196417146513116,
"grad_norm": 0.32624512910842896,
"learning_rate": 0.00017760716570697379,
"loss": 0.2543,
"step": 175
},
{
"epoch": 0.11516314779270634,
"grad_norm": 0.3013479709625244,
"learning_rate": 0.00017696737044145875,
"loss": 0.2662,
"step": 180
},
{
"epoch": 0.1183621241202815,
"grad_norm": 0.2553490102291107,
"learning_rate": 0.0001763275751759437,
"loss": 0.2479,
"step": 185
},
{
"epoch": 0.12156110044785669,
"grad_norm": 0.3384646773338318,
"learning_rate": 0.00017568777991042867,
"loss": 0.2979,
"step": 190
},
{
"epoch": 0.12476007677543186,
"grad_norm": 0.4001105725765228,
"learning_rate": 0.00017504798464491362,
"loss": 0.2629,
"step": 195
},
{
"epoch": 0.12795905310300704,
"grad_norm": 0.3194480836391449,
"learning_rate": 0.0001744081893793986,
"loss": 0.2489,
"step": 200
},
{
"epoch": 0.13115802943058222,
"grad_norm": 0.38252341747283936,
"learning_rate": 0.00017376839411388357,
"loss": 0.2356,
"step": 205
},
{
"epoch": 0.1343570057581574,
"grad_norm": 0.31561562418937683,
"learning_rate": 0.00017312859884836853,
"loss": 0.2581,
"step": 210
},
{
"epoch": 0.13755598208573255,
"grad_norm": 0.2712710499763489,
"learning_rate": 0.0001724888035828535,
"loss": 0.2577,
"step": 215
},
{
"epoch": 0.14075495841330773,
"grad_norm": 0.3084522485733032,
"learning_rate": 0.00017184900831733845,
"loss": 0.2304,
"step": 220
},
{
"epoch": 0.14395393474088292,
"grad_norm": 0.2699475884437561,
"learning_rate": 0.00017120921305182344,
"loss": 0.2685,
"step": 225
},
{
"epoch": 0.1471529110684581,
"grad_norm": 0.3236371874809265,
"learning_rate": 0.0001705694177863084,
"loss": 0.2588,
"step": 230
},
{
"epoch": 0.15035188739603328,
"grad_norm": 0.28885379433631897,
"learning_rate": 0.00016992962252079336,
"loss": 0.2574,
"step": 235
},
{
"epoch": 0.15355086372360843,
"grad_norm": 0.3263382315635681,
"learning_rate": 0.00016928982725527832,
"loss": 0.2649,
"step": 240
},
{
"epoch": 0.15674984005118361,
"grad_norm": 0.2647656500339508,
"learning_rate": 0.00016865003198976328,
"loss": 0.2522,
"step": 245
},
{
"epoch": 0.1599488163787588,
"grad_norm": 0.2796100676059723,
"learning_rate": 0.00016801023672424826,
"loss": 0.2255,
"step": 250
},
{
"epoch": 0.16314779270633398,
"grad_norm": 0.32718002796173096,
"learning_rate": 0.00016737044145873322,
"loss": 0.2623,
"step": 255
},
{
"epoch": 0.16634676903390916,
"grad_norm": 0.34518498182296753,
"learning_rate": 0.00016673064619321818,
"loss": 0.2452,
"step": 260
},
{
"epoch": 0.1695457453614843,
"grad_norm": 0.3195638358592987,
"learning_rate": 0.00016609085092770314,
"loss": 0.2638,
"step": 265
},
{
"epoch": 0.1727447216890595,
"grad_norm": 0.2617906928062439,
"learning_rate": 0.0001654510556621881,
"loss": 0.2294,
"step": 270
},
{
"epoch": 0.17594369801663468,
"grad_norm": 0.28182393312454224,
"learning_rate": 0.00016481126039667306,
"loss": 0.2611,
"step": 275
},
{
"epoch": 0.17914267434420986,
"grad_norm": 0.24791140854358673,
"learning_rate": 0.00016417146513115805,
"loss": 0.2178,
"step": 280
},
{
"epoch": 0.18234165067178504,
"grad_norm": 0.2434699386358261,
"learning_rate": 0.000163531669865643,
"loss": 0.2396,
"step": 285
},
{
"epoch": 0.1855406269993602,
"grad_norm": 0.3352366089820862,
"learning_rate": 0.00016289187460012797,
"loss": 0.2368,
"step": 290
},
{
"epoch": 0.18873960332693537,
"grad_norm": 0.32391369342803955,
"learning_rate": 0.00016225207933461293,
"loss": 0.2575,
"step": 295
},
{
"epoch": 0.19193857965451055,
"grad_norm": 0.33067309856414795,
"learning_rate": 0.00016161228406909789,
"loss": 0.2726,
"step": 300
},
{
"epoch": 0.19513755598208574,
"grad_norm": 0.3429821729660034,
"learning_rate": 0.00016097248880358287,
"loss": 0.2507,
"step": 305
},
{
"epoch": 0.19833653230966092,
"grad_norm": 0.3166626989841461,
"learning_rate": 0.00016033269353806783,
"loss": 0.2474,
"step": 310
},
{
"epoch": 0.20153550863723607,
"grad_norm": 0.2972988784313202,
"learning_rate": 0.0001596928982725528,
"loss": 0.2361,
"step": 315
},
{
"epoch": 0.20473448496481125,
"grad_norm": 0.2758512794971466,
"learning_rate": 0.00015905310300703775,
"loss": 0.2266,
"step": 320
},
{
"epoch": 0.20793346129238643,
"grad_norm": 0.2929127812385559,
"learning_rate": 0.0001584133077415227,
"loss": 0.2514,
"step": 325
},
{
"epoch": 0.21113243761996162,
"grad_norm": 0.3643186092376709,
"learning_rate": 0.0001577735124760077,
"loss": 0.2628,
"step": 330
},
{
"epoch": 0.2143314139475368,
"grad_norm": 0.3119208514690399,
"learning_rate": 0.00015713371721049266,
"loss": 0.2446,
"step": 335
},
{
"epoch": 0.21753039027511195,
"grad_norm": 0.2954186201095581,
"learning_rate": 0.00015649392194497762,
"loss": 0.248,
"step": 340
},
{
"epoch": 0.22072936660268713,
"grad_norm": 0.28279921412467957,
"learning_rate": 0.00015585412667946258,
"loss": 0.2256,
"step": 345
},
{
"epoch": 0.22392834293026231,
"grad_norm": 0.28081214427948,
"learning_rate": 0.00015521433141394754,
"loss": 0.2481,
"step": 350
},
{
"epoch": 0.2271273192578375,
"grad_norm": 0.36415377259254456,
"learning_rate": 0.00015457453614843252,
"loss": 0.2577,
"step": 355
},
{
"epoch": 0.23032629558541268,
"grad_norm": 0.3519168794155121,
"learning_rate": 0.00015393474088291748,
"loss": 0.2504,
"step": 360
},
{
"epoch": 0.23352527191298783,
"grad_norm": 0.3644118309020996,
"learning_rate": 0.00015329494561740244,
"loss": 0.2453,
"step": 365
},
{
"epoch": 0.236724248240563,
"grad_norm": 0.2938327491283417,
"learning_rate": 0.0001526551503518874,
"loss": 0.2378,
"step": 370
},
{
"epoch": 0.2399232245681382,
"grad_norm": 0.25993379950523376,
"learning_rate": 0.00015201535508637236,
"loss": 0.2387,
"step": 375
},
{
"epoch": 0.24312220089571338,
"grad_norm": 0.2894437313079834,
"learning_rate": 0.00015137555982085735,
"loss": 0.2524,
"step": 380
},
{
"epoch": 0.24632117722328856,
"grad_norm": 0.29153236746788025,
"learning_rate": 0.0001507357645553423,
"loss": 0.2351,
"step": 385
},
{
"epoch": 0.2495201535508637,
"grad_norm": 0.2899223268032074,
"learning_rate": 0.00015009596928982727,
"loss": 0.2493,
"step": 390
},
{
"epoch": 0.2527191298784389,
"grad_norm": 0.27078118920326233,
"learning_rate": 0.00014945617402431223,
"loss": 0.2543,
"step": 395
},
{
"epoch": 0.2559181062060141,
"grad_norm": 0.23355244100093842,
"learning_rate": 0.0001488163787587972,
"loss": 0.232,
"step": 400
},
{
"epoch": 0.2591170825335892,
"grad_norm": 0.29392296075820923,
"learning_rate": 0.00014817658349328217,
"loss": 0.2356,
"step": 405
},
{
"epoch": 0.26231605886116444,
"grad_norm": 0.3395833969116211,
"learning_rate": 0.00014753678822776713,
"loss": 0.2498,
"step": 410
},
{
"epoch": 0.2655150351887396,
"grad_norm": 0.2617484927177429,
"learning_rate": 0.00014689699296225207,
"loss": 0.2239,
"step": 415
},
{
"epoch": 0.2687140115163148,
"grad_norm": 0.29424986243247986,
"learning_rate": 0.00014625719769673705,
"loss": 0.2602,
"step": 420
},
{
"epoch": 0.27191298784388995,
"grad_norm": 0.251095175743103,
"learning_rate": 0.000145617402431222,
"loss": 0.2113,
"step": 425
},
{
"epoch": 0.2751119641714651,
"grad_norm": 0.31057092547416687,
"learning_rate": 0.000144977607165707,
"loss": 0.2388,
"step": 430
},
{
"epoch": 0.2783109404990403,
"grad_norm": 0.23010189831256866,
"learning_rate": 0.00014433781190019196,
"loss": 0.2294,
"step": 435
},
{
"epoch": 0.28150991682661547,
"grad_norm": 0.28975558280944824,
"learning_rate": 0.0001436980166346769,
"loss": 0.2433,
"step": 440
},
{
"epoch": 0.2847088931541907,
"grad_norm": 0.306234747171402,
"learning_rate": 0.00014305822136916188,
"loss": 0.2715,
"step": 445
},
{
"epoch": 0.28790786948176583,
"grad_norm": 0.3650927245616913,
"learning_rate": 0.00014241842610364684,
"loss": 0.2577,
"step": 450
},
{
"epoch": 0.291106845809341,
"grad_norm": 0.3464455306529999,
"learning_rate": 0.0001417786308381318,
"loss": 0.2514,
"step": 455
},
{
"epoch": 0.2943058221369162,
"grad_norm": 0.3354053199291229,
"learning_rate": 0.00014113883557261678,
"loss": 0.2479,
"step": 460
},
{
"epoch": 0.29750479846449135,
"grad_norm": 0.27853867411613464,
"learning_rate": 0.00014049904030710172,
"loss": 0.2454,
"step": 465
},
{
"epoch": 0.30070377479206656,
"grad_norm": 0.29874446988105774,
"learning_rate": 0.0001398592450415867,
"loss": 0.2271,
"step": 470
},
{
"epoch": 0.3039027511196417,
"grad_norm": 0.30210456252098083,
"learning_rate": 0.00013921944977607166,
"loss": 0.2532,
"step": 475
},
{
"epoch": 0.30710172744721687,
"grad_norm": 0.2790578305721283,
"learning_rate": 0.00013857965451055662,
"loss": 0.2254,
"step": 480
},
{
"epoch": 0.3103007037747921,
"grad_norm": 0.28601694107055664,
"learning_rate": 0.0001379398592450416,
"loss": 0.2396,
"step": 485
},
{
"epoch": 0.31349968010236723,
"grad_norm": 0.2820374667644501,
"learning_rate": 0.00013730006397952654,
"loss": 0.2405,
"step": 490
},
{
"epoch": 0.31669865642994244,
"grad_norm": 0.48833322525024414,
"learning_rate": 0.00013666026871401153,
"loss": 0.2405,
"step": 495
},
{
"epoch": 0.3198976327575176,
"grad_norm": 0.30562731623649597,
"learning_rate": 0.0001360204734484965,
"loss": 0.2486,
"step": 500
},
{
"epoch": 0.32309660908509275,
"grad_norm": 0.43254855275154114,
"learning_rate": 0.00013538067818298145,
"loss": 0.2541,
"step": 505
},
{
"epoch": 0.32629558541266795,
"grad_norm": 0.29583895206451416,
"learning_rate": 0.00013474088291746643,
"loss": 0.2442,
"step": 510
},
{
"epoch": 0.3294945617402431,
"grad_norm": 0.3376995325088501,
"learning_rate": 0.00013410108765195137,
"loss": 0.2484,
"step": 515
},
{
"epoch": 0.3326935380678183,
"grad_norm": 0.36025670170783997,
"learning_rate": 0.00013346129238643635,
"loss": 0.2506,
"step": 520
},
{
"epoch": 0.33589251439539347,
"grad_norm": 0.33433622121810913,
"learning_rate": 0.0001328214971209213,
"loss": 0.2201,
"step": 525
},
{
"epoch": 0.3390914907229686,
"grad_norm": 0.2948063910007477,
"learning_rate": 0.00013218170185540627,
"loss": 0.2192,
"step": 530
},
{
"epoch": 0.34229046705054383,
"grad_norm": 0.3154048025608063,
"learning_rate": 0.00013154190658989126,
"loss": 0.2563,
"step": 535
},
{
"epoch": 0.345489443378119,
"grad_norm": 0.3658270537853241,
"learning_rate": 0.0001309021113243762,
"loss": 0.262,
"step": 540
},
{
"epoch": 0.3486884197056942,
"grad_norm": 0.287977010011673,
"learning_rate": 0.00013026231605886118,
"loss": 0.2516,
"step": 545
},
{
"epoch": 0.35188739603326935,
"grad_norm": 0.32740944623947144,
"learning_rate": 0.00012962252079334614,
"loss": 0.2598,
"step": 550
},
{
"epoch": 0.3550863723608445,
"grad_norm": 0.3204440772533417,
"learning_rate": 0.0001289827255278311,
"loss": 0.2476,
"step": 555
},
{
"epoch": 0.3582853486884197,
"grad_norm": 0.2696886360645294,
"learning_rate": 0.00012834293026231608,
"loss": 0.2135,
"step": 560
},
{
"epoch": 0.36148432501599487,
"grad_norm": 0.24359866976737976,
"learning_rate": 0.00012770313499680102,
"loss": 0.2182,
"step": 565
},
{
"epoch": 0.3646833013435701,
"grad_norm": 0.2711159586906433,
"learning_rate": 0.000127063339731286,
"loss": 0.2381,
"step": 570
},
{
"epoch": 0.36788227767114523,
"grad_norm": 0.3088955283164978,
"learning_rate": 0.00012642354446577096,
"loss": 0.1968,
"step": 575
},
{
"epoch": 0.3710812539987204,
"grad_norm": 0.28893983364105225,
"learning_rate": 0.00012578374920025592,
"loss": 0.2539,
"step": 580
},
{
"epoch": 0.3742802303262956,
"grad_norm": 0.30569225549697876,
"learning_rate": 0.0001251439539347409,
"loss": 0.2491,
"step": 585
},
{
"epoch": 0.37747920665387075,
"grad_norm": 0.29621535539627075,
"learning_rate": 0.00012450415866922584,
"loss": 0.2148,
"step": 590
},
{
"epoch": 0.38067818298144596,
"grad_norm": 0.32468584179878235,
"learning_rate": 0.0001238643634037108,
"loss": 0.2395,
"step": 595
},
{
"epoch": 0.3838771593090211,
"grad_norm": 0.3553343415260315,
"learning_rate": 0.0001232245681381958,
"loss": 0.2083,
"step": 600
},
{
"epoch": 0.38707613563659626,
"grad_norm": 0.3482455611228943,
"learning_rate": 0.00012258477287268075,
"loss": 0.2406,
"step": 605
},
{
"epoch": 0.3902751119641715,
"grad_norm": 0.2828838527202606,
"learning_rate": 0.00012194497760716572,
"loss": 0.2455,
"step": 610
},
{
"epoch": 0.3934740882917466,
"grad_norm": 0.18922460079193115,
"learning_rate": 0.00012130518234165067,
"loss": 0.2402,
"step": 615
},
{
"epoch": 0.39667306461932184,
"grad_norm": 0.36727848649024963,
"learning_rate": 0.00012066538707613564,
"loss": 0.2348,
"step": 620
},
{
"epoch": 0.399872040946897,
"grad_norm": 0.29675740003585815,
"learning_rate": 0.00012002559181062061,
"loss": 0.2314,
"step": 625
},
{
"epoch": 0.40307101727447214,
"grad_norm": 0.3717605471611023,
"learning_rate": 0.00011938579654510557,
"loss": 0.2252,
"step": 630
},
{
"epoch": 0.40626999360204735,
"grad_norm": 0.3321962356567383,
"learning_rate": 0.00011874600127959055,
"loss": 0.2004,
"step": 635
},
{
"epoch": 0.4094689699296225,
"grad_norm": 0.2975044846534729,
"learning_rate": 0.00011810620601407549,
"loss": 0.2394,
"step": 640
},
{
"epoch": 0.4126679462571977,
"grad_norm": 0.5578257441520691,
"learning_rate": 0.00011746641074856047,
"loss": 0.2731,
"step": 645
},
{
"epoch": 0.41586692258477287,
"grad_norm": 0.3975297212600708,
"learning_rate": 0.00011682661548304543,
"loss": 0.2558,
"step": 650
},
{
"epoch": 0.419065898912348,
"grad_norm": 0.25277817249298096,
"learning_rate": 0.0001161868202175304,
"loss": 0.2312,
"step": 655
},
{
"epoch": 0.42226487523992323,
"grad_norm": 0.32298269867897034,
"learning_rate": 0.00011554702495201537,
"loss": 0.2448,
"step": 660
},
{
"epoch": 0.4254638515674984,
"grad_norm": 0.4133547246456146,
"learning_rate": 0.00011490722968650032,
"loss": 0.2408,
"step": 665
},
{
"epoch": 0.4286628278950736,
"grad_norm": 0.2923837900161743,
"learning_rate": 0.00011426743442098529,
"loss": 0.2268,
"step": 670
},
{
"epoch": 0.43186180422264875,
"grad_norm": 0.2599218785762787,
"learning_rate": 0.00011362763915547025,
"loss": 0.2255,
"step": 675
},
{
"epoch": 0.4350607805502239,
"grad_norm": 0.3489529490470886,
"learning_rate": 0.00011298784388995522,
"loss": 0.254,
"step": 680
},
{
"epoch": 0.4382597568777991,
"grad_norm": 0.30105629563331604,
"learning_rate": 0.0001123480486244402,
"loss": 0.269,
"step": 685
},
{
"epoch": 0.44145873320537427,
"grad_norm": 0.37994736433029175,
"learning_rate": 0.00011170825335892514,
"loss": 0.2191,
"step": 690
},
{
"epoch": 0.4446577095329495,
"grad_norm": 0.35344600677490234,
"learning_rate": 0.00011106845809341012,
"loss": 0.2566,
"step": 695
},
{
"epoch": 0.44785668586052463,
"grad_norm": 0.28387364745140076,
"learning_rate": 0.00011042866282789508,
"loss": 0.2277,
"step": 700
},
{
"epoch": 0.4510556621880998,
"grad_norm": 0.36002182960510254,
"learning_rate": 0.00010978886756238005,
"loss": 0.2447,
"step": 705
},
{
"epoch": 0.454254638515675,
"grad_norm": 0.30156582593917847,
"learning_rate": 0.00010914907229686502,
"loss": 0.234,
"step": 710
},
{
"epoch": 0.45745361484325014,
"grad_norm": 0.3468836843967438,
"learning_rate": 0.00010850927703134997,
"loss": 0.2466,
"step": 715
},
{
"epoch": 0.46065259117082535,
"grad_norm": 0.33030760288238525,
"learning_rate": 0.00010786948176583493,
"loss": 0.2745,
"step": 720
},
{
"epoch": 0.4638515674984005,
"grad_norm": 0.3344646096229553,
"learning_rate": 0.0001072296865003199,
"loss": 0.2154,
"step": 725
},
{
"epoch": 0.46705054382597566,
"grad_norm": 0.40286871790885925,
"learning_rate": 0.00010658989123480487,
"loss": 0.2227,
"step": 730
},
{
"epoch": 0.47024952015355087,
"grad_norm": 0.2880500257015228,
"learning_rate": 0.00010595009596928985,
"loss": 0.2219,
"step": 735
},
{
"epoch": 0.473448496481126,
"grad_norm": 0.28410083055496216,
"learning_rate": 0.0001053103007037748,
"loss": 0.2124,
"step": 740
},
{
"epoch": 0.47664747280870123,
"grad_norm": 0.2937108278274536,
"learning_rate": 0.00010467050543825975,
"loss": 0.2272,
"step": 745
},
{
"epoch": 0.4798464491362764,
"grad_norm": 0.33740442991256714,
"learning_rate": 0.00010403071017274473,
"loss": 0.2332,
"step": 750
},
{
"epoch": 0.48304542546385154,
"grad_norm": 0.2969924807548523,
"learning_rate": 0.0001033909149072297,
"loss": 0.2382,
"step": 755
},
{
"epoch": 0.48624440179142675,
"grad_norm": 0.25243079662323,
"learning_rate": 0.00010275111964171466,
"loss": 0.2266,
"step": 760
},
{
"epoch": 0.4894433781190019,
"grad_norm": 0.30554574728012085,
"learning_rate": 0.00010211132437619962,
"loss": 0.2276,
"step": 765
},
{
"epoch": 0.4926423544465771,
"grad_norm": 0.281543493270874,
"learning_rate": 0.00010147152911068458,
"loss": 0.2286,
"step": 770
},
{
"epoch": 0.49584133077415227,
"grad_norm": 0.2906375229358673,
"learning_rate": 0.00010083173384516955,
"loss": 0.2446,
"step": 775
},
{
"epoch": 0.4990403071017274,
"grad_norm": 0.27951693534851074,
"learning_rate": 0.00010019193857965453,
"loss": 0.2422,
"step": 780
},
{
"epoch": 0.5022392834293026,
"grad_norm": 0.3846909999847412,
"learning_rate": 9.955214331413948e-05,
"loss": 0.238,
"step": 785
},
{
"epoch": 0.5054382597568778,
"grad_norm": 0.27699944376945496,
"learning_rate": 9.891234804862444e-05,
"loss": 0.2299,
"step": 790
},
{
"epoch": 0.508637236084453,
"grad_norm": 0.2875959575176239,
"learning_rate": 9.82725527831094e-05,
"loss": 0.2293,
"step": 795
},
{
"epoch": 0.5118362124120281,
"grad_norm": 0.23799441754817963,
"learning_rate": 9.763275751759438e-05,
"loss": 0.2332,
"step": 800
},
{
"epoch": 0.5150351887396033,
"grad_norm": 0.3258192241191864,
"learning_rate": 9.699296225207935e-05,
"loss": 0.2137,
"step": 805
},
{
"epoch": 0.5182341650671785,
"grad_norm": 0.4027354121208191,
"learning_rate": 9.63531669865643e-05,
"loss": 0.2528,
"step": 810
},
{
"epoch": 0.5214331413947537,
"grad_norm": 0.5166566967964172,
"learning_rate": 9.571337172104927e-05,
"loss": 0.2171,
"step": 815
},
{
"epoch": 0.5246321177223289,
"grad_norm": 0.2721567451953888,
"learning_rate": 9.507357645553423e-05,
"loss": 0.2233,
"step": 820
},
{
"epoch": 0.527831094049904,
"grad_norm": 0.295593798160553,
"learning_rate": 9.44337811900192e-05,
"loss": 0.2302,
"step": 825
},
{
"epoch": 0.5310300703774792,
"grad_norm": 0.3672064244747162,
"learning_rate": 9.379398592450416e-05,
"loss": 0.225,
"step": 830
},
{
"epoch": 0.5342290467050543,
"grad_norm": 0.27494198083877563,
"learning_rate": 9.315419065898912e-05,
"loss": 0.2221,
"step": 835
},
{
"epoch": 0.5374280230326296,
"grad_norm": 0.4797211289405823,
"learning_rate": 9.25143953934741e-05,
"loss": 0.2439,
"step": 840
},
{
"epoch": 0.5406269993602048,
"grad_norm": 0.2528781592845917,
"learning_rate": 9.187460012795905e-05,
"loss": 0.2267,
"step": 845
},
{
"epoch": 0.5438259756877799,
"grad_norm": 0.435768723487854,
"learning_rate": 9.123480486244403e-05,
"loss": 0.2433,
"step": 850
},
{
"epoch": 0.5470249520153551,
"grad_norm": 0.3093527853488922,
"learning_rate": 9.059500959692899e-05,
"loss": 0.2519,
"step": 855
},
{
"epoch": 0.5502239283429302,
"grad_norm": 0.31521865725517273,
"learning_rate": 8.995521433141395e-05,
"loss": 0.2563,
"step": 860
},
{
"epoch": 0.5534229046705055,
"grad_norm": 0.2934703230857849,
"learning_rate": 8.931541906589892e-05,
"loss": 0.2576,
"step": 865
},
{
"epoch": 0.5566218809980806,
"grad_norm": 0.33386310935020447,
"learning_rate": 8.867562380038388e-05,
"loss": 0.2342,
"step": 870
},
{
"epoch": 0.5598208573256558,
"grad_norm": 0.3183169960975647,
"learning_rate": 8.803582853486885e-05,
"loss": 0.2325,
"step": 875
},
{
"epoch": 0.5630198336532309,
"grad_norm": 0.27800726890563965,
"learning_rate": 8.739603326935381e-05,
"loss": 0.2256,
"step": 880
},
{
"epoch": 0.5662188099808061,
"grad_norm": 0.31215038895606995,
"learning_rate": 8.675623800383877e-05,
"loss": 0.206,
"step": 885
},
{
"epoch": 0.5694177863083814,
"grad_norm": 0.31119444966316223,
"learning_rate": 8.611644273832375e-05,
"loss": 0.2447,
"step": 890
},
{
"epoch": 0.5726167626359565,
"grad_norm": 0.27355003356933594,
"learning_rate": 8.54766474728087e-05,
"loss": 0.2001,
"step": 895
},
{
"epoch": 0.5758157389635317,
"grad_norm": 0.3279555141925812,
"learning_rate": 8.483685220729366e-05,
"loss": 0.218,
"step": 900
},
{
"epoch": 0.5790147152911068,
"grad_norm": 0.33151063323020935,
"learning_rate": 8.419705694177864e-05,
"loss": 0.2206,
"step": 905
},
{
"epoch": 0.582213691618682,
"grad_norm": 0.34513482451438904,
"learning_rate": 8.35572616762636e-05,
"loss": 0.2615,
"step": 910
},
{
"epoch": 0.5854126679462572,
"grad_norm": 0.38724973797798157,
"learning_rate": 8.291746641074857e-05,
"loss": 0.2445,
"step": 915
},
{
"epoch": 0.5886116442738324,
"grad_norm": 0.30885693430900574,
"learning_rate": 8.227767114523353e-05,
"loss": 0.2131,
"step": 920
},
{
"epoch": 0.5918106206014075,
"grad_norm": 0.3177817463874817,
"learning_rate": 8.163787587971849e-05,
"loss": 0.2382,
"step": 925
},
{
"epoch": 0.5950095969289827,
"grad_norm": 0.2687653601169586,
"learning_rate": 8.099808061420346e-05,
"loss": 0.2352,
"step": 930
},
{
"epoch": 0.5982085732565579,
"grad_norm": 0.23862284421920776,
"learning_rate": 8.035828534868842e-05,
"loss": 0.2073,
"step": 935
},
{
"epoch": 0.6014075495841331,
"grad_norm": 0.39008885622024536,
"learning_rate": 7.97184900831734e-05,
"loss": 0.235,
"step": 940
},
{
"epoch": 0.6046065259117083,
"grad_norm": 0.310867041349411,
"learning_rate": 7.907869481765836e-05,
"loss": 0.222,
"step": 945
},
{
"epoch": 0.6078055022392834,
"grad_norm": 0.306083083152771,
"learning_rate": 7.843889955214332e-05,
"loss": 0.2255,
"step": 950
},
{
"epoch": 0.6110044785668586,
"grad_norm": 0.3439415991306305,
"learning_rate": 7.779910428662829e-05,
"loss": 0.2124,
"step": 955
},
{
"epoch": 0.6142034548944337,
"grad_norm": 0.2899746000766754,
"learning_rate": 7.715930902111325e-05,
"loss": 0.1934,
"step": 960
},
{
"epoch": 0.617402431222009,
"grad_norm": 0.3107975721359253,
"learning_rate": 7.651951375559822e-05,
"loss": 0.2277,
"step": 965
},
{
"epoch": 0.6206014075495841,
"grad_norm": 0.3492138385772705,
"learning_rate": 7.587971849008317e-05,
"loss": 0.2413,
"step": 970
},
{
"epoch": 0.6238003838771593,
"grad_norm": 0.27954694628715515,
"learning_rate": 7.523992322456814e-05,
"loss": 0.2164,
"step": 975
},
{
"epoch": 0.6269993602047345,
"grad_norm": 0.34748780727386475,
"learning_rate": 7.46001279590531e-05,
"loss": 0.2258,
"step": 980
},
{
"epoch": 0.6301983365323096,
"grad_norm": 0.35154399275779724,
"learning_rate": 7.396033269353807e-05,
"loss": 0.2175,
"step": 985
},
{
"epoch": 0.6333973128598849,
"grad_norm": 0.3060106039047241,
"learning_rate": 7.332053742802303e-05,
"loss": 0.2318,
"step": 990
},
{
"epoch": 0.63659628918746,
"grad_norm": 0.31797581911087036,
"learning_rate": 7.268074216250799e-05,
"loss": 0.2027,
"step": 995
},
{
"epoch": 0.6397952655150352,
"grad_norm": 0.33520376682281494,
"learning_rate": 7.204094689699297e-05,
"loss": 0.2312,
"step": 1000
},
{
"epoch": 0.6429942418426103,
"grad_norm": 0.21483170986175537,
"learning_rate": 7.140115163147793e-05,
"loss": 0.2058,
"step": 1005
},
{
"epoch": 0.6461932181701855,
"grad_norm": 0.33610713481903076,
"learning_rate": 7.07613563659629e-05,
"loss": 0.2415,
"step": 1010
},
{
"epoch": 0.6493921944977608,
"grad_norm": 0.26923608779907227,
"learning_rate": 7.012156110044786e-05,
"loss": 0.2208,
"step": 1015
},
{
"epoch": 0.6525911708253359,
"grad_norm": 0.3274904489517212,
"learning_rate": 6.948176583493282e-05,
"loss": 0.212,
"step": 1020
},
{
"epoch": 0.6557901471529111,
"grad_norm": 0.3245833218097687,
"learning_rate": 6.884197056941779e-05,
"loss": 0.2191,
"step": 1025
},
{
"epoch": 0.6589891234804862,
"grad_norm": 0.27330532670021057,
"learning_rate": 6.820217530390275e-05,
"loss": 0.1992,
"step": 1030
},
{
"epoch": 0.6621880998080614,
"grad_norm": 0.25868290662765503,
"learning_rate": 6.756238003838772e-05,
"loss": 0.2399,
"step": 1035
},
{
"epoch": 0.6653870761356366,
"grad_norm": 0.30688875913619995,
"learning_rate": 6.692258477287268e-05,
"loss": 0.2342,
"step": 1040
},
{
"epoch": 0.6685860524632118,
"grad_norm": 0.2929762899875641,
"learning_rate": 6.628278950735764e-05,
"loss": 0.2369,
"step": 1045
},
{
"epoch": 0.6717850287907869,
"grad_norm": 0.31221550703048706,
"learning_rate": 6.564299424184262e-05,
"loss": 0.2224,
"step": 1050
},
{
"epoch": 0.6749840051183621,
"grad_norm": 0.32099705934524536,
"learning_rate": 6.500319897632758e-05,
"loss": 0.2144,
"step": 1055
},
{
"epoch": 0.6781829814459372,
"grad_norm": 0.2826564908027649,
"learning_rate": 6.436340371081254e-05,
"loss": 0.2271,
"step": 1060
},
{
"epoch": 0.6813819577735125,
"grad_norm": 0.2807115316390991,
"learning_rate": 6.372360844529751e-05,
"loss": 0.202,
"step": 1065
},
{
"epoch": 0.6845809341010877,
"grad_norm": 0.3093265891075134,
"learning_rate": 6.308381317978247e-05,
"loss": 0.2328,
"step": 1070
},
{
"epoch": 0.6877799104286628,
"grad_norm": 0.3201155662536621,
"learning_rate": 6.244401791426744e-05,
"loss": 0.209,
"step": 1075
},
{
"epoch": 0.690978886756238,
"grad_norm": 0.27004414796829224,
"learning_rate": 6.18042226487524e-05,
"loss": 0.2278,
"step": 1080
},
{
"epoch": 0.6941778630838131,
"grad_norm": 0.27909284830093384,
"learning_rate": 6.116442738323736e-05,
"loss": 0.23,
"step": 1085
},
{
"epoch": 0.6973768394113884,
"grad_norm": 0.3248424828052521,
"learning_rate": 6.0524632117722334e-05,
"loss": 0.2309,
"step": 1090
},
{
"epoch": 0.7005758157389635,
"grad_norm": 0.31023091077804565,
"learning_rate": 5.9884836852207293e-05,
"loss": 0.2084,
"step": 1095
},
{
"epoch": 0.7037747920665387,
"grad_norm": 0.302664190530777,
"learning_rate": 5.924504158669226e-05,
"loss": 0.2367,
"step": 1100
},
{
"epoch": 0.7069737683941139,
"grad_norm": 0.328142374753952,
"learning_rate": 5.860524632117722e-05,
"loss": 0.2215,
"step": 1105
},
{
"epoch": 0.710172744721689,
"grad_norm": 0.34118109941482544,
"learning_rate": 5.796545105566219e-05,
"loss": 0.2077,
"step": 1110
},
{
"epoch": 0.7133717210492643,
"grad_norm": 0.2934885025024414,
"learning_rate": 5.732565579014716e-05,
"loss": 0.2226,
"step": 1115
},
{
"epoch": 0.7165706973768394,
"grad_norm": 0.306045800447464,
"learning_rate": 5.668586052463212e-05,
"loss": 0.1962,
"step": 1120
},
{
"epoch": 0.7197696737044146,
"grad_norm": 0.3221231997013092,
"learning_rate": 5.6046065259117085e-05,
"loss": 0.2097,
"step": 1125
},
{
"epoch": 0.7229686500319897,
"grad_norm": 0.29269224405288696,
"learning_rate": 5.5406269993602045e-05,
"loss": 0.2147,
"step": 1130
},
{
"epoch": 0.7261676263595649,
"grad_norm": 0.3249344229698181,
"learning_rate": 5.476647472808701e-05,
"loss": 0.1956,
"step": 1135
},
{
"epoch": 0.7293666026871402,
"grad_norm": 0.3102353811264038,
"learning_rate": 5.4126679462571984e-05,
"loss": 0.2092,
"step": 1140
},
{
"epoch": 0.7325655790147153,
"grad_norm": 0.38312363624572754,
"learning_rate": 5.3486884197056944e-05,
"loss": 0.2294,
"step": 1145
},
{
"epoch": 0.7357645553422905,
"grad_norm": 0.33595603704452515,
"learning_rate": 5.284708893154191e-05,
"loss": 0.242,
"step": 1150
},
{
"epoch": 0.7389635316698656,
"grad_norm": 0.29311510920524597,
"learning_rate": 5.220729366602687e-05,
"loss": 0.2212,
"step": 1155
},
{
"epoch": 0.7421625079974408,
"grad_norm": 0.2701033651828766,
"learning_rate": 5.1567498400511836e-05,
"loss": 0.2082,
"step": 1160
},
{
"epoch": 0.745361484325016,
"grad_norm": 0.3194945156574249,
"learning_rate": 5.092770313499681e-05,
"loss": 0.235,
"step": 1165
},
{
"epoch": 0.7485604606525912,
"grad_norm": 0.25952160358428955,
"learning_rate": 5.028790786948176e-05,
"loss": 0.2405,
"step": 1170
},
{
"epoch": 0.7517594369801663,
"grad_norm": 0.3131108283996582,
"learning_rate": 4.9648112603966736e-05,
"loss": 0.2118,
"step": 1175
},
{
"epoch": 0.7549584133077415,
"grad_norm": 0.40070056915283203,
"learning_rate": 4.9008317338451695e-05,
"loss": 0.2386,
"step": 1180
},
{
"epoch": 0.7581573896353166,
"grad_norm": 0.38076481223106384,
"learning_rate": 4.836852207293666e-05,
"loss": 0.229,
"step": 1185
},
{
"epoch": 0.7613563659628919,
"grad_norm": 0.26312530040740967,
"learning_rate": 4.772872680742163e-05,
"loss": 0.2029,
"step": 1190
},
{
"epoch": 0.7645553422904671,
"grad_norm": 0.365788996219635,
"learning_rate": 4.7088931541906594e-05,
"loss": 0.2215,
"step": 1195
},
{
"epoch": 0.7677543186180422,
"grad_norm": 0.246324360370636,
"learning_rate": 4.644913627639156e-05,
"loss": 0.2067,
"step": 1200
},
{
"epoch": 0.7709532949456174,
"grad_norm": 0.2862643003463745,
"learning_rate": 4.580934101087652e-05,
"loss": 0.1955,
"step": 1205
},
{
"epoch": 0.7741522712731925,
"grad_norm": 0.3309938609600067,
"learning_rate": 4.516954574536149e-05,
"loss": 0.227,
"step": 1210
},
{
"epoch": 0.7773512476007678,
"grad_norm": 0.3223839998245239,
"learning_rate": 4.4529750479846447e-05,
"loss": 0.2292,
"step": 1215
},
{
"epoch": 0.780550223928343,
"grad_norm": 0.3467869460582733,
"learning_rate": 4.388995521433142e-05,
"loss": 0.209,
"step": 1220
},
{
"epoch": 0.7837492002559181,
"grad_norm": 0.34096330404281616,
"learning_rate": 4.325015994881638e-05,
"loss": 0.2041,
"step": 1225
},
{
"epoch": 0.7869481765834933,
"grad_norm": 0.33046209812164307,
"learning_rate": 4.2610364683301346e-05,
"loss": 0.1974,
"step": 1230
},
{
"epoch": 0.7901471529110684,
"grad_norm": 0.4353679418563843,
"learning_rate": 4.197056941778631e-05,
"loss": 0.1992,
"step": 1235
},
{
"epoch": 0.7933461292386437,
"grad_norm": 0.35748305916786194,
"learning_rate": 4.133077415227127e-05,
"loss": 0.2237,
"step": 1240
},
{
"epoch": 0.7965451055662188,
"grad_norm": 0.3174164891242981,
"learning_rate": 4.0690978886756245e-05,
"loss": 0.2136,
"step": 1245
},
{
"epoch": 0.799744081893794,
"grad_norm": 0.3202930688858032,
"learning_rate": 4.0051183621241205e-05,
"loss": 0.2301,
"step": 1250
},
{
"epoch": 0.8029430582213691,
"grad_norm": 0.2494242787361145,
"learning_rate": 3.941138835572617e-05,
"loss": 0.1783,
"step": 1255
},
{
"epoch": 0.8061420345489443,
"grad_norm": 0.35551324486732483,
"learning_rate": 3.877159309021113e-05,
"loss": 0.2031,
"step": 1260
},
{
"epoch": 0.8093410108765196,
"grad_norm": 0.2531558573246002,
"learning_rate": 3.81317978246961e-05,
"loss": 0.2365,
"step": 1265
},
{
"epoch": 0.8125399872040947,
"grad_norm": 0.43033191561698914,
"learning_rate": 3.7492002559181063e-05,
"loss": 0.2217,
"step": 1270
},
{
"epoch": 0.8157389635316699,
"grad_norm": 0.3271211087703705,
"learning_rate": 3.685220729366603e-05,
"loss": 0.2198,
"step": 1275
},
{
"epoch": 0.818937939859245,
"grad_norm": 0.2692057490348816,
"learning_rate": 3.6212412028150996e-05,
"loss": 0.2099,
"step": 1280
},
{
"epoch": 0.8221369161868202,
"grad_norm": 0.2919643223285675,
"learning_rate": 3.5572616762635956e-05,
"loss": 0.2148,
"step": 1285
},
{
"epoch": 0.8253358925143954,
"grad_norm": 0.3477243185043335,
"learning_rate": 3.493282149712092e-05,
"loss": 0.2045,
"step": 1290
},
{
"epoch": 0.8285348688419706,
"grad_norm": 0.3863447308540344,
"learning_rate": 3.429302623160589e-05,
"loss": 0.2481,
"step": 1295
},
{
"epoch": 0.8317338451695457,
"grad_norm": 0.3201524019241333,
"learning_rate": 3.3653230966090855e-05,
"loss": 0.2121,
"step": 1300
},
{
"epoch": 0.8349328214971209,
"grad_norm": 0.3398021459579468,
"learning_rate": 3.3013435700575815e-05,
"loss": 0.2251,
"step": 1305
},
{
"epoch": 0.838131797824696,
"grad_norm": 0.29826006293296814,
"learning_rate": 3.237364043506078e-05,
"loss": 0.1768,
"step": 1310
},
{
"epoch": 0.8413307741522713,
"grad_norm": 0.3297532796859741,
"learning_rate": 3.173384516954575e-05,
"loss": 0.2026,
"step": 1315
},
{
"epoch": 0.8445297504798465,
"grad_norm": 0.3907334804534912,
"learning_rate": 3.1094049904030714e-05,
"loss": 0.2032,
"step": 1320
},
{
"epoch": 0.8477287268074216,
"grad_norm": 0.27420273423194885,
"learning_rate": 3.0454254638515677e-05,
"loss": 0.1845,
"step": 1325
},
{
"epoch": 0.8509277031349968,
"grad_norm": 0.36319416761398315,
"learning_rate": 2.981445937300064e-05,
"loss": 0.2143,
"step": 1330
},
{
"epoch": 0.8541266794625719,
"grad_norm": 0.26537972688674927,
"learning_rate": 2.9174664107485606e-05,
"loss": 0.2228,
"step": 1335
},
{
"epoch": 0.8573256557901472,
"grad_norm": 0.3364832103252411,
"learning_rate": 2.853486884197057e-05,
"loss": 0.1889,
"step": 1340
},
{
"epoch": 0.8605246321177223,
"grad_norm": 0.32128915190696716,
"learning_rate": 2.789507357645554e-05,
"loss": 0.2023,
"step": 1345
},
{
"epoch": 0.8637236084452975,
"grad_norm": 0.3016292452812195,
"learning_rate": 2.7255278310940502e-05,
"loss": 0.247,
"step": 1350
},
{
"epoch": 0.8669225847728727,
"grad_norm": 0.3532962501049042,
"learning_rate": 2.6615483045425465e-05,
"loss": 0.22,
"step": 1355
},
{
"epoch": 0.8701215611004478,
"grad_norm": 0.26920193433761597,
"learning_rate": 2.5975687779910428e-05,
"loss": 0.1888,
"step": 1360
},
{
"epoch": 0.8733205374280231,
"grad_norm": 0.27570822834968567,
"learning_rate": 2.533589251439539e-05,
"loss": 0.2278,
"step": 1365
},
{
"epoch": 0.8765195137555982,
"grad_norm": 0.2538580596446991,
"learning_rate": 2.4696097248880358e-05,
"loss": 0.1739,
"step": 1370
},
{
"epoch": 0.8797184900831734,
"grad_norm": 0.2816247045993805,
"learning_rate": 2.4056301983365324e-05,
"loss": 0.2169,
"step": 1375
},
{
"epoch": 0.8829174664107485,
"grad_norm": 0.3473341763019562,
"learning_rate": 2.341650671785029e-05,
"loss": 0.226,
"step": 1380
},
{
"epoch": 0.8861164427383237,
"grad_norm": 0.2831457257270813,
"learning_rate": 2.2776711452335254e-05,
"loss": 0.2125,
"step": 1385
},
{
"epoch": 0.889315419065899,
"grad_norm": 0.30768829584121704,
"learning_rate": 2.213691618682022e-05,
"loss": 0.1819,
"step": 1390
},
{
"epoch": 0.8925143953934741,
"grad_norm": 0.31688612699508667,
"learning_rate": 2.1497120921305183e-05,
"loss": 0.2194,
"step": 1395
},
{
"epoch": 0.8957133717210493,
"grad_norm": 0.3418295383453369,
"learning_rate": 2.0857325655790146e-05,
"loss": 0.2169,
"step": 1400
},
{
"epoch": 0.8989123480486244,
"grad_norm": 0.37767553329467773,
"learning_rate": 2.0217530390275112e-05,
"loss": 0.2323,
"step": 1405
},
{
"epoch": 0.9021113243761996,
"grad_norm": 0.26787883043289185,
"learning_rate": 1.957773512476008e-05,
"loss": 0.2024,
"step": 1410
},
{
"epoch": 0.9053103007037748,
"grad_norm": 0.26510849595069885,
"learning_rate": 1.8937939859245045e-05,
"loss": 0.2327,
"step": 1415
},
{
"epoch": 0.90850927703135,
"grad_norm": 0.2409944236278534,
"learning_rate": 1.8298144593730008e-05,
"loss": 0.1867,
"step": 1420
},
{
"epoch": 0.9117082533589251,
"grad_norm": 0.29780739545822144,
"learning_rate": 1.765834932821497e-05,
"loss": 0.1885,
"step": 1425
},
{
"epoch": 0.9149072296865003,
"grad_norm": 0.3131493031978607,
"learning_rate": 1.7018554062699938e-05,
"loss": 0.1984,
"step": 1430
},
{
"epoch": 0.9181062060140754,
"grad_norm": 0.30493828654289246,
"learning_rate": 1.63787587971849e-05,
"loss": 0.2028,
"step": 1435
},
{
"epoch": 0.9213051823416507,
"grad_norm": 0.4079281985759735,
"learning_rate": 1.5738963531669867e-05,
"loss": 0.2297,
"step": 1440
},
{
"epoch": 0.9245041586692259,
"grad_norm": 0.3364027440547943,
"learning_rate": 1.5099168266154832e-05,
"loss": 0.2007,
"step": 1445
},
{
"epoch": 0.927703134996801,
"grad_norm": 0.29711103439331055,
"learning_rate": 1.4459373000639795e-05,
"loss": 0.2201,
"step": 1450
},
{
"epoch": 0.9309021113243762,
"grad_norm": 0.3361244797706604,
"learning_rate": 1.3819577735124761e-05,
"loss": 0.2118,
"step": 1455
},
{
"epoch": 0.9341010876519513,
"grad_norm": 0.2692669630050659,
"learning_rate": 1.3179782469609724e-05,
"loss": 0.2036,
"step": 1460
},
{
"epoch": 0.9373000639795266,
"grad_norm": 0.2790112793445587,
"learning_rate": 1.253998720409469e-05,
"loss": 0.1977,
"step": 1465
},
{
"epoch": 0.9404990403071017,
"grad_norm": 0.3373515009880066,
"learning_rate": 1.1900191938579655e-05,
"loss": 0.1959,
"step": 1470
},
{
"epoch": 0.9436980166346769,
"grad_norm": 0.27667558193206787,
"learning_rate": 1.126039667306462e-05,
"loss": 0.21,
"step": 1475
},
{
"epoch": 0.946896992962252,
"grad_norm": 0.342916876077652,
"learning_rate": 1.0620601407549585e-05,
"loss": 0.225,
"step": 1480
},
{
"epoch": 0.9500959692898272,
"grad_norm": 0.31250983476638794,
"learning_rate": 9.98080614203455e-06,
"loss": 0.2517,
"step": 1485
},
{
"epoch": 0.9532949456174025,
"grad_norm": 0.29028651118278503,
"learning_rate": 9.341010876519514e-06,
"loss": 0.2276,
"step": 1490
},
{
"epoch": 0.9564939219449776,
"grad_norm": 0.3296051621437073,
"learning_rate": 8.701215611004479e-06,
"loss": 0.1873,
"step": 1495
},
{
"epoch": 0.9596928982725528,
"grad_norm": 0.36832043528556824,
"learning_rate": 8.061420345489444e-06,
"loss": 0.2064,
"step": 1500
},
{
"epoch": 0.9628918746001279,
"grad_norm": 0.3034568130970001,
"learning_rate": 7.421625079974409e-06,
"loss": 0.2035,
"step": 1505
},
{
"epoch": 0.9660908509277031,
"grad_norm": 0.26890963315963745,
"learning_rate": 6.781829814459373e-06,
"loss": 0.2019,
"step": 1510
},
{
"epoch": 0.9692898272552783,
"grad_norm": 0.39937588572502136,
"learning_rate": 6.142034548944338e-06,
"loss": 0.2026,
"step": 1515
},
{
"epoch": 0.9724888035828535,
"grad_norm": 0.2571321427822113,
"learning_rate": 5.502239283429303e-06,
"loss": 0.1928,
"step": 1520
},
{
"epoch": 0.9756877799104287,
"grad_norm": 0.30180391669273376,
"learning_rate": 4.862444017914268e-06,
"loss": 0.2086,
"step": 1525
},
{
"epoch": 0.9788867562380038,
"grad_norm": 0.4258427321910858,
"learning_rate": 4.222648752399233e-06,
"loss": 0.1849,
"step": 1530
},
{
"epoch": 0.982085732565579,
"grad_norm": 0.29561159014701843,
"learning_rate": 3.5828534868841974e-06,
"loss": 0.2151,
"step": 1535
},
{
"epoch": 0.9852847088931542,
"grad_norm": 0.2644422948360443,
"learning_rate": 2.943058221369162e-06,
"loss": 0.189,
"step": 1540
},
{
"epoch": 0.9884836852207294,
"grad_norm": 0.30060943961143494,
"learning_rate": 2.303262955854127e-06,
"loss": 0.2164,
"step": 1545
},
{
"epoch": 0.9916826615483045,
"grad_norm": 0.3195488154888153,
"learning_rate": 1.6634676903390916e-06,
"loss": 0.2519,
"step": 1550
},
{
"epoch": 0.9948816378758797,
"grad_norm": 0.30431658029556274,
"learning_rate": 1.0236724248240563e-06,
"loss": 0.231,
"step": 1555
},
{
"epoch": 0.9980806142034548,
"grad_norm": 0.34465327858924866,
"learning_rate": 3.838771593090211e-07,
"loss": 0.2068,
"step": 1560
}
],
"logging_steps": 5,
"max_steps": 1563,
"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": 8.925037329289052e+17,
"train_batch_size": 10,
"trial_name": null,
"trial_params": null
}