Error: module 'torch._C' has no attribute '_cuda_resetPeakMemoryStats'

[rmurphy2718]

I tried to run the code in the Quick Start. When I got to this step,

optimizer = torch.optim.Adam(task.parameters(), lr=1e-3)
solver = core.Engine(task, train_set, valid_set, test_set, optimizer,
                     batch_size=1024)
solver.train(num_epoch=5)

I got the bug

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-8-46cb357a6598> in <module>
      2 solver = core.Engine(task, train_set, valid_set, test_set, optimizer,
      3                      batch_size=1024)
----> 4 solver.train(num_epoch=5)

~/torch_drug_test/torchdrug/torchdrug/core/engine.py in train(self, num_epoch, batch_per_epoch)
    129         model.train()
    130 
--> 131         for epoch in self.meter(num_epoch):
    132             sampler.set_epoch(epoch)
    133 

~/torch_drug_test/torchdrug/torchdrug/core/meter.py in __call__(self, num_epoch)
    100                 logger.warning(pretty.separator)
    101                 logger.warning("Epoch %d end" % epoch)
--> 102             self.step()

~/torch_drug_test/torchdrug/torchdrug/core/meter.py in step(self)
     82         logger.warning("ETA: %s" % pretty.time(eta))
     83         logger.warning("max GPU memory: %.1f MiB" % (torch.cuda.max_memory_allocated() / 1e6))
---> 84         torch.cuda.reset_peak_memory_stats()
     85 
     86         logger.warning(pretty.line)

~/anaconda3/envs/torchdrug/lib/python3.8/site-packages/torch/cuda/memory.py in reset_peak_memory_stats(device)
    236     """
    237     device = _get_device_index(device, optional=True)
--> 238     return torch._C._cuda_resetPeakMemoryStats(device)
    239 
    240 

AttributeError: module 'torch._C' has no attribute '_cuda_resetPeakMemoryStats'

It's likely because I'm using the CPU install, and it was easy enough for me to comment out the line

torch.cuda.reset_peak_memory_stats()

in
/torchdrug/core/meter.py. So perhaps one could just add an if statement to see if the GPU is enabled?

Broadly speaking, is this code repo robust for CPU users or is it targeted at GPU only?

[KiddoZhu]

That's a good suggestion! The library is designed for both CPUs and GPUs, but we only tested on CPUs for debug purpose. I think we need to cover more real test scenarios for CPUs.

[KiddoZhu]

Update

Fixed. Now it tests well on basic training and inference. Please tell us if you find new issues about CPU installation.

[rmurphy2718]

@KiddoZhu, I have confirmed that the CPU bug was fixed with a fresh install.

However, when I was installing freshly, I still encountered some problems with Conda. Before going into them (note they were different from the other issue report I raised #1), where is the best place to document them? I feel this thread is inappropriate. Should I open a brand new thread? Whatever you prefer.

For reference, I managed to install using the "from source" instructions.

conda create -n torchdrug2 python=3.7
conda activate torchdrug2
conda install pytorch torchvision torchaudio cpuonly -c pytorch-lts

conda install -c conda-forge rdkit

cd ~
git clone https://github.com/DeepGraphLearning/torchdrug
cd torchdrug

pip install -r requirements.txt
python setup.py install

Then I ran through the test steps in my original installation tutorial
except this time without needing to manually comment out "torch.cuda.reset_peak_memory_stats()". It worked!

[KiddoZhu]

@rmurphy2718 Sorry for missing the reply. You can just open a new thread for that.

If you find some bugs in installation and have some (relatively) elegant solutions, you can also directly open a pull request to modify the document for installation : )