Windows 10 WSL Run Fails

[kiansheik]

If someone else has a guide on getting this to run on WSL that would be helpful! or just regular windows.

But I am running this in WSL and I get this error when I run the command imagine "a dog delivering pizza"

Traceback (most recent call last):
  File "/usr/local/bin/imagine", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/cli.py", line 111, in main
    fire.Fire(train)
  File "/usr/local/lib/python3.8/dist-packages/fire/core.py", line 141, in Fire
    component_trace = _Fire(component, args, parsed_flag_args, context, name)
  File "/usr/local/lib/python3.8/dist-packages/fire/core.py", line 466, in _Fire
    component, remaining_args = _CallAndUpdateTrace(
  File "/usr/local/lib/python3.8/dist-packages/fire/core.py", line 681, in _CallAndUpdateTrace
    component = fn(*varargs, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/cli.py", line 73, in train
    imagine = Imagine(
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/deep_daze.py", line 286, in __init__
    self.clip_encoding = self.create_clip_encoding(text=text, img=img, encoding=clip_encoding)
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/deep_daze.py", line 309, in create_clip_encoding
    encoding = self.create_text_encoding(text)
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/deep_daze.py", line 317, in create_text_encoding
    text_encoding = perceptor.encode_text(tokenized_text).detach()
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/clip.py", line 526, in encode_text
    x = self.transformer(x)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/clip.py", line 381, in forward
    return self.resblocks(x)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/container.py", line 119, in forward
    input = module(input)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/clip.py", line 368, in forward
    x = x + self.attention(self.ln_1(x))
  File "/usr/local/lib/python3.8/dist-packages/deep_daze/clip.py", line 365, in attention
    return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/activation.py", line 980, in forward
    return F.multi_head_attention_forward(
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/functional.py", line 4633, in multi_head_attention_forward
    q, k, v = linear(query, in_proj_weight, in_proj_bias).chunk(3, dim=-1)
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/functional.py", line 1753, in linear
    return torch._C._nn.linear(input, weight, bias)
RuntimeError: CUDA error: CUBLAS_STATUS_INTERNAL_ERROR when calling `cublasCreate(handle)`

[eensander]

I am getting the same error using Python 3.6 on Ubuntu 18.04.5.

...
  File "/home/sander/.local/lib/python3.6/site-packages/torch/nn/functional.py", line 1753, in linear
    return torch._C._nn.linear(input, weight, bias)
RuntimeError: CUDA error: CUBLAS_STATUS_INTERNAL_ERROR when calling `cublasCreate(handle)`

Key details of $ nvidia-smi -q:

Driver Version                        : 450.102.04
CUDA Version                          : 11.0
Product Name                          : GeForce RTX 2060 SUPER

[ashleyha]

I'm also getting the same error with Python 3.7.6 on Ubuntu 18.04.5. CUDA version 11.2, 2x GeForce RTX 2080 Ti.

So far I've tried deleting the on-disk cache with sudo rm -rf ~/.nv and running imagine with the low memory recommendations, but still having the issue.