Inference on Image and CUDA out of memory

[SanKumSan]

#DVK
!python test.py -opt options/df2k/test_df2k.yml

{'name': 'Track1', 'suffix': None, 'model': 'srgan', 'distortion': 'sr', 'scale': 4, 'crop_border': None, 'gpu_ids': [0], 'datasets': {'test_1': {'name': 'DIV2K', 'mode': 'LR', 'dataroot_LR': '/content/drive/MyDrive/RealSR/Real-SR/codes/testimages/images', 'phase': 'test', 'scale': 4, 'data_type': 'img'}}, 'network_G': {'which_model_G': 'RRDBNet', 'in_nc': 3, 'out_nc': 3, 'nf': 64, 'nb': 23, 'upscale': 4, 'scale': 4}, 'path': {'pretrain_model_G': '/content/drive/MyDrive/RealSR/Real-SR/codes/pretrained_model/DF2K.pth', 'results_root': '/content/drive/MyDrive/RealSR/Real-SR/results/Track1', 'root': '/content/drive/MyDrive/RealSR/Real-SR', 'log': '/content/drive/MyDrive/RealSR/Real-SR/results/Track1'}, 'is_train': False}
22-06-09 05:38:07.840 - INFO: name: Track1
suffix: None
model: srgan
distortion: sr
scale: 4
crop_border: None
gpu_ids: [0]
datasets:[
test_1:[
name: DIV2K
mode: LR
dataroot_LR: /content/drive/MyDrive/RealSR/Real-SR/codes/testimages/images
phase: test
scale: 4
data_type: img
]
]
network_G:[
which_model_G: RRDBNet
in_nc: 3
out_nc: 3
nf: 64
nb: 23
upscale: 4
scale: 4
]
path:[
pretrain_model_G: /content/drive/MyDrive/RealSR/Real-SR/codes/pretrained_model/DF2K.pth
results_root: /content/drive/MyDrive/RealSR/Real-SR/results/Track1
root: /content/drive/MyDrive/RealSR/Real-SR
log: /content/drive/MyDrive/RealSR/Real-SR/results/Track1
]
is_train: False

22-06-09 05:38:09.095 - INFO: Dataset [LRDataset - DIV2K] is created.
22-06-09 05:38:09.096 - INFO: Number of test images in [DIV2K]: 1
22-06-09 05:38:30.196 - INFO: Loading model for G [/content/drive/MyDrive/RealSR/Real-SR/codes/pretrained_model/DF2K.pth] ...
22-06-09 05:38:33.049 - INFO: Model [SRGANModel] is created.
22-06-09 05:38:33.049 - INFO:
Testing [DIV2K]...
Traceback (most recent call last):
File "test.py", line 61, in
model.test()
File "/content/drive/MyDrive/RealSR/Real-SR/codes/models/SRGAN_model.py", line 205, in test
self.fake_H = self.netG(self.var_L)
File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/usr/local/lib/python3.7/dist-packages/torch/nn/parallel/data_parallel.py", line 166, in forward
return self.module(*inputs[0], **kwargs[0])
File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/content/drive/MyDrive/RealSR/Real-SR/codes/models/modules/RRDBNet_arch.py", line 69, in forward
fea = self.lrelu(self.upconv2(F.interpolate(fea, scale_factor=2, mode='nearest')))
File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py", line 447, in forward
return self._conv_forward(input, self.weight, self.bias)
File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py", line 444, in _conv_forward
self.padding, self.dilation, self.groups)
RuntimeError: CUDA out of memory. Tried to allocate 7.97 GiB (GPU 0; 14.76 GiB total capacity; 10.54 GiB already allocated; 2.23 GiB free; 11.29 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF