t1msan's picture
End of training
418438e verified
{
"best_metric": 0.000785744923632592,
"best_model_checkpoint": "swin-base-patch4-window7-224-in22k-Kontur-competition-1.3K/checkpoint-334",
"epoch": 9.865470852017937,
"eval_steps": 500,
"global_step": 550,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.18,
"grad_norm": 3.5358009338378906,
"learning_rate": 9.090909090909091e-06,
"loss": 0.674,
"step": 10
},
{
"epoch": 0.36,
"grad_norm": 3.1279354095458984,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.4594,
"step": 20
},
{
"epoch": 0.54,
"grad_norm": 2.273749351501465,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.1591,
"step": 30
},
{
"epoch": 0.72,
"grad_norm": 16.241113662719727,
"learning_rate": 3.6363636363636364e-05,
"loss": 0.0193,
"step": 40
},
{
"epoch": 0.9,
"grad_norm": 5.422853469848633,
"learning_rate": 4.545454545454546e-05,
"loss": 0.0593,
"step": 50
},
{
"epoch": 0.99,
"eval_loss": 0.029350830242037773,
"eval_runtime": 35.1661,
"eval_samples_per_second": 44.986,
"eval_steps_per_second": 0.711,
"step": 55
},
{
"epoch": 1.08,
"grad_norm": 6.060277938842773,
"learning_rate": 4.94949494949495e-05,
"loss": 0.0433,
"step": 60
},
{
"epoch": 1.26,
"grad_norm": 0.930535614490509,
"learning_rate": 4.848484848484849e-05,
"loss": 0.018,
"step": 70
},
{
"epoch": 1.43,
"grad_norm": 4.823692321777344,
"learning_rate": 4.7474747474747476e-05,
"loss": 0.0436,
"step": 80
},
{
"epoch": 1.61,
"grad_norm": 1.3981133699417114,
"learning_rate": 4.6464646464646464e-05,
"loss": 0.0184,
"step": 90
},
{
"epoch": 1.79,
"grad_norm": 0.6671331524848938,
"learning_rate": 4.545454545454546e-05,
"loss": 0.0201,
"step": 100
},
{
"epoch": 1.97,
"grad_norm": 1.8355345726013184,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.0098,
"step": 110
},
{
"epoch": 1.99,
"eval_loss": 0.03146681934595108,
"eval_runtime": 28.3147,
"eval_samples_per_second": 55.872,
"eval_steps_per_second": 0.883,
"step": 111
},
{
"epoch": 2.15,
"grad_norm": 0.02300291508436203,
"learning_rate": 4.343434343434344e-05,
"loss": 0.0131,
"step": 120
},
{
"epoch": 2.33,
"grad_norm": 7.2036871910095215,
"learning_rate": 4.242424242424243e-05,
"loss": 0.0215,
"step": 130
},
{
"epoch": 2.51,
"grad_norm": 1.0436112880706787,
"learning_rate": 4.141414141414142e-05,
"loss": 0.0224,
"step": 140
},
{
"epoch": 2.69,
"grad_norm": 2.077420949935913,
"learning_rate": 4.0404040404040405e-05,
"loss": 0.0133,
"step": 150
},
{
"epoch": 2.87,
"grad_norm": 0.17279785871505737,
"learning_rate": 3.939393939393939e-05,
"loss": 0.0066,
"step": 160
},
{
"epoch": 3.0,
"eval_loss": 0.03221270069479942,
"eval_runtime": 27.4288,
"eval_samples_per_second": 57.677,
"eval_steps_per_second": 0.911,
"step": 167
},
{
"epoch": 3.05,
"grad_norm": 0.07281704246997833,
"learning_rate": 3.838383838383838e-05,
"loss": 0.0034,
"step": 170
},
{
"epoch": 3.23,
"grad_norm": 0.06530273705720901,
"learning_rate": 3.7373737373737376e-05,
"loss": 0.0038,
"step": 180
},
{
"epoch": 3.41,
"grad_norm": 0.2904812693595886,
"learning_rate": 3.6363636363636364e-05,
"loss": 0.0045,
"step": 190
},
{
"epoch": 3.59,
"grad_norm": 0.01948358118534088,
"learning_rate": 3.535353535353535e-05,
"loss": 0.0052,
"step": 200
},
{
"epoch": 3.77,
"grad_norm": 0.24817952513694763,
"learning_rate": 3.434343434343435e-05,
"loss": 0.004,
"step": 210
},
{
"epoch": 3.95,
"grad_norm": 1.0436078310012817,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.0179,
"step": 220
},
{
"epoch": 4.0,
"eval_loss": 0.006767116021364927,
"eval_runtime": 27.0156,
"eval_samples_per_second": 58.559,
"eval_steps_per_second": 0.925,
"step": 223
},
{
"epoch": 4.13,
"grad_norm": 0.4218502342700958,
"learning_rate": 3.232323232323233e-05,
"loss": 0.011,
"step": 230
},
{
"epoch": 4.3,
"grad_norm": 1.2248862981796265,
"learning_rate": 3.131313131313132e-05,
"loss": 0.0043,
"step": 240
},
{
"epoch": 4.48,
"grad_norm": 3.450435161590576,
"learning_rate": 3.0303030303030306e-05,
"loss": 0.0115,
"step": 250
},
{
"epoch": 4.66,
"grad_norm": 0.10841131955385208,
"learning_rate": 2.9292929292929294e-05,
"loss": 0.0189,
"step": 260
},
{
"epoch": 4.84,
"grad_norm": 2.503843307495117,
"learning_rate": 2.8282828282828282e-05,
"loss": 0.0078,
"step": 270
},
{
"epoch": 4.99,
"eval_loss": 0.0032722819596529007,
"eval_runtime": 27.2058,
"eval_samples_per_second": 58.149,
"eval_steps_per_second": 0.919,
"step": 278
},
{
"epoch": 5.02,
"grad_norm": 1.6022576093673706,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.0069,
"step": 280
},
{
"epoch": 5.2,
"grad_norm": 0.23544766008853912,
"learning_rate": 2.6262626262626268e-05,
"loss": 0.0077,
"step": 290
},
{
"epoch": 5.38,
"grad_norm": 1.4383817911148071,
"learning_rate": 2.5252525252525256e-05,
"loss": 0.002,
"step": 300
},
{
"epoch": 5.56,
"grad_norm": 0.105356365442276,
"learning_rate": 2.4242424242424244e-05,
"loss": 0.0017,
"step": 310
},
{
"epoch": 5.74,
"grad_norm": 0.7489631175994873,
"learning_rate": 2.3232323232323232e-05,
"loss": 0.0012,
"step": 320
},
{
"epoch": 5.92,
"grad_norm": 1.5391371250152588,
"learning_rate": 2.2222222222222223e-05,
"loss": 0.0015,
"step": 330
},
{
"epoch": 5.99,
"eval_loss": 0.000785744923632592,
"eval_runtime": 26.43,
"eval_samples_per_second": 59.856,
"eval_steps_per_second": 0.946,
"step": 334
},
{
"epoch": 6.1,
"grad_norm": 0.05875357240438461,
"learning_rate": 2.1212121212121215e-05,
"loss": 0.002,
"step": 340
},
{
"epoch": 6.28,
"grad_norm": 2.8288419246673584,
"learning_rate": 2.0202020202020203e-05,
"loss": 0.0036,
"step": 350
},
{
"epoch": 6.46,
"grad_norm": 0.018499189987778664,
"learning_rate": 1.919191919191919e-05,
"loss": 0.0007,
"step": 360
},
{
"epoch": 6.64,
"grad_norm": 0.00614353409036994,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.0036,
"step": 370
},
{
"epoch": 6.82,
"grad_norm": 0.01476567517966032,
"learning_rate": 1.7171717171717173e-05,
"loss": 0.0041,
"step": 380
},
{
"epoch": 7.0,
"grad_norm": 0.16370485723018646,
"learning_rate": 1.6161616161616165e-05,
"loss": 0.0017,
"step": 390
},
{
"epoch": 7.0,
"eval_loss": 0.007814141921699047,
"eval_runtime": 26.7049,
"eval_samples_per_second": 59.24,
"eval_steps_per_second": 0.936,
"step": 390
},
{
"epoch": 7.17,
"grad_norm": 0.09164857864379883,
"learning_rate": 1.5151515151515153e-05,
"loss": 0.0004,
"step": 400
},
{
"epoch": 7.35,
"grad_norm": 0.19655057787895203,
"learning_rate": 1.4141414141414141e-05,
"loss": 0.0005,
"step": 410
},
{
"epoch": 7.53,
"grad_norm": 0.0026813277509063482,
"learning_rate": 1.3131313131313134e-05,
"loss": 0.0003,
"step": 420
},
{
"epoch": 7.71,
"grad_norm": 0.7142350673675537,
"learning_rate": 1.2121212121212122e-05,
"loss": 0.0005,
"step": 430
},
{
"epoch": 7.89,
"grad_norm": 0.004237661603838205,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.0008,
"step": 440
},
{
"epoch": 8.0,
"eval_loss": 0.002696491777896881,
"eval_runtime": 26.1158,
"eval_samples_per_second": 60.576,
"eval_steps_per_second": 0.957,
"step": 446
},
{
"epoch": 8.07,
"grad_norm": 0.010493074543774128,
"learning_rate": 1.0101010101010101e-05,
"loss": 0.0001,
"step": 450
},
{
"epoch": 8.25,
"grad_norm": 0.015726672485470772,
"learning_rate": 9.090909090909091e-06,
"loss": 0.0003,
"step": 460
},
{
"epoch": 8.43,
"grad_norm": 2.051889657974243,
"learning_rate": 8.080808080808082e-06,
"loss": 0.0023,
"step": 470
},
{
"epoch": 8.61,
"grad_norm": 0.001357908477075398,
"learning_rate": 7.0707070707070704e-06,
"loss": 0.0018,
"step": 480
},
{
"epoch": 8.79,
"grad_norm": 0.0037935995496809483,
"learning_rate": 6.060606060606061e-06,
"loss": 0.0001,
"step": 490
},
{
"epoch": 8.97,
"grad_norm": 0.014398468658328056,
"learning_rate": 5.050505050505051e-06,
"loss": 0.0019,
"step": 500
},
{
"epoch": 8.99,
"eval_loss": 0.0011395993642508984,
"eval_runtime": 26.5634,
"eval_samples_per_second": 59.556,
"eval_steps_per_second": 0.941,
"step": 501
},
{
"epoch": 9.15,
"grad_norm": 0.04990648478269577,
"learning_rate": 4.040404040404041e-06,
"loss": 0.0008,
"step": 510
},
{
"epoch": 9.33,
"grad_norm": 0.009485547430813313,
"learning_rate": 3.0303030303030305e-06,
"loss": 0.0024,
"step": 520
},
{
"epoch": 9.51,
"grad_norm": 0.007252832874655724,
"learning_rate": 2.0202020202020206e-06,
"loss": 0.0007,
"step": 530
},
{
"epoch": 9.69,
"grad_norm": 0.007266469299793243,
"learning_rate": 1.0101010101010103e-06,
"loss": 0.0003,
"step": 540
},
{
"epoch": 9.87,
"grad_norm": 0.01849411241710186,
"learning_rate": 0.0,
"loss": 0.0014,
"step": 550
},
{
"epoch": 9.87,
"eval_loss": 0.0035754255950450897,
"eval_runtime": 37.9782,
"eval_samples_per_second": 41.655,
"eval_steps_per_second": 0.658,
"step": 550
},
{
"epoch": 9.87,
"step": 550,
"total_flos": 1.100143961239951e+19,
"train_loss": 0.03171633874193173,
"train_runtime": 4576.6164,
"train_samples_per_second": 31.093,
"train_steps_per_second": 0.12
}
],
"logging_steps": 10,
"max_steps": 550,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 500,
"total_flos": 1.100143961239951e+19,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}