-
I encountered an error with the adjoint of `getindex` of a [special
matrix](https://docs.julialang.org/en/v1/stdlib/LinearAlgebra/#Special-matrices-1). The issue is that the current `getindex` adjoin…
-
### Describe the bug
I followed the instructions to train the dreambooth model, but the RuntimeError: CUDA error: invalid argument occured at accelerator.backward(loss)
My device is RTX-3090, and I …
-
Running train_text_to_image_lora.py with the following command:
['C:\\Python3109\\python.exe', 'train_text_to_image_lora.py', '--pretrained_model_name_or_path=stabilityai/stable-diffusion-2-1', '--…
-
PR #565
While trying to convert the compact convolution transformer to backend agnostic, I am facing issue working with the PyTorch backend.
The model works with tensorflow/jax backends but throws a…
-
## Describe the bug
`ot.optim.generic_conditional_gradient` uses the following cost function for line search (when `reg2` is provided):
```python
def cost(G):
return nx.sum(M * G) + reg1 *…
-
bitsandbytes does not support Windows by default.
I guess that's why I ran into below issue:
```
RuntimeError:
CUDA Setup failed despite GPU being available. Please run the following co…
-
I have two questions around the current codebase as it stands.
1. How does chunked cross entropy loss work? How do we scale the value we get from that to ordinary Cross entropy loss?
2. When using…
-
SciPy sparse matrices (i.e. `scipy.sparse.*matrix`) are tested but their sparse arrays counterpart (i.e. `scipy.sparse.*array`) aren't yet will become ubiquitous (see #26418).
Tests and their param…
-
I have finsh the torch2onnx, and the value is same with torch version, like this
torch result
![企业微信截图_16685886725453](https://user-images.githubusercontent.com/26165705/202218941-33a8c280-8f50-4b23…
-
**Describe the bug**
After building the lib and building the quick example the map is not visible.
**To Reproduce**
Steps to reproduce the behavior:
1. Built the lib with: ~/Qt6/6.5.3/gcc_64/bin…