Text Generation
Safetensors
English
Russian
conversational
zephyr-python-ru / trainer_state.json
MexIvanov's picture
Upload 15 files
f5768e8
{
"best_metric": 0.4947710931301117,
"best_model_checkpoint": "zephyr_7B_ruPython/checkpoint-3466",
"epoch": 1.9997115245925285,
"eval_steps": 500,
"global_step": 3466,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 7.692307692307694e-06,
"loss": 1.0975,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 1.5384615384615387e-05,
"loss": 1.083,
"step": 40
},
{
"epoch": 0.03,
"learning_rate": 2.307692307692308e-05,
"loss": 1.0227,
"step": 60
},
{
"epoch": 0.05,
"learning_rate": 3.0769230769230774e-05,
"loss": 0.9524,
"step": 80
},
{
"epoch": 0.06,
"learning_rate": 3.846153846153846e-05,
"loss": 0.8904,
"step": 100
},
{
"epoch": 0.07,
"learning_rate": 4.615384615384616e-05,
"loss": 0.8482,
"step": 120
},
{
"epoch": 0.08,
"learning_rate": 5.384615384615385e-05,
"loss": 0.8015,
"step": 140
},
{
"epoch": 0.09,
"learning_rate": 6.153846153846155e-05,
"loss": 0.8032,
"step": 160
},
{
"epoch": 0.1,
"learning_rate": 6.923076923076924e-05,
"loss": 0.7839,
"step": 180
},
{
"epoch": 0.12,
"learning_rate": 7.692307692307693e-05,
"loss": 0.7964,
"step": 200
},
{
"epoch": 0.13,
"learning_rate": 8.461538461538461e-05,
"loss": 0.7531,
"step": 220
},
{
"epoch": 0.14,
"learning_rate": 9.230769230769232e-05,
"loss": 0.7508,
"step": 240
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 0.7531,
"step": 260
},
{
"epoch": 0.16,
"learning_rate": 0.0001076923076923077,
"loss": 0.7543,
"step": 280
},
{
"epoch": 0.17,
"learning_rate": 0.00011538461538461538,
"loss": 0.7426,
"step": 300
},
{
"epoch": 0.18,
"learning_rate": 0.0001230769230769231,
"loss": 0.7222,
"step": 320
},
{
"epoch": 0.2,
"learning_rate": 0.00013076923076923077,
"loss": 0.7456,
"step": 340
},
{
"epoch": 0.21,
"learning_rate": 0.00013846153846153847,
"loss": 0.7447,
"step": 360
},
{
"epoch": 0.22,
"learning_rate": 0.00014615384615384615,
"loss": 0.7297,
"step": 380
},
{
"epoch": 0.23,
"learning_rate": 0.00015384615384615385,
"loss": 0.7339,
"step": 400
},
{
"epoch": 0.24,
"learning_rate": 0.00016153846153846155,
"loss": 0.7418,
"step": 420
},
{
"epoch": 0.25,
"learning_rate": 0.00016923076923076923,
"loss": 0.7315,
"step": 440
},
{
"epoch": 0.27,
"learning_rate": 0.00017692307692307693,
"loss": 0.7104,
"step": 460
},
{
"epoch": 0.28,
"learning_rate": 0.00018461538461538463,
"loss": 0.717,
"step": 480
},
{
"epoch": 0.29,
"learning_rate": 0.00019230769230769233,
"loss": 0.7307,
"step": 500
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.728,
"step": 520
},
{
"epoch": 0.31,
"learning_rate": 0.0001991451164778799,
"loss": 0.7171,
"step": 540
},
{
"epoch": 0.32,
"learning_rate": 0.0001982902329557598,
"loss": 0.7137,
"step": 560
},
{
"epoch": 0.33,
"learning_rate": 0.0001974353494336397,
"loss": 0.6911,
"step": 580
},
{
"epoch": 0.35,
"learning_rate": 0.00019658046591151956,
"loss": 0.6992,
"step": 600
},
{
"epoch": 0.36,
"learning_rate": 0.00019572558238939946,
"loss": 0.6959,
"step": 620
},
{
"epoch": 0.37,
"learning_rate": 0.00019487069886727932,
"loss": 0.7046,
"step": 640
},
{
"epoch": 0.38,
"learning_rate": 0.00019401581534515922,
"loss": 0.6976,
"step": 660
},
{
"epoch": 0.39,
"learning_rate": 0.0001931609318230391,
"loss": 0.7042,
"step": 680
},
{
"epoch": 0.4,
"learning_rate": 0.000192306048300919,
"loss": 0.6908,
"step": 700
},
{
"epoch": 0.42,
"learning_rate": 0.0001914511647787989,
"loss": 0.6984,
"step": 720
},
{
"epoch": 0.43,
"learning_rate": 0.0001905962812566788,
"loss": 0.6999,
"step": 740
},
{
"epoch": 0.44,
"learning_rate": 0.0001897413977345587,
"loss": 0.7094,
"step": 760
},
{
"epoch": 0.45,
"learning_rate": 0.00018888651421243856,
"loss": 0.7003,
"step": 780
},
{
"epoch": 0.46,
"learning_rate": 0.00018803163069031845,
"loss": 0.6809,
"step": 800
},
{
"epoch": 0.47,
"learning_rate": 0.00018717674716819835,
"loss": 0.7135,
"step": 820
},
{
"epoch": 0.48,
"learning_rate": 0.00018632186364607822,
"loss": 0.6913,
"step": 840
},
{
"epoch": 0.5,
"learning_rate": 0.0001854669801239581,
"loss": 0.6932,
"step": 860
},
{
"epoch": 0.51,
"learning_rate": 0.000184612096601838,
"loss": 0.6832,
"step": 880
},
{
"epoch": 0.52,
"learning_rate": 0.0001837572130797179,
"loss": 0.6927,
"step": 900
},
{
"epoch": 0.53,
"learning_rate": 0.0001829023295575978,
"loss": 0.6729,
"step": 920
},
{
"epoch": 0.54,
"learning_rate": 0.00018204744603547766,
"loss": 0.6742,
"step": 940
},
{
"epoch": 0.55,
"learning_rate": 0.00018119256251335756,
"loss": 0.6718,
"step": 960
},
{
"epoch": 0.57,
"learning_rate": 0.00018033767899123745,
"loss": 0.6868,
"step": 980
},
{
"epoch": 0.58,
"learning_rate": 0.00017948279546911735,
"loss": 0.6869,
"step": 1000
},
{
"epoch": 0.59,
"learning_rate": 0.00017862791194699724,
"loss": 0.6633,
"step": 1020
},
{
"epoch": 0.6,
"learning_rate": 0.0001777730284248771,
"loss": 0.6967,
"step": 1040
},
{
"epoch": 0.61,
"learning_rate": 0.000176918144902757,
"loss": 0.6758,
"step": 1060
},
{
"epoch": 0.62,
"learning_rate": 0.0001760632613806369,
"loss": 0.6918,
"step": 1080
},
{
"epoch": 0.63,
"learning_rate": 0.0001752083778585168,
"loss": 0.6718,
"step": 1100
},
{
"epoch": 0.65,
"learning_rate": 0.00017435349433639666,
"loss": 0.6722,
"step": 1120
},
{
"epoch": 0.66,
"learning_rate": 0.00017349861081427656,
"loss": 0.684,
"step": 1140
},
{
"epoch": 0.67,
"learning_rate": 0.00017264372729215645,
"loss": 0.6619,
"step": 1160
},
{
"epoch": 0.68,
"learning_rate": 0.00017178884377003635,
"loss": 0.6789,
"step": 1180
},
{
"epoch": 0.69,
"learning_rate": 0.00017093396024791624,
"loss": 0.677,
"step": 1200
},
{
"epoch": 0.7,
"learning_rate": 0.00017007907672579614,
"loss": 0.6836,
"step": 1220
},
{
"epoch": 0.72,
"learning_rate": 0.000169266937379782,
"loss": 0.6823,
"step": 1240
},
{
"epoch": 0.73,
"learning_rate": 0.0001684120538576619,
"loss": 0.6936,
"step": 1260
},
{
"epoch": 0.74,
"learning_rate": 0.0001675571703355418,
"loss": 0.6646,
"step": 1280
},
{
"epoch": 0.75,
"learning_rate": 0.0001667022868134217,
"loss": 0.6733,
"step": 1300
},
{
"epoch": 0.76,
"learning_rate": 0.00016584740329130159,
"loss": 0.6753,
"step": 1320
},
{
"epoch": 0.77,
"learning_rate": 0.00016499251976918145,
"loss": 0.677,
"step": 1340
},
{
"epoch": 0.78,
"learning_rate": 0.00016413763624706135,
"loss": 0.6692,
"step": 1360
},
{
"epoch": 0.8,
"learning_rate": 0.00016328275272494122,
"loss": 0.656,
"step": 1380
},
{
"epoch": 0.81,
"learning_rate": 0.0001624278692028211,
"loss": 0.6767,
"step": 1400
},
{
"epoch": 0.82,
"learning_rate": 0.000161572985680701,
"loss": 0.6485,
"step": 1420
},
{
"epoch": 0.83,
"learning_rate": 0.0001607181021585809,
"loss": 0.679,
"step": 1440
},
{
"epoch": 0.84,
"learning_rate": 0.0001598632186364608,
"loss": 0.6701,
"step": 1460
},
{
"epoch": 0.85,
"learning_rate": 0.0001590083351143407,
"loss": 0.6632,
"step": 1480
},
{
"epoch": 0.87,
"learning_rate": 0.00015815345159222056,
"loss": 0.6707,
"step": 1500
},
{
"epoch": 0.88,
"learning_rate": 0.00015729856807010045,
"loss": 0.6738,
"step": 1520
},
{
"epoch": 0.89,
"learning_rate": 0.00015644368454798035,
"loss": 0.68,
"step": 1540
},
{
"epoch": 0.9,
"learning_rate": 0.00015558880102586024,
"loss": 0.6748,
"step": 1560
},
{
"epoch": 0.91,
"learning_rate": 0.00015473391750374014,
"loss": 0.6628,
"step": 1580
},
{
"epoch": 0.92,
"learning_rate": 0.00015387903398162,
"loss": 0.6674,
"step": 1600
},
{
"epoch": 0.93,
"learning_rate": 0.0001530241504594999,
"loss": 0.6572,
"step": 1620
},
{
"epoch": 0.95,
"learning_rate": 0.0001521692669373798,
"loss": 0.6547,
"step": 1640
},
{
"epoch": 0.96,
"learning_rate": 0.0001513143834152597,
"loss": 0.6599,
"step": 1660
},
{
"epoch": 0.97,
"learning_rate": 0.00015045949989313955,
"loss": 0.6665,
"step": 1680
},
{
"epoch": 0.98,
"learning_rate": 0.00014960461637101945,
"loss": 0.6687,
"step": 1700
},
{
"epoch": 0.99,
"learning_rate": 0.00014874973284889934,
"loss": 0.6655,
"step": 1720
},
{
"epoch": 1.0,
"eval_loss": 0.5166412591934204,
"eval_runtime": 3665.2848,
"eval_samples_per_second": 1.081,
"eval_steps_per_second": 0.36,
"step": 1733
},
{
"epoch": 1.0,
"learning_rate": 0.00014789484932677924,
"loss": 0.673,
"step": 1740
},
{
"epoch": 1.02,
"learning_rate": 0.00014703996580465913,
"loss": 0.6603,
"step": 1760
},
{
"epoch": 1.03,
"learning_rate": 0.00014618508228253903,
"loss": 0.6548,
"step": 1780
},
{
"epoch": 1.04,
"learning_rate": 0.0001453301987604189,
"loss": 0.6708,
"step": 1800
},
{
"epoch": 1.05,
"learning_rate": 0.0001444753152382988,
"loss": 0.6668,
"step": 1820
},
{
"epoch": 1.06,
"learning_rate": 0.00014362043171617866,
"loss": 0.6901,
"step": 1840
},
{
"epoch": 1.07,
"learning_rate": 0.00014276554819405855,
"loss": 0.6648,
"step": 1860
},
{
"epoch": 1.08,
"learning_rate": 0.00014191066467193845,
"loss": 0.6431,
"step": 1880
},
{
"epoch": 1.1,
"learning_rate": 0.00014105578114981834,
"loss": 0.6653,
"step": 1900
},
{
"epoch": 1.11,
"learning_rate": 0.00014020089762769824,
"loss": 0.6693,
"step": 1920
},
{
"epoch": 1.12,
"learning_rate": 0.00013934601410557813,
"loss": 0.6732,
"step": 1940
},
{
"epoch": 1.13,
"learning_rate": 0.00013849113058345803,
"loss": 0.6434,
"step": 1960
},
{
"epoch": 1.14,
"learning_rate": 0.00013763624706133792,
"loss": 0.6429,
"step": 1980
},
{
"epoch": 1.15,
"learning_rate": 0.0001367813635392178,
"loss": 0.6619,
"step": 2000
},
{
"epoch": 1.17,
"learning_rate": 0.00013592648001709766,
"loss": 0.6561,
"step": 2020
},
{
"epoch": 1.18,
"learning_rate": 0.00013507159649497755,
"loss": 0.6518,
"step": 2040
},
{
"epoch": 1.19,
"learning_rate": 0.00013421671297285745,
"loss": 0.6358,
"step": 2060
},
{
"epoch": 1.2,
"learning_rate": 0.00013336182945073734,
"loss": 0.6711,
"step": 2080
},
{
"epoch": 1.21,
"learning_rate": 0.00013250694592861724,
"loss": 0.6543,
"step": 2100
},
{
"epoch": 1.22,
"learning_rate": 0.00013165206240649713,
"loss": 0.6517,
"step": 2120
},
{
"epoch": 1.23,
"learning_rate": 0.00013079717888437703,
"loss": 0.6549,
"step": 2140
},
{
"epoch": 1.25,
"learning_rate": 0.0001299422953622569,
"loss": 0.6558,
"step": 2160
},
{
"epoch": 1.26,
"learning_rate": 0.0001290874118401368,
"loss": 0.6585,
"step": 2180
},
{
"epoch": 1.27,
"learning_rate": 0.00012823252831801668,
"loss": 0.6444,
"step": 2200
},
{
"epoch": 1.28,
"learning_rate": 0.00012737764479589655,
"loss": 0.642,
"step": 2220
},
{
"epoch": 1.29,
"learning_rate": 0.00012652276127377645,
"loss": 0.6708,
"step": 2240
},
{
"epoch": 1.3,
"learning_rate": 0.00012566787775165634,
"loss": 0.653,
"step": 2260
},
{
"epoch": 1.32,
"learning_rate": 0.00012481299422953623,
"loss": 0.6438,
"step": 2280
},
{
"epoch": 1.33,
"learning_rate": 0.00012395811070741613,
"loss": 0.6296,
"step": 2300
},
{
"epoch": 1.34,
"learning_rate": 0.00012310322718529602,
"loss": 0.6376,
"step": 2320
},
{
"epoch": 1.35,
"learning_rate": 0.0001222483436631759,
"loss": 0.6284,
"step": 2340
},
{
"epoch": 1.36,
"learning_rate": 0.00012139346014105579,
"loss": 0.6373,
"step": 2360
},
{
"epoch": 1.37,
"learning_rate": 0.00012053857661893568,
"loss": 0.6289,
"step": 2380
},
{
"epoch": 1.38,
"learning_rate": 0.00011968369309681558,
"loss": 0.6421,
"step": 2400
},
{
"epoch": 1.4,
"learning_rate": 0.00011882880957469544,
"loss": 0.6385,
"step": 2420
},
{
"epoch": 1.41,
"learning_rate": 0.00011797392605257534,
"loss": 0.6312,
"step": 2440
},
{
"epoch": 1.42,
"learning_rate": 0.00011711904253045522,
"loss": 0.6454,
"step": 2460
},
{
"epoch": 1.43,
"learning_rate": 0.00011626415900833511,
"loss": 0.6476,
"step": 2480
},
{
"epoch": 1.44,
"learning_rate": 0.00011540927548621501,
"loss": 0.6373,
"step": 2500
},
{
"epoch": 1.45,
"learning_rate": 0.0001145543919640949,
"loss": 0.647,
"step": 2520
},
{
"epoch": 1.47,
"learning_rate": 0.00011369950844197479,
"loss": 0.6319,
"step": 2540
},
{
"epoch": 1.48,
"learning_rate": 0.00011284462491985468,
"loss": 0.6506,
"step": 2560
},
{
"epoch": 1.49,
"learning_rate": 0.00011198974139773457,
"loss": 0.637,
"step": 2580
},
{
"epoch": 1.5,
"learning_rate": 0.00011113485787561446,
"loss": 0.6339,
"step": 2600
},
{
"epoch": 1.51,
"learning_rate": 0.00011027997435349434,
"loss": 0.6316,
"step": 2620
},
{
"epoch": 1.52,
"learning_rate": 0.00010942509083137422,
"loss": 0.6325,
"step": 2640
},
{
"epoch": 1.53,
"learning_rate": 0.00010857020730925411,
"loss": 0.6202,
"step": 2660
},
{
"epoch": 1.55,
"learning_rate": 0.00010771532378713401,
"loss": 0.6219,
"step": 2680
},
{
"epoch": 1.56,
"learning_rate": 0.00010686044026501389,
"loss": 0.6367,
"step": 2700
},
{
"epoch": 1.57,
"learning_rate": 0.00010600555674289378,
"loss": 0.6374,
"step": 2720
},
{
"epoch": 1.58,
"learning_rate": 0.00010515067322077368,
"loss": 0.6235,
"step": 2740
},
{
"epoch": 1.59,
"learning_rate": 0.00010429578969865357,
"loss": 0.626,
"step": 2760
},
{
"epoch": 1.6,
"learning_rate": 0.00010344090617653345,
"loss": 0.6325,
"step": 2780
},
{
"epoch": 1.62,
"learning_rate": 0.00010258602265441335,
"loss": 0.6425,
"step": 2800
},
{
"epoch": 1.63,
"learning_rate": 0.00010173113913229324,
"loss": 0.6344,
"step": 2820
},
{
"epoch": 1.64,
"learning_rate": 0.00010087625561017311,
"loss": 0.6208,
"step": 2840
},
{
"epoch": 1.65,
"learning_rate": 0.000100021372088053,
"loss": 0.6339,
"step": 2860
},
{
"epoch": 1.66,
"learning_rate": 9.916648856593289e-05,
"loss": 0.6348,
"step": 2880
},
{
"epoch": 1.67,
"learning_rate": 9.831160504381278e-05,
"loss": 0.6193,
"step": 2900
},
{
"epoch": 1.68,
"learning_rate": 9.745672152169268e-05,
"loss": 0.6283,
"step": 2920
},
{
"epoch": 1.7,
"learning_rate": 9.660183799957256e-05,
"loss": 0.6314,
"step": 2940
},
{
"epoch": 1.71,
"learning_rate": 9.574695447745245e-05,
"loss": 0.6373,
"step": 2960
},
{
"epoch": 1.72,
"learning_rate": 9.489207095533233e-05,
"loss": 0.6437,
"step": 2980
},
{
"epoch": 1.73,
"learning_rate": 9.403718743321223e-05,
"loss": 0.6433,
"step": 3000
},
{
"epoch": 1.74,
"learning_rate": 9.318230391109212e-05,
"loss": 0.6238,
"step": 3020
},
{
"epoch": 1.75,
"learning_rate": 9.2327420388972e-05,
"loss": 0.6237,
"step": 3040
},
{
"epoch": 1.77,
"learning_rate": 9.14725368668519e-05,
"loss": 0.6388,
"step": 3060
},
{
"epoch": 1.78,
"learning_rate": 9.06176533447318e-05,
"loss": 0.6422,
"step": 3080
},
{
"epoch": 1.79,
"learning_rate": 8.976276982261168e-05,
"loss": 0.5997,
"step": 3100
},
{
"epoch": 1.8,
"learning_rate": 8.890788630049156e-05,
"loss": 0.6265,
"step": 3120
},
{
"epoch": 1.81,
"learning_rate": 8.805300277837145e-05,
"loss": 0.6298,
"step": 3140
},
{
"epoch": 1.82,
"learning_rate": 8.719811925625135e-05,
"loss": 0.622,
"step": 3160
},
{
"epoch": 1.83,
"learning_rate": 8.634323573413123e-05,
"loss": 0.6232,
"step": 3180
},
{
"epoch": 1.85,
"learning_rate": 8.548835221201111e-05,
"loss": 0.6337,
"step": 3200
},
{
"epoch": 1.86,
"learning_rate": 8.4633468689891e-05,
"loss": 0.6187,
"step": 3220
},
{
"epoch": 1.87,
"learning_rate": 8.37785851677709e-05,
"loss": 0.6446,
"step": 3240
},
{
"epoch": 1.88,
"learning_rate": 8.292370164565079e-05,
"loss": 0.6298,
"step": 3260
},
{
"epoch": 1.89,
"learning_rate": 8.206881812353067e-05,
"loss": 0.6364,
"step": 3280
},
{
"epoch": 1.9,
"learning_rate": 8.121393460141055e-05,
"loss": 0.6314,
"step": 3300
},
{
"epoch": 1.92,
"learning_rate": 8.035905107929045e-05,
"loss": 0.633,
"step": 3320
},
{
"epoch": 1.93,
"learning_rate": 7.950416755717034e-05,
"loss": 0.6163,
"step": 3340
},
{
"epoch": 1.94,
"learning_rate": 7.864928403505023e-05,
"loss": 0.6351,
"step": 3360
},
{
"epoch": 1.95,
"learning_rate": 7.779440051293012e-05,
"loss": 0.6038,
"step": 3380
},
{
"epoch": 1.96,
"learning_rate": 7.693951699081e-05,
"loss": 0.6245,
"step": 3400
},
{
"epoch": 1.97,
"learning_rate": 7.60846334686899e-05,
"loss": 0.6311,
"step": 3420
},
{
"epoch": 1.98,
"learning_rate": 7.522974994656978e-05,
"loss": 0.6391,
"step": 3440
},
{
"epoch": 2.0,
"learning_rate": 7.437486642444967e-05,
"loss": 0.6205,
"step": 3460
},
{
"epoch": 2.0,
"eval_loss": 0.4947710931301117,
"eval_runtime": 3649.3488,
"eval_samples_per_second": 1.086,
"eval_steps_per_second": 0.362,
"step": 3466
}
],
"logging_steps": 20,
"max_steps": 5199,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 3.6485939432993587e+18,
"trial_name": null,
"trial_params": null
}