-
I was wondering if torchtune could support apple silicon. It looks like all the examples are for CUDA.
-
### Describe the feature request
Currently there is no MPS (apple Metal Performance Shaders GPU) provider. There is CoreML provider but from testing I did on my MacBook with arm64 M3 and MPS It does …
-
Appears to be caused by this:
```
/Users/vikhyat/Coding/moondream/.venv/lib/python3.12/site-packages/transformers/generation/utils.py:1513: UserWarning: The operator 'aten::isin.Tensor_Tensor_out'…
-
####################
logs:
`using mps requires --mps-root to be specified.`
####################
The contents of the `nvidia-device-plugin.yml` file are as follows:
```
...
env:
- name: CONFIG…
-
**Describe the bug** 🪲
I made a KVM connection to a connected device, and then clicked the IDE-Redirection function, but the upload of the .iso file failed, KVM would interrupt the connection, an…
-
### Describe the bug
Running Kolors with MPS, the returned result is always black image.
### Reproduction
Run this source code. Output image returns black image
```python
pipe = KolorsPipeline.f…
-
### 🚀 The feature, motivation and pitch
I would like to be able to use `torch.sparse` on Metal, which currently breaks:
```
import torch
assert(torch.backends.mps.is_available())
mps_device = t…
-
### 🐛 Describe the bug
`torch.layer_norm()` gives wrong results on MPS if applied on a slice of tensor.
For example if I do `torch.layer_norm(a_tensor[123:125])` on MPS, what it actually calculat…
-
**LocalAI version:2.16.0
**Environment, CPU architecture, OS, and Version:**
mac studio M2 Ultra
**Describe the bug**
using backend transformers for glm4, trust_remote_code: true not c…
-
Hi,
Congrats on the launch!
Is MPS (Apple Silicon) or MLX support planned?
Thank you!