search

dearleiii / PIRM-2018-SISR-Challenge

Super Resolution
https://www.pirm2018.org/PIRM-SR.html
2 stars 0 forks source link

Index out of range for model resuming and testing #17

Closed dearleiii closed 6 years ago

dearleiii commented 6 years ago 
  (1): LeakyReLU(negative_slope=0.2, inplace)
  (2): Conv2d(8, 16, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1))
  (3): LeakyReLU(negative_slope=0.2, inplace)
  (4): Conv2d(16, 32, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
  (5): LeakyReLU(negative_slope=0.2, inplace)
)
(regressor): Sequential(
  (0): Linear(in_features=1048576, out_features=256, bias=True)
  (1): LeakyReLU(negative_slope=0.01)
  (2): Linear(in_features=256, out_features=1, bias=True)
)

) ) cintpus: torch.Size([300, 3, 1020, 2040]) scores: torch.Size([300, 1]) Traceback (most recent call last): File "load_model_test.py", line 65, in outputs = model1(inputs) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/modules/module.py", line 491, in call result = self.forward(*input, *kwargs) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/data_parallel.py", line 110, in forward inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/data_parallel.py", line 121, in scatter return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/scatter_gather.py", line 36, in scatter_kwargs inputs = scatter(inputs, target_gpus, dim) if inputs else [] File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/scatter_gather.py", line 29, in scatter return scatter_map(inputs) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/scatter_gather.py", line 16, in scatter_map return list(zip(map(scatter_map, obj))) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/scatter_gather.py", line 14, in scatter_map return Scatter.apply(target_gpus, None, dim, obj) File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/_functions.py", line 73, in forward streams = [_get_stream(device) for device in ctx.target_gpus] File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/_functions.py", line 73, in streams = [_get_stream(device) for device in ctx.target_gpus] File "/home/home2/leichen/.local/lib/python3.5/site-packages/torch/nn/parallel/_functions.py", line 100, in _get_stream if _streams[device] is None: IndexError: list index out of range leichen@gpu-compute3$ python3 load_model_test.py

dearleiii commented 6 years ago

Recommended approach for saving a model

There are two main approaches for serializing and restoring a model.

The first (recommended) saves and loads only the model parameters:

torch.save(the_model.state_dict(), PATH) Then later:

the_model = TheModelClass(*args, **kwargs) the_model.load_state_dict(torch.load(PATH)) The second saves and loads the entire model:

torch.save(the_model, PATH) Then later:

the_model = torch.load(PATH) However in this case, the serialized data is bound to the specific classes and the exact directory structure used, so it can break in various ways when used in other projects, or after some serious refactors

dearleiii commented 6 years ago

Seriellization Semantics