theophilusowiti/Caracal_GPT
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
main
Caracal_GPT/checkpoints/checkpoint-21630/trainer_state.json

3058 lines
70 KiB
JSON
Raw Permalink Normal View History

初始化项目，由ModelHub XC社区提供模型 Model: theophilusowiti/Caracal_GPT Source: Original Platform
2026-06-12 01:44:18 +08:00
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.999953768983611,
"eval_steps": 500,
"global_step": 21630,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004623101638889531,
"grad_norm": 10.25,
"learning_rate": 2e-05,
"loss": 5.7331,
"step": 50
},
{
"epoch": 0.009246203277779062,
"grad_norm": 8.375,
"learning_rate": 4e-05,
"loss": 5.0317,
"step": 100
},
{
"epoch": 0.013869304916668592,
"grad_norm": 7.65625,
"learning_rate": 6e-05,
"loss": 4.6245,
"step": 150
},
{
"epoch": 0.018492406555558123,
"grad_norm": 6.34375,
"learning_rate": 8e-05,
"loss": 4.4327,
"step": 200
},
{
"epoch": 0.023115508194447656,
"grad_norm": 6.75,
"learning_rate": 0.0001,
"loss": 4.2418,
"step": 250
},
{
"epoch": 0.027738609833337185,
"grad_norm": 5.75,
"learning_rate": 0.00012,
"loss": 4.1714,
"step": 300
},
{
"epoch": 0.03236171147222672,
"grad_norm": 5.5,
"learning_rate": 0.00014,
"loss": 4.0904,
"step": 350
},
{
"epoch": 0.036984813111116246,
"grad_norm": 6.90625,
"learning_rate": 0.00016,
"loss": 4.1346,
"step": 400
},
{
"epoch": 0.041607914750005776,
"grad_norm": 5.625,
"learning_rate": 0.00018,
"loss": 4.1936,
"step": 450
},
{
"epoch": 0.04623101638889531,
"grad_norm": 5.15625,
"learning_rate": 0.0002,
"loss": 4.167,
"step": 500
},
{
"epoch": 0.05085411802778484,
"grad_norm": 5.0625,
"learning_rate": 0.0001999993252640082,
"loss": 4.1719,
"step": 550
},
{
"epoch": 0.05547721966667437,
"grad_norm": 5.59375,
"learning_rate": 0.0001999973010651382,
"loss": 4.0257,
"step": 600
},
{
"epoch": 0.060100321305563906,
"grad_norm": 4.84375,
"learning_rate": 0.00019999392743070592,
"loss": 4.0536,
"step": 650
},
{
"epoch": 0.06472342294445343,
"grad_norm": 4.875,
"learning_rate": 0.00019998920440623765,
"loss": 3.9901,
"step": 700
},
{
"epoch": 0.06934652458334296,
"grad_norm": 4.78125,
"learning_rate": 0.0001999831320554693,
"loss": 3.9791,
"step": 750
},
{
"epoch": 0.07396962622223249,
"grad_norm": 4.71875,
"learning_rate": 0.00019997571046034555,
"loss": 3.9548,
"step": 800
},
{
"epoch": 0.07859272786112202,
"grad_norm": 4.40625,
"learning_rate": 0.0001999669397210187,
"loss": 3.8842,
"step": 850
},
{
"epoch": 0.08321582950001155,
"grad_norm": 4.84375,
"learning_rate": 0.00019995681995584746,
"loss": 3.8699,
"step": 900
},
{
"epoch": 0.0878389311389011,
"grad_norm": 5.09375,
"learning_rate": 0.0001999453513013952,
"loss": 3.8166,
"step": 950
},
{
"epoch": 0.09246203277779062,
"grad_norm": 4.4375,
"learning_rate": 0.00019993253391242826,
"loss": 3.785,
"step": 1000
},
{
"epoch": 0.09708513441668015,
"grad_norm": 4.96875,
"learning_rate": 0.00019991836796191364,
"loss": 3.7597,
"step": 1050
},
{
"epoch": 0.10170823605556968,
"grad_norm": 4.75,
"learning_rate": 0.00019990285364101688,
"loss": 3.7262,
"step": 1100
},
{
"epoch": 0.10633133769445921,
"grad_norm": 4.25,
"learning_rate": 0.00019988599115909942,
"loss": 3.7565,
"step": 1150
},
{
"epoch": 0.11095443933334874,
"grad_norm": 4.25,
"learning_rate": 0.00019986778074371575,
"loss": 3.6486,
"step": 1200
},
{
"epoch": 0.11557754097223827,
"grad_norm": 5.09375,
"learning_rate": 0.00019984822264061024,
"loss": 3.6558,
"step": 1250
},
{
"epoch": 0.12020064261112781,
"grad_norm": 4.1875,
"learning_rate": 0.0001998273171137141,
"loss": 3.602,
"step": 1300
},
{
"epoch": 0.12482374425001734,
"grad_norm": 4.3125,
"learning_rate": 0.00019980506444514155,
"loss": 3.6141,
"step": 1350
},
{
"epoch": 0.12944684588890687,
"grad_norm": 3.953125,
"learning_rate": 0.00019978146493518608,
"loss": 3.5908,
"step": 1400
},
{
"epoch": 0.13406994752779638,
"grad_norm": 4.375,
"learning_rate": 0.00019975651890231645,
"loss": 3.5565,
"step": 1450
},
{
"epoch": 0.13869304916668593,
"grad_norm": 4.09375,
"learning_rate": 0.00019973022668317244,
"loss": 3.5753,
"step": 1500
},
{
"epoch": 0.14331615080557547,
"grad_norm": 4.0625,
"learning_rate": 0.00019970258863256015,
"loss": 3.5305,
"step": 1550
},
{
"epoch": 0.14793925244446499,
"grad_norm": 4.09375,
"learning_rate": 0.00019967360512344732,
"loss": 3.5222,
"step": 1600
},
{
"epoch": 0.15256235408335453,
"grad_norm": 4.0,
"learning_rate": 0.00019964327654695835,
"loss": 3.5663,
"step": 1650
},
{
"epoch": 0.15718545572224404,
"grad_norm": 4.25,
"learning_rate": 0.00019961160331236877,
"loss": 3.5668,
"step": 1700
},
{
"epoch": 0.1618085573611336,
"grad_norm": 4.0625,
"learning_rate": 0.00019957858584710009,
"loss": 3.445,
"step": 1750
},
{
"epoch": 0.1664316590000231,
"grad_norm": 4.15625,
"learning_rate": 0.00019954422459671377,
"loss": 3.4294,
"step": 1800
},
{
"epoch": 0.17105476063891265,
"grad_norm": 4.0,
"learning_rate": 0.00019950852002490518,
"loss": 3.4634,
"step": 1850
},
{
"epoch": 0.1756778622778022,
"grad_norm": 3.859375,
"learning_rate": 0.00019947147261349755,
"loss": 3.4504,
"step": 1900
},
{
"epoch": 0.1803009639166917,
"grad_norm": 4.15625,
"learning_rate": 0.00019943308286243534,
"loss": 3.463,
"step": 1950
},
{
"epoch": 0.18492406555558125,
"grad_norm": 3.5,
"learning_rate": 0.00019939335128977748,
"loss": 3.4113,
"step": 2000
},
{
"epoch": 0.18954716719447076,
"grad_norm": 4.28125,
"learning_rate": 0.00019935227843169035,
"loss": 3.4378,
"step": 2050
},
{
"epoch": 0.1941702688333603,
"grad_norm": 4.40625,
"learning_rate": 0.00019930986484244077,
"loss": 3.3705,
"step": 2100
},
{
"epoch": 0.19879337047224982,
"grad_norm": 4.1875,
"learning_rate": 0.00019926611109438816,
"loss": 3.4306,
"step": 2150
},
{
"epoch": 0.20341647211113936,
"grad_norm": 4.0625,
"learning_rate": 0.0001992210177779771,
"loss": 3.366,
"step": 2200
},
{
"epoch": 0.2080395737500289,
"grad_norm": 3.703125,
"learning_rate": 0.00019917458550172928,
"loss": 3.3909,
"step": 2250
},
{
"epoch": 0.21266267538891842,
"grad_norm": 4.09375,
"learning_rate": 0.0001991268148922353,
"loss": 3.3814,
"step": 2300
},
{
"epoch": 0.21728577702780796,
"grad_norm": 3.796875,
"learning_rate": 0.00019907770659414607,
"loss": 3.3391,
"step": 2350
},
{
"epoch": 0.22190887866669748,
"grad_norm": 4.09375,
"learning_rate": 0.0001990272612701644,
"loss": 3.3336,
"step": 2400
},
{
"epoch": 0.22653198030558702,
"grad_norm": 5.0625,
"learning_rate": 0.0001989754796010357,
"loss": 3.3146,
"step": 2450
},
{
"epoch": 0.23115508194447654,
"grad_norm": 4.21875,
"learning_rate": 0.00019892236228553916,
"loss": 3.3387,
"step": 2500
},
{
"epoch": 0.23577818358336608,
"grad_norm": 4.09375,
"learning_rate": 0.0001988679100404781,
"loss": 3.3373,
"step": 2550
},
{
"epoch": 0.24040128522225562,
"grad_norm": 4.125,
"learning_rate": 0.0001988121236006702,
"loss": 3.3327,
"step": 2600
},
{
"epoch": 0.24502438686114514,
"grad_norm": 4.15625,
"learning_rate": 0.00019875500371893795,
"loss": 3.2672,
"step": 2650
},
{
"epoch": 0.24964748850003468,
"grad_norm": 4.0,
"learning_rate": 0.00019869655116609808,
"loss": 3.2931,
"step": 2700
},
{
"epoch": 0.2542705901389242,
"grad_norm": 4.1875,
"learning_rate": 0.0001986367667309515,
"loss": 3.2913,
"step": 2750
},
{
"epoch": 0.25889369177781374,
"grad_norm": 4.0625,
"learning_rate": 0.00019857565122027228,
"loss": 3.2848,
"step": 2800
},
{
"epoch": 0.26351679341670325,
"grad_norm": 4.3125,
"learning_rate": 0.0001985132054587972,
"loss": 3.2729,
"step": 2850
},
{
"epoch": 0.26813989505559277,
"grad_norm": 4.25,
"learning_rate": 0.00019844943028921432,
"loss": 3.3057,
"step": 2900
},
{
"epoch": 0.27276299669448234,
"grad_norm": 4.34375,
"learning_rate": 0.00019838432657215166,
"loss": 3.2535,
"step": 2950
},
{
"epoch": 0.27738609833337186,
"grad_norm": 4.59375,
"learning_rate": 0.00019831789518616567,
"loss": 3.2543,
"step": 3000
},
{
"epoch": 0.28200919997226137,
"grad_norm": 4.75,
"learning_rate": 0.00019825013702772925,
"loss": 3.2435,
"step": 3050
},
{
"epoch": 0.28663230161115094,
"grad_norm": 4.40625,
"learning_rate": 0.00019818105301121978,
"loss": 3.2287,
"step": 3100
},
{
"epoch": 0.29125540325004046,
"grad_norm": 4.5,
"learning_rate": 0.00019811064406890673,
"loss": 3.2054,
"step": 3150
},
{
"epoch": 0.29587850488892997,
"grad_norm": 4.5,
"learning_rate": 0.00019803891115093903,
"loss": 3.2044,
"step": 3200
},
{
"epoch": 0.3005016065278195,
"grad_norm": 4.40625,
"learning_rate": 0.00019796585522533232,
"loss": 3.2105,
"step": 3250
},
{
"epoch": 0.30512470816670906,
"grad_norm": 4.5625,
"learning_rate": 0.00019789147727795588,
"loss": 3.2122,
"step": 3300
},
{
"epoch": 0.3097478098055986,
"grad_norm": 4.21875,
"learning_rate": 0.00019781577831251917,
"loss": 3.2451,
"step": 3350
},
{
"epoch": 0.3143709114444881,
"grad_norm": 4.40625,
"learning_rate": 0.00019773875935055862,
"loss": 3.232,
"step": 3400
},
{
"epoch": 0.31899401308337766,
"grad_norm": 4.53125,
"learning_rate": 0.00019766042143142353,
"loss": 3.2113,
"step": 3450
},
{
"epoch": 0.3236171147222672,
"grad_norm": 4.125,
"learning_rate": 0.00019758076561226213,
"loss": 3.2058,
"step": 3500
},
{
"epoch": 0.3282402163611567,
"grad_norm": 4.34375,
"learning_rate": 0.0001974997929680074,
"loss": 3.2143,
"step": 3550
},
{
"epoch": 0.3328633180000462,
"grad_norm": 4.59375,
"learning_rate": 0.00019741750459136253,
"loss": 3.2039,
"step": 3600
},
{
"epoch": 0.3374864196389358,
"grad_norm": 4.15625,
"learning_rate": 0.00019733390159278605,
"loss": 3.224,
"step": 3650
},
{
"epoch": 0.3421095212778253,
"grad_norm": 4.5625,
"learning_rate": 0.00019724898510047701,
"loss": 3.1668,
"step": 3700
},
{
"epoch": 0.3467326229167148,
"grad_norm": 4.59375,
"learning_rate": 0.00019716275626035972,
"loss": 3.1734,
"step": 3750
},
{
"epoch": 0.3513557245556044,
"grad_norm": 4.40625,
"learning_rate": 0.0001970752162360682,
"loss": 3.1776,
"step": 3800
},
{
"epoch": 0.3559788261944939,
"grad_norm": 4.125,
"learning_rate": 0.00019698636620893054,
"loss": 3.1227,
"step": 3850
},
{
"epoch": 0.3606019278333834,
"grad_norm": 4.53125,
"learning_rate": 0.00019689620737795295,
"loss": 3.1555,
"step": 3900
},
{
"epoch": 0.3652250294722729,
"grad_norm": 4.5625,
"learning_rate": 0.00019680474095980365,
"loss": 3.1237,
"step": 3950
},
{
"epoch": 0.3698481311111625,
"grad_norm": 4.40625,
"learning_rate": 0.0001967119681887963,
"loss": 3.14,
"step": 4000
},
{
"epoch": 0.374471232750052,
"grad_norm": 4.03125,
"learning_rate": 0.00019661789031687343,
"loss": 3.1183,
"step": 4050
},
{
"epoch": 0.3790943343889415,
"grad_norm": 4.5625,
"learning_rate": 0.0001965225086135896,
"loss": 3.1166,
"step": 4100
},
{
"epoch": 0.3837174360278311,
"grad_norm": 4.21875,
"learning_rate": 0.00019642582436609415,
"loss": 3.1272,
"step": 4150
},
{
"epoch": 0.3883405376667206,
"grad_norm": 4.78125,
"learning_rate": 0.00019632783887911393,
"loss": 3.1201,
"step": 4200
},
{
"epoch": 0.3929636393056101,
"grad_norm": 4.03125,
"learning_rate": 0.0001962285534749356,
"loss": 3.1078,
"step": 4250
},
{
"epoch": 0.39758674094449964,
"grad_norm": 4.71875,
"learning_rate": 0.00019612796949338793,
"loss": 3.124,
"step": 4300
},
{
"epoch": 0.4022098425833892,
"grad_norm": 4.4375,
"learning_rate": 0.00019602608829182353,
"loss": 3.0682,
"step": 4350
},
{
"epoch": 0.4068329442222787,
"grad_norm": 4.4375,
"learning_rate": 0.00019592291124510066,
"loss": 3.0967,
"step": 4400
},
{
"epoch": 0.41145604586116824,
"grad_norm": 4.15625,
"learning_rate": 0.0001958184397455647,
"loss": 3.0829,
"step": 4450
},
{
"epoch": 0.4160791475000578,
"grad_norm": 4.625,
"learning_rate": 0.00019571267520302922,
"loss": 3.0902,
"step": 4500
},
{
"epoch": 0.4207022491389473,
"grad_norm": 4.40625,
"learning_rate": 0.00019560561904475714,
"loss": 3.0657,
"step": 4550
},
{
"epoch": 0.42532535077783684,
"grad_norm": 4.28125,
"learning_rate": 0.00019549727271544128,
"loss": 3.0467,
"step": 4600
},
{
"epoch": 0.42994845241672636,
"grad_norm": 4.21875,
"learning_rate": 0.00019538763767718505,
"loss": 3.0887,
"step": 4650
},
{
"epoch": 0.4345715540556159,
"grad_norm": 4.375,
"learning_rate": 0.0001952767154094825,
"loss": 3.0263,
"step": 4700
},
{
"epoch": 0.43919465569450544,
"grad_norm": 4.09375,
"learning_rate": 0.00019516450740919865,
"loss": 3.0493,
"step": 4750
},
{
"epoch": 0.44381775733339496,
"grad_norm": 4.3125,
"learning_rate": 0.00019505101519054895,
"loss": 3.0268,
"step": 4800
},
{
"epoch": 0.44844085897228453,
"grad_norm": 4.28125,
"learning_rate": 0.00019493624028507912,
"loss": 3.0399,
"step": 4850
},
{
"epoch": 0.45306396061117404,
"grad_norm": 4.4375,
"learning_rate": 0.00019482018424164436,
"loss": 3.0722,
"step": 4900
},
{
"epoch": 0.45768706225006356,
"grad_norm": 4.21875,
"learning_rate": 0.00019470284862638845,
"loss": 3.0155,
"step": 4950
},
{
"epoch": 0.4623101638889531,
"grad_norm": 4.46875,
"learning_rate": 0.00019458423502272267,
"loss": 3.0341,
"step": 5000
},
{
"epoch": 0.46693326552784264,
"grad_norm": 4.0625,
"learning_rate": 0.00019446434503130432,
"loss": 3.0222,
"step": 5050
},
{
"epoch": 0.47155636716673216,
"grad_norm": 4.53125,
"learning_rate": 0.0001943431802700153,
"loss": 3.0176,
"step": 5100
},
{
"epoch": 0.4761794688056217,
"grad_norm": 4.0,
"learning_rate": 0.00019422074237394007,
"loss": 3.0232,
"step": 5150
},
{
"epoch": 0.48080257044451125,
"grad_norm": 4.4375,
"learning_rate": 0.0001940970329953438,
"loss": 3.0065,
"step": 5200
},
{
"epoch": 0.48542567208340076,
"grad_norm": 4.3125,
"learning_rate": 0.00019397205380364984,
"loss": 3.0149,
"step": 5250
},
{
"epoch": 0.4900487737222903,
"grad_norm": 4.25,
"learning_rate": 0.00019384580648541737,
"loss": 3.0091,
"step": 5300
},
{
"epoch": 0.4946718753611798,
"grad_norm": 4.21875,
"learning_rate": 0.0001937182927443186,
"loss": 3.0174,
"step": 5350
},
{
"epoch": 0.49929497700006936,
"grad_norm": 3.90625,
"learning_rate": 0.00019358951430111573,
"loss": 3.0074,
"step": 5400
},
{
"epoch": 0.5039180786389589,
"grad_norm": 4.46875,
"learning_rate": 0.00019345947289363776,
"loss": 2.9757,
"step": 5450
},
{
"epoch": 0.5085411802778484,
"grad_norm": 4.09375,
"learning_rate": 0.0001933281702767571,
"loss": 2.9508,
"step": 5500
},
{
"epoch": 0.5131642819167379,
"grad_norm": 4.1875,
"learning_rate": 0.0001931956082223657,
"loss": 2.95,
"step": 5550
},
{
"epoch": 0.5177873835556275,
"grad_norm": 4.125,
"learning_rate": 0.00019306178851935144,
"loss": 2.9755,
"step": 5600
},
{
"epoch": 0.522410485194517,
"grad_norm": 4.21875,
"learning_rate": 0.00019292671297357363,
"loss": 2.9631,
"step": 5650
},
{
"epoch": 0.5270335868334065,
"grad_norm": 4.3125,
"learning_rate": 0.00019279038340783898,
"loss": 2.9755,
"step": 5700
},
{
"epoch": 0.5316566884722961,
"grad_norm": 4.0,
"learning_rate": 0.00019265280166187674,
"loss": 2.9691,
"step": 5750
},
{
"epoch": 0.5362797901111855,
"grad_norm": 4.03125,
"learning_rate": 0.00019251396959231408,
"loss": 2.9681,
"step": 5800
},
{
"epoch": 0.5409028917500751,
"grad_norm": 4.03125,
"learning_rate": 0.00019237388907265078,
"loss": 2.9503,
"step": 5850
},
{
"epoch": 0.5455259933889647,
"grad_norm": 4.59375,
"learning_rate": 0.00019223256199323436,
"loss": 2.9646,
"step": 5900
},
{
"epoch": 0.5501490950278541,
"grad_norm": 4.0,
"learning_rate": 0.00019208999026123404,
"loss": 2.9603,
"step": 5950
},
{
"epoch": 0.5547721966667437,
"grad_norm": 3.9375,
"learning_rate": 0.00019194617580061543,
"loss": 2.9419,
"step": 6000
},
{
"epoch": 0.5593952983056333,
"grad_norm": 4.1875,
"learning_rate": 0.00019180112055211445,
"loss": 2.9279,
"step": 6050
},
{
"epoch": 0.5640183999445227,
"grad_norm": 4.03125,
"learning_rate": 0.00019165482647321095,
"loss": 2.9468,
"step": 6100
},
{
"epoch": 0.5686415015834123,
"grad_norm": 4.5,
"learning_rate": 0.00019150729553810256,
"loss": 2.9412,
"step": 6150
},
{
"epoch": 0.5732646032223019,
"grad_norm": 4.40625,
"learning_rate": 0.00019135852973767794,
"loss": 2.9177,
"step": 6200
},
{
"epoch": 0.5778877048611913,
"grad_norm": 4.40625,
"learning_rate": 0.00019120853107948985,
"loss": 2.9285,
"step": 6250
},
{
"epoch": 0.5825108065000809,
"grad_norm": 4.625,
"learning_rate": 0.00019105730158772824,
"loss": 2.9534,
"step": 6300
},
{
"epoch": 0.5871339081389705,
"grad_norm": 4.1875,
"learning_rate": 0.00019090484330319262,
"loss": 2.9199,
"step": 6350
},
{
"epoch": 0.5917570097778599,
"grad_norm": 4.09375,
"learning_rate": 0.00019075115828326493,
"loss": 2.8957,
"step": 6400
},
{
"epoch": 0.5963801114167495,
"grad_norm": 4.25,
"learning_rate": 0.0001905962486018814,
"loss": 2.9327,
"step": 6450
},
{
"epoch": 0.601003213055639,
"grad_norm": 4.125,
"learning_rate": 0.0001904401163495048,
"loss": 2.873,
"step": 6500
},
{
"epoch": 0.6056263146945285,
"grad_norm": 4.3125,
"learning_rate": 0.00019028276363309608,
"loss": 2.8886,
"step": 6550
},
{
"epoch": 0.6102494163334181,
"grad_norm": 4.28125,
"learning_rate": 0.00019012419257608614,
"loss": 2.9023,
"step": 6600
},
{
"epoch": 0.6148725179723076,
"grad_norm": 4.40625,
"learning_rate": 0.0001899644053183469,
"loss": 2.9093,
"step": 6650
},
{
"epoch": 0.6194956196111971,
"grad_norm": 4.46875,
"learning_rate": 0.00018980340401616273,
"loss": 2.9233,
"step": 6700
},
{
"epoch": 0.6241187212500867,
"grad_norm": 4.15625,
"learning_rate": 0.00018964119084220103,
"loss": 2.8934,
"step": 6750
},
{
"epoch": 0.6287418228889762,
"grad_norm": 4.4375,
"learning_rate": 0.0001894777679854831,
"loss": 2.8878,
"step": 6800
},
{
"epoch": 0.6333649245278657,
"grad_norm": 4.0,
"learning_rate": 0.00018931313765135468,
"loss": 2.8924,
"step": 6850
},
{
"epoch": 0.6379880261667553,
"grad_norm": 4.40625,
"learning_rate": 0.00018914730206145598,
"loss": 2.9166,
"step": 6900
},
{
"epoch": 0.6426111278056448,
"grad_norm": 4.15625,
"learning_rate": 0.0001889802634536918,
"loss": 2.8957,
"step": 6950
},
{
"epoch": 0.6472342294445343,
"grad_norm": 4.40625,
"learning_rate": 0.00018881202408220138,
"loss": 2.8684,
"step": 7000
},
{
"epoch": 0.6518573310834239,
"grad_norm": 3.984375,
"learning_rate": 0.00018864258621732792,
"loss": 2.9041,
"step": 7050
},
{
"epoch": 0.6564804327223134,
"grad_norm": 4.125,
"learning_rate": 0.0001884719521455879,
"loss": 2.8904,
"step": 7100
},
{
"epoch": 0.661103534361203,
"grad_norm": 3.953125,
"learning_rate": 0.0001883001241696403,
"loss": 2.8749,
"step": 7150
},
{
"epoch": 0.6657266360000924,
"grad_norm": 4.5,
"learning_rate": 0.0001881271046082556,
"loss": 2.852,
"step": 7200
},
{
"epoch": 0.670349737638982,
"grad_norm": 4.125,
"learning_rate": 0.0001879528957962842,
"loss": 2.8679,
"step": 7250
},
{
"epoch": 0.6749728392778716,
"grad_norm": 3.96875,
"learning_rate": 0.00018777750008462532,
"loss": 2.8695,
"step": 7300
},
{
"epoch": 0.679595940916761,
"grad_norm": 4.0625,
"learning_rate": 0.00018760091984019482,
"loss": 2.8536,
"step": 7350
},
{
"epoch": 0.6842190425556506,
"grad_norm": 4.40625,
"learning_rate": 0.00018742315744589374,
"loss": 2.8616,
"step": 7400
},
{
"epoch": 0.6888421441945402,
"grad_norm": 4.125,
"learning_rate": 0.00018724421530057572,
"loss": 2.8568,
"step": 7450
},
{
"epoch": 0.6934652458334296,
"grad_norm": 4.125,
"learning_rate": 0.0001870640958190149,
"loss": 2.8603,
"step": 7500
},
{
"epoch": 0.6980883474723192,
"grad_norm": 4.15625,
"learning_rate": 0.00018688280143187326,
"loss": 2.888,
"step": 7550
},
{
"epoch": 0.7027114491112088,
"grad_norm": 4.0625,
"learning_rate": 0.0001867003345856677,
"loss": 2.823,
"step": 7600
},
{
"epoch": 0.7073345507500982,
"grad_norm": 4.375,
"learning_rate": 0.00018651669774273722,
"loss": 2.8504,
"step": 7650
},
{
"epoch": 0.7119576523889878,
"grad_norm": 4.125,
"learning_rate": 0.00018633189338120956,
"loss": 2.837,
"step": 7700
},
{
"epoch": 0.7165807540278774,
"grad_norm": 3.859375,
"learning_rate": 0.0001861459239949678,
"loss": 2.8477,
"step": 7750
},
{
"epoch": 0.7212038556667668,
"grad_norm": 4.15625,
"learning_rate": 0.00018595879209361673,
"loss": 2.8518,
"step": 7800
},
{
"epoch": 0.7258269573056564,
"grad_norm": 4.15625,
"learning_rate": 0.00018577050020244888,
"loss": 2.815,
"step": 7850
},
{
"epoch": 0.7304500589445458,
"grad_norm": 4.3125,
"learning_rate": 0.00018558105086241062,
"loss": 2.8459,
"step": 7900
},
{
"epoch": 0.7350731605834354,
"grad_norm": 3.921875,
"learning_rate": 0.00018539044663006769,
"loss": 2.8051,
"step": 7950
},
{
"epoch": 0.739696262222325,
"grad_norm": 3.9375,
"learning_rate": 0.0001851986900775708,
"loss": 2.8473,
"step": 8000
},
{
"epoch": 0.7443193638612144,
"grad_norm": 4.09375,
"learning_rate": 0.00018500578379262094,
"loss": 2.8471,
"step": 8050
},
{
"epoch": 0.748942465500104,
"grad_norm": 4.28125,
"learning_rate": 0.00018481173037843436,
"loss": 2.8341,
"step": 8100
},
{
"epoch": 0.7535655671389936,
"grad_norm": 4.3125,
"learning_rate": 0.0001846165324537075,
"loss": 2.8392,
"step": 8150
},
{
"epoch": 0.758188668777883,
"grad_norm": 4.21875,
"learning_rate": 0.00018442019265258166,
"loss": 2.8026,
"step": 8200
},
{
"epoch": 0.7628117704167726,
"grad_norm": 4.4375,
"learning_rate": 0.00018422271362460749,
"loss": 2.7998,
"step": 8250
},
{
"epoch": 0.7674348720556622,
"grad_norm": 3.8125,
"learning_rate": 0.00018402409803470912,
"loss": 2.8222,
"step": 8300
},
{
"epoch": 0.7720579736945516,
"grad_norm": 3.8125,
"learning_rate": 0.0001838243485631483,
"loss": 2.794,
"step": 8350
},
{
"epoch": 0.7766810753334412,
"grad_norm": 4.03125,
"learning_rate": 0.0001836234679054882,
"loss": 2.8043,
"step": 8400
},
{
"epoch": 0.7813041769723308,
"grad_norm": 4.09375,
"learning_rate": 0.00018342145877255694,
"loss": 2.8087,
"step": 8450
},
{
"epoch": 0.7859272786112202,
"grad_norm": 4.03125,
"learning_rate": 0.00018321832389041126,
"loss": 2.7922,
"step": 8500
},
{
"epoch": 0.7905503802501098,
"grad_norm": 4.15625,
"learning_rate": 0.00018301406600029947,
"loss": 2.7813,
"step": 8550
},
{
"epoch": 0.7951734818889993,
"grad_norm": 3.921875,
"learning_rate": 0.00018280868785862454,
"loss": 2.7775,
"step": 8600
},
{
"epoch": 0.7997965835278888,
"grad_norm": 4.03125,
"learning_rate": 0.00018260219223690697,
"loss": 2.8344,
"step": 8650
},
{
"epoch": 0.8044196851667784,
"grad_norm": 4.09375,
"learning_rate": 0.0001823945819217473,
"loss": 2.7897,
"step": 8700
},
{
"epoch": 0.8090427868056679,
"grad_norm": 4.0625,
"learning_rate": 0.00018218585971478864,
"loss": 2.8006,
"step": 8750
},
{
"epoch": 0.8136658884445575,
"grad_norm": 4.4375,
"learning_rate": 0.0001819760284326786,
"loss": 2.7929,
"step": 8800
},
{
"epoch": 0.818288990083447,
"grad_norm": 3.875,
"learning_rate": 0.0001817650909070316,
"loss": 2.7931,
"step": 8850
},
{
"epoch": 0.8229120917223365,
"grad_norm": 4.09375,
"learning_rate": 0.00018155304998439045,
"loss": 2.7972,
"step": 8900
},
{
"epoch": 0.827535193361226,
"grad_norm": 4.0625,
"learning_rate": 0.00018133990852618796,
"loss": 2.7644,
"step": 8950
},
{
"epoch": 0.8321582950001156,
"grad_norm": 4.65625,
"learning_rate": 0.00018112566940870842,
"loss": 2.785,
"step": 9000
},
{
"epoch": 0.8367813966390051,
"grad_norm": 4.375,
"learning_rate": 0.00018091033552304872,
"loss": 2.8149,
"step": 9050
},
{
"epoch": 0.8414044982778947,
"grad_norm": 3.921875,
"learning_rate": 0.0001806939097750793,
"loss": 2.7606,
"step": 9100
},
{
"epoch": 0.8460275999167842,
"grad_norm": 4.0,
"learning_rate": 0.00018047639508540498,
"loss": 2.7607,
"step": 9150
},
{
"epoch": 0.8506507015556737,
"grad_norm": 3.8125,
"learning_rate": 0.00018025779438932555,
"loss": 2.7729,
"step": 9200
},
{
"epoch": 0.8552738031945633,
"grad_norm": 3.796875,
"learning_rate": 0.00018003811063679618,
"loss": 2.804,
"step": 9250
},
{
"epoch": 0.8598969048334527,
"grad_norm": 4.0625,
"learning_rate": 0.0001798173467923876,
"loss": 2.7478,
"step": 9300
},
{
"epoch": 0.8645200064723423,
"grad_norm": 4.1875,
"learning_rate": 0.00017959550583524595,
"loss": 2.7585,
"step": 9350
},
{
"epoch": 0.8691431081112319,
"grad_norm": 4.34375,
"learning_rate": 0.00017937259075905289,
"loss": 2.7708,
"step": 9400
},
{
"epoch": 0.8737662097501213,
"grad_norm": 3.953125,
"learning_rate": 0.00017914860457198483,
"loss": 2.8107,
"step": 9450
},
{
"epoch": 0.8783893113890109,
"grad_norm": 4.25,
"learning_rate": 0.0001789235502966727,
"loss": 2.7751,
"step": 9500
},
{
"epoch": 0.8830124130279005,
"grad_norm": 4.25,
"learning_rate": 0.00017869743097016083,
"loss": 2.7921,
"step": 9550
},
{
"epoch": 0.8876355146667899,
"grad_norm": 4.0,
"learning_rate": 0.00017847024964386617,
"loss": 2.7672,
"step": 9600
},
{
"epoch": 0.8922586163056795,
"grad_norm": 4.21875,
"learning_rate": 0.00017824200938353717,
"loss": 2.7566,
"step": 9650
},
{
"epoch": 0.8968817179445691,
"grad_norm": 4.3125,
"learning_rate": 0.00017801271326921207,
"loss": 2.7537,
"step": 9700
},
{
"epoch": 0.9015048195834585,
"grad_norm": 4.1875,
"learning_rate": 0.0001777823643951778,
"loss": 2.757,
"step": 9750
},
{
"epoch": 0.9061279212223481,
"grad_norm": 4.09375,
"learning_rate": 0.0001775509658699278,
"loss": 2.7451,
"step": 9800
},
{
"epoch": 0.9107510228612377,
"grad_norm": 4.0,
"learning_rate": 0.00017731852081612036,
"loss": 2.7302,
"step": 9850
},
{
"epoch": 0.9153741245001271,
"grad_norm": 4.1875,
"learning_rate": 0.00017708503237053638,
"loss": 2.7548,
"step": 9900
},
{
"epoch": 0.9199972261390167,
"grad_norm": 4.125,
"learning_rate": 0.000176850503684037,
"loss": 2.761,
"step": 9950
},
{
"epoch": 0.9246203277779061,
"grad_norm": 3.953125,
"learning_rate": 0.00017661493792152117,
"loss": 2.755,
"step": 10000
},
{
"epoch": 0.9292434294167957,
"grad_norm": 3.90625,
"learning_rate": 0.00017637833826188282,
"loss": 2.7341,
"step": 10050
},
{
"epoch": 0.9338665310556853,
"grad_norm": 4.34375,
"learning_rate": 0.00017614070789796809,
"loss": 2.7432,
"step": 10100
},
{
"epoch": 0.9384896326945747,
"grad_norm": 3.8125,
"learning_rate": 0.00017590205003653213,
"loss": 2.7364,
"step": 10150
},
{
"epoch": 0.9431127343334643,
"grad_norm": 4.15625,
"learning_rate": 0.00017566236789819597,
"loss": 2.7333,
"step": 10200
},
{
"epoch": 0.9477358359723539,
"grad_norm": 4.09375,
"learning_rate": 0.00017542166471740288,
"loss": 2.7549,
"step": 10250
},
{
"epoch": 0.9523589376112434,
"grad_norm": 4.0625,
"learning_rate": 0.00017517994374237488,
"loss": 2.7483,
"step": 10300
},
{
"epoch": 0.9569820392501329,
"grad_norm": 4.1875,
"learning_rate": 0.0001749372082350688,
"loss": 2.741,
"step": 10350
},
{
"epoch": 0.9616051408890225,
"grad_norm": 3.8125,
"learning_rate": 0.00017469346147113224,
"loss": 2.7154,
"step": 10400
},
{
"epoch": 0.966228242527912,
"grad_norm": 3.765625,
"learning_rate": 0.00017444870673985957,
"loss": 2.7204,
"step": 10450
},
{
"epoch": 0.9708513441668015,
"grad_norm": 3.875,
"learning_rate": 0.00017420294734414728,
"loss": 2.7356,
"step": 10500
},
{
"epoch": 0.9754744458056911,
"grad_norm": 4.03125,
"learning_rate": 0.0001739561866004496,
"loss": 2.7299,
"step": 10550
},
{
"epoch": 0.9800975474445806,
"grad_norm": 4.0,
"learning_rate": 0.00017370842783873358,
"loss": 2.7336,
"step": 10600
},
{
"epoch": 0.9847206490834701,
"grad_norm": 4.34375,
"learning_rate": 0.0001734596744024343,
"loss": 2.7171,
"step": 10650
},
{
"epoch": 0.9893437507223596,
"grad_norm": 4.15625,
"learning_rate": 0.00017320992964840975,
"loss": 2.74,
"step": 10700
},
{
"epoch": 0.9939668523612492,
"grad_norm": 4.0,
"learning_rate": 0.00017295919694689538,
"loss": 2.7389,
"step": 10750
},
{
"epoch": 0.9985899540001387,
"grad_norm": 4.03125,
"learning_rate": 0.00017270747968145872,
"loss": 2.72,
"step": 10800
},
{
"epoch": 1.0032130556390282,
"grad_norm": 1.53125,
"learning_rate": 0.00017245478124895377,
"loss": 2.5214,
"step": 10850
},
{
"epoch": 1.0078361572779178,
"grad_norm": 1.5546875,
"learning_rate": 0.00017220110505947514,
"loss": 2.498,
"step": 10900
},
{
"epoch": 1.0124592589168073,
"grad_norm": 1.578125,
"learning_rate": 0.00017194645453631182,
"loss": 2.5162,
"step": 10950
},
{
"epoch": 1.017082360555697,
"grad_norm": 1.5625,
"learning_rate": 0.00017169083311590138,
"loss": 2.5105,
"step": 11000
},
{
"epoch": 1.0217054621945862,
"grad_norm": 1.609375,
"learning_rate": 0.00017143424424778317,
"loss": 2.5043,
"step": 11050
},
{
"epoch": 1.0263285638334758,
"grad_norm": 1.5625,
"learning_rate": 0.0001711766913945522,
"loss": 2.5069,
"step": 11100
},
{
"epoch": 1.0309516654723654,
"grad_norm": 1.5625,
"learning_rate": 0.00017091817803181197,
"loss": 2.5159,
"step": 11150
},
{
"epoch": 1.035574767111255,
"grad_norm": 1.625,
"learning_rate": 0.0001706587076481279,
"loss": 2.5123,
"step": 11200
},
{
"epoch": 1.0401978687501445,
"grad_norm": 1.546875,
"learning_rate": 0.00017039828374498014,
"loss": 2.5271,
"step": 11250
},
{
"epoch": 1.044820970389034,
"grad_norm": 1.6171875,
"learning_rate": 0.00017013690983671634,
"loss": 2.5232,
"step": 11300
},
{
"epoch": 1.0494440720279234,
"grad_norm": 1.5625,
"learning_rate": 0.00016987458945050412,
"loss": 2.5279,
"step": 11350
},
{
"epoch": 1.054067173666813,
"grad_norm": 1.59375,
"learning_rate": 0.0001696113261262836,
"loss": 2.5241,
"step": 11400
},
{
"epoch": 1.0586902753057026,
"grad_norm": 1.578125,
"learning_rate": 0.0001693471234167196,
"loss": 2.5013,
"step": 11450
},
{
"epoch": 1.0633133769445922,
"grad_norm": 1.546875,
"learning_rate": 0.00016908198488715362,
"loss": 2.5146,
"step": 11500
},
{
"epoch": 1.0679364785834817,
"grad_norm": 1.5703125,
"learning_rate": 0.0001688159141155559,
"loss": 2.4962,
"step": 11550
},
{
"epoch": 1.072559580222371,
"grad_norm": 1.640625,
"learning_rate": 0.00016854891469247693,
"loss": 2.5287,
"step": 11600
},
{
"epoch": 1.0771826818612606,
"grad_norm": 1.5859375,
"learning_rate": 0.00016828099022099912,
"loss": 2.5173,
"step": 11650
},
{
"epoch": 1.0818057835001502,
"grad_norm": 1.59375,
"learning_rate": 0.00016801214431668812,
"loss": 2.5403,
"step": 11700
},
{
"epoch": 1.0864288851390398,
"grad_norm": 1.5703125,
"learning_rate": 0.00016774238060754417,
"loss": 2.5175,
"step": 11750
},
{
"epoch": 1.0910519867779294,
"grad_norm": 1.5625,
"learning_rate": 0.00016747170273395285,
"loss": 2.5131,
"step": 11800
},
{
"epoch": 1.095675088416819,
"grad_norm": 1.6328125,
"learning_rate": 0.0001672001143486363,
"loss": 2.5176,
"step": 11850
},
{
"epoch": 1.1002981900557083,
"grad_norm": 1.59375,
"learning_rate": 0.00016692761911660364,
"loss": 2.4988,
"step": 11900
},
{
"epoch": 1.1049212916945979,
"grad_norm": 1.6328125,
"learning_rate": 0.00016665422071510173,
"loss": 2.4954,
"step": 11950
},
{
"epoch": 1.1095443933334874,
"grad_norm": 1.53125,
"learning_rate": 0.00016637992283356537,
"loss": 2.514,
"step": 12000
},
{
"epoch": 1.114167494972377,
"grad_norm": 1.5625,
"learning_rate": 0.00016610472917356764,
"loss": 2.5236,
"step": 12050
},
{
"epoch": 1.1187905966112666,
"grad_norm": 1.6015625,
"learning_rate": 0.00016582864344876988,
"loss": 2.4954,
"step": 12100
},
{
"epoch": 1.1234136982501561,
"grad_norm": 1.5546875,
"learning_rate": 0.00016555166938487156,
"loss": 2.5275,
"step": 12150
},
{
"epoch": 1.1280367998890455,
"grad_norm": 1.6171875,
"learning_rate": 0.00016527381071956013,
"loss": 2.5166,
"step": 12200
},
{
"epoch": 1.132659901527935,
"grad_norm": 1.578125,
"learning_rate": 0.0001649950712024604,
"loss": 2.4762,
"step": 12250
},
{
"epoch": 1.1372830031668246,
"grad_norm": 1.6171875,
"learning_rate": 0.00016471545459508405,
"loss": 2.5207,
"step": 12300
},
{
"epoch": 1.1419061048057142,
"grad_norm": 1.6484375,
"learning_rate": 0.00016443496467077892,
"loss": 2.4879,
"step": 12350
},
{
"epoch": 1.1465292064446038,
"grad_norm": 1.5859375,
"learning_rate": 0.0001641536052146779,
"loss": 2.5061,
"step": 12400
},
{
"epoch": 1.1511523080834931,
"grad_norm": 1.5703125,
"learning_rate": 0.00016387138002364802,
"loss": 2.4956,
"step": 12450
},
{
"epoch": 1.1557754097223827,
"grad_norm": 1.625,
"learning_rate": 0.00016358829290623921,
"loss": 2.5048,
"step": 12500
},
{
"epoch": 1.1603985113612723,
"grad_norm": 1.5703125,
"learning_rate": 0.00016330434768263273,
"loss": 2.5197,
"step": 12550
},
{
"epoch": 1.1650216130001618,
"grad_norm": 1.578125,
"learning_rate": 0.0001630195481845899,
"loss": 2.5013,
"step": 12600
},
{
"epoch": 1.1696447146390514,
"grad_norm": 1.6328125,
"learning_rate": 0.00016273389825540013,
"loss": 2.4961,
"step": 12650
},
{
"epoch": 1.1742678162779407,
"grad_norm": 1.5703125,
"learning_rate": 0.00016244740174982915,
"loss": 2.5246,
"step": 12700
},
{
"epoch": 1.1788909179168303,
"grad_norm": 1.5859375,
"learning_rate": 0.00016216006253406707,
"loss": 2.4902,
"step": 12750
},
{
"epoch": 1.1835140195557199,
"grad_norm": 1.59375,
"learning_rate": 0.00016187188448567614,
"loss": 2.4813,
"step": 12800
},
{
"epoch": 1.1881371211946095,
"grad_norm": 1.6640625,
"learning_rate": 0.00016158287149353834,
"loss": 2.5187,
"step": 12850
},
{
"epoch": 1.192760222833499,
"grad_norm": 1.609375,
"learning_rate": 0.000161293027457803,
"loss": 2.4897,
"step": 12900
},
{
"epoch": 1.1973833244723886,
"grad_norm": 1.625,
"learning_rate": 0.00016100235628983427,
"loss": 2.5023,
"step": 12950
},
{
"epoch": 1.2020064261112782,
"grad_norm": 1.578125,
"learning_rate": 0.00016071086191215802,
"loss": 2.4894,
"step": 13000
},
{
"epoch": 1.2066295277501675,
"grad_norm": 1.5859375,
"learning_rate": 0.00016041854825840925,
"loss": 2.4818,
"step": 13050
},
{
"epoch": 1.211252629389057,
"grad_norm": 1.59375,
"learning_rate": 0.00016012541927327889,
"loss": 2.4782,
"step": 13100
},
{
"epoch": 1.2158757310279467,
"grad_norm": 1.625,
"learning_rate": 0.00015983147891246034,
"loss": 2.487,
"step": 13150
},
{
"epoch": 1.2204988326668362,
"grad_norm": 1.59375,
"learning_rate": 0.00015953673114259655,
"loss": 2.4863,
"step": 13200
},
{
"epoch": 1.2251219343057258,
"grad_norm": 1.5859375,
"learning_rate": 0.000159241179941226,
"loss": 2.5009,
"step": 13250
},
{
"epoch": 1.2297450359446152,
"grad_norm": 1.5703125,
"learning_rate": 0.0001589448292967294,
"loss": 2.4921,
"step": 13300
},
{
"epoch": 1.2343681375835047,
"grad_norm": 1.6015625,
"learning_rate": 0.00015864768320827567,
"loss": 2.4804,
"step": 13350
},
{
"epoch": 1.2389912392223943,
"grad_norm": 1.578125,
"learning_rate": 0.000158349745685768,
"loss": 2.4838,
"step": 13400
},
{
"epoch": 1.2436143408612839,
"grad_norm": 1.5703125,
"learning_rate": 0.00015805102074978978,
"loss": 2.4727,
"step": 13450
},
{
"epoch": 1.2482374425001734,
"grad_norm": 1.6328125,
"learning_rate": 0.00015775151243155037,
"loss": 2.4635,
"step": 13500
},
{
"epoch": 1.2528605441390628,
"grad_norm": 1.5546875,
"learning_rate": 0.00015745122477283058,
"loss": 2.4711,
"step": 13550
},
{
"epoch": 1.2574836457779524,
"grad_norm": 1.6015625,
"learning_rate": 0.00015715016182592825,
"loss": 2.4705,
"step": 13600
},
{
"epoch": 1.262106747416842,
"grad_norm": 1.5625,
"learning_rate": 0.00015684832765360345,
"loss": 2.4649,
"step": 13650
},
{
"epoch": 1.2667298490557315,
"grad_norm": 1.5625,
"learning_rate": 0.00015654572632902388,
"loss": 2.489,
"step": 13700
},
{
"epoch": 1.271352950694621,
"grad_norm": 1.5859375,
"learning_rate": 0.00015624236193570952,
"loss": 2.4898,
"step": 13750
},
{
"epoch": 1.2759760523335106,
"grad_norm": 1.6171875,
"learning_rate": 0.00015593823856747794,
"loss": 2.4757,
"step": 13800
},
{
"epoch": 1.2805991539724002,
"grad_norm": 1.546875,
"learning_rate": 0.0001556333603283888,
"loss": 2.4623,
"step": 13850
},
{
"epoch": 1.2852222556112896,
"grad_norm": 1.5625,
"learning_rate": 0.0001553277313326885,
"loss": 2.4997,
"step": 13900
},
{
"epoch": 1.2898453572501791,
"grad_norm": 1.5859375,
"learning_rate": 0.00015502135570475464,
"loss": 2.4782,
"step": 13950
},
{
"epoch": 1.2944684588890687,
"grad_norm": 1.59375,
"learning_rate": 0.00015471423757904058,
"loss": 2.454,
"step": 14000
},
{
"epoch": 1.2990915605279583,
"grad_norm": 1.59375,
"learning_rate": 0.00015440638110001932,
"loss": 2.4702,
"step": 14050
},
{
"epoch": 1.3037146621668478,
"grad_norm": 1.5625,
"learning_rate": 0.00015409779042212785,
"loss": 2.5136,
"step": 14100
},
{
"epoch": 1.3083377638057372,
"grad_norm": 1.6328125,
"learning_rate": 0.00015378846970971088,
"loss": 2.459,
"step": 14150
},
{
"epoch": 1.3129608654446268,
"grad_norm": 1.578125,
"learning_rate": 0.00015347842313696475,
"loss": 2.4711,
"step": 14200
},
{
"epoch": 1.3175839670835163,
"grad_norm": 1.546875,
"learning_rate": 0.00015316765488788115,
"loss": 2.4729,
"step": 14250
},
{
"epoch": 1.322207068722406,
"grad_norm": 1.5546875,
"learning_rate": 0.00015285616915619045,
"loss": 2.4373,
"step": 14300
},
{
"epoch": 1.3268301703612955,
"grad_norm": 1.578125,
"learning_rate": 0.00015254397014530545,
"loss": 2.4782,
"step": 14350
},
{
"epoch": 1.3314532720001848,
"grad_norm": 1.625,
"learning_rate": 0.0001522310620682642,
"loss": 2.452,
"step": 14400
},
{
"epoch": 1.3360763736390744,
"grad_norm": 1.6875,
"learning_rate": 0.00015191744914767362,
"loss": 2.4553,
"step": 14450
},
{
"epoch": 1.340699475277964,
"grad_norm": 1.5703125,
"learning_rate": 0.0001516031356156522,
"loss": 2.4566,
"step": 14500
},
{
"epoch": 1.3453225769168535,
"grad_norm": 1.578125,
"learning_rate": 0.00015128812571377296,
"loss": 2.467,
"step": 14550
},
{
"epoch": 1.349945678555743,
"grad_norm": 1.5703125,
"learning_rate": 0.00015097242369300633,
"loss": 2.4664,
"step": 14600
},
{
"epoch": 1.3545687801946324,
"grad_norm": 1.6328125,
"learning_rate": 0.00015065603381366258,
"loss": 2.4439,
"step": 14650
},
{
"epoch": 1.3591918818335222,
"grad_norm": 1.578125,
"learning_rate": 0.0001503389603453345,
"loss": 2.4654,
"step": 14700
},
{
"epoch": 1.3638149834724116,
"grad_norm": 1.5625,
"learning_rate": 0.00015002120756683974,
"loss": 2.4445,
"step": 14750
},
{
"epoch": 1.3684380851113012,
"grad_norm": 1.609375,
"learning_rate": 0.00014970277976616297,
"loss": 2.4913,
"step": 14800
},
{
"epoch": 1.3730611867501907,
"grad_norm": 1.578125,
"learning_rate": 0.00014938368124039824,
"loss": 2.4439,
"step": 14850
},
{
"epoch": 1.3776842883890803,
"grad_norm": 1.6328125,
"learning_rate": 0.00014906391629569064,
"loss": 2.4388,
"step": 14900
},
{
"epoch": 1.3823073900279699,
"grad_norm": 1.6015625,
"learning_rate": 0.0001487434892471786,
"loss": 2.4716,
"step": 14950
},
{
"epoch": 1.3869304916668592,
"grad_norm": 1.59375,
"learning_rate": 0.00014842240441893535,
"loss": 2.4324,
"step": 15000
},
{
"epoch": 1.3915535933057488,
"grad_norm": 1.6484375,
"learning_rate": 0.00014810066614391067,
"loss": 2.4636,
"step": 15050
},
{
"epoch": 1.3961766949446384,
"grad_norm": 1.625,
"learning_rate": 0.00014777827876387244,
"loss": 2.4547,
"step": 15100
},
{
"epoch": 1.400799796583528,
"grad_norm": 1.59375,
"learning_rate": 0.00014745524662934805,
"loss": 2.4286,
"step": 15150
},
{
"epoch": 1.4054228982224175,
"grad_norm": 1.6015625,
"learning_rate": 0.00014713157409956563,
"loss": 2.432,
"step": 15200
},
{
"epoch": 1.4100459998613069,
"grad_norm": 1.625,
"learning_rate": 0.00014680726554239533,
"loss": 2.4503,
"step": 15250
},
{
"epoch": 1.4146691015001964,
"grad_norm": 1.5625,
"learning_rate": 0.00014648232533429024,
"loss": 2.4482,
"step": 15300
},
{
"epoch": 1.419292203139086,
"grad_norm": 1.59375,
"learning_rate": 0.00014615675786022742,
"loss": 2.4261,
"step": 15350
},
{
"epoch": 1.4239153047779756,
"grad_norm": 1.6015625,
"learning_rate": 0.00014583056751364876,
"loss": 2.4349,
"step": 15400
},
{
"epoch": 1.4285384064168651,
"grad_norm": 1.5859375,
"learning_rate": 0.00014550375869640155,
"loss": 2.4276,
"step": 15450
},
{
"epoch": 1.4331615080557545,
"grad_norm": 1.609375,
"learning_rate": 0.00014517633581867927,
"loss": 2.4324,
"step": 15500
},
{
"epoch": 1.437784609694644,
"grad_norm": 1.59375,
"learning_rate": 0.0001448483032989619,
"loss": 2.4295,
"step": 15550
},
{
"epoch": 1.4424077113335336,
"grad_norm": 1.6171875,
"learning_rate": 0.00014451966556395634,
"loss": 2.4481,
"step": 15600
},
{
"epoch": 1.4470308129724232,
"grad_norm": 1.59375,
"learning_rate": 0.00014419042704853685,
"loss": 2.4232,
"step": 15650
},
{
"epoch": 1.4516539146113128,
"grad_norm": 1.6015625,
"learning_rate": 0.00014386059219568487,
"loss": 2.4199,
"step": 15700
},
{
"epoch": 1.4562770162502023,
"grad_norm": 1.6171875,
"learning_rate": 0.00014353016545642938,
"loss": 2.4452,
"step": 15750
},
{
"epoch": 1.460900117889092,
"grad_norm": 1.5234375,
"learning_rate": 0.00014319915128978666,
"loss": 2.4472,
"step": 15800
},
{
"epoch": 1.4655232195279813,
"grad_norm": 1.5859375,
"learning_rate": 0.00014286755416270013,
"loss": 2.3967,
"step": 15850
},
{
"epoch": 1.4701463211668708,
"grad_norm": 1.5703125,
"learning_rate": 0.00014253537854998008,
"loss": 2.4349,
"step": 15900
},
{
"epoch": 1.4747694228057604,
"grad_norm": 1.5625,
"learning_rate": 0.00014220262893424338,
"loss": 2.4176,
"step": 15950
},
{
"epoch": 1.47939252444465,
"grad_norm": 1.5625,
"learning_rate": 0.00014186930980585288,
"loss": 2.4103,
"step": 16000
},
{
"epoch": 1.4840156260835395,
"grad_norm": 1.578125,
"learning_rate": 0.00014153542566285683,
"loss": 2.4375,
"step": 16050
},
{
"epoch": 1.488638727722429,
"grad_norm": 1.5859375,
"learning_rate": 0.0001412009810109282,
"loss": 2.4011,
"step": 16100
},
{
"epoch": 1.4932618293613185,
"grad_norm": 1.6640625,
"learning_rate": 0.0001408659803633038,
"loss": 2.4458,
"step": 16150
},
{
"epoch": 1.497884931000208,
"grad_norm": 1.59375,
"learning_rate": 0.00014053042824072366,
"loss": 2.4333,
"step": 16200
},
{
"epoch": 1.5025080326390976,
"grad_norm": 1.578125,
"learning_rate": 0.0001401943291713695,
"loss": 2.4149,
"step": 16250
},
{
"epoch": 1.5071311342779872,
"grad_norm": 1.625,
"learning_rate": 0.00013985768769080422,
"loss": 2.4228,
"step": 16300
},
{
"epoch": 1.5117542359168765,
"grad_norm": 1.6484375,
"learning_rate": 0.00013952050834191022,
"loss": 2.4181,
"step": 16350
},
{
"epoch": 1.5163773375557663,
"grad_norm": 1.6171875,
"learning_rate": 0.00013918279567482836,
"loss": 2.3965,
"step": 16400
},
{
"epoch": 1.5210004391946557,
"grad_norm": 1.5859375,
"learning_rate": 0.00013884455424689648,
"loss": 2.4236,
"step": 16450
},
{
"epoch": 1.5256235408335452,
"grad_norm": 1.625,
"learning_rate": 0.0001385057886225879,
"loss": 2.4047,
"step": 16500
},
{
"epoch": 1.5302466424724348,
"grad_norm": 1.59375,
"learning_rate": 0.00013816650337344978,
"loss": 2.4175,
"step": 16550
},
{
"epoch": 1.5348697441113242,
"grad_norm": 1.578125,
"learning_rate": 0.0001378267030780415,
"loss": 2.4103,
"step": 16600
},
{
"epoch": 1.539492845750214,
"grad_norm": 1.625,
"learning_rate": 0.0001374863923218729,
"loss": 2.4134,
"step": 16650
},
{
"epoch": 1.5441159473891033,
"grad_norm": 1.65625,
"learning_rate": 0.0001371455756973422,
"loss": 2.383,
"step": 16700
},
{
"epoch": 1.5487390490279929,
"grad_norm": 1.578125,
"learning_rate": 0.00013680425780367435,
"loss": 2.4155,
"step": 16750
},
{
"epoch": 1.5533621506668824,
"grad_norm": 1.609375,
"learning_rate": 0.00013646244324685866,
"loss": 2.4177,
"step": 16800
},
{
"epoch": 1.5579852523057718,
"grad_norm": 1.609375,
"learning_rate": 0.00013612013663958682,
"loss": 2.4139,
"step": 16850
},
{
"epoch": 1.5626083539446616,
"grad_norm": 1.6171875,
"learning_rate": 0.00013577734260119057,
"loss": 2.3825,
"step": 16900
},
{
"epoch": 1.567231455583551,
"grad_norm": 1.6015625,
"learning_rate": 0.00013543406575757945,
"loss": 2.4286,
"step": 16950
},
{
"epoch": 1.5718545572224405,
"grad_norm": 1.6640625,
"learning_rate": 0.0001350903107411783,
"loss": 2.4163,
"step": 17000
},
{
"epoch": 1.57647765886133,
"grad_norm": 1.6875,
"learning_rate": 0.0001347460821908647,
"loss": 2.4182,
"step": 17050
},
{
"epoch": 1.5811007605002196,
"grad_norm": 1.640625,
"learning_rate": 0.00013440138475190658,
"loss": 2.4172,
"step": 17100
},
{
"epoch": 1.5857238621391092,
"grad_norm": 1.625,
"learning_rate": 0.00013405622307589922,
"loss": 2.4159,
"step": 17150
},
{
"epoch": 1.5903469637779986,
"grad_norm": 1.5859375,
"learning_rate": 0.00013371060182070278,
"loss": 2.3812,
"step": 17200
},
{
"epoch": 1.5949700654168883,
"grad_norm": 1.625,
"learning_rate": 0.00013336452565037928,
"loss": 2.4175,
"step": 17250
},
{
"epoch": 1.5995931670557777,
"grad_norm": 1.640625,
"learning_rate": 0.00013301799923512967,
"loss": 2.3963,
"step": 17300
},
{
"epoch": 1.6042162686946673,
"grad_norm": 1.609375,
"learning_rate": 0.00013267102725123082,
"loss": 2.392,
"step": 17350
},
{
"epoch": 1.6088393703335568,
"grad_norm": 1.65625,
"learning_rate": 0.00013232361438097247,
"loss": 2.4052,
"step": 17400
},
{
"epoch": 1.6134624719724462,
"grad_norm": 1.5703125,
"learning_rate": 0.00013197576531259397,
"loss": 2.4092,
"step": 17450
},
{
"epoch": 1.618085573611336,
"grad_norm": 1.6484375,
"learning_rate": 0.000131627484740221,
"loss": 2.4137,
"step": 17500
},
{
"epoch": 1.6227086752502253,
"grad_norm": 1.6015625,
"learning_rate": 0.0001312787773638024,
"loss": 2.3949,
"step": 17550
},
{
"epoch": 1.627331776889115,
"grad_norm": 1.6015625,
"learning_rate": 0.00013092964788904637,
"loss": 2.3977,
"step": 17600
},
{
"epoch": 1.6319548785280045,
"grad_norm": 1.6328125,
"learning_rate": 0.00013058010102735753,
"loss": 2.3846,
"step": 17650
},
{
"epoch": 1.6365779801668938,
"grad_norm": 1.5703125,
"learning_rate": 0.00013023014149577273,
"loss": 2.3914,
"step": 17700
},
{
"epoch": 1.6412010818057836,
"grad_norm": 1.6640625,
"learning_rate": 0.00012987977401689787,
"loss": 2.3723,
"step": 17750
},
{
"epoch": 1.645824183444673,
"grad_norm": 1.609375,
"learning_rate": 0.00012952900331884384,
"loss": 2.3723,
"step": 17800
},
{
"epoch": 1.6504472850835625,
"grad_norm": 1.609375,
"learning_rate": 0.00012917783413516303,
"loss": 2.3774,
"step": 17850
},
{
"epoch": 1.655070386722452,
"grad_norm": 1.6328125,
"learning_rate": 0.00012882627120478513,
"loss": 2.3885,
"step": 17900
},
{
"epoch": 1.6596934883613417,
"grad_norm": 1.59375,
"learning_rate": 0.00012847431927195336,
"loss": 2.3877,
"step": 17950
},
{
"epoch": 1.6643165900002312,
"grad_norm": 1.5703125,
"learning_rate": 0.00012812198308616058,
"loss": 2.3833,
"step": 18000
},
{
"epoch": 1.6689396916391206,
"grad_norm": 1.578125,
"learning_rate": 0.00012776926740208476,
"loss": 2.3785,
"step": 18050
},
{
"epoch": 1.6735627932780102,
"grad_norm": 1.59375,
"learning_rate": 0.0001274161769795253,
"loss": 2.3713,
"step": 18100
},
{
"epoch": 1.6781858949168997,
"grad_norm": 1.640625,
"learning_rate": 0.00012706271658333854,
"loss": 2.3791,
"step": 18150
},
{
"epoch": 1.6828089965557893,
"grad_norm": 1.640625,
"learning_rate": 0.00012670889098337346,
"loss": 2.3984,
"step": 18200
},
{
"epoch": 1.6874320981946789,
"grad_norm": 1.609375,
"learning_rate": 0.00012635470495440745,
"loss": 2.3546,
"step": 18250
},
{
"epoch": 1.6920551998335682,
"grad_norm": 1.625,
"learning_rate": 0.0001260001632760817,
"loss": 2.375,
"step": 18300
},
{
"epoch": 1.696678301472458,
"grad_norm": 1.6015625,
"learning_rate": 0.00012564527073283689,
"loss": 2.3862,
"step": 18350
},
{
"epoch": 1.7013014031113474,
"grad_norm": 1.59375,
"learning_rate": 0.00012529003211384835,
"loss": 2.3727,
"step": 18400
},
{
"epoch": 1.705924504750237,
"grad_norm": 1.5859375,
"learning_rate": 0.00012493445221296187,
"loss": 2.3674,
"step": 18450
},
{
"epoch": 1.7105476063891265,
"grad_norm": 1.578125,
"learning_rate": 0.00012457853582862846,
"loss": 2.3544,
"step": 18500
},
{
"epoch": 1.7151707080280159,
"grad_norm": 1.609375,
"learning_rate": 0.00012422228776384005,
"loss": 2.3813,
"step": 18550
},
{
"epoch": 1.7197938096669056,
"grad_norm": 1.640625,
"learning_rate": 0.00012386571282606453,
"loss": 2.3698,
"step": 18600
},
{
"epoch": 1.724416911305795,
"grad_norm": 1.6171875,
"learning_rate": 0.0001235088158271807,
"loss": 2.3722,
"step": 18650
},
{
"epoch": 1.7290400129446846,
"grad_norm": 1.6875,
"learning_rate": 0.00012315160158341362,
"loss": 2.3633,
"step": 18700
},
{
"epoch": 1.7336631145835741,
"grad_norm": 1.671875,
"learning_rate": 0.0001227940749152694,
"loss": 2.3592,
"step": 18750
},
{
"epoch": 1.7382862162224635,
"grad_norm": 1.6171875,
"learning_rate": 0.0001224362406474703,
"loss": 2.3437,
"step": 18800
},
{
"epoch": 1.7429093178613533,
"grad_norm": 1.65625,
"learning_rate": 0.00012207810360888945,
"loss": 2.399,
"step": 18850
},
{
"epoch": 1.7475324195002426,
"grad_norm": 1.6640625,
"learning_rate": 0.00012171966863248592,
"loss": 2.3765,
"step": 18900
},
{
"epoch": 1.7521555211391322,
"grad_norm": 1.578125,
"learning_rate": 0.00012136094055523926,
"loss": 2.3579,
"step": 18950
},
{
"epoch": 1.7567786227780218,
"grad_norm": 1.671875,
"learning_rate": 0.00012100192421808436,
"loss": 2.3734,
"step": 19000
},
{
"epoch": 1.7614017244169113,
"grad_norm": 1.5625,
"learning_rate": 0.00012064262446584611,
"loss": 2.3618,
"step": 19050
},
{
"epoch": 1.766024826055801,
"grad_norm": 1.6484375,
"learning_rate": 0.00012028304614717405,
"loss": 2.3521,
"step": 19100
},
{
"epoch": 1.7706479276946903,
"grad_norm": 1.6171875,
"learning_rate": 0.0001199231941144768,
"loss": 2.3685,
"step": 19150
},
{
"epoch": 1.77527102933358,
"grad_norm": 1.59375,
"learning_rate": 0.00011956307322385671,
"loss": 2.3938,
"step": 19200
},
{
"epoch": 1.7798941309724694,
"grad_norm": 1.65625,
"learning_rate": 0.00011920268833504436,
"loss": 2.3281,
"step": 19250
},
{
"epoch": 1.784517232611359,
"grad_norm": 1.625,
"learning_rate": 0.00011884204431133279,
"loss": 2.3791,
"step": 19300
},
{
"epoch": 1.7891403342502485,
"grad_norm": 1.65625,
"learning_rate": 0.00011848114601951215,
"loss": 2.3382,
"step": 19350
},
{
"epoch": 1.793763435889138,
"grad_norm": 1.6484375,
"learning_rate": 0.0001181199983298037,
"loss": 2.3673,
"step": 19400
},
{
"epoch": 1.7983865375280277,
"grad_norm": 1.5859375,
"learning_rate": 0.00011775860611579436,
"loss": 2.3179,
"step": 19450
},
{
"epoch": 1.803009639166917,
"grad_norm": 1.6328125,
"learning_rate": 0.0001173969742543708,
"loss": 2.3498,
"step": 19500
},
{
"epoch": 1.8076327408058066,
"grad_norm": 1.65625,
"learning_rate": 0.00011703510762565373,
"loss": 2.3468,
"step": 19550
},
{
"epoch": 1.8122558424446962,
"grad_norm": 1.671875,
"learning_rate": 0.00011667301111293184,
"loss": 2.3609,
"step": 19600
},
{
"epoch": 1.8168789440835855,
"grad_norm": 1.671875,
"learning_rate": 0.00011631068960259614,
"loss": 2.3475,
"step": 19650
},
{
"epoch": 1.8215020457224753,
"grad_norm": 1.6171875,
"learning_rate": 0.00011594814798407394,
"loss": 2.3439,
"step": 19700
},
{
"epoch": 1.8261251473613647,
"grad_norm": 1.6171875,
"learning_rate": 0.00011558539114976279,
"loss": 2.3575,
"step": 19750
},
{
"epoch": 1.8307482490002542,
"grad_norm": 1.5703125,
"learning_rate": 0.00011522242399496455,
"loss": 2.3465,
"step": 19800
},
{
"epoch": 1.8353713506391438,
"grad_norm": 1.6484375,
"learning_rate": 0.00011485925141781925,
"loss": 2.3366,
"step": 19850
},
{
"epoch": 1.8399944522780334,
"grad_norm": 1.640625,
"learning_rate": 0.00011449587831923911,
"loss": 2.3546,
"step": 19900
},
{
"epoch": 1.844617553916923,
"grad_norm": 1.6015625,
"learning_rate": 0.00011413230960284225,
"loss": 2.3136,
"step": 19950
},
{
"epoch": 1.8492406555558123,
"grad_norm": 1.625,
"learning_rate": 0.0001137685501748867,
"loss": 2.3212,
"step": 20000
},
{
"epoch": 1.853863757194702,
"grad_norm": 1.640625,
"learning_rate": 0.00011340460494420397,
"loss": 2.3274,
"step": 20050
},
{
"epoch": 1.8584868588335914,
"grad_norm": 1.6484375,
"learning_rate": 0.000113040478822133,
"loss": 2.3365,
"step": 20100
},
{
"epoch": 1.863109960472481,
"grad_norm": 1.6015625,
"learning_rate": 0.00011267617672245383,
"loss": 2.3208,
"step": 20150
},
{
"epoch": 1.8677330621113706,
"grad_norm": 1.6796875,
"learning_rate": 0.00011231170356132114,
"loss": 2.355,
"step": 20200
},
{
"epoch": 1.87235616375026,
"grad_norm": 1.6640625,
"learning_rate": 0.00011194706425719823,
"loss": 2.3337,
"step": 20250
},
{
"epoch": 1.8769792653891497,
"grad_norm": 1.6640625,
"learning_rate": 0.0001115822637307903,
"loss": 2.3295,
"step": 20300
},
{
"epoch": 1.881602367028039,
"grad_norm": 1.6796875,
"learning_rate": 0.00011121730690497824,
"loss": 2.354,
"step": 20350
},
{
"epoch": 1.8862254686669286,
"grad_norm": 1.59375,
"learning_rate": 0.00011085219870475218,
"loss": 2.3407,
"step": 20400
},
{
"epoch": 1.8908485703058182,
"grad_norm": 1.59375,
"learning_rate": 0.00011048694405714504,
"loss": 2.337,
"step": 20450
},
{
"epoch": 1.8954716719447076,
"grad_norm": 1.671875,
"learning_rate": 0.00011012154789116587,
"loss": 2.3333,
"step": 20500
},
{
"epoch": 1.9000947735835974,
"grad_norm": 1.6875,
"learning_rate": 0.00010975601513773362,
"loss": 2.2976,
"step": 20550
},
{
"epoch": 1.9047178752224867,
"grad_norm": 1.671875,
"learning_rate": 0.00010939035072961037,
"loss": 2.3282,
"step": 20600
},
{
"epoch": 1.9093409768613763,
"grad_norm": 1.6015625,
"learning_rate": 0.00010902455960133484,
"loss": 2.2868,
"step": 20650
},
{
"epoch": 1.9139640785002658,
"grad_norm": 1.671875,
"learning_rate": 0.00010865864668915587,
"loss": 2.3214,
"step": 20700
},
{
"epoch": 1.9185871801391554,
"grad_norm": 1.5859375,
"learning_rate": 0.00010829261693096564,
"loss": 2.3309,
"step": 20750
},
{
"epoch": 1.923210281778045,
"grad_norm": 1.546875,
"learning_rate": 0.00010792647526623325,
"loss": 2.3206,
"step": 20800
},
{
"epoch": 1.9278333834169343,
"grad_norm": 1.625,
"learning_rate": 0.00010756022663593783,
"loss": 2.3218,
"step": 20850
},
{
"epoch": 1.932456485055824,
"grad_norm": 1.6796875,
"learning_rate": 0.0001071938759825021,
"loss": 2.3191,
"step": 20900
},
{
"epoch": 1.9370795866947135,
"grad_norm": 1.59375,
"learning_rate": 0.00010682742824972541,
"loss": 2.3302,
"step": 20950
},
{
"epoch": 1.941702688333603,
"grad_norm": 1.59375,
"learning_rate": 0.0001064608883827173,
"loss": 2.3299,
"step": 21000
},
{
"epoch": 1.9463257899724926,
"grad_norm": 1.671875,
"learning_rate": 0.0001060942613278306,
"loss": 2.3193,
"step": 21050
},
{
"epoch": 1.950948891611382,
"grad_norm": 1.625,
"learning_rate": 0.00010572755203259463,
"loss": 2.3078,
"step": 21100
},
{
"epoch": 1.9555719932502718,
"grad_norm": 1.65625,
"learning_rate": 0.00010536076544564866,
"loss": 2.3217,
"step": 21150
},
{
"epoch": 1.960195094889161,
"grad_norm": 1.640625,
"learning_rate": 0.00010499390651667489,
"loss": 2.2886,
"step": 21200
},
{
"epoch": 1.9648181965280507,
"grad_norm": 1.640625,
"learning_rate": 0.00010462698019633179,
"loss": 2.3196,
"step": 21250
},
{
"epoch": 1.9694412981669402,
"grad_norm": 1.6796875,
"learning_rate": 0.00010425999143618725,
"loss": 2.3266,
"step": 21300
},
{
"epoch": 1.9740643998058296,
"grad_norm": 1.6484375,
"learning_rate": 0.00010389294518865181,
"loss": 2.311,
"step": 21350
},
{
"epoch": 1.9786875014447194,
"grad_norm": 1.671875,
"learning_rate": 0.00010352584640691165,
"loss": 2.302,
"step": 21400
},
{
"epoch": 1.9833106030836087,
"grad_norm": 1.6875,
"learning_rate": 0.00010315870004486208,
"loss": 2.326,
"step": 21450
},
{
"epoch": 1.9879337047224983,
"grad_norm": 1.6015625,
"learning_rate": 0.00010279151105704037,
"loss": 2.3165,
"step": 21500
},
{
"epoch": 1.9925568063613879,
"grad_norm": 1.6328125,
"learning_rate": 0.00010242428439855898,
"loss": 2.3102,
"step": 21550
},
{
"epoch": 1.9971799080002772,
"grad_norm": 1.625,
"learning_rate": 0.00010205702502503886,
"loss": 2.3078,
"step": 21600
}
],
"logging_steps": 50,
"max_steps": 43260,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.1248877358188421e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}
Reference in New Issue Copy Permalink