-
Currently the tensor tracking algorithms do an unweighted linear least squares fit on the trilinear-interpolated diffusion data to obtain the tensor at each scanner space point.
Instead, the `tensor_…
-
Dear J. George,
In [the atomate2 elastic workflow example](https://github.com/JaGeo/TutorialAtomate2Forcefields/blob/main/atomate2_workflow_tutorial/elastic.ipynb), you obtained [the following elas…
-
I was struggling with this for a while to get the base64 image to show. I ended up having to change the `send_images` https://github.com/Acly/comfyui-tooling-nodes/blob/ef5ccfa98f035fd1d9fd00a21ac4ab7…
-
As the title says, it may be a bug in nccl on the H20 GPU.
Here are some discussions: https://github.com/pytorch/pytorch/issues/138811
I have implemented a pure cpp version and can reproduce the pro…
-
I'm currently working with the code and I'm having some trouble OVERRIDE forward where the self.module.image_newline attribute is set or initialized in the model.
I've traced the model through the …
-
When attempting to use the code to infer with a 512x320 images + video, I get this error:
```
File "/workspace/ToonCrafter_with_SketchGuidance/cldm/cldm.py", line 339, in forward
h += gui…
-
### Issue
We propose fine-grain speculative execution on metal. Specifically, we introduce a new type of parallelism where devices are used to pre-execute ops from speculative output.
- Speculative…
-
On f82029c5cf6367a1070e33832b31d2c988c6c6aa commit LoRa works fine.
But on 388b70134b10a71d8666a6c1c2749d7ad896467f commit and all commits after it LoRa produces blurry low quality images.
I can't…
-
On tiling `linalg` ops using the tiling interface, we create `extract_slice` ops that retrieve a slice of the input tensors. These correspond to the footprint of one tile of iterations on the tensor. …
-
Running with GPU=1:
```
PYTHONPATH=":/data/openpilot/tinygrad_repo" GPU=1 python3 test_onnx.py "https://github.com/commaai/openpilot/raw/13d4b5f8bb79c06811f4bee2c3542aca65292aac/selfdrive/modeld/mo…