2020-1-capstone-design2 / 2016104167

Go to a project
Toggle navigation Toggle navigation pinning
조현아
d96c0c41 1 parent dad343e2

update logs

Inline Side-by-side
Showing 7 changed files with 291 additions and 0 deletions
......@@ -9,6 +9,8 @@ from torch.utils.tensorboard import SummaryWriter
from utils import *
# command
# python "eval.py" --model_path='logs/'
def eval(model_path):
print('\n[+] Parse arguments')
......
{"use_cuda": true, "network": "resnet50", "dataset": "BraTS", "optimizer": "adam", "fast_auto_augment": true, "learning_rate": 0.0001, "seed": null, "num_workers": 4, "print_step": 100, "val_step": 100, "scheduler": "exp", "batch_size": 32, "start_step": 0, "max_step": 500, "augment_path": null}
\ No newline at end of file
DEFALUT_CANDIDATES = [
ShearXY,
TranslateXY,
# Rotate,
# AutoContrast,
# Invert,
Equalize,
Solarize,
Posterize,
# Contrast,
# Color,
Brightness,
Sharpness,
Cutout,
# SamplePairing,
]
[+] Parse arguments
Args(augment_path=None, batch_size=32, dataset='BraTS', fast_auto_augment=True, learning_rate=0.0001, max_step=500, network='resnet50', num_workers=4, optimizer='adam', print_step=100, scheduler='exp', seed=None, start_step=0, use_cuda=True, val_step=100)
[+] Create log dir
[+] Create network
[+] Load dataset
[+] Child 0 training started (GPU: 0)
[+] Training step: 0/500 Elapsed time: 0.04min Learning rate: 9.999283e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 0.000%
Acc@5 : 0.000%
Loss : 7.5784010887146
[+] Training step: 100/500 Elapsed time: 0.48min Learning rate: 9.927842001747633e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 78.125%
Acc@5 : 100.000%
Loss : 0.4084218442440033
[+] Training step: 200/500 Elapsed time: 0.96min Learning rate: 9.856911421715387e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 93.750%
Acc@5 : 100.000%
Loss : 0.2725507915019989
[+] Training step: 300/500 Elapsed time: 1.42min Learning rate: 9.786487613163069e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 90.625%
Acc@5 : 100.000%
Loss : 0.20991499722003937
[+] Training step: 400/500 Elapsed time: 1.88min Learning rate: 9.716566955405027e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 93.750%
Acc@5 : 100.000%
Loss : 0.2204296737909317
100%|????????????????????????????????| 1/1 [00:01<00:00, 1.58s/trial, best loss: 0.8958249092102051]
100%|?????????????????????????????????| 1/1 [00:01<00:00, 1.39s/trial, best loss: 1.509151816368103]
[+] Child 1 training started (GPU: 0)
[+] Training step: 0/500 Elapsed time: 0.03min Learning rate: 9.999283e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 0.000%
Acc@5 : 0.000%
Loss : 7.634987831115723
[+] Training step: 100/500 Elapsed time: 0.48min Learning rate: 9.927842001747633e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 87.500%
Acc@5 : 100.000%
Loss : 0.29290342330932617
[+] Training step: 200/500 Elapsed time: 0.96min Learning rate: 9.856911421715387e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 90.625%
Acc@5 : 100.000%
Loss : 0.28638142347335815
[+] Training step: 300/500 Elapsed time: 1.42min Learning rate: 9.786487613163069e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 96.875%
Acc@5 : 100.000%
Loss : 0.06958930194377899
[+] Training step: 400/500 Elapsed time: 1.88min Learning rate: 9.716566955405027e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 100.000%
Acc@5 : 100.000%
Loss : 0.030036240816116333
100%|??????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.54s/trial, best loss: 2.1128218173980713]
100%|??????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.50s/trial, best loss: 1.9411643743515015]
[+] Child 2 training started (GPU: 0)
[+] Training step: 0/500 Elapsed time: 0.03min Learning rate: 9.999283e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 0.000%
Acc@5 : 0.000%
Loss : 7.582807540893555
[+] Training step: 100/500 Elapsed time: 0.49min Learning rate: 9.927842001747633e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 75.000%
Acc@5 : 100.000%
Loss : 0.5312898755073547
[+] Training step: 200/500 Elapsed time: 0.98min Learning rate: 9.856911421715387e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 84.375%
Acc@5 : 100.000%
Loss : 0.4784519672393799
[+] Training step: 300/500 Elapsed time: 1.45min Learning rate: 9.786487613163069e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 100.000%
Acc@5 : 100.000%
Loss : 0.03968067467212677
[+] Training step: 400/500 Elapsed time: 1.89min Learning rate: 9.716566955405027e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 100.000%
Acc@5 : 100.000%
Loss : 0.025451302528381348
100%|??????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.53s/trial, best loss: 2.5077414512634277]
100%|???????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.59s/trial, best loss: 4.707443714141846]
[+] Child 3 training started (GPU: 0)
[+] Training step: 0/500 Elapsed time: 0.03min Learning rate: 9.999283e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 0.000%
Acc@5 : 0.000%
Loss : 7.614710807800293
[+] Training step: 100/500 Elapsed time: 0.49min Learning rate: 9.927842001747633e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 75.000%
Acc@5 : 100.000%
Loss : 0.46335405111312866
[+] Training step: 200/500 Elapsed time: 0.96min Learning rate: 9.856911421715387e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 90.625%
Acc@5 : 100.000%
Loss : 0.16135810315608978
[+] Training step: 300/500 Elapsed time: 1.44min Learning rate: 9.786487613163069e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 84.375%
Acc@5 : 100.000%
Loss : 0.4632360339164734
[+] Training step: 400/500 Elapsed time: 1.90min Learning rate: 9.716566955405027e-05 Device name: GeForce GTX 1080 Ti
Acc@1 : 100.000%
Acc@5 : 100.000%
Loss : 0.04105471074581146
100%|???????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.55s/trial, best loss: 2.492347240447998]
100%|??????????????????????????????????????????????????????????| 1/1 [00:01<00:00, 1.56s/trial, best loss: 2.6143996715545654]
RandomChoice(
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Brightness(prob=0.47, magnitude=0.06)
Sharpness(prob=0.52, magnitude=0.28)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Solarize(prob=0.70, magnitude=0.03)
Sharpness(prob=0.98, magnitude=0.62)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Posterize(prob=0.08, magnitude=0.88)
Solarize(prob=0.98, magnitude=0.76)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Posterize(prob=0.37, magnitude=0.60)
Cutout(prob=0.75, magnitude=0.83)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
ShearXY(prob=0.56, magnitude=0.86)
Cutout(prob=0.37, magnitude=0.00)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Sharpness(prob=0.09, magnitude=0.75)
Equalize(prob=0.70, magnitude=0.90)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
TranslateXY(prob=0.67, magnitude=0.95)
Posterize(prob=0.31, magnitude=0.92)
ToTensor()
)
Compose(
Pad(padding=4, fill=0, padding_mode=constant)
RandomCrop(size=(32, 32), padding=None)
RandomHorizontalFlip(p=0.5)
Equalize(prob=0.32, magnitude=0.07)
Posterize(prob=0.83, magnitude=0.82)
ToTensor()
)
)
[+] Start training
[+] Use 1 GPUs
[+] Using GPU: GeForce GTX 1080 Ti
/opt/conda/lib/python3.7/site-packages/torch/nn/functional.py:2457: UserWarning: nn.functional.upsample is deprecated. Use nn.functional.interpolate instead.
warnings.warn("nn.functional.upsample is deprecated. Use nn.functional.interpolate instead.")
/opt/conda/lib/python3.7/site-packages/torch/nn/functional.py:2539: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details.
"See the documentation of nn.Upsample for details.".format(mode))
[+] Training step: 0/500 Training epoch: 0/64 Elapsed time: 0.08min Learning rate: 9.999283e-05
Acc@1 : 0.000%
Acc@5 : 0.000%
Loss : 7.677338123321533
FW Time : 200.173ms
BW Time : 22.200ms
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 17.083898544311523
[+] Model saved
[+] Training step: 100/500 Training epoch: 0/64 Elapsed time: 0.88min Learning rate: 9.927842001747633e-05
Acc@1 : 34.375%
Acc@5 : 100.000%
Loss : 0.9044204950332642
FW Time : 20.693ms
BW Time : 19.411ms
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 10.917157173156738
[+] Model saved
[+] Training step: 200/500 Training epoch: 0/64 Elapsed time: 1.74min Learning rate: 9.856911421715387e-05
Acc@1 : 34.375%
Acc@5 : 100.000%
Loss : 0.7641889452934265
FW Time : 20.088ms
BW Time : 9.569ms
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 26.895051956176758
[+] Model saved
[+] Training step: 300/500 Training epoch: 0/64 Elapsed time: 2.57min Learning rate: 9.786487613163069e-05
Acc@1 : 56.250%
Acc@5 : 100.000%
Loss : 0.8696596622467041
FW Time : 19.580ms
BW Time : 9.993ms
OMP: Warning #190: Forking a process while a parallel region is active is potentially unsafe.
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 11.694602966308594
[+] Model saved
[+] Training step: 400/500 Training epoch: 0/64 Elapsed time: 3.43min Learning rate: 9.716566955405027e-05
Acc@1 : 71.875%
Acc@5 : 100.000%
Loss : 0.7189279198646545
FW Time : 19.634ms
BW Time : 16.867ms
OMP: Warning #190: Forking a process while a parallel region is active is potentially unsafe.
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 12.062773704528809
[+] Valid results
Acc@1 : 47.767%
Acc@5 : 100.000%
Loss : 12.063
Infer Time(per image) : 2.722ms
\ No newline at end of file