[4/N] Enable intel GPU for unsloth #2621
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gujinghui
reviewed
May 26, 2025
unsloth/models/_utils.py
Outdated
| torch_amp_custom_bwd = torch.amp.custom_bwd(device_type = "cuda") | ||
| pass | ||
| elif DEVICE_TYPE == "xpu": | ||
| if Version(torch_version) < Version("2.4.0"): |
There was a problem hiding this comment.
Suggested change
| if Version(torch_version) < Version("2.4.0"): | |
| if Version(torch_version) < Version("2.6.0"): |
gujinghui
approved these changes
Jun 3, 2025
zejun-chen
reviewed
Jun 3, 2025
unsloth/models/_utils.py
Outdated
| if DEVICE_TYPE == "cuda": | ||
| major_version, minor_version = torch.cuda.get_device_capability() | ||
|
|
||
|
|
zejun-chen
approved these changes
Jun 3, 2025
|
@danielhanchen, @shimmyshimmer |
danielhanchen
requested changes
Jun 24, 2025
| try: vllm_version = f" vLLM: {importlib_version('vllm')}." | ||
| except: vllm_version = "" | ||
|
|
||
| statistics = \ |
Contributor
There was a problem hiding this comment.
On printouts, it's best not to make 2 - instead define new variables like device_count = torch.xpu.device_count() if DEVICE_TYPE == "xpu" else torch.cuda.device_count()
Contributor
Author
|
Closed pr as duplicated with |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Hi unsloth, we are going to support unsloth intel GPU with several prs and this is the third pr.
For the first step we are aiming to support several models with LoRA, and increase our feature in the future (including BNB, FlashAttention, xformers).
For this PR, we resolve device specific API for cuda and Intel GPU(XPU) for model utils and llama model
For cuda specific path, we didn't change the logics, only add check and tab to pass python grammar.
cc: