imdatta0's picture
End of training
8b1b653 verified
raw
history blame
28 kB
{
"best_metric": 2.0125675201416016,
"best_model_checkpoint": "/home/datta0/models/lora_final/Qwen2-7B_pct_reverse/checkpoint-8",
"epoch": 0.9996779388083736,
"eval_steps": 8,
"global_step": 388,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0025764895330112722,
"grad_norm": 5.605652332305908,
"learning_rate": 3.75e-05,
"loss": 2.0287,
"step": 1
},
{
"epoch": 0.010305958132045089,
"grad_norm": 3.3669660091400146,
"learning_rate": 0.00015,
"loss": 2.1016,
"step": 4
},
{
"epoch": 0.020611916264090178,
"grad_norm": 3.2068347930908203,
"learning_rate": 0.0003,
"loss": 2.1146,
"step": 8
},
{
"epoch": 0.020611916264090178,
"eval_loss": 2.0125675201416016,
"eval_runtime": 10.2437,
"eval_samples_per_second": 23.917,
"eval_steps_per_second": 3.026,
"step": 8
},
{
"epoch": 0.030917874396135265,
"grad_norm": 4.330184459686279,
"learning_rate": 0.0002999179886011389,
"loss": 2.0531,
"step": 12
},
{
"epoch": 0.041223832528180356,
"grad_norm": 3.5907206535339355,
"learning_rate": 0.00029967204408281613,
"loss": 2.1241,
"step": 16
},
{
"epoch": 0.041223832528180356,
"eval_loss": 2.0399129390716553,
"eval_runtime": 10.2349,
"eval_samples_per_second": 23.938,
"eval_steps_per_second": 3.029,
"step": 16
},
{
"epoch": 0.05152979066022544,
"grad_norm": 4.796765327453613,
"learning_rate": 0.0002992624353817517,
"loss": 2.0772,
"step": 20
},
{
"epoch": 0.06183574879227053,
"grad_norm": 5.596073627471924,
"learning_rate": 0.00029868961039904624,
"loss": 2.1297,
"step": 24
},
{
"epoch": 0.06183574879227053,
"eval_loss": 2.0522220134735107,
"eval_runtime": 10.2415,
"eval_samples_per_second": 23.922,
"eval_steps_per_second": 3.027,
"step": 24
},
{
"epoch": 0.07214170692431562,
"grad_norm": 4.186759948730469,
"learning_rate": 0.00029795419551040833,
"loss": 2.1422,
"step": 28
},
{
"epoch": 0.08244766505636071,
"grad_norm": 3.64916729927063,
"learning_rate": 0.0002970569948812214,
"loss": 2.0751,
"step": 32
},
{
"epoch": 0.08244766505636071,
"eval_loss": 2.0552845001220703,
"eval_runtime": 10.2352,
"eval_samples_per_second": 23.937,
"eval_steps_per_second": 3.029,
"step": 32
},
{
"epoch": 0.0927536231884058,
"grad_norm": 3.8289270401000977,
"learning_rate": 0.0002959989895872009,
"loss": 2.1251,
"step": 36
},
{
"epoch": 0.10305958132045089,
"grad_norm": 2.9287831783294678,
"learning_rate": 0.0002947813365416023,
"loss": 2.1182,
"step": 40
},
{
"epoch": 0.10305958132045089,
"eval_loss": 2.06056547164917,
"eval_runtime": 10.2327,
"eval_samples_per_second": 23.943,
"eval_steps_per_second": 3.029,
"step": 40
},
{
"epoch": 0.11336553945249597,
"grad_norm": 2.9628751277923584,
"learning_rate": 0.0002934053672301536,
"loss": 2.06,
"step": 44
},
{
"epoch": 0.12367149758454106,
"grad_norm": 4.328015327453613,
"learning_rate": 0.00029187258625509513,
"loss": 2.1169,
"step": 48
},
{
"epoch": 0.12367149758454106,
"eval_loss": 2.0665557384490967,
"eval_runtime": 10.2388,
"eval_samples_per_second": 23.929,
"eval_steps_per_second": 3.028,
"step": 48
},
{
"epoch": 0.13397745571658615,
"grad_norm": 3.2509281635284424,
"learning_rate": 0.0002901846696899191,
"loss": 2.084,
"step": 52
},
{
"epoch": 0.14428341384863125,
"grad_norm": 3.328118085861206,
"learning_rate": 0.0002883434632466077,
"loss": 2.1253,
"step": 56
},
{
"epoch": 0.14428341384863125,
"eval_loss": 2.0729715824127197,
"eval_runtime": 10.2357,
"eval_samples_per_second": 23.936,
"eval_steps_per_second": 3.029,
"step": 56
},
{
"epoch": 0.15458937198067632,
"grad_norm": 3.165828227996826,
"learning_rate": 0.00028635098025737434,
"loss": 2.1193,
"step": 60
},
{
"epoch": 0.16489533011272142,
"grad_norm": 2.8887364864349365,
"learning_rate": 0.0002842093994731145,
"loss": 2.1136,
"step": 64
},
{
"epoch": 0.16489533011272142,
"eval_loss": 2.071445941925049,
"eval_runtime": 10.2149,
"eval_samples_per_second": 23.985,
"eval_steps_per_second": 3.035,
"step": 64
},
{
"epoch": 0.1752012882447665,
"grad_norm": 2.8911757469177246,
"learning_rate": 0.00028192106268097334,
"loss": 2.1775,
"step": 68
},
{
"epoch": 0.1855072463768116,
"grad_norm": 4.058267116546631,
"learning_rate": 0.0002794884721436361,
"loss": 2.127,
"step": 72
},
{
"epoch": 0.1855072463768116,
"eval_loss": 2.083888292312622,
"eval_runtime": 10.2223,
"eval_samples_per_second": 23.967,
"eval_steps_per_second": 3.033,
"step": 72
},
{
"epoch": 0.19581320450885667,
"grad_norm": 3.4438698291778564,
"learning_rate": 0.0002769142878631403,
"loss": 2.1427,
"step": 76
},
{
"epoch": 0.20611916264090177,
"grad_norm": 2.7667691707611084,
"learning_rate": 0.000274201324672203,
"loss": 2.1835,
"step": 80
},
{
"epoch": 0.20611916264090177,
"eval_loss": 2.080258369445801,
"eval_runtime": 10.2065,
"eval_samples_per_second": 24.004,
"eval_steps_per_second": 3.037,
"step": 80
},
{
"epoch": 0.21642512077294687,
"grad_norm": 3.604095220565796,
"learning_rate": 0.0002713525491562421,
"loss": 2.2259,
"step": 84
},
{
"epoch": 0.22673107890499195,
"grad_norm": 3.442725896835327,
"learning_rate": 0.00026837107640945905,
"loss": 2.1556,
"step": 88
},
{
"epoch": 0.22673107890499195,
"eval_loss": 2.090332269668579,
"eval_runtime": 10.1927,
"eval_samples_per_second": 24.037,
"eval_steps_per_second": 3.041,
"step": 88
},
{
"epoch": 0.23703703703703705,
"grad_norm": 2.6903796195983887,
"learning_rate": 0.00026526016662852886,
"loss": 2.1015,
"step": 92
},
{
"epoch": 0.24734299516908212,
"grad_norm": 3.1680166721343994,
"learning_rate": 0.0002620232215476231,
"loss": 2.1396,
"step": 96
},
{
"epoch": 0.24734299516908212,
"eval_loss": 2.0886850357055664,
"eval_runtime": 10.1623,
"eval_samples_per_second": 24.109,
"eval_steps_per_second": 3.05,
"step": 96
},
{
"epoch": 0.2576489533011272,
"grad_norm": 2.664879560470581,
"learning_rate": 0.00025866378071866334,
"loss": 2.1838,
"step": 100
},
{
"epoch": 0.2679549114331723,
"grad_norm": 3.213966131210327,
"learning_rate": 0.00025518551764087326,
"loss": 2.1656,
"step": 104
},
{
"epoch": 0.2679549114331723,
"eval_loss": 2.092794418334961,
"eval_runtime": 10.1287,
"eval_samples_per_second": 24.189,
"eval_steps_per_second": 3.061,
"step": 104
},
{
"epoch": 0.2782608695652174,
"grad_norm": 3.1961517333984375,
"learning_rate": 0.00025159223574386114,
"loss": 2.1535,
"step": 108
},
{
"epoch": 0.2885668276972625,
"grad_norm": 2.4013512134552,
"learning_rate": 0.00024788786422862526,
"loss": 2.0821,
"step": 112
},
{
"epoch": 0.2885668276972625,
"eval_loss": 2.0880165100097656,
"eval_runtime": 32.1246,
"eval_samples_per_second": 7.627,
"eval_steps_per_second": 0.965,
"step": 112
},
{
"epoch": 0.29887278582930754,
"grad_norm": 2.3859596252441406,
"learning_rate": 0.00024407645377103054,
"loss": 2.1672,
"step": 116
},
{
"epoch": 0.30917874396135264,
"grad_norm": 2.644300699234009,
"learning_rate": 0.00024016217209245374,
"loss": 2.1287,
"step": 120
},
{
"epoch": 0.30917874396135264,
"eval_loss": 2.092282772064209,
"eval_runtime": 32.2089,
"eval_samples_per_second": 7.607,
"eval_steps_per_second": 0.962,
"step": 120
},
{
"epoch": 0.31948470209339774,
"grad_norm": 2.3782858848571777,
"learning_rate": 0.0002361492994024415,
"loss": 2.1545,
"step": 124
},
{
"epoch": 0.32979066022544284,
"grad_norm": 2.4614486694335938,
"learning_rate": 0.00023204222371836405,
"loss": 2.1298,
"step": 128
},
{
"epoch": 0.32979066022544284,
"eval_loss": 2.096928834915161,
"eval_runtime": 32.3953,
"eval_samples_per_second": 7.563,
"eval_steps_per_second": 0.957,
"step": 128
},
{
"epoch": 0.34009661835748795,
"grad_norm": 2.94144606590271,
"learning_rate": 0.00022784543606718227,
"loss": 2.1464,
"step": 132
},
{
"epoch": 0.350402576489533,
"grad_norm": 2.5649967193603516,
"learning_rate": 0.0002235635255745762,
"loss": 2.1317,
"step": 136
},
{
"epoch": 0.350402576489533,
"eval_loss": 2.096112012863159,
"eval_runtime": 32.6415,
"eval_samples_per_second": 7.506,
"eval_steps_per_second": 0.95,
"step": 136
},
{
"epoch": 0.3607085346215781,
"grad_norm": 2.5300047397613525,
"learning_rate": 0.00021920117444680317,
"loss": 2.107,
"step": 140
},
{
"epoch": 0.3710144927536232,
"grad_norm": 2.83783221244812,
"learning_rate": 0.0002147631528507739,
"loss": 2.1325,
"step": 144
},
{
"epoch": 0.3710144927536232,
"eval_loss": 2.0906081199645996,
"eval_runtime": 32.902,
"eval_samples_per_second": 7.446,
"eval_steps_per_second": 0.942,
"step": 144
},
{
"epoch": 0.3813204508856683,
"grad_norm": 3.4615397453308105,
"learning_rate": 0.0002102543136979454,
"loss": 2.1376,
"step": 148
},
{
"epoch": 0.39162640901771334,
"grad_norm": 2.862887144088745,
"learning_rate": 0.0002056795873377331,
"loss": 2.1398,
"step": 152
},
{
"epoch": 0.39162640901771334,
"eval_loss": 2.090609312057495,
"eval_runtime": 32.9273,
"eval_samples_per_second": 7.441,
"eval_steps_per_second": 0.941,
"step": 152
},
{
"epoch": 0.40193236714975844,
"grad_norm": 2.8273372650146484,
"learning_rate": 0.00020104397616624645,
"loss": 2.1514,
"step": 156
},
{
"epoch": 0.41223832528180354,
"grad_norm": 2.578927516937256,
"learning_rate": 0.0001963525491562421,
"loss": 2.1569,
"step": 160
},
{
"epoch": 0.41223832528180354,
"eval_loss": 2.088578701019287,
"eval_runtime": 33.0266,
"eval_samples_per_second": 7.418,
"eval_steps_per_second": 0.939,
"step": 160
},
{
"epoch": 0.42254428341384864,
"grad_norm": 2.9263362884521484,
"learning_rate": 0.00019161043631427666,
"loss": 2.1761,
"step": 164
},
{
"epoch": 0.43285024154589374,
"grad_norm": 2.4806456565856934,
"learning_rate": 0.00018682282307111987,
"loss": 2.195,
"step": 168
},
{
"epoch": 0.43285024154589374,
"eval_loss": 2.086169481277466,
"eval_runtime": 10.2506,
"eval_samples_per_second": 23.901,
"eval_steps_per_second": 3.024,
"step": 168
},
{
"epoch": 0.4431561996779388,
"grad_norm": 2.7219581604003906,
"learning_rate": 0.00018199494461156203,
"loss": 2.1314,
"step": 172
},
{
"epoch": 0.4534621578099839,
"grad_norm": 3.6275038719177246,
"learning_rate": 0.00017713208014981648,
"loss": 2.1038,
"step": 176
},
{
"epoch": 0.4534621578099839,
"eval_loss": 2.0899009704589844,
"eval_runtime": 10.2299,
"eval_samples_per_second": 23.949,
"eval_steps_per_second": 3.03,
"step": 176
},
{
"epoch": 0.463768115942029,
"grad_norm": 3.646092176437378,
"learning_rate": 0.00017223954715677627,
"loss": 2.1985,
"step": 180
},
{
"epoch": 0.4740740740740741,
"grad_norm": 3.2789456844329834,
"learning_rate": 0.00016732269554543794,
"loss": 2.1159,
"step": 184
},
{
"epoch": 0.4740740740740741,
"eval_loss": 2.0844767093658447,
"eval_runtime": 10.2405,
"eval_samples_per_second": 23.924,
"eval_steps_per_second": 3.027,
"step": 184
},
{
"epoch": 0.48438003220611914,
"grad_norm": 2.5345401763916016,
"learning_rate": 0.00016238690182084986,
"loss": 2.1215,
"step": 188
},
{
"epoch": 0.49468599033816424,
"grad_norm": 2.662414073944092,
"learning_rate": 0.00015743756320098332,
"loss": 2.1605,
"step": 192
},
{
"epoch": 0.49468599033816424,
"eval_loss": 2.080495595932007,
"eval_runtime": 10.2407,
"eval_samples_per_second": 23.924,
"eval_steps_per_second": 3.027,
"step": 192
},
{
"epoch": 0.5049919484702093,
"grad_norm": 3.060046672821045,
"learning_rate": 0.00015248009171495378,
"loss": 2.0823,
"step": 196
},
{
"epoch": 0.5152979066022544,
"grad_norm": 2.4873805046081543,
"learning_rate": 0.00014751990828504622,
"loss": 2.0894,
"step": 200
},
{
"epoch": 0.5152979066022544,
"eval_loss": 2.0765280723571777,
"eval_runtime": 10.227,
"eval_samples_per_second": 23.956,
"eval_steps_per_second": 3.031,
"step": 200
},
{
"epoch": 0.5256038647342995,
"grad_norm": 2.515195608139038,
"learning_rate": 0.00014256243679901663,
"loss": 2.1746,
"step": 204
},
{
"epoch": 0.5359098228663446,
"grad_norm": 2.422018051147461,
"learning_rate": 0.00013761309817915014,
"loss": 2.1368,
"step": 208
},
{
"epoch": 0.5359098228663446,
"eval_loss": 2.074774742126465,
"eval_runtime": 10.2435,
"eval_samples_per_second": 23.918,
"eval_steps_per_second": 3.026,
"step": 208
},
{
"epoch": 0.5462157809983897,
"grad_norm": 2.5775578022003174,
"learning_rate": 0.00013267730445456208,
"loss": 2.1201,
"step": 212
},
{
"epoch": 0.5565217391304348,
"grad_norm": 2.872375249862671,
"learning_rate": 0.00012776045284322368,
"loss": 2.1626,
"step": 216
},
{
"epoch": 0.5565217391304348,
"eval_loss": 2.0714917182922363,
"eval_runtime": 10.2294,
"eval_samples_per_second": 23.951,
"eval_steps_per_second": 3.03,
"step": 216
},
{
"epoch": 0.5668276972624798,
"grad_norm": 3.076042652130127,
"learning_rate": 0.00012286791985018355,
"loss": 2.1489,
"step": 220
},
{
"epoch": 0.577133655394525,
"grad_norm": 2.7333149909973145,
"learning_rate": 0.00011800505538843798,
"loss": 2.0765,
"step": 224
},
{
"epoch": 0.577133655394525,
"eval_loss": 2.0629920959472656,
"eval_runtime": 10.2284,
"eval_samples_per_second": 23.953,
"eval_steps_per_second": 3.031,
"step": 224
},
{
"epoch": 0.58743961352657,
"grad_norm": 2.5191431045532227,
"learning_rate": 0.00011317717692888012,
"loss": 2.1448,
"step": 228
},
{
"epoch": 0.5977455716586151,
"grad_norm": 2.869346857070923,
"learning_rate": 0.00010838956368572334,
"loss": 2.0879,
"step": 232
},
{
"epoch": 0.5977455716586151,
"eval_loss": 2.0677056312561035,
"eval_runtime": 10.2216,
"eval_samples_per_second": 23.969,
"eval_steps_per_second": 3.033,
"step": 232
},
{
"epoch": 0.6080515297906602,
"grad_norm": 2.3626813888549805,
"learning_rate": 0.0001036474508437579,
"loss": 2.1169,
"step": 236
},
{
"epoch": 0.6183574879227053,
"grad_norm": 2.478994369506836,
"learning_rate": 9.895602383375353e-05,
"loss": 2.0851,
"step": 240
},
{
"epoch": 0.6183574879227053,
"eval_loss": 2.0554275512695312,
"eval_runtime": 10.1901,
"eval_samples_per_second": 24.043,
"eval_steps_per_second": 3.042,
"step": 240
},
{
"epoch": 0.6286634460547504,
"grad_norm": 2.3692219257354736,
"learning_rate": 9.432041266226686e-05,
"loss": 2.0926,
"step": 244
},
{
"epoch": 0.6389694041867955,
"grad_norm": 2.496814012527466,
"learning_rate": 8.97456863020546e-05,
"loss": 2.0731,
"step": 248
},
{
"epoch": 0.6389694041867955,
"eval_loss": 2.054917812347412,
"eval_runtime": 10.1928,
"eval_samples_per_second": 24.037,
"eval_steps_per_second": 3.041,
"step": 248
},
{
"epoch": 0.6492753623188405,
"grad_norm": 2.299753427505493,
"learning_rate": 8.523684714922608e-05,
"loss": 2.0903,
"step": 252
},
{
"epoch": 0.6595813204508857,
"grad_norm": 2.7625904083251953,
"learning_rate": 8.079882555319684e-05,
"loss": 2.113,
"step": 256
},
{
"epoch": 0.6595813204508857,
"eval_loss": 2.0516955852508545,
"eval_runtime": 10.1568,
"eval_samples_per_second": 24.122,
"eval_steps_per_second": 3.052,
"step": 256
},
{
"epoch": 0.6698872785829307,
"grad_norm": 2.4187095165252686,
"learning_rate": 7.643647442542382e-05,
"loss": 2.1152,
"step": 260
},
{
"epoch": 0.6801932367149759,
"grad_norm": 2.1766324043273926,
"learning_rate": 7.215456393281776e-05,
"loss": 2.0796,
"step": 264
},
{
"epoch": 0.6801932367149759,
"eval_loss": 2.0484488010406494,
"eval_runtime": 10.1383,
"eval_samples_per_second": 24.166,
"eval_steps_per_second": 3.058,
"step": 264
},
{
"epoch": 0.6904991948470209,
"grad_norm": 2.2651991844177246,
"learning_rate": 6.795777628163599e-05,
"loss": 2.1202,
"step": 268
},
{
"epoch": 0.700805152979066,
"grad_norm": 1.8196227550506592,
"learning_rate": 6.385070059755846e-05,
"loss": 2.1406,
"step": 272
},
{
"epoch": 0.700805152979066,
"eval_loss": 2.0456597805023193,
"eval_runtime": 32.4805,
"eval_samples_per_second": 7.543,
"eval_steps_per_second": 0.954,
"step": 272
},
{
"epoch": 0.7111111111111111,
"grad_norm": 2.1418817043304443,
"learning_rate": 5.983782790754623e-05,
"loss": 2.0593,
"step": 276
},
{
"epoch": 0.7214170692431562,
"grad_norm": 2.5641324520111084,
"learning_rate": 5.592354622896944e-05,
"loss": 2.0454,
"step": 280
},
{
"epoch": 0.7214170692431562,
"eval_loss": 2.042100667953491,
"eval_runtime": 32.5534,
"eval_samples_per_second": 7.526,
"eval_steps_per_second": 0.952,
"step": 280
},
{
"epoch": 0.7317230273752013,
"grad_norm": 2.242624282836914,
"learning_rate": 5.211213577137469e-05,
"loss": 2.1192,
"step": 284
},
{
"epoch": 0.7420289855072464,
"grad_norm": 2.441951274871826,
"learning_rate": 4.840776425613886e-05,
"loss": 2.1278,
"step": 288
},
{
"epoch": 0.7420289855072464,
"eval_loss": 2.037829637527466,
"eval_runtime": 32.37,
"eval_samples_per_second": 7.569,
"eval_steps_per_second": 0.958,
"step": 288
},
{
"epoch": 0.7523349436392914,
"grad_norm": 2.0311145782470703,
"learning_rate": 4.481448235912671e-05,
"loss": 2.1032,
"step": 292
},
{
"epoch": 0.7626409017713366,
"grad_norm": 2.180769205093384,
"learning_rate": 4.133621928133665e-05,
"loss": 2.0616,
"step": 296
},
{
"epoch": 0.7626409017713366,
"eval_loss": 2.035311222076416,
"eval_runtime": 32.434,
"eval_samples_per_second": 7.554,
"eval_steps_per_second": 0.956,
"step": 296
},
{
"epoch": 0.7729468599033816,
"grad_norm": 2.1823863983154297,
"learning_rate": 3.797677845237696e-05,
"loss": 2.1213,
"step": 300
},
{
"epoch": 0.7832528180354267,
"grad_norm": 2.1388301849365234,
"learning_rate": 3.473983337147118e-05,
"loss": 2.0697,
"step": 304
},
{
"epoch": 0.7832528180354267,
"eval_loss": 2.033958673477173,
"eval_runtime": 32.6119,
"eval_samples_per_second": 7.513,
"eval_steps_per_second": 0.951,
"step": 304
},
{
"epoch": 0.7935587761674718,
"grad_norm": 2.2005107402801514,
"learning_rate": 3.162892359054098e-05,
"loss": 2.029,
"step": 308
},
{
"epoch": 0.8038647342995169,
"grad_norm": 2.344057321548462,
"learning_rate": 2.8647450843757897e-05,
"loss": 2.0557,
"step": 312
},
{
"epoch": 0.8038647342995169,
"eval_loss": 2.029961585998535,
"eval_runtime": 31.6467,
"eval_samples_per_second": 7.742,
"eval_steps_per_second": 0.98,
"step": 312
},
{
"epoch": 0.814170692431562,
"grad_norm": 2.2549006938934326,
"learning_rate": 2.5798675327796993e-05,
"loss": 2.085,
"step": 316
},
{
"epoch": 0.8244766505636071,
"grad_norm": 2.0386688709259033,
"learning_rate": 2.3085712136859668e-05,
"loss": 2.0954,
"step": 320
},
{
"epoch": 0.8244766505636071,
"eval_loss": 2.030390501022339,
"eval_runtime": 32.6796,
"eval_samples_per_second": 7.497,
"eval_steps_per_second": 0.949,
"step": 320
},
{
"epoch": 0.8347826086956521,
"grad_norm": 1.9037587642669678,
"learning_rate": 2.0511527856363912e-05,
"loss": 2.114,
"step": 324
},
{
"epoch": 0.8450885668276973,
"grad_norm": 1.6996440887451172,
"learning_rate": 1.8078937319026654e-05,
"loss": 2.094,
"step": 328
},
{
"epoch": 0.8450885668276973,
"eval_loss": 2.0296928882598877,
"eval_runtime": 10.2274,
"eval_samples_per_second": 23.955,
"eval_steps_per_second": 3.031,
"step": 328
},
{
"epoch": 0.8553945249597423,
"grad_norm": 2.2876534461975098,
"learning_rate": 1.579060052688548e-05,
"loss": 2.0518,
"step": 332
},
{
"epoch": 0.8657004830917875,
"grad_norm": 2.0373549461364746,
"learning_rate": 1.3649019742625623e-05,
"loss": 2.0539,
"step": 336
},
{
"epoch": 0.8657004830917875,
"eval_loss": 2.0265724658966064,
"eval_runtime": 10.2427,
"eval_samples_per_second": 23.919,
"eval_steps_per_second": 3.027,
"step": 336
},
{
"epoch": 0.8760064412238325,
"grad_norm": 2.0661487579345703,
"learning_rate": 1.1656536753392287e-05,
"loss": 2.0917,
"step": 340
},
{
"epoch": 0.8863123993558776,
"grad_norm": 2.622814655303955,
"learning_rate": 9.815330310080887e-06,
"loss": 2.0866,
"step": 344
},
{
"epoch": 0.8863123993558776,
"eval_loss": 2.0249547958374023,
"eval_runtime": 10.2333,
"eval_samples_per_second": 23.941,
"eval_steps_per_second": 3.029,
"step": 344
},
{
"epoch": 0.8966183574879227,
"grad_norm": 1.9814549684524536,
"learning_rate": 8.127413744904804e-06,
"loss": 2.0342,
"step": 348
},
{
"epoch": 0.9069243156199678,
"grad_norm": 2.146690607070923,
"learning_rate": 6.594632769846353e-06,
"loss": 2.061,
"step": 352
},
{
"epoch": 0.9069243156199678,
"eval_loss": 2.0226657390594482,
"eval_runtime": 10.24,
"eval_samples_per_second": 23.926,
"eval_steps_per_second": 3.027,
"step": 352
},
{
"epoch": 0.9172302737520129,
"grad_norm": 2.2478368282318115,
"learning_rate": 5.218663458397715e-06,
"loss": 2.1308,
"step": 356
},
{
"epoch": 0.927536231884058,
"grad_norm": 1.9740926027297974,
"learning_rate": 4.001010412799138e-06,
"loss": 2.126,
"step": 360
},
{
"epoch": 0.927536231884058,
"eval_loss": 2.0219550132751465,
"eval_runtime": 10.2389,
"eval_samples_per_second": 23.928,
"eval_steps_per_second": 3.028,
"step": 360
},
{
"epoch": 0.937842190016103,
"grad_norm": 2.0188748836517334,
"learning_rate": 2.9430051187785962e-06,
"loss": 2.0966,
"step": 364
},
{
"epoch": 0.9481481481481482,
"grad_norm": 2.021127939224243,
"learning_rate": 2.0458044895916513e-06,
"loss": 2.0616,
"step": 368
},
{
"epoch": 0.9481481481481482,
"eval_loss": 2.022151470184326,
"eval_runtime": 10.2382,
"eval_samples_per_second": 23.93,
"eval_steps_per_second": 3.028,
"step": 368
},
{
"epoch": 0.9584541062801932,
"grad_norm": 2.2153139114379883,
"learning_rate": 1.3103896009537207e-06,
"loss": 2.0186,
"step": 372
},
{
"epoch": 0.9687600644122383,
"grad_norm": 2.1602697372436523,
"learning_rate": 7.375646182482875e-07,
"loss": 2.106,
"step": 376
},
{
"epoch": 0.9687600644122383,
"eval_loss": 2.0219268798828125,
"eval_runtime": 10.2317,
"eval_samples_per_second": 23.945,
"eval_steps_per_second": 3.03,
"step": 376
},
{
"epoch": 0.9790660225442834,
"grad_norm": 2.1220812797546387,
"learning_rate": 3.2795591718381975e-07,
"loss": 2.0164,
"step": 380
},
{
"epoch": 0.9893719806763285,
"grad_norm": 2.2961068153381348,
"learning_rate": 8.201139886109264e-08,
"loss": 2.0596,
"step": 384
},
{
"epoch": 0.9893719806763285,
"eval_loss": 2.021911859512329,
"eval_runtime": 10.235,
"eval_samples_per_second": 23.937,
"eval_steps_per_second": 3.029,
"step": 384
},
{
"epoch": 0.9996779388083736,
"grad_norm": 2.0847766399383545,
"learning_rate": 0.0,
"loss": 2.1024,
"step": 388
}
],
"logging_steps": 4,
"max_steps": 388,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 8,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8.611644452078223e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}