url string | repository_url string | labels_url string | comments_url string | events_url string | html_url string | id int64 | node_id string | number int64 | title string | user dict | labels list | state string | locked bool | assignee dict | assignees list | milestone null | comments list | created_at timestamp[ms] | updated_at timestamp[ms] | closed_at timestamp[ms] | author_association string | type dict | active_lock_reason null | draft bool | pull_request dict | body string | closed_by dict | reactions dict | timeline_url string | performed_via_github_app null | state_reason string | sub_issues_summary dict | issue_dependencies_summary dict | is_pull_request bool | is_closed bool |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/huggingface/transformers/issues/40340 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40340/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40340/comments | https://api.github.com/repos/huggingface/transformers/issues/40340/events | https://github.com/huggingface/transformers/pull/40340 | 3,341,100,849 | PR_kwDOCUB6oc6kpy1m | 40,340 | Fix AutoAWQ version check for module skipping (Fixes #39798) | {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T09:57:11 | 2025-08-21T10:55:43 | 2025-08-21T10:55:43 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40340",
"html_url": "https://github.com/huggingface/transformers/pull/40340",
"diff_url": "https://github.com/huggingface/transformers/pull/40340.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40340.patch",
"merged_at": null
} | # What does this PR do?
Fixes AutoAWQ version checking logic that was causing AttributeError when version attributes are missing, preventing legitimate users with correct versions from loading AWQ models.
Fixes #39798
## Root Cause & Solution
**Problem**: Version checking logic failed when AutoAWQ module didn't have expected version attributes, causing AttributeError even for users with correct versions.
**Solution**: Implemented robust version checking with multiple fallbacks:
1. Primary: `importlib.metadata.version()`
2. Fallback: Module `__version__` attribute
3. Final fallback: Module `version` attribute
4. Warning when version cannot be determined but module imports successfully
## Testing
- Added 6 comprehensive test cases covering all scenarios
- All tests pass locally
- Verified fix works with original reproduction case
## Files Changed
- `src/transformers/utils/quantization_config.py`: Robust version checking logic
- `tests/quantization/autoawq/test_awq_version_check.py`: Comprehensive test suite
## Backward Compatibility
✅ No breaking changes - existing workflows unaffected
## Before submitting
- [x] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [x] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [x] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [x] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@SunMarc @MekkCyber @Rocketknight1
| {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40340/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40340/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40339 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40339/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40339/comments | https://api.github.com/repos/huggingface/transformers/issues/40339/events | https://github.com/huggingface/transformers/pull/40339 | 3,340,839,527 | PR_kwDOCUB6oc6ko7_U | 40,339 | [processor] move commonalities to mixin | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T08:41:13 | 2025-08-22T11:04:44 | 2025-08-22T11:04:44 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40339",
"html_url": "https://github.com/huggingface/transformers/pull/40339",
"diff_url": "https://github.com/huggingface/transformers/pull/40339.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40339.patch",
"merged_at": "2025-08-22T11:04:44"
} | # What does this PR do?
As per title, moves `batch_decode, decode, model_input_names` to the Mixin so we stop duplicating the same code in all processors. Partially related to code unbloating since I am planning to modularize processors slightly in the next PR. This one is the first clean-up | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40339/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40339/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40338 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40338/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40338/comments | https://api.github.com/repos/huggingface/transformers/issues/40338/events | https://github.com/huggingface/transformers/pull/40338 | 3,340,740,683 | PR_kwDOCUB6oc6komzO | 40,338 | [Kosmos 2.5] Rename checkpoints | {
"login": "NielsRogge",
"id": 48327001,
"node_id": "MDQ6VXNlcjQ4MzI3MDAx",
"avatar_url": "https://avatars.githubusercontent.com/u/48327001?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/NielsRogge",
"html_url": "https://github.com/NielsRogge",
"followers_url": "https://api.github.com/users/NielsRogge/followers",
"following_url": "https://api.github.com/users/NielsRogge/following{/other_user}",
"gists_url": "https://api.github.com/users/NielsRogge/gists{/gist_id}",
"starred_url": "https://api.github.com/users/NielsRogge/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/NielsRogge/subscriptions",
"organizations_url": "https://api.github.com/users/NielsRogge/orgs",
"repos_url": "https://api.github.com/users/NielsRogge/repos",
"events_url": "https://api.github.com/users/NielsRogge/events{/privacy}",
"received_events_url": "https://api.github.com/users/NielsRogge/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T08:07:58 | 2025-08-28T14:09:47 | 2025-08-28T11:30:41 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40338",
"html_url": "https://github.com/huggingface/transformers/pull/40338",
"diff_url": "https://github.com/huggingface/transformers/pull/40338.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40338.patch",
"merged_at": "2025-08-28T11:30:41"
} | # What does this PR do?
This PR renames all instances of `ydshieh/kosmos-2.5` to `microsoft/kosmos-2.5`.
cc @ydshieh feel free to take over this PR
To do:
- [x] add KOSMOS-2.5 chat to the docs
| {
"login": "NielsRogge",
"id": 48327001,
"node_id": "MDQ6VXNlcjQ4MzI3MDAx",
"avatar_url": "https://avatars.githubusercontent.com/u/48327001?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/NielsRogge",
"html_url": "https://github.com/NielsRogge",
"followers_url": "https://api.github.com/users/NielsRogge/followers",
"following_url": "https://api.github.com/users/NielsRogge/following{/other_user}",
"gists_url": "https://api.github.com/users/NielsRogge/gists{/gist_id}",
"starred_url": "https://api.github.com/users/NielsRogge/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/NielsRogge/subscriptions",
"organizations_url": "https://api.github.com/users/NielsRogge/orgs",
"repos_url": "https://api.github.com/users/NielsRogge/repos",
"events_url": "https://api.github.com/users/NielsRogge/events{/privacy}",
"received_events_url": "https://api.github.com/users/NielsRogge/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40338/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40338/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40337 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40337/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40337/comments | https://api.github.com/repos/huggingface/transformers/issues/40337/events | https://github.com/huggingface/transformers/pull/40337 | 3,340,732,993 | PR_kwDOCUB6oc6kolLP | 40,337 | Remove more PyTorch 2.2 compatible code | {
"login": "cyyever",
"id": 17618148,
"node_id": "MDQ6VXNlcjE3NjE4MTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/17618148?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cyyever",
"html_url": "https://github.com/cyyever",
"followers_url": "https://api.github.com/users/cyyever/followers",
"following_url": "https://api.github.com/users/cyyever/following{/other_user}",
"gists_url": "https://api.github.com/users/cyyever/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cyyever/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cyyever/subscriptions",
"organizations_url": "https://api.github.com/users/cyyever/orgs",
"repos_url": "https://api.github.com/users/cyyever/repos",
"events_url": "https://api.github.com/users/cyyever/events{/privacy}",
"received_events_url": "https://api.github.com/users/cyyever/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T08:05:16 | 2025-08-21T15:21:45 | 2025-08-21T15:19:54 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40337",
"html_url": "https://github.com/huggingface/transformers/pull/40337",
"diff_url": "https://github.com/huggingface/transformers/pull/40337.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40337.patch",
"merged_at": "2025-08-21T15:19:54"
} | # What does this PR do?
Follows the recent changes | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40337/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40337/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40336 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40336/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40336/comments | https://api.github.com/repos/huggingface/transformers/issues/40336/events | https://github.com/huggingface/transformers/pull/40336 | 3,340,644,256 | PR_kwDOCUB6oc6koSYI | 40,336 | Fix qwen-omni processor text only mode | {
"login": "yuekaizhang",
"id": 28823245,
"node_id": "MDQ6VXNlcjI4ODIzMjQ1",
"avatar_url": "https://avatars.githubusercontent.com/u/28823245?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuekaizhang",
"html_url": "https://github.com/yuekaizhang",
"followers_url": "https://api.github.com/users/yuekaizhang/followers",
"following_url": "https://api.github.com/users/yuekaizhang/following{/other_user}",
"gists_url": "https://api.github.com/users/yuekaizhang/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yuekaizhang/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yuekaizhang/subscriptions",
"organizations_url": "https://api.github.com/users/yuekaizhang/orgs",
"repos_url": "https://api.github.com/users/yuekaizhang/repos",
"events_url": "https://api.github.com/users/yuekaizhang/events{/privacy}",
"received_events_url": "https://api.github.com/users/yuekaizhang/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T07:34:03 | 2025-08-21T09:57:32 | 2025-08-21T09:57:32 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40336",
"html_url": "https://github.com/huggingface/transformers/pull/40336",
"diff_url": "https://github.com/huggingface/transformers/pull/40336.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40336.patch",
"merged_at": "2025-08-21T09:57:32"
} | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Hi @BakerBunker and @zucchini-nlp,
This PR enables the Qwen2.5-Omni-Processor to handle text-only inputs, without requiring corresponding audio, video, or image data.
This feature is a prerequisite for the VLLM implementation, when uses multimodal embeddings as input.
```
from transformers import Qwen2_5OmniProcessor
text = "<|im_start|>system\nYou are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.<|im_end|>\n<|im_start|>user\n<|vision_bos|><|IMAGE|><|vision_eos|><prompt><|im_end|>\n<|im_start|>assistant\n"
processor = Qwen2_5OmniProcessor.from_pretrained("Qwen/Qwen2.5-Omni-7B")
inputs = processor(text=text)
print(inputs)
```
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40336/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40336/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40335 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40335/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40335/comments | https://api.github.com/repos/huggingface/transformers/issues/40335/events | https://github.com/huggingface/transformers/issues/40335 | 3,340,627,387 | I_kwDOCUB6oc7HHe27 | 40,335 | repetitio/flan-t5-small is not longer exists | {
"login": "jiqing-feng",
"id": 107918818,
"node_id": "U_kgDOBm614g",
"avatar_url": "https://avatars.githubusercontent.com/u/107918818?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jiqing-feng",
"html_url": "https://github.com/jiqing-feng",
"followers_url": "https://api.github.com/users/jiqing-feng/followers",
"following_url": "https://api.github.com/users/jiqing-feng/following{/other_user}",
"gists_url": "https://api.github.com/users/jiqing-feng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jiqing-feng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jiqing-feng/subscriptions",
"organizations_url": "https://api.github.com/users/jiqing-feng/orgs",
"repos_url": "https://api.github.com/users/jiqing-feng/repos",
"events_url": "https://api.github.com/users/jiqing-feng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jiqing-feng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-21T07:28:05 | 2025-09-01T09:44:00 | 2025-09-01T09:44:00 | CONTRIBUTOR | null | null | null | null | ### System Info
latest main branch of transformers and pytorch 2.8
### Who can help?
@SunMarc
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
RUN_SLOW=1 pytest tests/quantization/ggml/test_ggml.py::GgufModelTests::test_t5_f16
error:
```
OSError: repetitio/flan-t5-small is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models'
E If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `hf auth login` or by passing `token=<your_token>`
```
### Expected behavior
Cannot find the model `repetitio/flan-t5-small` in HF hub. | {
"login": "SunMarc",
"id": 57196510,
"node_id": "MDQ6VXNlcjU3MTk2NTEw",
"avatar_url": "https://avatars.githubusercontent.com/u/57196510?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SunMarc",
"html_url": "https://github.com/SunMarc",
"followers_url": "https://api.github.com/users/SunMarc/followers",
"following_url": "https://api.github.com/users/SunMarc/following{/other_user}",
"gists_url": "https://api.github.com/users/SunMarc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SunMarc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SunMarc/subscriptions",
"organizations_url": "https://api.github.com/users/SunMarc/orgs",
"repos_url": "https://api.github.com/users/SunMarc/repos",
"events_url": "https://api.github.com/users/SunMarc/events{/privacy}",
"received_events_url": "https://api.github.com/users/SunMarc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40335/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40335/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40334 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40334/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40334/comments | https://api.github.com/repos/huggingface/transformers/issues/40334/events | https://github.com/huggingface/transformers/issues/40334 | 3,340,444,650 | I_kwDOCUB6oc7HGyPq | 40,334 | It takes too long time to load GGUF tokenizer | {
"login": "jiqing-feng",
"id": 107918818,
"node_id": "U_kgDOBm614g",
"avatar_url": "https://avatars.githubusercontent.com/u/107918818?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jiqing-feng",
"html_url": "https://github.com/jiqing-feng",
"followers_url": "https://api.github.com/users/jiqing-feng/followers",
"following_url": "https://api.github.com/users/jiqing-feng/following{/other_user}",
"gists_url": "https://api.github.com/users/jiqing-feng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jiqing-feng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jiqing-feng/subscriptions",
"organizations_url": "https://api.github.com/users/jiqing-feng/orgs",
"repos_url": "https://api.github.com/users/jiqing-feng/repos",
"events_url": "https://api.github.com/users/jiqing-feng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jiqing-feng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2392046359,
"node_id": "MDU6TGFiZWwyMzkyMDQ2MzU5",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Good%20Second%20Issue",
"name": "Good Second Issue",
"color": "dd935a",
"default": false,
"description": "Issues that are more difficult to do than \"Good First\" issues - give it a try if you want!"
},
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-21T06:20:31 | 2025-08-28T12:10:50 | 2025-08-28T12:10:50 | CONTRIBUTOR | null | null | null | null | ### System Info
Latest main branch of transformers and PyTorch 2.8 released
### Who can help?
@SunMarc @ArthurZucker
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
The test `tests/quantization/ggml/test_ggml.py::GgufModelTests::test_nemotron_q6_k` takes too long time because of loading the tokenizer:
<img width="1711" height="233" alt="Image" src="https://github.com/user-attachments/assets/1861c237-10ae-460b-8ef7-fdc8ce1b77d6" />
It takes me almost 50 minutes to load a tokenizer in the test which is quite annoying.
### Expected behavior
I uploaded the tokenizer so we can avoid the process. Check [here](https://huggingface.co/bartowski/Nemotron-Mini-4B-Instruct-GGUF/discussions/3) | {
"login": "SunMarc",
"id": 57196510,
"node_id": "MDQ6VXNlcjU3MTk2NTEw",
"avatar_url": "https://avatars.githubusercontent.com/u/57196510?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SunMarc",
"html_url": "https://github.com/SunMarc",
"followers_url": "https://api.github.com/users/SunMarc/followers",
"following_url": "https://api.github.com/users/SunMarc/following{/other_user}",
"gists_url": "https://api.github.com/users/SunMarc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SunMarc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SunMarc/subscriptions",
"organizations_url": "https://api.github.com/users/SunMarc/orgs",
"repos_url": "https://api.github.com/users/SunMarc/repos",
"events_url": "https://api.github.com/users/SunMarc/events{/privacy}",
"received_events_url": "https://api.github.com/users/SunMarc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40334/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40334/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40333 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40333/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40333/comments | https://api.github.com/repos/huggingface/transformers/issues/40333/events | https://github.com/huggingface/transformers/pull/40333 | 3,340,129,377 | PR_kwDOCUB6oc6kmlJH | 40,333 | fix qwen25-vl grad acc | {
"login": "iMountTai",
"id": 35353688,
"node_id": "MDQ6VXNlcjM1MzUzNjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/35353688?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/iMountTai",
"html_url": "https://github.com/iMountTai",
"followers_url": "https://api.github.com/users/iMountTai/followers",
"following_url": "https://api.github.com/users/iMountTai/following{/other_user}",
"gists_url": "https://api.github.com/users/iMountTai/gists{/gist_id}",
"starred_url": "https://api.github.com/users/iMountTai/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/iMountTai/subscriptions",
"organizations_url": "https://api.github.com/users/iMountTai/orgs",
"repos_url": "https://api.github.com/users/iMountTai/repos",
"events_url": "https://api.github.com/users/iMountTai/events{/privacy}",
"received_events_url": "https://api.github.com/users/iMountTai/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T03:14:03 | 2025-08-26T09:30:36 | 2025-08-26T09:30:07 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40333",
"html_url": "https://github.com/huggingface/transformers/pull/40333",
"diff_url": "https://github.com/huggingface/transformers/pull/40333.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40333.patch",
"merged_at": "2025-08-26T09:30:07"
} | # What does this PR do?
Currently, there are still many models that do not accept loss_kwargs (such as Qwen25-VL). Therefore, this PR is proposed with reference to the gemma3 PR. Now, the training loss and grad_norm are normal.
## reference
#37208
| {
"login": "SunMarc",
"id": 57196510,
"node_id": "MDQ6VXNlcjU3MTk2NTEw",
"avatar_url": "https://avatars.githubusercontent.com/u/57196510?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SunMarc",
"html_url": "https://github.com/SunMarc",
"followers_url": "https://api.github.com/users/SunMarc/followers",
"following_url": "https://api.github.com/users/SunMarc/following{/other_user}",
"gists_url": "https://api.github.com/users/SunMarc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SunMarc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SunMarc/subscriptions",
"organizations_url": "https://api.github.com/users/SunMarc/orgs",
"repos_url": "https://api.github.com/users/SunMarc/repos",
"events_url": "https://api.github.com/users/SunMarc/events{/privacy}",
"received_events_url": "https://api.github.com/users/SunMarc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40333/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40333/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40332 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40332/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40332/comments | https://api.github.com/repos/huggingface/transformers/issues/40332/events | https://github.com/huggingface/transformers/pull/40332 | 3,340,080,199 | PR_kwDOCUB6oc6kma-I | 40,332 | fix spelling error and inconsistencies | {
"login": "TianqiTang1117",
"id": 47551755,
"node_id": "MDQ6VXNlcjQ3NTUxNzU1",
"avatar_url": "https://avatars.githubusercontent.com/u/47551755?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/TianqiTang1117",
"html_url": "https://github.com/TianqiTang1117",
"followers_url": "https://api.github.com/users/TianqiTang1117/followers",
"following_url": "https://api.github.com/users/TianqiTang1117/following{/other_user}",
"gists_url": "https://api.github.com/users/TianqiTang1117/gists{/gist_id}",
"starred_url": "https://api.github.com/users/TianqiTang1117/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/TianqiTang1117/subscriptions",
"organizations_url": "https://api.github.com/users/TianqiTang1117/orgs",
"repos_url": "https://api.github.com/users/TianqiTang1117/repos",
"events_url": "https://api.github.com/users/TianqiTang1117/events{/privacy}",
"received_events_url": "https://api.github.com/users/TianqiTang1117/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T02:39:14 | 2025-08-21T11:06:46 | 2025-08-21T11:06:46 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40332",
"html_url": "https://github.com/huggingface/transformers/pull/40332",
"diff_url": "https://github.com/huggingface/transformers/pull/40332.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40332.patch",
"merged_at": null
} | Summary
- Corrected a spelling mistake in the speech recognition CTC example so that inputs and labels are described as having different “lengths” instead of “lenghts,” improving clarity in the batching logic
- Fixed dtype selection in the data collators by replacing strict type(...) is int checks with isinstance(..., int), ensuring both PyTorch and NumPy paths handle integer subclasses correctly
- Removed a duplicated word in the TrainingArguments documentation to clarify how repositories are synced when pushing to the Hub
- Enhanced the data collator test to verify correct handling of NumPy integer label IDs, strengthening coverage for mixed input types
Testing
- pytest tests/trainer/test_data_collator.py::DataCollatorIntegrationTest::test_default_with_dict -q (failed: ModuleNotFoundError: No module named 'numpy')
Notes
- NumPy installation failed due to proxy restrictions, preventing the test from running successfully. | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40332/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40332/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40331 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40331/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40331/comments | https://api.github.com/repos/huggingface/transformers/issues/40331/events | https://github.com/huggingface/transformers/pull/40331 | 3,340,077,556 | PR_kwDOCUB6oc6kmabi | 40,331 | Update quantization overview for XPU | {
"login": "jiqing-feng",
"id": 107918818,
"node_id": "U_kgDOBm614g",
"avatar_url": "https://avatars.githubusercontent.com/u/107918818?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jiqing-feng",
"html_url": "https://github.com/jiqing-feng",
"followers_url": "https://api.github.com/users/jiqing-feng/followers",
"following_url": "https://api.github.com/users/jiqing-feng/following{/other_user}",
"gists_url": "https://api.github.com/users/jiqing-feng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jiqing-feng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jiqing-feng/subscriptions",
"organizations_url": "https://api.github.com/users/jiqing-feng/orgs",
"repos_url": "https://api.github.com/users/jiqing-feng/repos",
"events_url": "https://api.github.com/users/jiqing-feng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jiqing-feng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-21T02:37:10 | 2025-08-28T09:53:36 | 2025-08-28T09:53:00 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40331",
"html_url": "https://github.com/huggingface/transformers/pull/40331",
"diff_url": "https://github.com/huggingface/transformers/pull/40331.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40331.patch",
"merged_at": "2025-08-28T09:52:59"
} | Update quantization overview for XPU.
Keep in draft until optimum-quanto PR: [395](https://github.com/huggingface/optimum-quanto/pull/395) merged. | {
"login": "MekkCyber",
"id": 93391238,
"node_id": "U_kgDOBZEJhg",
"avatar_url": "https://avatars.githubusercontent.com/u/93391238?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MekkCyber",
"html_url": "https://github.com/MekkCyber",
"followers_url": "https://api.github.com/users/MekkCyber/followers",
"following_url": "https://api.github.com/users/MekkCyber/following{/other_user}",
"gists_url": "https://api.github.com/users/MekkCyber/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MekkCyber/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MekkCyber/subscriptions",
"organizations_url": "https://api.github.com/users/MekkCyber/orgs",
"repos_url": "https://api.github.com/users/MekkCyber/repos",
"events_url": "https://api.github.com/users/MekkCyber/events{/privacy}",
"received_events_url": "https://api.github.com/users/MekkCyber/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40331/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40331/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40330 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40330/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40330/comments | https://api.github.com/repos/huggingface/transformers/issues/40330/events | https://github.com/huggingface/transformers/issues/40330 | 3,340,056,210 | I_kwDOCUB6oc7HFTaS | 40,330 | open-qwen2vl-base | {
"login": "olccihyeon",
"id": 36918246,
"node_id": "MDQ6VXNlcjM2OTE4MjQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/36918246?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/olccihyeon",
"html_url": "https://github.com/olccihyeon",
"followers_url": "https://api.github.com/users/olccihyeon/followers",
"following_url": "https://api.github.com/users/olccihyeon/following{/other_user}",
"gists_url": "https://api.github.com/users/olccihyeon/gists{/gist_id}",
"starred_url": "https://api.github.com/users/olccihyeon/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/olccihyeon/subscriptions",
"organizations_url": "https://api.github.com/users/olccihyeon/orgs",
"repos_url": "https://api.github.com/users/olccihyeon/repos",
"events_url": "https://api.github.com/users/olccihyeon/events{/privacy}",
"received_events_url": "https://api.github.com/users/olccihyeon/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1843244711,
"node_id": "MDU6TGFiZWwxODQzMjQ0NzEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model",
"name": "New model",
"color": "fbca04",
"default": false,
"description": ""
}
] | open | false | null | [] | null | [] | 2025-08-21T02:24:01 | 2025-08-23T10:18:28 | null | NONE | null | null | null | null | ### Model description
is there any plan to add open-qwen2vl-base model?
### Open source status
- [x] The model implementation is available
- [x] The model weights are available
### Provide useful links for the implementation
_No response_ | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40330/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40330/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40329 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40329/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40329/comments | https://api.github.com/repos/huggingface/transformers/issues/40329/events | https://github.com/huggingface/transformers/pull/40329 | 3,339,829,035 | PR_kwDOCUB6oc6klmgt | 40,329 | [EfficientLoFTR] LRU Cached embedding computation at inference | {
"login": "sbucaille",
"id": 24275548,
"node_id": "MDQ6VXNlcjI0Mjc1NTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/24275548?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sbucaille",
"html_url": "https://github.com/sbucaille",
"followers_url": "https://api.github.com/users/sbucaille/followers",
"following_url": "https://api.github.com/users/sbucaille/following{/other_user}",
"gists_url": "https://api.github.com/users/sbucaille/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sbucaille/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sbucaille/subscriptions",
"organizations_url": "https://api.github.com/users/sbucaille/orgs",
"repos_url": "https://api.github.com/users/sbucaille/repos",
"events_url": "https://api.github.com/users/sbucaille/events{/privacy}",
"received_events_url": "https://api.github.com/users/sbucaille/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T23:51:41 | 2025-09-02T03:14:53 | 2025-08-27T14:05:08 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40329",
"html_url": "https://github.com/huggingface/transformers/pull/40329",
"diff_url": "https://github.com/huggingface/transformers/pull/40329.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40329.patch",
"merged_at": "2025-08-27T14:05:08"
} | # What does this PR do?
Reverted EfficientLoFTR RoPE computation back to inference time to allow different image sizes usage
Decorated with LRU Cache similarly to DinoV3 implementation
## Who can review?
@qubvel
| {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40329/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40329/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40328 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40328/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40328/comments | https://api.github.com/repos/huggingface/transformers/issues/40328/events | https://github.com/huggingface/transformers/pull/40328 | 3,339,651,061 | PR_kwDOCUB6oc6klAnB | 40,328 | [rfc] Prototype to make torch.compile work with DynamicCache | {
"login": "anijain2305",
"id": 13822661,
"node_id": "MDQ6VXNlcjEzODIyNjYx",
"avatar_url": "https://avatars.githubusercontent.com/u/13822661?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anijain2305",
"html_url": "https://github.com/anijain2305",
"followers_url": "https://api.github.com/users/anijain2305/followers",
"following_url": "https://api.github.com/users/anijain2305/following{/other_user}",
"gists_url": "https://api.github.com/users/anijain2305/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anijain2305/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anijain2305/subscriptions",
"organizations_url": "https://api.github.com/users/anijain2305/orgs",
"repos_url": "https://api.github.com/users/anijain2305/repos",
"events_url": "https://api.github.com/users/anijain2305/events{/privacy}",
"received_events_url": "https://api.github.com/users/anijain2305/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1834056761,
"node_id": "MDU6TGFiZWwxODM0MDU2NzYx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Modeling",
"name": "Core: Modeling",
"color": "FF8446",
"default": false,
"description": "Internals of the library; Models."
},
{
"id": 6625174977,
"node_id": "LA_kwDOCUB6oc8AAAABiuQlwQ",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Compilation",
"name": "Compilation",
"color": "58E75E",
"default": false,
"description": "Issues related to torchdynamo and torchinductor"
}
] | open | false | null | [] | null | [] | 2025-08-20T21:57:09 | 2025-09-16T14:16:37 | null | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40328",
"html_url": "https://github.com/huggingface/transformers/pull/40328",
"diff_url": "https://github.com/huggingface/transformers/pull/40328.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40328.patch",
"merged_at": null
} | # What does this PR do?
## Background on StaticCache vs DynamicCache
A static KV cache must be preallocated for the model’s maximum context window (e.g., 32k tokens), even if you only generate a few tokens. That leads to large memory pressure and wastage. A dynamic cache grows with the sequence actually in use: with P prefill tokens and D decode steps, memory scales with P + D instead of max_seq_length. For example, if you prefill 2,000 tokens and generate 50, the dynamic cache holds ~2,050 tokens’ worth of K/V per layer, versus 32k for a static allocation, significantly reducing memory pressure.
Lets do a quick experiment for Qwen3
<img width="724" height="120" alt="image" src="https://github.com/user-attachments/assets/1d37a49e-98a7-4447-8d20-ed4e43fb11d0" />
In all cases, the output generated text was exactly the same.
As expected, StaticCache has a higher memory footprint. But the generation time is also higher. This is because the main tensors involved in [SDPA](https://github.com/huggingface/transformers/blob/242bb2cafccec9f90479f5f688bca9d240b1031f/src/transformers/integrations/sdpa_attention.py#L83) operation - (key, value, attention_mask) have max seq length shape. Even though attention_mask fires up only a portion of the actual compute, there is still more work to be done here compared to Dynamic KV cache. (Side chat - Driss and Boyuan told me that we might be able to use better eager code or flex decoding to make this better, but thats not the focus of this doc.)
## torch.compile and Dynamic Cache
transformers introduced StaticCache just for torch.compile. For Qwen3, torch.compile latency is 20 seconds, achieving 1.2x speedup and 25 seconds of compile time ([tlparse](https://manifold.edge.x2p.facebook.net/v0/read/tree/logs/.tmpifaWBm/index.html?bucketName=tlparse_reports&apiKey=tlparse_reports-key&withPayload=1&timeoutMsec=10000)).
But as pointed out earlier, DynamicCache is a better default setting for eager PyTorch. So, we will look at the changes required to make torch.compile work with DynamicCache. WIth this PR, I see torch.compile improve **generation latency to 3.6 (while eager was 8.88 seconds), leading to 2.4x speedup** and a cold start latency of 36 seconds ([tlparse](https://manifold.edge.x2p.facebook.net/v0/read/tree/logs/.tmpq4CUzG/index.html?bucketName=tlparse_reports&apiKey=tlparse_reports-key&withPayload=1&timeoutMsec=10000)) and warm start of 11 seconds.
For this RFC
* We will use fullgraph=True and do full model compilation (no regional compilation yet).
* We will compile only the decode step to keep the discussion simple.
There are two main changes
**Avoid recompilations (symbolic seq length)** : Tell torch.compile to treat sequence length as dynamic for all KV-cache–related tensors - key, value, and attention_mask, using torch._dynamo.mark_dynamic. To ensure a single shared symbolic size is used everywhere (since mark_dynamic alone can yield distinct symbols), wrap the model and insert explicit size-equality checks (e.g., via torch._check) that tie the symbols together. This nudges the compiler to unify on one seq-len symbol, improving inductor codegen and in future compile time.
**CUDA graphs**: Disable cudagraphs for DynamicCache. Because the sequence length changes step-to-step, cudagraphs would constantly re-record, which undercuts their benefit and can hurt latency. (They remain useful when shapes are fixed, e.g., with a static cache.)
### Update
The PR now extends torch.compile to `DynamicSlidingWindowLayer`. There is one extra recompilation on the integer value of `cumulative_length` that I hope can be prevented by converting the integer into a scalar tensor.
Speedup - Pytorch eager has 6.66 seconds, compile takes 5.10 seconds for Gemma2. Although the compile time is high (currently due to 2 compilations). | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40328/reactions",
"total_count": 7,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 2,
"confused": 0,
"heart": 5,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40328/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40327 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40327/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40327/comments | https://api.github.com/repos/huggingface/transformers/issues/40327/events | https://github.com/huggingface/transformers/issues/40327 | 3,339,157,959 | I_kwDOCUB6oc7HB4HH | 40,327 | ValueError when loading 4-bit quantized models with sharded weights on Transformers >=4.54.0 | {
"login": "rice8y",
"id": 146926884,
"node_id": "U_kgDOCMHtJA",
"avatar_url": "https://avatars.githubusercontent.com/u/146926884?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rice8y",
"html_url": "https://github.com/rice8y",
"followers_url": "https://api.github.com/users/rice8y/followers",
"following_url": "https://api.github.com/users/rice8y/following{/other_user}",
"gists_url": "https://api.github.com/users/rice8y/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rice8y/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rice8y/subscriptions",
"organizations_url": "https://api.github.com/users/rice8y/orgs",
"repos_url": "https://api.github.com/users/rice8y/repos",
"events_url": "https://api.github.com/users/rice8y/events{/privacy}",
"received_events_url": "https://api.github.com/users/rice8y/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-20T18:36:38 | 2025-10-06T08:02:56 | 2025-10-06T08:02:56 | NONE | null | null | null | null | ### System Info
- **transformers version**: 4.54.0 (fails), 4.51.3 (works)
- **Platform**: Linux-5.19.0-32-generic-x86_64-with-glibc2.35
- **Python version**: 3.12
- **Huggingface_hub version**: 0.34.4
- **Safetensors version**: 0.6.2
- **Accelerate version**: 1.10.0
- **PyTorch version**: 2.8.0
- **GPU**: NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
- **CUDA**: 12.9
- **BitsAndBytes**: 0.47.0
- **OS**: Ubuntu 22.04.2 LTS
### Who can help?
@SunMarc @MekkCyber (quantization - bitsandbytes)
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
### Bug Description
When trying to load large 4-bit quantized models (72B parameters) with sharded weights using Transformers >=4.54.0, a `ValueError` is raised during model loading. The error specifically occurs with models that have many parameter shards and use immediate 4-bit quantization.
### Code to Reproduce
```python
from unsloth import FastVisionModel
# This fails on Transformers >=4.54.0
model, tokenizer = FastVisionModel.from_pretrained(
model_name = "unsloth/Qwen2.5-VL-72B-Instruct-unsloth-bnb-4bit",
max_seq_length = 65536,
dtype = None,
load_in_4bit = True,
)
```
### Expected behavior
The 72B 4-bit quantized model should load successfully, similar to smaller models (32B) which work without issues.
---
## Additional Details
### Error Message
```
ValueError: Supplied state dict for model.language_model.layers.7.self_attn.o_proj.weight does not contain `bitsandbytes__*` and possibly other `quantized_stats` components.
```
### Full Error Traceback
```
Traceback (most recent call last):
File "/path/to/project/sft_all.py", line 254, in <module>
main()
File "/path/to/project/sft_all.py", line 151, in main
model, tokenizer = FastVisionModel.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/unsloth/models/loader.py", line 855, in from_pretrained
model, tokenizer = FastBaseModel.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/unsloth/models/vision.py", line 444, in from_pretrained
model = auto_model.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/models/auto/auto_factory.py", line 600, in from_pretrained
return model_class.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/modeling_utils.py", line 315, in _wrapper
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/modeling_utils.py", line 5001, in from_pretrained
) = cls._load_pretrained_model(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/modeling_utils.py", line 5459, in _load_pretrained_model
_error_msgs, disk_offload_index, cpu_offload_index = load_shard_file(args)
^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/modeling_utils.py", line 937, in load_shard_file
disk_offload_index, cpu_offload_index = _load_state_dict_into_meta_model(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/path/to/venv/lib/python3.12/site-packages/transformers/modeling_utils.py", line 852, in _load_state_dict_into_meta_model
hf_quantizer.create_quantized_param(
File "/path/to/venv/lib/python3.12/site-packages/transformers/quantizers/quantizer_bnb_4bit.py", line 219, in create_quantized_param
raise ValueError(
ValueError: Supplied state dict for model.language_model.layers.7.self_attn.o_proj.weight does not contain `bitsandbytes__*` and possibly other `quantized_stats` components.
```
### Root Cause Analysis
The issue appears to be related to the sharded streaming load implementation with immediate 4-bit quantization (`create_quantized_param`) that was introduced in Transformers v4.54.0.
**Key observations:**
1. **Model size dependency**: Only affects large models (72B parameters) with many shards, while smaller models (32B) load successfully
2. **Version dependency**: Works on Transformers v4.51.3, fails on v4.54.0+
3. **Timing issue**: The quantization metadata (`bitsandbytes__*` components) is not yet available when `create_quantized_param` is called during the sharded loading process
**Hypothesis:**
In large sharded models, the parameter creation and quantization happen before all the necessary quantization metadata is loaded from the sharded state dict, causing the validation in `create_quantized_param` to fail.
### Workaround
Downgrading Transformers to v4.51.3 resolves the issue:
```bash
pip install transformers==4.51.3
```
### Additional Context
- This issue was initially encountered when using Unsloth's FastVisionModel wrapper, but the root cause is in the Transformers library's quantization loading mechanism
- The error occurs consistently across different GPU architectures
- Other quantized models of similar size likely experience the same issue | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40327/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40327/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40326 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40326/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40326/comments | https://api.github.com/repos/huggingface/transformers/issues/40326/events | https://github.com/huggingface/transformers/pull/40326 | 3,339,075,604 | PR_kwDOCUB6oc6kjEIZ | 40,326 | Fix an infinite loop bug in recursive search of relative imports | {
"login": "eladsegal",
"id": 13485709,
"node_id": "MDQ6VXNlcjEzNDg1NzA5",
"avatar_url": "https://avatars.githubusercontent.com/u/13485709?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/eladsegal",
"html_url": "https://github.com/eladsegal",
"followers_url": "https://api.github.com/users/eladsegal/followers",
"following_url": "https://api.github.com/users/eladsegal/following{/other_user}",
"gists_url": "https://api.github.com/users/eladsegal/gists{/gist_id}",
"starred_url": "https://api.github.com/users/eladsegal/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/eladsegal/subscriptions",
"organizations_url": "https://api.github.com/users/eladsegal/orgs",
"repos_url": "https://api.github.com/users/eladsegal/repos",
"events_url": "https://api.github.com/users/eladsegal/events{/privacy}",
"received_events_url": "https://api.github.com/users/eladsegal/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T18:02:50 | 2025-08-21T11:41:57 | 2025-08-21T11:39:44 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40326",
"html_url": "https://github.com/huggingface/transformers/pull/40326",
"diff_url": "https://github.com/huggingface/transformers/pull/40326.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40326.patch",
"merged_at": "2025-08-21T11:39:43"
} | # What does this PR do?
This PR fixes a bug in `get_relative_import_files` in `dynamic_module_utils`.
The mechanism to exclude already-found imports from the next iteration had a wrong inclusion check, as it was missing a ".py" suffix.
This means that even if an import was in `all_relative_imports`, it wasn't excluded from `files_to_check`.
In most cases it was a silent bug, but for files with circular imports it resulted in an infinite loop.
## Who can review?
Maybe @Rocketknight1, @Cyrilvallez ?
| {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40326/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40326/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40325 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40325/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40325/comments | https://api.github.com/repos/huggingface/transformers/issues/40325/events | https://github.com/huggingface/transformers/pull/40325 | 3,339,038,156 | PR_kwDOCUB6oc6ki7_t | 40,325 | Feature/mean token accuracy trainer #40303 | {
"login": "badwriter123",
"id": 91039605,
"node_id": "MDQ6VXNlcjkxMDM5NjA1",
"avatar_url": "https://avatars.githubusercontent.com/u/91039605?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/badwriter123",
"html_url": "https://github.com/badwriter123",
"followers_url": "https://api.github.com/users/badwriter123/followers",
"following_url": "https://api.github.com/users/badwriter123/following{/other_user}",
"gists_url": "https://api.github.com/users/badwriter123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/badwriter123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/badwriter123/subscriptions",
"organizations_url": "https://api.github.com/users/badwriter123/orgs",
"repos_url": "https://api.github.com/users/badwriter123/repos",
"events_url": "https://api.github.com/users/badwriter123/events{/privacy}",
"received_events_url": "https://api.github.com/users/badwriter123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T17:48:49 | 2025-10-15T15:52:49 | 2025-10-15T15:52:48 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40325",
"html_url": "https://github.com/huggingface/transformers/pull/40325",
"diff_url": "https://github.com/huggingface/transformers/pull/40325.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40325.patch",
"merged_at": null
} | # What does this PR do?
Feature : Add mean token accuracy to Trainer
Fixes #40303
### Description
1. For the file --> src/transformers/training_args.py Added a new boolean argument to allow users to toggle the mean token accuracy computation.
2. For the file --> src/transformers/trainer.py. Added the new method (_compute_mean_token_accuracy) . Improvised for both Language Models and Seq2seq Models.
The feature is now ready to be used by setting compute_mean_token_accuracy=True in TrainingArguments. | {
"login": "badwriter123",
"id": 91039605,
"node_id": "MDQ6VXNlcjkxMDM5NjA1",
"avatar_url": "https://avatars.githubusercontent.com/u/91039605?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/badwriter123",
"html_url": "https://github.com/badwriter123",
"followers_url": "https://api.github.com/users/badwriter123/followers",
"following_url": "https://api.github.com/users/badwriter123/following{/other_user}",
"gists_url": "https://api.github.com/users/badwriter123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/badwriter123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/badwriter123/subscriptions",
"organizations_url": "https://api.github.com/users/badwriter123/orgs",
"repos_url": "https://api.github.com/users/badwriter123/repos",
"events_url": "https://api.github.com/users/badwriter123/events{/privacy}",
"received_events_url": "https://api.github.com/users/badwriter123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40325/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40325/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40324 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40324/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40324/comments | https://api.github.com/repos/huggingface/transformers/issues/40324/events | https://github.com/huggingface/transformers/pull/40324 | 3,338,935,298 | PR_kwDOCUB6oc6kilJy | 40,324 | Fix chunked attention mask with left-padding | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T17:18:01 | 2025-08-21T08:52:50 | 2025-08-21T08:52:49 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40324",
"html_url": "https://github.com/huggingface/transformers/pull/40324",
"diff_url": "https://github.com/huggingface/transformers/pull/40324.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40324.patch",
"merged_at": "2025-08-21T08:52:49"
} | # What does this PR do?
When we go beyond the chunk size in prefill or decoding with some sequences not spanning the chunk size yet, and we have batched inputs with padding, the padded sequences need some offset to start the chunk at the correct index. Fixes https://github.com/huggingface/transformers/issues/40214
Note that the fix only applies for recent torch versions - otherwise a warning is raised to ask to upgrade for support
| {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40324/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40324/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40323 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40323/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40323/comments | https://api.github.com/repos/huggingface/transformers/issues/40323/events | https://github.com/huggingface/transformers/issues/40323 | 3,338,684,984 | I_kwDOCUB6oc7HAEo4 | 40,323 | Is there a plan to add DINOv3 into AutoBackbone? | {
"login": "Farenweh",
"id": 57565535,
"node_id": "MDQ6VXNlcjU3NTY1NTM1",
"avatar_url": "https://avatars.githubusercontent.com/u/57565535?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Farenweh",
"html_url": "https://github.com/Farenweh",
"followers_url": "https://api.github.com/users/Farenweh/followers",
"following_url": "https://api.github.com/users/Farenweh/following{/other_user}",
"gists_url": "https://api.github.com/users/Farenweh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Farenweh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Farenweh/subscriptions",
"organizations_url": "https://api.github.com/users/Farenweh/orgs",
"repos_url": "https://api.github.com/users/Farenweh/repos",
"events_url": "https://api.github.com/users/Farenweh/events{/privacy}",
"received_events_url": "https://api.github.com/users/Farenweh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2648621985,
"node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request",
"name": "Feature request",
"color": "FBCA04",
"default": false,
"description": "Request for a new feature"
},
{
"id": 5769473378,
"node_id": "LA_kwDOCUB6oc8AAAABV-MtYg",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Vision",
"name": "Vision",
"color": "C079EF",
"default": false,
"description": ""
}
] | open | false | null | [] | null | [] | 2025-08-20T16:02:45 | 2025-08-24T21:23:15 | null | NONE | null | null | null | null | ### Feature request
Is there a plan to add DINOv3 to AutoBackbone. At present, DINOv2 is already inside, and I think DINOv3 should be able to inherit it directly. Appreciate a lot.
### Motivation
For the convenience of use
### Your contribution
DINOv3 should be able to inherit from DINOv2 directly. | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40323/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40323/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40322 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40322/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40322/comments | https://api.github.com/repos/huggingface/transformers/issues/40322/events | https://github.com/huggingface/transformers/pull/40322 | 3,338,588,534 | PR_kwDOCUB6oc6khYck | 40,322 | Fix: Only call Trainer.align_special_tokens if model has "config" attribute | {
"login": "tomaarsen",
"id": 37621491,
"node_id": "MDQ6VXNlcjM3NjIxNDkx",
"avatar_url": "https://avatars.githubusercontent.com/u/37621491?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tomaarsen",
"html_url": "https://github.com/tomaarsen",
"followers_url": "https://api.github.com/users/tomaarsen/followers",
"following_url": "https://api.github.com/users/tomaarsen/following{/other_user}",
"gists_url": "https://api.github.com/users/tomaarsen/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tomaarsen/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tomaarsen/subscriptions",
"organizations_url": "https://api.github.com/users/tomaarsen/orgs",
"repos_url": "https://api.github.com/users/tomaarsen/repos",
"events_url": "https://api.github.com/users/tomaarsen/events{/privacy}",
"received_events_url": "https://api.github.com/users/tomaarsen/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T15:28:30 | 2025-08-21T11:06:46 | 2025-08-21T11:06:42 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40322",
"html_url": "https://github.com/huggingface/transformers/pull/40322",
"diff_url": "https://github.com/huggingface/transformers/pull/40322.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40322.patch",
"merged_at": "2025-08-21T11:06:42"
} | Resolves https://github.com/huggingface/transformers/pull/38441#issuecomment-3206817330
# What does this PR do?
#38441 introduced a function on the `Trainer` that calls `self.model.config`, but this only works if the trained model has a `config` parameter. This isn't always the case, especially for third parties like Sentence Transformers and SetFit.
We've encountered this before in the Trainer, and we explicitly check whether the `self.model` has this `config` attribute:
https://github.com/huggingface/transformers/blob/4977ec2ae81d868fcd96ad4c7b1bffdc8c57f52c/src/transformers/trainer.py#L4708-L4711
The very straight-forward fix is simply to only call this function if `config` exists on `self.model`.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@SunMarc @gante
## TODO
@gante I'm open to advice re. tests to prevent future regressions here. I've not worked with the trainer tests before.
- Tom Aarsen | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40322/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40322/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40321 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40321/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40321/comments | https://api.github.com/repos/huggingface/transformers/issues/40321/events | https://github.com/huggingface/transformers/pull/40321 | 3,338,554,061 | PR_kwDOCUB6oc6khRCC | 40,321 | [WIP] Add Ovis2.5 | {
"login": "thisisiron",
"id": 23303033,
"node_id": "MDQ6VXNlcjIzMzAzMDMz",
"avatar_url": "https://avatars.githubusercontent.com/u/23303033?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thisisiron",
"html_url": "https://github.com/thisisiron",
"followers_url": "https://api.github.com/users/thisisiron/followers",
"following_url": "https://api.github.com/users/thisisiron/following{/other_user}",
"gists_url": "https://api.github.com/users/thisisiron/gists{/gist_id}",
"starred_url": "https://api.github.com/users/thisisiron/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/thisisiron/subscriptions",
"organizations_url": "https://api.github.com/users/thisisiron/orgs",
"repos_url": "https://api.github.com/users/thisisiron/repos",
"events_url": "https://api.github.com/users/thisisiron/events{/privacy}",
"received_events_url": "https://api.github.com/users/thisisiron/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | open | false | null | [] | null | [] | 2025-08-20T15:17:54 | 2025-09-16T14:45:30 | null | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40321",
"html_url": "https://github.com/huggingface/transformers/pull/40321",
"diff_url": "https://github.com/huggingface/transformers/pull/40321.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40321.patch",
"merged_at": null
} | # What does this PR do?
This PR adds support for Ovis2.5, a multimodal vision-language model designed for native-resolution visual perception and reflective reasoning.
Ovis2.5 is released in two variants:
- [Ovis2.5-9B](https://huggingface.co/AIDC-AI/Ovis2.5-9B)
- [Ovis2.5-2B](https://huggingface.co/AIDC-AI/Ovis2.5-2B)
This is still a draft PR. Further integration and testing to follow.
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
| null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40321/reactions",
"total_count": 2,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 2,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40321/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40320 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40320/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40320/comments | https://api.github.com/repos/huggingface/transformers/issues/40320/events | https://github.com/huggingface/transformers/pull/40320 | 3,338,372,449 | PR_kwDOCUB6oc6kgpUl | 40,320 | [RoPE] explicit factor > implicit factor in YaRN | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T14:27:43 | 2025-08-26T23:53:20 | 2025-08-26T13:58:28 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40320",
"html_url": "https://github.com/huggingface/transformers/pull/40320",
"diff_url": "https://github.com/huggingface/transformers/pull/40320.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40320.patch",
"merged_at": "2025-08-26T13:58:28"
} | # What does this PR do?
Fixes #38224
This PR:
- Makes the explicit RoPE factor take precedence over the implicit RoPE factor
- On YaRN validation, throw warnings if it has missing/bad parameterization (related to the implicit RoPE factor) | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40320/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40320/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40319 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40319/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40319/comments | https://api.github.com/repos/huggingface/transformers/issues/40319/events | https://github.com/huggingface/transformers/pull/40319 | 3,338,350,622 | PR_kwDOCUB6oc6kgkjF | 40,319 | add type hints | {
"login": "wirthual",
"id": 2640499,
"node_id": "MDQ6VXNlcjI2NDA0OTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/2640499?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/wirthual",
"html_url": "https://github.com/wirthual",
"followers_url": "https://api.github.com/users/wirthual/followers",
"following_url": "https://api.github.com/users/wirthual/following{/other_user}",
"gists_url": "https://api.github.com/users/wirthual/gists{/gist_id}",
"starred_url": "https://api.github.com/users/wirthual/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/wirthual/subscriptions",
"organizations_url": "https://api.github.com/users/wirthual/orgs",
"repos_url": "https://api.github.com/users/wirthual/repos",
"events_url": "https://api.github.com/users/wirthual/events{/privacy}",
"received_events_url": "https://api.github.com/users/wirthual/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T14:22:00 | 2025-08-21T11:20:00 | 2025-08-21T11:20:00 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40319",
"html_url": "https://github.com/huggingface/transformers/pull/40319",
"diff_url": "https://github.com/huggingface/transformers/pull/40319.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40319.patch",
"merged_at": "2025-08-21T11:20:00"
} | # What does this PR do?
Add type hints to `import_utils.py`
Based on these [docs](https://mypy.readthedocs.io/en/stable/running_mypy.html#missing-library-stubs-or-py-typed-marker), this change should avoid errors like:
```
infinity_emb/inference/loading_strategy.py:10: error: Skipping analyzing "transformers.utils.import_utils": module is installed, but missing library stubs or py.typed marker [import-untyped]
```
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
## Before submitting
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
| {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40319/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40319/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40318 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40318/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40318/comments | https://api.github.com/repos/huggingface/transformers/issues/40318/events | https://github.com/huggingface/transformers/issues/40318 | 3,338,185,349 | I_kwDOCUB6oc7G-KqF | 40,318 | TypeError: 'builtins.safe_open' object is not iterable / ValueError: Exception encountered when calling layer 'vit' (type TFViTMainLayer). | {
"login": "ngvananh2508",
"id": 130348585,
"node_id": "U_kgDOB8T2KQ",
"avatar_url": "https://avatars.githubusercontent.com/u/130348585?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ngvananh2508",
"html_url": "https://github.com/ngvananh2508",
"followers_url": "https://api.github.com/users/ngvananh2508/followers",
"following_url": "https://api.github.com/users/ngvananh2508/following{/other_user}",
"gists_url": "https://api.github.com/users/ngvananh2508/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ngvananh2508/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ngvananh2508/subscriptions",
"organizations_url": "https://api.github.com/users/ngvananh2508/orgs",
"repos_url": "https://api.github.com/users/ngvananh2508/repos",
"events_url": "https://api.github.com/users/ngvananh2508/events{/privacy}",
"received_events_url": "https://api.github.com/users/ngvananh2508/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-20T13:38:18 | 2025-08-22T13:52:19 | 2025-08-22T13:52:19 | NONE | null | null | null | null | ### System Info
Transformer version: 4.55.2, TF: 2.19.0, Python: 3.11.11, Ubuntu: 24.04,
### Who can help?
@amyeroberts @qubvel
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
```python
from transformers import ViTFeatureExtractor, TFViTModel
base_model = TFViTModel.from_pretrained('google/vit-base-patch16-224-in21k')
```
```
TypeError Traceback (most recent call last)
[/usr/local/lib/python3.12/dist-packages/transformers/modeling_tf_pytorch_utils.py](https://localhost:8080/#) in load_pytorch_state_dict_in_tf2_model(tf_model, pt_state_dict, tf_inputs, allow_missing_keys, output_loading_info, _prefix, tf_to_pt_weight_rename, ignore_mismatched_sizes, skip_logger_warnings)
331 # Convert old format to new format if needed from a PyTorch state_dict
332 tf_keys_to_pt_keys = {}
--> 333 for key in pt_state_dict:
334 new_key = None
335 if "gamma" in key:
TypeError: 'builtins.safe_open' object is not iterable,
```
When I use parameter: use_safetensors=False, I downloaded the model successfully but I had another issue:
```py
from tensorflow.keras.layers import Input
inputs = Input(shape=(3,224,224))
outputs = base_model.vit(x)
hf_model = Model(inputs=inputs, outputs=outputs)
```
```
ValueError: Exception encountered when calling layer 'vit' (type TFViTMainLayer).
Data of type <class 'keras.src.backend.common.keras_tensor.KerasTensor'> is not allowed only (<class 'tensorflow.python.framework.tensor.Tensor'>, <class 'bool'>, <class 'int'>, <class 'transformers.utils.generic.ModelOutput'>, <class 'tuple'>, <class 'list'>, <class 'dict'>, <class 'numpy.ndarray'>) is accepted for pixel_values.
Call arguments received by layer 'vit' (type TFViTMainLayer):
• pixel_values=<KerasTensor shape=(None, 3, 224, 224), dtype=float32, sparse=False, ragged=False, name=keras_tensor_10>
• head_mask=None
• output_attentions=None
• output_hidden_states=None
• interpolate_pos_encoding=None
• return_dict=None
• training=False
```
### Expected behavior
Download the model successfully
| {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40318/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40318/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40317 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40317/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40317/comments | https://api.github.com/repos/huggingface/transformers/issues/40317/events | https://github.com/huggingface/transformers/pull/40317 | 3,338,019,327 | PR_kwDOCUB6oc6kfcv3 | 40,317 | byebye torch 2.1 | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T12:51:36 | 2025-08-20T14:03:48 | 2025-08-20T14:03:46 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40317",
"html_url": "https://github.com/huggingface/transformers/pull/40317",
"diff_url": "https://github.com/huggingface/transformers/pull/40317.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40317.patch",
"merged_at": "2025-08-20T14:03:46"
} | Torch 2.1 is almost 2 years old now and breaks in a lot of cases, so it's time to bump the minimum torch version to 2.2.
Because we now support SDPA in all cases, this PR also removes `require_torch_sdpa` in tests, and `is_torch_sdpa_available` is now mostly a dummy function that we retain for compatibility with remote code only. | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40317/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40317/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40316 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40316/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40316/comments | https://api.github.com/repos/huggingface/transformers/issues/40316/events | https://github.com/huggingface/transformers/pull/40316 | 3,337,906,951 | PR_kwDOCUB6oc6kfEd7 | 40,316 | Make cache_config not mandatory | {
"login": "remi-or",
"id": 83456801,
"node_id": "MDQ6VXNlcjgzNDU2ODAx",
"avatar_url": "https://avatars.githubusercontent.com/u/83456801?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/remi-or",
"html_url": "https://github.com/remi-or",
"followers_url": "https://api.github.com/users/remi-or/followers",
"following_url": "https://api.github.com/users/remi-or/following{/other_user}",
"gists_url": "https://api.github.com/users/remi-or/gists{/gist_id}",
"starred_url": "https://api.github.com/users/remi-or/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/remi-or/subscriptions",
"organizations_url": "https://api.github.com/users/remi-or/orgs",
"repos_url": "https://api.github.com/users/remi-or/repos",
"events_url": "https://api.github.com/users/remi-or/events{/privacy}",
"received_events_url": "https://api.github.com/users/remi-or/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T12:16:50 | 2025-08-26T10:06:18 | 2025-08-26T10:06:18 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40316",
"html_url": "https://github.com/huggingface/transformers/pull/40316",
"diff_url": "https://github.com/huggingface/transformers/pull/40316.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40316.patch",
"merged_at": "2025-08-26T10:06:18"
} | With recent changes introduced in #39836 the tests related to `executorch.py` such as `test_export_text_only_with_hybrid_cache` will fail if the model `generation_config` has no `cache_config` or if any of the required attribute is missing from `cache_config` . This is the case for models such has https://huggingface.co/google/gemma-3-4b-it which is already used in `gemma3` models test.
This PR aims to remove the requirement of `cache_config` by restoring old default values, albeit while throwing a warning so that users are aware these attributes are missing from the config. This will ensure backwards-compatibility for old models without having to edit their hub config.
cc. @jackzhxng @zucchini-nlp as this relates to #39836 | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40316/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40316/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40315 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40315/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40315/comments | https://api.github.com/repos/huggingface/transformers/issues/40315/events | https://github.com/huggingface/transformers/pull/40315 | 3,337,881,958 | PR_kwDOCUB6oc6ke-7N | 40,315 | Remove unnecessary contiguous calls for modern torch | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T12:09:06 | 2025-08-20T12:24:44 | 2025-08-20T12:24:14 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40315",
"html_url": "https://github.com/huggingface/transformers/pull/40315",
"diff_url": "https://github.com/huggingface/transformers/pull/40315.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40315.patch",
"merged_at": "2025-08-20T12:24:14"
} | `sdpa_integration.py` calls `contiguous()` on the three attention tensors to work around a bug that was present before `torch 2.2`. However, our minimum version is now >= 2.2 anyway, so this bug should not longer be relevant, and we can remove the workaround to boost performance.
Pytorch issue link: https://github.com/pytorch/pytorch/issues/112577 | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40315/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40315/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40314 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40314/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40314/comments | https://api.github.com/repos/huggingface/transformers/issues/40314/events | https://github.com/huggingface/transformers/pull/40314 | 3,337,662,198 | PR_kwDOCUB6oc6keN_a | 40,314 | Update notification service amd_daily_ci_workflows definition | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T11:04:49 | 2025-08-20T15:49:49 | 2025-08-20T15:49:47 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40314",
"html_url": "https://github.com/huggingface/transformers/pull/40314",
"diff_url": "https://github.com/huggingface/transformers/pull/40314.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40314.patch",
"merged_at": "2025-08-20T15:49:47"
} | Add mi355 wf to amd_daily_ci_workflows | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40314/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40314/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40313 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40313/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40313/comments | https://api.github.com/repos/huggingface/transformers/issues/40313/events | https://github.com/huggingface/transformers/pull/40313 | 3,337,660,898 | PR_kwDOCUB6oc6keNtY | 40,313 | T5 test and target device fixes | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T11:04:24 | 2025-08-21T12:07:29 | 2025-08-21T12:07:29 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40313",
"html_url": "https://github.com/huggingface/transformers/pull/40313",
"diff_url": "https://github.com/huggingface/transformers/pull/40313.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40313.patch",
"merged_at": "2025-08-21T12:07:29"
} | # What does this PR do?
This PR fixes the tests for the T5 model. I encountered some actual target device issues in `executorch` which I've tried to fix. By the looks of it, there are no regressions, but it requires a proper review.
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40313/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40313/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40312 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40312/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40312/comments | https://api.github.com/repos/huggingface/transformers/issues/40312/events | https://github.com/huggingface/transformers/pull/40312 | 3,337,599,638 | PR_kwDOCUB6oc6kd_7Y | 40,312 | Fix GOT-OCR2 and Cohere2Vision image processor patches caculation | {
"login": "Isotr0py",
"id": 41363108,
"node_id": "MDQ6VXNlcjQxMzYzMTA4",
"avatar_url": "https://avatars.githubusercontent.com/u/41363108?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Isotr0py",
"html_url": "https://github.com/Isotr0py",
"followers_url": "https://api.github.com/users/Isotr0py/followers",
"following_url": "https://api.github.com/users/Isotr0py/following{/other_user}",
"gists_url": "https://api.github.com/users/Isotr0py/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Isotr0py/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Isotr0py/subscriptions",
"organizations_url": "https://api.github.com/users/Isotr0py/orgs",
"repos_url": "https://api.github.com/users/Isotr0py/repos",
"events_url": "https://api.github.com/users/Isotr0py/events{/privacy}",
"received_events_url": "https://api.github.com/users/Isotr0py/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T10:48:01 | 2025-08-20T12:32:29 | 2025-08-20T11:13:58 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40312",
"html_url": "https://github.com/huggingface/transformers/pull/40312",
"diff_url": "https://github.com/huggingface/transformers/pull/40312.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40312.patch",
"merged_at": "2025-08-20T11:13:58"
} | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
- Related issue: https://github.com/vllm-project/vllm/pull/23245#issue-3337120245
- Fix incorrect num_patches calculation for GOT-OCR2 and Cohere2Vision image processor
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40312/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40312/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40311 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40311/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40311/comments | https://api.github.com/repos/huggingface/transformers/issues/40311/events | https://github.com/huggingface/transformers/pull/40311 | 3,337,447,775 | PR_kwDOCUB6oc6kdeWr | 40,311 | [docs] remove flax references from `/en/model_doc` | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T10:05:28 | 2025-08-21T10:00:35 | 2025-08-21T09:52:55 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40311",
"html_url": "https://github.com/huggingface/transformers/pull/40311",
"diff_url": "https://github.com/huggingface/transformers/pull/40311.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40311.patch",
"merged_at": "2025-08-21T09:52:55"
} | # What does this PR do?
In all files in `/en/model_doc` (i.e. transformers model docs):
1. finds all files with `"flax"` references
2. deletes all `"flax"` and `"tf"` references in those files
3. (minor nits were also fixed in those files along the way, like missing terminations of code blocks or missing paper reference)
P.S.: the push doc builder CI got significantly faster :o This PR shaves off >1 min from it (~9mins now) | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40311/reactions",
"total_count": 4,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 2,
"confused": 0,
"heart": 2,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40311/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40310 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40310/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40310/comments | https://api.github.com/repos/huggingface/transformers/issues/40310/events | https://github.com/huggingface/transformers/pull/40310 | 3,337,387,482 | PR_kwDOCUB6oc6kdRQh | 40,310 | Fix links in Glm4vMoe configuration classes to point to the correct H… | {
"login": "vvvdwbvvv",
"id": 56776714,
"node_id": "MDQ6VXNlcjU2Nzc2NzE0",
"avatar_url": "https://avatars.githubusercontent.com/u/56776714?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vvvdwbvvv",
"html_url": "https://github.com/vvvdwbvvv",
"followers_url": "https://api.github.com/users/vvvdwbvvv/followers",
"following_url": "https://api.github.com/users/vvvdwbvvv/following{/other_user}",
"gists_url": "https://api.github.com/users/vvvdwbvvv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vvvdwbvvv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vvvdwbvvv/subscriptions",
"organizations_url": "https://api.github.com/users/vvvdwbvvv/orgs",
"repos_url": "https://api.github.com/users/vvvdwbvvv/repos",
"events_url": "https://api.github.com/users/vvvdwbvvv/events{/privacy}",
"received_events_url": "https://api.github.com/users/vvvdwbvvv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T09:47:05 | 2025-08-21T11:43:14 | 2025-08-21T11:42:54 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40310",
"html_url": "https://github.com/huggingface/transformers/pull/40310",
"diff_url": "https://github.com/huggingface/transformers/pull/40310.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40310.patch",
"merged_at": "2025-08-21T11:42:54"
} |
# What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes the path import GLM-4.5V to `zai-org/GLM-4.5V` in `modular_glm4v_moe.py`
## Before submitting
- [x] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [x] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40310/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40310/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40309 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40309/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40309/comments | https://api.github.com/repos/huggingface/transformers/issues/40309/events | https://github.com/huggingface/transformers/pull/40309 | 3,337,384,382 | PR_kwDOCUB6oc6kdQkV | 40,309 | Change multimodal data links to HF hub | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T09:46:17 | 2025-08-22T09:50:04 | 2025-08-22T09:50:04 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40309",
"html_url": "https://github.com/huggingface/transformers/pull/40309",
"diff_url": "https://github.com/huggingface/transformers/pull/40309.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40309.patch",
"merged_at": "2025-08-22T09:50:04"
} | # What does this PR do?
As per title, it was reported to be flaky due to excess requests in CI
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40309/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40309/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40308 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40308/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40308/comments | https://api.github.com/repos/huggingface/transformers/issues/40308/events | https://github.com/huggingface/transformers/pull/40308 | 3,337,347,658 | PR_kwDOCUB6oc6kdIZX | 40,308 | Qwen2.5-VL test fixes for ROCm | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T09:36:28 | 2025-08-21T16:13:07 | 2025-08-21T16:13:07 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40308",
"html_url": "https://github.com/huggingface/transformers/pull/40308",
"diff_url": "https://github.com/huggingface/transformers/pull/40308.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40308.patch",
"merged_at": "2025-08-21T16:13:07"
} | # What does this PR do?
This PR adjusts the output expectations for the ROCm platform in testing.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40308/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40308/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40307 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40307/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40307/comments | https://api.github.com/repos/huggingface/transformers/issues/40307/events | https://github.com/huggingface/transformers/pull/40307 | 3,337,263,330 | PR_kwDOCUB6oc6kc15R | 40,307 | Qwen2.5-Omni test fixes | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T09:13:30 | 2025-08-20T12:48:30 | 2025-08-20T12:48:30 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40307",
"html_url": "https://github.com/huggingface/transformers/pull/40307",
"diff_url": "https://github.com/huggingface/transformers/pull/40307.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40307.patch",
"merged_at": "2025-08-20T12:48:30"
} | # What does this PR do?
This PR makes the test suite for Qwen2.5-Omni green: updates output expectations and fixes the model parallel test.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40307/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40307/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40306 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40306/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40306/comments | https://api.github.com/repos/huggingface/transformers/issues/40306/events | https://github.com/huggingface/transformers/pull/40306 | 3,337,047,796 | PR_kwDOCUB6oc6kcHf- | 40,306 | Update CI with nightly torch workflow file | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T08:11:42 | 2025-08-20T14:59:03 | 2025-08-20T14:59:01 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40306",
"html_url": "https://github.com/huggingface/transformers/pull/40306",
"diff_url": "https://github.com/huggingface/transformers/pull/40306.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40306.patch",
"merged_at": "2025-08-20T14:59:01"
} | # What does this PR do?
Fix CI with nightly torch + make it produce the new failed tests reports (compared to the Nvidia CI with stable torch).
See comments in `Files changed` | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40306/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40306/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40305 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40305/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40305/comments | https://api.github.com/repos/huggingface/transformers/issues/40305/events | https://github.com/huggingface/transformers/pull/40305 | 3,336,849,445 | PR_kwDOCUB6oc6kbcyQ | 40,305 | Remove OTel SDK dependencies | {
"login": "anuraaga",
"id": 198344,
"node_id": "MDQ6VXNlcjE5ODM0NA==",
"avatar_url": "https://avatars.githubusercontent.com/u/198344?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anuraaga",
"html_url": "https://github.com/anuraaga",
"followers_url": "https://api.github.com/users/anuraaga/followers",
"following_url": "https://api.github.com/users/anuraaga/following{/other_user}",
"gists_url": "https://api.github.com/users/anuraaga/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anuraaga/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anuraaga/subscriptions",
"organizations_url": "https://api.github.com/users/anuraaga/orgs",
"repos_url": "https://api.github.com/users/anuraaga/repos",
"events_url": "https://api.github.com/users/anuraaga/events{/privacy}",
"received_events_url": "https://api.github.com/users/anuraaga/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-20T07:08:14 | 2025-08-20T10:31:44 | 2025-08-20T10:31:44 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40305",
"html_url": "https://github.com/huggingface/transformers/pull/40305",
"diff_url": "https://github.com/huggingface/transformers/pull/40305.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40305.patch",
"merged_at": "2025-08-20T10:31:44"
} | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
Removes unused dependencies on otel SDK.
I noticed #39422 removed the usage but not the dependencies.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [X] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@McPatate @ArthurZucker
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40305/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40305/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40304 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40304/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40304/comments | https://api.github.com/repos/huggingface/transformers/issues/40304/events | https://github.com/huggingface/transformers/pull/40304 | 3,336,428,667 | PR_kwDOCUB6oc6kaDiM | 40,304 | Gpt oss optim | {
"login": "jiqing-feng",
"id": 107918818,
"node_id": "U_kgDOBm614g",
"avatar_url": "https://avatars.githubusercontent.com/u/107918818?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jiqing-feng",
"html_url": "https://github.com/jiqing-feng",
"followers_url": "https://api.github.com/users/jiqing-feng/followers",
"following_url": "https://api.github.com/users/jiqing-feng/following{/other_user}",
"gists_url": "https://api.github.com/users/jiqing-feng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jiqing-feng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jiqing-feng/subscriptions",
"organizations_url": "https://api.github.com/users/jiqing-feng/orgs",
"repos_url": "https://api.github.com/users/jiqing-feng/repos",
"events_url": "https://api.github.com/users/jiqing-feng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jiqing-feng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 7510456769,
"node_id": "LA_kwDOCUB6oc8AAAABv6h5wQ",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Mixture%20of%20Experts",
"name": "Mixture of Experts",
"color": "DDB5D0",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2025-08-20T03:22:47 | 2025-08-29T06:32:22 | 2025-08-25T12:36:33 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40304",
"html_url": "https://github.com/huggingface/transformers/pull/40304",
"diff_url": "https://github.com/huggingface/transformers/pull/40304.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40304.patch",
"merged_at": "2025-08-25T12:36:33"
} | Enabling fast indexing for CPU. This optimization can bring 3x speed-up for [lmsys/gpt-oss-20b-bf16](https://huggingface.co/lmsys/gpt-oss-20b-bf16) on Intel 6th Gen Xeon. | {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40304/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40304/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40303 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40303/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40303/comments | https://api.github.com/repos/huggingface/transformers/issues/40303/events | https://github.com/huggingface/transformers/issues/40303 | 3,336,345,426 | I_kwDOCUB6oc7G3JdS | 40,303 | Add mean_token_accuracy from TRL's `SFTTrainer` to `Trainer` | {
"login": "RylanSchaeffer",
"id": 8942987,
"node_id": "MDQ6VXNlcjg5NDI5ODc=",
"avatar_url": "https://avatars.githubusercontent.com/u/8942987?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RylanSchaeffer",
"html_url": "https://github.com/RylanSchaeffer",
"followers_url": "https://api.github.com/users/RylanSchaeffer/followers",
"following_url": "https://api.github.com/users/RylanSchaeffer/following{/other_user}",
"gists_url": "https://api.github.com/users/RylanSchaeffer/gists{/gist_id}",
"starred_url": "https://api.github.com/users/RylanSchaeffer/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/RylanSchaeffer/subscriptions",
"organizations_url": "https://api.github.com/users/RylanSchaeffer/orgs",
"repos_url": "https://api.github.com/users/RylanSchaeffer/repos",
"events_url": "https://api.github.com/users/RylanSchaeffer/events{/privacy}",
"received_events_url": "https://api.github.com/users/RylanSchaeffer/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2648621985,
"node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request",
"name": "Feature request",
"color": "FBCA04",
"default": false,
"description": "Request for a new feature"
}
] | open | false | null | [] | null | [] | 2025-08-20T02:32:47 | 2025-08-21T03:18:42 | null | NONE | null | null | null | null | ### Feature request
TRL's `SFTTrainer` has a metric called `mean_token_accuracy`: https://github.com/huggingface/trl/blob/main/trl/trainer/sft_trainer.py#L1095-L1123:
```
shift_logits = outputs.logits[..., :-1, :].contiguous()
shift_labels = inputs["labels"][..., 1:].contiguous()
# When using Prompt Tuning, skip the virtual tokens in logits before accuracy computation, since they do
# not correspond to actual input labels.
shift_logits = shift_logits[:, self.num_virtual_tokens :, :]
# Get predictions
predictions = shift_logits.argmax(dim=-1)
# Create mask for non-padding tokens (assuming ignore_index is -100)
mask = shift_labels != -100
# Calculate accuracy only on non-padding tokens
correct_predictions = (predictions == shift_labels) & mask
total_tokens = mask.sum()
correct_tokens = correct_predictions.sum()
# Gather the correct_tokens and total_tokens across all processes
correct_tokens = self.accelerator.gather_for_metrics(correct_tokens)
total_tokens = self.accelerator.gather_for_metrics(total_tokens)
# Compute the mean token accuracy and log it
total_sum = total_tokens.sum()
accuracy = (correct_tokens.sum() / total_sum).item() if total_sum > 0 else 0.0
self._metrics[mode]["mean_token_accuracy"].append(accuracy)
```
I like this metric and I'd like for this metric to be optionally be added to `transformer`'s `Trainer`.
### Motivation
(Mean) Token Accuracy is a useful additional metric that I'd like the option of additionally logging.
| null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40303/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40303/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40302 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40302/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40302/comments | https://api.github.com/repos/huggingface/transformers/issues/40302/events | https://github.com/huggingface/transformers/issues/40302 | 3,335,997,066 | I_kwDOCUB6oc7G10aK | 40,302 | Add audio-text-to-text task | {
"login": "jackzhxng",
"id": 32371937,
"node_id": "MDQ6VXNlcjMyMzcxOTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/32371937?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jackzhxng",
"html_url": "https://github.com/jackzhxng",
"followers_url": "https://api.github.com/users/jackzhxng/followers",
"following_url": "https://api.github.com/users/jackzhxng/following{/other_user}",
"gists_url": "https://api.github.com/users/jackzhxng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jackzhxng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jackzhxng/subscriptions",
"organizations_url": "https://api.github.com/users/jackzhxng/orgs",
"repos_url": "https://api.github.com/users/jackzhxng/repos",
"events_url": "https://api.github.com/users/jackzhxng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jackzhxng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2648621985,
"node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request",
"name": "Feature request",
"color": "FBCA04",
"default": false,
"description": "Request for a new feature"
}
] | open | false | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [] | 2025-08-19T23:18:30 | 2025-09-09T20:52:48 | null | CONTRIBUTOR | null | null | null | null | ### Feature request
Add `audio-text-to-text`, for models like Voxtral. https://github.com/huggingface/transformers/blob/main/utils/update_metadata.py#L71, or something like `multimodal-text-generation` works too.
@Rocketknight1
### Motivation
Needed to enable Voxtral on Optimum ET cleanly
### Your contribution
N/A - I'd contribute but I'm not sure where the touch points are to add a task like this or if it's even okay to add a new task | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40302/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40302/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40301 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40301/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40301/comments | https://api.github.com/repos/huggingface/transformers/issues/40301/events | https://github.com/huggingface/transformers/issues/40301 | 3,335,943,934 | I_kwDOCUB6oc7G1nb- | 40,301 | Label smoothing for multi-label classification | {
"login": "andreabosisio",
"id": 79710398,
"node_id": "MDQ6VXNlcjc5NzEwMzk4",
"avatar_url": "https://avatars.githubusercontent.com/u/79710398?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/andreabosisio",
"html_url": "https://github.com/andreabosisio",
"followers_url": "https://api.github.com/users/andreabosisio/followers",
"following_url": "https://api.github.com/users/andreabosisio/following{/other_user}",
"gists_url": "https://api.github.com/users/andreabosisio/gists{/gist_id}",
"starred_url": "https://api.github.com/users/andreabosisio/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/andreabosisio/subscriptions",
"organizations_url": "https://api.github.com/users/andreabosisio/orgs",
"repos_url": "https://api.github.com/users/andreabosisio/repos",
"events_url": "https://api.github.com/users/andreabosisio/events{/privacy}",
"received_events_url": "https://api.github.com/users/andreabosisio/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2648621985,
"node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request",
"name": "Feature request",
"color": "FBCA04",
"default": false,
"description": "Request for a new feature"
}
] | open | false | null | [] | null | [] | 2025-08-19T22:57:37 | 2025-08-22T19:58:44 | null | NONE | null | null | null | null | ### Feature request
Implement a method to apply label smoothing for multi-label classification tasks
### Motivation
I tried applying label smoothing to my multi-label classification problem—given that my dataset is noisy and unbalanced, I thought it might help—but I ran into issue #40258
### Your contribution
I recommend implementing the method outlined in this Reddit post: [[R] Simple Correction for Label Smoothing with Multi-Label Classification](https://www.reddit.com/r/MachineLearning/comments/oh1ed6/r_simple_correction_for_label_smoothing_with/), as it seems to provide a valid solution | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40301/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40301/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40300 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40300/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40300/comments | https://api.github.com/repos/huggingface/transformers/issues/40300/events | https://github.com/huggingface/transformers/pull/40300 | 3,335,941,661 | PR_kwDOCUB6oc6kYaLl | 40,300 | fix: use consistent dtype for DETR sine positional embeddings | {
"login": "agkphysics",
"id": 5217766,
"node_id": "MDQ6VXNlcjUyMTc3NjY=",
"avatar_url": "https://avatars.githubusercontent.com/u/5217766?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/agkphysics",
"html_url": "https://github.com/agkphysics",
"followers_url": "https://api.github.com/users/agkphysics/followers",
"following_url": "https://api.github.com/users/agkphysics/following{/other_user}",
"gists_url": "https://api.github.com/users/agkphysics/gists{/gist_id}",
"starred_url": "https://api.github.com/users/agkphysics/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/agkphysics/subscriptions",
"organizations_url": "https://api.github.com/users/agkphysics/orgs",
"repos_url": "https://api.github.com/users/agkphysics/repos",
"events_url": "https://api.github.com/users/agkphysics/events{/privacy}",
"received_events_url": "https://api.github.com/users/agkphysics/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T22:56:47 | 2025-08-21T14:49:57 | 2025-08-21T14:49:57 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40300",
"html_url": "https://github.com/huggingface/transformers/pull/40300",
"diff_url": "https://github.com/huggingface/transformers/pull/40300.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40300.patch",
"merged_at": "2025-08-21T14:49:57"
} | # What does this PR do?
Allows the use of non-float32 types, such as float16, with DETR models.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@amyeroberts, @qubvel | {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40300/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40300/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40299 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40299/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40299/comments | https://api.github.com/repos/huggingface/transformers/issues/40299/events | https://github.com/huggingface/transformers/pull/40299 | 3,335,879,092 | PR_kwDOCUB6oc6kYMA4 | 40,299 | Remove deprecated max_size parameter from ConditionalDetr image processors | {
"login": "Rishirandhawa",
"id": 18483046,
"node_id": "MDQ6VXNlcjE4NDgzMDQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/18483046?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rishirandhawa",
"html_url": "https://github.com/Rishirandhawa",
"followers_url": "https://api.github.com/users/Rishirandhawa/followers",
"following_url": "https://api.github.com/users/Rishirandhawa/following{/other_user}",
"gists_url": "https://api.github.com/users/Rishirandhawa/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rishirandhawa/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rishirandhawa/subscriptions",
"organizations_url": "https://api.github.com/users/Rishirandhawa/orgs",
"repos_url": "https://api.github.com/users/Rishirandhawa/repos",
"events_url": "https://api.github.com/users/Rishirandhawa/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rishirandhawa/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | open | false | null | [] | null | [] | 2025-08-19T22:31:50 | 2025-08-20T11:21:35 | null | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40299",
"html_url": "https://github.com/huggingface/transformers/pull/40299",
"diff_url": "https://github.com/huggingface/transformers/pull/40299.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40299.patch",
"merged_at": null
} | ## Summary
This PR removes the deprecated `max_size` parameter from both `ConditionalDetrImageProcessor` and `ConditionalDetrImageProcessorFast` classes.
## Fixes
Closes #37939
## Changes
- **Removed** `max_size` parameter handling from `ConditionalDetrImageProcessor.__init__`
- **Removed** `max_size` parameter handling from `ConditionalDetrImageProcessorFast.__init__`
- **Removed** `max_size` parameter handling from `from_dict` methods in both classes
- **Removed** `max_size` parameter handling from `preprocess`/`resize` methods
- **Removed** all deprecation warnings that referenced v4.26 removal
## Breaking Change
⚠️ **This is a breaking change** - the `max_size` parameter is no longer accepted.
Users should migrate to using `size['longest_edge']` instead:
```python
# Before (deprecated, now removed):
processor = ConditionalDetrImageProcessor(max_size=800)
# After (correct usage):
processor = ConditionalDetrImageProcessor(size={'shortest_edge': 600, 'longest_edge': 800})
```
## Justification
The `max_size` parameter was marked for deprecation and scheduled for removal in v4.26. The current version is v4.56, so this removal is long overdue. The deprecation warnings have been showing for many versions, giving users ample time to migrate their code.
## Testing
- Verified that `max_size` parameter is no longer accepted and raises `TypeError`
- Verified that proper `size` parameter still works correctly
- Confirmed no deprecation warnings are emitted for normal usage | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40299/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40299/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40298 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40298/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40298/comments | https://api.github.com/repos/huggingface/transformers/issues/40298/events | https://github.com/huggingface/transformers/pull/40298 | 3,335,828,787 | PR_kwDOCUB6oc6kYBDs | 40,298 | [3/3] make docs device agnostic, all en docs for existing models done | {
"login": "yao-matrix",
"id": 7245027,
"node_id": "MDQ6VXNlcjcyNDUwMjc=",
"avatar_url": "https://avatars.githubusercontent.com/u/7245027?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yao-matrix",
"html_url": "https://github.com/yao-matrix",
"followers_url": "https://api.github.com/users/yao-matrix/followers",
"following_url": "https://api.github.com/users/yao-matrix/following{/other_user}",
"gists_url": "https://api.github.com/users/yao-matrix/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yao-matrix/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yao-matrix/subscriptions",
"organizations_url": "https://api.github.com/users/yao-matrix/orgs",
"repos_url": "https://api.github.com/users/yao-matrix/repos",
"events_url": "https://api.github.com/users/yao-matrix/events{/privacy}",
"received_events_url": "https://api.github.com/users/yao-matrix/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T22:08:41 | 2025-08-20T15:17:10 | 2025-08-20T04:01:28 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40298",
"html_url": "https://github.com/huggingface/transformers/pull/40298",
"diff_url": "https://github.com/huggingface/transformers/pull/40298.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40298.patch",
"merged_at": "2025-08-20T04:01:28"
} | Following https://github.com/huggingface/transformers/pull/40143 and https://github.com/huggingface/transformers/pull/40256, this is the last PR to make docs device agnostic, we've made all english docs for existing models device agnostic.
@stevhliu , pls help review, thx very much. | {
"login": "stevhliu",
"id": 59462357,
"node_id": "MDQ6VXNlcjU5NDYyMzU3",
"avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevhliu",
"html_url": "https://github.com/stevhliu",
"followers_url": "https://api.github.com/users/stevhliu/followers",
"following_url": "https://api.github.com/users/stevhliu/following{/other_user}",
"gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions",
"organizations_url": "https://api.github.com/users/stevhliu/orgs",
"repos_url": "https://api.github.com/users/stevhliu/repos",
"events_url": "https://api.github.com/users/stevhliu/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevhliu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40298/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40298/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40297 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40297/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40297/comments | https://api.github.com/repos/huggingface/transformers/issues/40297/events | https://github.com/huggingface/transformers/pull/40297 | 3,335,788,032 | PR_kwDOCUB6oc6kX4YV | 40,297 | Add Voxtral to AutoModelForSpeechSeq2Seq | {
"login": "jackzhxng",
"id": 32371937,
"node_id": "MDQ6VXNlcjMyMzcxOTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/32371937?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jackzhxng",
"html_url": "https://github.com/jackzhxng",
"followers_url": "https://api.github.com/users/jackzhxng/followers",
"following_url": "https://api.github.com/users/jackzhxng/following{/other_user}",
"gists_url": "https://api.github.com/users/jackzhxng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jackzhxng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jackzhxng/subscriptions",
"organizations_url": "https://api.github.com/users/jackzhxng/orgs",
"repos_url": "https://api.github.com/users/jackzhxng/repos",
"events_url": "https://api.github.com/users/jackzhxng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jackzhxng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T21:47:39 | 2025-08-19T23:18:47 | 2025-08-19T23:18:47 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40297",
"html_url": "https://github.com/huggingface/transformers/pull/40297",
"diff_url": "https://github.com/huggingface/transformers/pull/40297.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40297.patch",
"merged_at": null
} | # What does this PR do?
Add Voxtral to AutoModelForSpeechSeq2Seq, needed for enabling this model in Optimum ET (specifically for the CLI).
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@Rocketknight1
| {
"login": "jackzhxng",
"id": 32371937,
"node_id": "MDQ6VXNlcjMyMzcxOTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/32371937?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jackzhxng",
"html_url": "https://github.com/jackzhxng",
"followers_url": "https://api.github.com/users/jackzhxng/followers",
"following_url": "https://api.github.com/users/jackzhxng/following{/other_user}",
"gists_url": "https://api.github.com/users/jackzhxng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jackzhxng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jackzhxng/subscriptions",
"organizations_url": "https://api.github.com/users/jackzhxng/orgs",
"repos_url": "https://api.github.com/users/jackzhxng/repos",
"events_url": "https://api.github.com/users/jackzhxng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jackzhxng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40297/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40297/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40296 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40296/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40296/comments | https://api.github.com/repos/huggingface/transformers/issues/40296/events | https://github.com/huggingface/transformers/pull/40296 | 3,335,733,216 | PR_kwDOCUB6oc6kXshh | 40,296 | Fix label smoothing incompatibility with multi-label classification | {
"login": "avchauzov",
"id": 21357563,
"node_id": "MDQ6VXNlcjIxMzU3NTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/21357563?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/avchauzov",
"html_url": "https://github.com/avchauzov",
"followers_url": "https://api.github.com/users/avchauzov/followers",
"following_url": "https://api.github.com/users/avchauzov/following{/other_user}",
"gists_url": "https://api.github.com/users/avchauzov/gists{/gist_id}",
"starred_url": "https://api.github.com/users/avchauzov/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/avchauzov/subscriptions",
"organizations_url": "https://api.github.com/users/avchauzov/orgs",
"repos_url": "https://api.github.com/users/avchauzov/repos",
"events_url": "https://api.github.com/users/avchauzov/events{/privacy}",
"received_events_url": "https://api.github.com/users/avchauzov/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T21:19:53 | 2025-08-25T21:21:05 | 2025-08-25T14:23:31 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40296",
"html_url": "https://github.com/huggingface/transformers/pull/40296",
"diff_url": "https://github.com/huggingface/transformers/pull/40296.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40296.patch",
"merged_at": "2025-08-25T14:23:31"
} | # What does this PR do?
Fixes a RuntimeError when using `label_smoothing_factor > 0` with multi-label classification.
Previously, users would get a confusing error:
`RuntimeError: gather(): Expected dtype int32/int64 for index`
This PR adds a clear validation check in `LabelSmoother` that detects this incompatible combination and raises a helpful error message instead:
```
ValueError: Label smoothing is not compatible with multi-label classification.
Multi-label classification uses one-hot encoded labels (float tensors),
but label smoothing requires integer class indices.
Please set label_smoothing_factor=0.0 in your TrainingArguments.
```
**Changes:**
- Add validation in `LabelSmoother.__call__()` to detect multi-label classification
- Raise clear `ValueError` with actionable solution
- Add test case to prevent regression
**Testing:**
- ✅ All existing tests pass
- ✅ New test covers the fix
- ✅ Original reproduction case now shows clear error
Fixes #40258
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [x] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [x] Did you write any new necessary tests?
## Who can review?
@zach-huggingface @SunMarc @qgallouedec - trainer related changes | {
"login": "SunMarc",
"id": 57196510,
"node_id": "MDQ6VXNlcjU3MTk2NTEw",
"avatar_url": "https://avatars.githubusercontent.com/u/57196510?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SunMarc",
"html_url": "https://github.com/SunMarc",
"followers_url": "https://api.github.com/users/SunMarc/followers",
"following_url": "https://api.github.com/users/SunMarc/following{/other_user}",
"gists_url": "https://api.github.com/users/SunMarc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SunMarc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SunMarc/subscriptions",
"organizations_url": "https://api.github.com/users/SunMarc/orgs",
"repos_url": "https://api.github.com/users/SunMarc/repos",
"events_url": "https://api.github.com/users/SunMarc/events{/privacy}",
"received_events_url": "https://api.github.com/users/SunMarc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40296/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40296/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40295 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40295/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40295/comments | https://api.github.com/repos/huggingface/transformers/issues/40295/events | https://github.com/huggingface/transformers/pull/40295 | 3,335,583,949 | PR_kwDOCUB6oc6kXL5o | 40,295 | [`FA`] Fix dtype in varlen with position ids | {
"login": "vasqu",
"id": 73884904,
"node_id": "MDQ6VXNlcjczODg0OTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/73884904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vasqu",
"html_url": "https://github.com/vasqu",
"followers_url": "https://api.github.com/users/vasqu/followers",
"following_url": "https://api.github.com/users/vasqu/following{/other_user}",
"gists_url": "https://api.github.com/users/vasqu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vasqu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vasqu/subscriptions",
"organizations_url": "https://api.github.com/users/vasqu/orgs",
"repos_url": "https://api.github.com/users/vasqu/repos",
"events_url": "https://api.github.com/users/vasqu/events{/privacy}",
"received_events_url": "https://api.github.com/users/vasqu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T20:23:01 | 2025-08-21T18:04:10 | 2025-08-20T09:15:55 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40295",
"html_url": "https://github.com/huggingface/transformers/pull/40295",
"diff_url": "https://github.com/huggingface/transformers/pull/40295.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40295.patch",
"merged_at": "2025-08-20T09:15:55"
} | Fixes #40293
`cu_seqlens` preparation changed in #40060 and led to int64 casts.
cc @ArthurZucker @Cyrilvallez @sfc-gh-sbekman
| {
"login": "vasqu",
"id": 73884904,
"node_id": "MDQ6VXNlcjczODg0OTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/73884904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vasqu",
"html_url": "https://github.com/vasqu",
"followers_url": "https://api.github.com/users/vasqu/followers",
"following_url": "https://api.github.com/users/vasqu/following{/other_user}",
"gists_url": "https://api.github.com/users/vasqu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vasqu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vasqu/subscriptions",
"organizations_url": "https://api.github.com/users/vasqu/orgs",
"repos_url": "https://api.github.com/users/vasqu/repos",
"events_url": "https://api.github.com/users/vasqu/events{/privacy}",
"received_events_url": "https://api.github.com/users/vasqu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40295/reactions",
"total_count": 2,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 2
} | https://api.github.com/repos/huggingface/transformers/issues/40295/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40294 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40294/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40294/comments | https://api.github.com/repos/huggingface/transformers/issues/40294/events | https://github.com/huggingface/transformers/pull/40294 | 3,335,513,210 | PR_kwDOCUB6oc6kW8gt | 40,294 | Add Higgs Audio V2 Model | {
"login": "szhengac",
"id": 3960020,
"node_id": "MDQ6VXNlcjM5NjAwMjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/3960020?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/szhengac",
"html_url": "https://github.com/szhengac",
"followers_url": "https://api.github.com/users/szhengac/followers",
"following_url": "https://api.github.com/users/szhengac/following{/other_user}",
"gists_url": "https://api.github.com/users/szhengac/gists{/gist_id}",
"starred_url": "https://api.github.com/users/szhengac/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/szhengac/subscriptions",
"organizations_url": "https://api.github.com/users/szhengac/orgs",
"repos_url": "https://api.github.com/users/szhengac/repos",
"events_url": "https://api.github.com/users/szhengac/events{/privacy}",
"received_events_url": "https://api.github.com/users/szhengac/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1843244711,
"node_id": "MDU6TGFiZWwxODQzMjQ0NzEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model",
"name": "New model",
"color": "fbca04",
"default": false,
"description": ""
},
{
"id": 6470596964,
"node_id": "LA_kwDOCUB6oc8AAAABga15ZA",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Audio",
"name": "Audio",
"color": "760453",
"default": false,
"description": ""
}
] | open | false | null | [] | null | [] | 2025-08-19T19:56:16 | 2025-10-07T17:00:26 | null | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40294",
"html_url": "https://github.com/huggingface/transformers/pull/40294",
"diff_url": "https://github.com/huggingface/transformers/pull/40294.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40294.patch",
"merged_at": null
} | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "szhengac",
"id": 3960020,
"node_id": "MDQ6VXNlcjM5NjAwMjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/3960020?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/szhengac",
"html_url": "https://github.com/szhengac",
"followers_url": "https://api.github.com/users/szhengac/followers",
"following_url": "https://api.github.com/users/szhengac/following{/other_user}",
"gists_url": "https://api.github.com/users/szhengac/gists{/gist_id}",
"starred_url": "https://api.github.com/users/szhengac/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/szhengac/subscriptions",
"organizations_url": "https://api.github.com/users/szhengac/orgs",
"repos_url": "https://api.github.com/users/szhengac/repos",
"events_url": "https://api.github.com/users/szhengac/events{/privacy}",
"received_events_url": "https://api.github.com/users/szhengac/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40294/reactions",
"total_count": 4,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 3,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40294/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40293 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40293/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40293/comments | https://api.github.com/repos/huggingface/transformers/issues/40293/events | https://github.com/huggingface/transformers/issues/40293 | 3,335,405,941 | I_kwDOCUB6oc7GzkF1 | 40,293 | main seems to break the FA2 interface, ok with 4.55.2 | {
"login": "sfc-gh-sbekman",
"id": 196988264,
"node_id": "U_kgDOC73NaA",
"avatar_url": "https://avatars.githubusercontent.com/u/196988264?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sfc-gh-sbekman",
"html_url": "https://github.com/sfc-gh-sbekman",
"followers_url": "https://api.github.com/users/sfc-gh-sbekman/followers",
"following_url": "https://api.github.com/users/sfc-gh-sbekman/following{/other_user}",
"gists_url": "https://api.github.com/users/sfc-gh-sbekman/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sfc-gh-sbekman/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sfc-gh-sbekman/subscriptions",
"organizations_url": "https://api.github.com/users/sfc-gh-sbekman/orgs",
"repos_url": "https://api.github.com/users/sfc-gh-sbekman/repos",
"events_url": "https://api.github.com/users/sfc-gh-sbekman/events{/privacy}",
"received_events_url": "https://api.github.com/users/sfc-gh-sbekman/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T19:10:57 | 2025-08-20T09:15:56 | 2025-08-20T09:15:56 | CONTRIBUTOR | null | null | null | null | ### System Info
FA2 varlen forward works fine with latest release `4.55.2`, breaks with the latest `main`
torch==2.8.0+cu129
flash-attn==2.8.3
### Who can help?
@ArthurZucker
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
Just doing llama-3 varlen forward with FA2
```
[rank0]: File "/code/users/stas/github/transformers/src/transformers/models/llama/modeling_llama.py", line 395, in forward
[rank0]: hidden_states = decoder_layer(
[rank0]: ^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/modeling_layers.py", line 93, in __call__
[rank0]: return self._gradient_checkpointing_func(partial(super().__call__, **kwargs), *args)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_compile.py", line 53, in inner
[rank0]: return disable_fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_dynamo/eval_frame.py", line 929, in _fn
[rank0]: return fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/utils/checkpoint.py", line 488, in checkpoint
[rank0]: return CheckpointFunction.apply(function, preserve, *args)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/autograd/function.py", line 576, in apply
[rank0]: return super().apply(*args, **kwargs) # type: ignore[misc]
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/utils/checkpoint.py", line 262, in forward
[rank0]: outputs = run_function(*args)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
[rank0]: return self._call_impl(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1879, in _call_impl
[rank0]: return inner()
[rank0]: ^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1827, in inner
[rank0]: result = forward_call(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/utils/deprecation.py", line 172, in wrapped_func
[rank0]: return func(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/models/llama/modeling_llama.py", line 294, in forward
[rank0]: hidden_states, _ = self.self_attn(
[rank0]: ^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
[rank0]: return self._call_impl(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1879, in _call_impl
[rank0]: return inner()
[rank0]: ^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1827, in inner
[rank0]: result = forward_call(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/utils/deprecation.py", line 172, in wrapped_func
[rank0]: return func(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/models/llama/modeling_llama.py", line 252, in forward
[rank0]: attn_output, attn_weights = attention_interface(
[rank0]: ^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/integrations/flash_attention.py", line 66, in flash_attention_forward
[rank0]: attn_output = _flash_attention_forward(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/code/users/stas/github/transformers/src/transformers/modeling_flash_attention_utils.py", line 671, in _flash_attention_forward
[rank0]: out = flash_varlen_fn(
[rank0]: ^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/flash_attn/flash_attn_interface.py", line 1443, in flash_attn_varlen_func
[rank0]: return FlashAttnVarlenFunc.apply(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/autograd/function.py", line 576, in apply
[rank0]: return super().apply(*args, **kwargs) # type: ignore[misc]
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/flash_attn/flash_attn_interface.py", line 925, in forward
[rank0]: out_padded, softmax_lse, S_dmask, rng_state = _wrapped_flash_attn_varlen_forward(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_ops.py", line 1243, in __call__
[rank0]: return self._op(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_library/autograd.py", line 111, in autograd_impl
[rank0]: result = forward_no_grad(*args, Metadata(keyset, keyword_only_args))
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_library/autograd.py", line 40, in forward_no_grad
[rank0]: result = op.redispatch(keyset & _C._after_autograd_keyset, *args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_ops.py", line 836, in redispatch
[rank0]: return self._handle.redispatch_boxed(keyset, *args, **kwargs) # type: ignore[return-value]
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_library/custom_ops.py", line 344, in backend_impl
[rank0]: result = self._backend_fns[device_type](*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_compile.py", line 53, in inner
[rank0]: return disable_fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_dynamo/eval_frame.py", line 929, in _fn
[rank0]: return fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/torch/_library/custom_ops.py", line 377, in wrapped_fn
[rank0]: return fn(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^
[rank0]: File "/home/yak/miniconda3/envs/dev/lib/python3.12/site-packages/flash_attn/flash_attn_interface.py", line 165, in _flash_attn_varlen_forward
[rank0]: out, softmax_lse, S_dmask, rng_state = flash_attn_gpu.varlen_fwd(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: RuntimeError: cu_seqlens_q must have dtype int32
```
### Expected behavior
not break :) | {
"login": "vasqu",
"id": 73884904,
"node_id": "MDQ6VXNlcjczODg0OTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/73884904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vasqu",
"html_url": "https://github.com/vasqu",
"followers_url": "https://api.github.com/users/vasqu/followers",
"following_url": "https://api.github.com/users/vasqu/following{/other_user}",
"gists_url": "https://api.github.com/users/vasqu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vasqu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vasqu/subscriptions",
"organizations_url": "https://api.github.com/users/vasqu/orgs",
"repos_url": "https://api.github.com/users/vasqu/repos",
"events_url": "https://api.github.com/users/vasqu/events{/privacy}",
"received_events_url": "https://api.github.com/users/vasqu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40293/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 1
} | https://api.github.com/repos/huggingface/transformers/issues/40293/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40292 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40292/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40292/comments | https://api.github.com/repos/huggingface/transformers/issues/40292/events | https://github.com/huggingface/transformers/issues/40292 | 3,335,218,340 | I_kwDOCUB6oc7Gy2Sk | 40,292 | transformers==4.55.2 still imports TFPreTrainedModel even with TRANSFORMERS_NO_TF=1" | {
"login": "Yanshuptl18",
"id": 199244790,
"node_id": "U_kgDOC-A79g",
"avatar_url": "https://avatars.githubusercontent.com/u/199244790?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Yanshuptl18",
"html_url": "https://github.com/Yanshuptl18",
"followers_url": "https://api.github.com/users/Yanshuptl18/followers",
"following_url": "https://api.github.com/users/Yanshuptl18/following{/other_user}",
"gists_url": "https://api.github.com/users/Yanshuptl18/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Yanshuptl18/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Yanshuptl18/subscriptions",
"organizations_url": "https://api.github.com/users/Yanshuptl18/orgs",
"repos_url": "https://api.github.com/users/Yanshuptl18/repos",
"events_url": "https://api.github.com/users/Yanshuptl18/events{/privacy}",
"received_events_url": "https://api.github.com/users/Yanshuptl18/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T18:02:00 | 2025-10-08T19:39:09 | 2025-08-25T19:12:58 | NONE | null | null | null | null | ### System Info
**Describe the bug**
When using transformers==4.55.2 with TensorFlow disabled (TRANSFORMERS_NO_TF=1), the library still attempts to import TFPreTrainedModel internally, causing an ImportError even though only PyTorch is being used.
**Environment info**
-> Python version: 3.13 (Windows 11, x86_64-pc-windows-msvc)
-> Transformers version: 4.55.2
-> Accelerate version: 1.10.0
-> Torch version: 2.5.0+cpu
-> TensorFlow: not installed (intentionally, since PyTorch is used)
**Error message**
ImportError: cannot import name 'TFPreTrainedModel' from 'transformers.modeling_tf_utils'
**Additional context**
This bug prevents PyTorch-only workflows from working with Transformers 4.55.2.
TensorFlow ≥2.16.1 is also not installable on Python 3.13, which makes this issue more severe.
**Related issue :** This may be caused by a regression in recent refactoring of modeling_tf_utils imports.
**Suggested fix : ** Respect TRANSFORMERS_NO_TF=1 when importing, and avoid hard imports of TensorFlow modules if TF is disabled.
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
1. Install transformers==4.55.2 with PyTorch only (no TensorFlow).
2. Set environment variable:
export TRANSFORMERS_NO_TF=1
(or set TRANSFORMERS_NO_TF=1 on Windows)
3. Run a script that only uses PyTorch models (e.g., Trainer with BertForSequenceClassification).
### Expected behavior
The library should not attempt to import TensorFlow components (TFPreTrainedModel) when TRANSFORMERS_NO_TF=1 is set, and no TensorFlow should be required. | {
"login": "Rocketknight1",
"id": 12866554,
"node_id": "MDQ6VXNlcjEyODY2NTU0",
"avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rocketknight1",
"html_url": "https://github.com/Rocketknight1",
"followers_url": "https://api.github.com/users/Rocketknight1/followers",
"following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}",
"gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions",
"organizations_url": "https://api.github.com/users/Rocketknight1/orgs",
"repos_url": "https://api.github.com/users/Rocketknight1/repos",
"events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}",
"received_events_url": "https://api.github.com/users/Rocketknight1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40292/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40292/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40291 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40291/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40291/comments | https://api.github.com/repos/huggingface/transformers/issues/40291/events | https://github.com/huggingface/transformers/pull/40291 | 3,334,990,442 | PR_kwDOCUB6oc6kVLjq | 40,291 | Fix Command-A-Vision get_number_of_image_patches | {
"login": "dongluw",
"id": 108290936,
"node_id": "U_kgDOBnRjeA",
"avatar_url": "https://avatars.githubusercontent.com/u/108290936?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dongluw",
"html_url": "https://github.com/dongluw",
"followers_url": "https://api.github.com/users/dongluw/followers",
"following_url": "https://api.github.com/users/dongluw/following{/other_user}",
"gists_url": "https://api.github.com/users/dongluw/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dongluw/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dongluw/subscriptions",
"organizations_url": "https://api.github.com/users/dongluw/orgs",
"repos_url": "https://api.github.com/users/dongluw/repos",
"events_url": "https://api.github.com/users/dongluw/events{/privacy}",
"received_events_url": "https://api.github.com/users/dongluw/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T16:30:31 | 2025-08-20T13:58:04 | 2025-08-20T13:58:04 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40291",
"html_url": "https://github.com/huggingface/transformers/pull/40291",
"diff_url": "https://github.com/huggingface/transformers/pull/40291.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40291.patch",
"merged_at": null
} | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
The number of patches calculated is inconsistent between `get_number_of_image_patches` and `crop_image_to_patches`, this will affect the corresponding implementation in vLLM.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "dongluw",
"id": 108290936,
"node_id": "U_kgDOBnRjeA",
"avatar_url": "https://avatars.githubusercontent.com/u/108290936?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dongluw",
"html_url": "https://github.com/dongluw",
"followers_url": "https://api.github.com/users/dongluw/followers",
"following_url": "https://api.github.com/users/dongluw/following{/other_user}",
"gists_url": "https://api.github.com/users/dongluw/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dongluw/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dongluw/subscriptions",
"organizations_url": "https://api.github.com/users/dongluw/orgs",
"repos_url": "https://api.github.com/users/dongluw/repos",
"events_url": "https://api.github.com/users/dongluw/events{/privacy}",
"received_events_url": "https://api.github.com/users/dongluw/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40291/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40291/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40290 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40290/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40290/comments | https://api.github.com/repos/huggingface/transformers/issues/40290/events | https://github.com/huggingface/transformers/pull/40290 | 3,334,954,106 | PR_kwDOCUB6oc6kVD1a | 40,290 | [models] Add AudioFlamingo3 integration | {
"login": "lashahub",
"id": 26011196,
"node_id": "MDQ6VXNlcjI2MDExMTk2",
"avatar_url": "https://avatars.githubusercontent.com/u/26011196?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lashahub",
"html_url": "https://github.com/lashahub",
"followers_url": "https://api.github.com/users/lashahub/followers",
"following_url": "https://api.github.com/users/lashahub/following{/other_user}",
"gists_url": "https://api.github.com/users/lashahub/gists{/gist_id}",
"starred_url": "https://api.github.com/users/lashahub/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/lashahub/subscriptions",
"organizations_url": "https://api.github.com/users/lashahub/orgs",
"repos_url": "https://api.github.com/users/lashahub/repos",
"events_url": "https://api.github.com/users/lashahub/events{/privacy}",
"received_events_url": "https://api.github.com/users/lashahub/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1843244711,
"node_id": "MDU6TGFiZWwxODQzMjQ0NzEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model",
"name": "New model",
"color": "fbca04",
"default": false,
"description": ""
},
{
"id": 6470596964,
"node_id": "LA_kwDOCUB6oc8AAAABga15ZA",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Audio",
"name": "Audio",
"color": "760453",
"default": false,
"description": ""
}
] | open | false | null | [] | null | [] | 2025-08-19T16:17:29 | 2025-10-29T15:51:09 | null | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40290",
"html_url": "https://github.com/huggingface/transformers/pull/40290",
"diff_url": "https://github.com/huggingface/transformers/pull/40290.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40290.patch",
"merged_at": null
} | This PR adds support for **AudioFlamingo3 (AF3)** — NVIDIA’s open large audio language model capable of reasoning over speech, sounds, and music.
- **Paper**: [Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models](https://arxiv.org/abs/2507.08128)
- **Model Weights**: [nvidia/audio-flamingo-3 on Hugging Face](https://huggingface.co/nvidia/audio-flamingo-3)
- **Original Code**: [NVIDIA Audio Flamingo GitHub (branch: audio_flamingo_3)](https://github.com/NVIDIA/audio-flamingo/tree/audio_flamingo_3)
It introduces the following components:
- `AudioFlamingo3` model class
- `AudioFlamingo3Processor` for preprocessing text + audio
- Configuration, modeling, and processing utilities
- Example usage
With this integration, AF3 can be loaded directly from the Hugging Face Hub:
```python
from transformers import AudioFlamingo3Processor, AudioFlamingo3
processor = AudioFlamingo3Processor.from_pretrained("nvidia/audio-flamingo-3")
model = AudioFlamingo3.from_pretrained("nvidia/audio-flamingo-3")
prompt = "What is happening in the audio?"
audio = "clap.wav"
input_ids, media, media_meta = processor(prompt, audio)
output_ids = model.generate(
input_ids=input_ids,
media=media,
media_meta=media_meta,
generation_config=model.default_generation_config,
)
print(processor.decode(output_ids))
# Example output: "A crowd is applauding and cheering."
``` | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40290/reactions",
"total_count": 4,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 4,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40290/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40289 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40289/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40289/comments | https://api.github.com/repos/huggingface/transformers/issues/40289/events | https://github.com/huggingface/transformers/pull/40289 | 3,334,858,913 | PR_kwDOCUB6oc6kUvjK | 40,289 | [docs] delete more TF/Flax docs | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T15:45:30 | 2025-08-20T10:19:58 | 2025-08-20T09:44:14 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40289",
"html_url": "https://github.com/huggingface/transformers/pull/40289",
"diff_url": "https://github.com/huggingface/transformers/pull/40289.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40289.patch",
"merged_at": "2025-08-20T09:44:14"
} | # What does this PR do?
Deletes a few more traces of TF/Flax from our docs 👼 This PR required modifying the repo consistency script, so that it is okay to have undocumented TF/Flax classes.
major folders to be touched in follow-up PRs:
- `model_docs` (many TF/Flax classes documented here)
- `tasks` (several TF examples) | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40289/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40289/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40288 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40288/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40288/comments | https://api.github.com/repos/huggingface/transformers/issues/40288/events | https://github.com/huggingface/transformers/pull/40288 | 3,334,810,032 | PR_kwDOCUB6oc6kUlKo | 40,288 | [`GPT OSS`] Refactor the tests as it was not properly checking the outputs | {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T15:29:06 | 2025-08-20T14:47:43 | 2025-08-20T14:47:41 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40288",
"html_url": "https://github.com/huggingface/transformers/pull/40288",
"diff_url": "https://github.com/huggingface/transformers/pull/40288.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40288.patch",
"merged_at": "2025-08-20T14:47:41"
} | * Changed the format of `integration_tests.json` from a list of dictionaries to a dictionary keyed by configuration strings, making lookups and validation more direct and maintainable.
| {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40288/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40288/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40287 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40287/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40287/comments | https://api.github.com/repos/huggingface/transformers/issues/40287/events | https://github.com/huggingface/transformers/pull/40287 | 3,334,711,055 | PR_kwDOCUB6oc6kUP-q | 40,287 | No more `natten` | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T14:58:30 | 2025-08-20T14:10:18 | 2025-08-20T14:10:16 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40287",
"html_url": "https://github.com/huggingface/transformers/pull/40287",
"diff_url": "https://github.com/huggingface/transformers/pull/40287.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40287.patch",
"merged_at": "2025-08-20T14:10:16"
} | # What does this PR do?
This is required by `nat` (already depreciated) and `dinat`.
We have to pin it to a very old version to avoid `natten2dav` issue, but now we get
> ImportError: cannot import name '_device_t' from 'torch.cuda' (/usr/local/lib/python3.10/dist-packages/torch/cuda/__init__.py)
with torch 2.8, which fails the job of `tests/utils` where `from transformers import *` is used.
Let's simply remove this natten from docker CI image (and no tests for `dinat` for which I think reasonable). | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40287/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40287/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40286 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40286/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40286/comments | https://api.github.com/repos/huggingface/transformers/issues/40286/events | https://github.com/huggingface/transformers/pull/40286 | 3,334,697,963 | PR_kwDOCUB6oc6kUNKY | 40,286 | Add MOSS-TTSD with XY-Tokenizer | {
"login": "gaoyang07",
"id": 32762712,
"node_id": "MDQ6VXNlcjMyNzYyNzEy",
"avatar_url": "https://avatars.githubusercontent.com/u/32762712?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gaoyang07",
"html_url": "https://github.com/gaoyang07",
"followers_url": "https://api.github.com/users/gaoyang07/followers",
"following_url": "https://api.github.com/users/gaoyang07/following{/other_user}",
"gists_url": "https://api.github.com/users/gaoyang07/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gaoyang07/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gaoyang07/subscriptions",
"organizations_url": "https://api.github.com/users/gaoyang07/orgs",
"repos_url": "https://api.github.com/users/gaoyang07/repos",
"events_url": "https://api.github.com/users/gaoyang07/events{/privacy}",
"received_events_url": "https://api.github.com/users/gaoyang07/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 6470596964,
"node_id": "LA_kwDOCUB6oc8AAAABga15ZA",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Audio",
"name": "Audio",
"color": "760453",
"default": false,
"description": ""
}
] | open | false | null | [] | null | [] | 2025-08-19T14:54:30 | 2025-09-12T08:47:55 | null | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40286",
"html_url": "https://github.com/huggingface/transformers/pull/40286",
"diff_url": "https://github.com/huggingface/transformers/pull/40286.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40286.patch",
"merged_at": null
} | # What does this PR do?
## MOSS-TTSD 🪐
[[Files and versions]](https://huggingface.co/fnlp/MOSS-TTSD-v0.5/tree/main) [[Community]](https://huggingface.co/fnlp/MOSS-TTSD-v0.5/discussions)
### Overview
MOSS-TTSD (text to spoken dialogue) is an open-source bilingual spoken dialogue synthesis model that supports both Chinese and English. It can transform dialogue scripts between two speakers into natural, expressive conversational speech. MOSS-TTSD supports voice cloning and single-session speech generation of up to 960 seconds, making it ideal for AI podcast production.
#### Highlights
- Highly Expressive Dialogue Speech: Built on unified semantic-acoustic neural audio codec, a pre-trained large language model, millions of hours of TTS data, and 400k hours synthetic and real conversational speech, MOSS-TTSD generates highly expressive, human-like dialogue speech with natural conversational prosody.
- Two-Speaker Voice Cloning: MOSS-TTSD supports zero-shot two speakers voice cloning and can generate conversational speech with accurate speaker swithcing based on dialogue scripts.
- Chinese-English Bilingual Support: MOSS-TTSD enables highly expressive speech generation in both Chinese and English.
- Long-Form Speech Generation (up to 960 seconds): Thanks to low-bitrate codec and training framework optimization, MOSS-TTSD has been trained for long speech generation, enabling single-session speech generation of up to 960 seconds.
- Fully Open Source & Commercial-Ready: MOSS-TTSD and its future updates will be fully open-source and support free commercial use.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40286/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40286/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40285 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40285/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40285/comments | https://api.github.com/repos/huggingface/transformers/issues/40285/events | https://github.com/huggingface/transformers/pull/40285 | 3,334,669,261 | PR_kwDOCUB6oc6kUG9m | 40,285 | Fix attention vizualizer | {
"login": "molbap",
"id": 39954772,
"node_id": "MDQ6VXNlcjM5OTU0Nzcy",
"avatar_url": "https://avatars.githubusercontent.com/u/39954772?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/molbap",
"html_url": "https://github.com/molbap",
"followers_url": "https://api.github.com/users/molbap/followers",
"following_url": "https://api.github.com/users/molbap/following{/other_user}",
"gists_url": "https://api.github.com/users/molbap/gists{/gist_id}",
"starred_url": "https://api.github.com/users/molbap/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/molbap/subscriptions",
"organizations_url": "https://api.github.com/users/molbap/orgs",
"repos_url": "https://api.github.com/users/molbap/repos",
"events_url": "https://api.github.com/users/molbap/events{/privacy}",
"received_events_url": "https://api.github.com/users/molbap/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T14:45:55 | 2025-08-21T13:13:36 | 2025-08-21T13:13:35 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40285",
"html_url": "https://github.com/huggingface/transformers/pull/40285",
"diff_url": "https://github.com/huggingface/transformers/pull/40285.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40285.patch",
"merged_at": "2025-08-21T13:13:35"
} | # What does this PR do?
I noticed a few models were broken with the visualizer, because it relied on `_update_causal_mask`, and we changed the interface quite a bit. I tested it, added a test for PaliGemma and Llama3.2 as well.
Also added a test to make the tool a tad more durable! | {
"login": "molbap",
"id": 39954772,
"node_id": "MDQ6VXNlcjM5OTU0Nzcy",
"avatar_url": "https://avatars.githubusercontent.com/u/39954772?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/molbap",
"html_url": "https://github.com/molbap",
"followers_url": "https://api.github.com/users/molbap/followers",
"following_url": "https://api.github.com/users/molbap/following{/other_user}",
"gists_url": "https://api.github.com/users/molbap/gists{/gist_id}",
"starred_url": "https://api.github.com/users/molbap/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/molbap/subscriptions",
"organizations_url": "https://api.github.com/users/molbap/orgs",
"repos_url": "https://api.github.com/users/molbap/repos",
"events_url": "https://api.github.com/users/molbap/events{/privacy}",
"received_events_url": "https://api.github.com/users/molbap/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40285/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40285/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40284 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40284/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40284/comments | https://api.github.com/repos/huggingface/transformers/issues/40284/events | https://github.com/huggingface/transformers/pull/40284 | 3,334,581,112 | PR_kwDOCUB6oc6kTz9f | 40,284 | Update `test_spm_converter_bytefallback_warning` | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T14:19:33 | 2025-08-21T12:09:29 | 2025-08-21T12:09:28 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40284",
"html_url": "https://github.com/huggingface/transformers/pull/40284",
"diff_url": "https://github.com/huggingface/transformers/pull/40284.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40284.patch",
"merged_at": "2025-08-21T12:09:28"
} | # What does this PR do?
not sure why, but now we get
```
{message : DeprecationWarning('builtin type SwigPyObject has no __module__ attribute'), category : 'DeprecationWarning', filename : '<frozen importlib._bootstrap>', lineno : 241, line : None}
```
which is unrelated to the test goal | {
"login": "ydshieh",
"id": 2521628,
"node_id": "MDQ6VXNlcjI1MjE2Mjg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ydshieh",
"html_url": "https://github.com/ydshieh",
"followers_url": "https://api.github.com/users/ydshieh/followers",
"following_url": "https://api.github.com/users/ydshieh/following{/other_user}",
"gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions",
"organizations_url": "https://api.github.com/users/ydshieh/orgs",
"repos_url": "https://api.github.com/users/ydshieh/repos",
"events_url": "https://api.github.com/users/ydshieh/events{/privacy}",
"received_events_url": "https://api.github.com/users/ydshieh/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40284/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40284/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40283 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40283/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40283/comments | https://api.github.com/repos/huggingface/transformers/issues/40283/events | https://github.com/huggingface/transformers/pull/40283 | 3,334,552,595 | PR_kwDOCUB6oc6kTt8- | 40,283 | fix which routing method | {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T14:12:18 | 2025-08-19T14:38:18 | 2025-08-19T14:35:14 | COLLABORATOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40283",
"html_url": "https://github.com/huggingface/transformers/pull/40283",
"diff_url": "https://github.com/huggingface/transformers/pull/40283.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40283.patch",
"merged_at": "2025-08-19T14:35:14"
} | # What does this PR do?
Fixes #40031! Tested on both TP and DP
Run output:
```python
arthur transformers fix-dp-gpt-oss ≢ ~1 cd /fsx/arthur/transformers ; /usr/bin/env /fsx/arthur/.venv/bin/python /admin/home/arthur/.vscode-server/extensions/ms-python.debugpy-2025.10.0-linux-x64/bundled/libs/debugpy/adapter/../../debugpy/launcher 40835 -- -m torch.distributed.run --nproc_per_node=4 /fsx/arthur/test_dequantize.py
*****************************************
Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
*****************************************
Loading checkpoint shards: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:05<00:00, 1.64it/s]
Loading checkpoint shards: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:05<00:00, 1.53it/s]
['Once upon a time in a kingdom faraway, there was a beautiful prince named Rohan. He had a magical garden']
Loading checkpoint shards: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:06<00:00, 1.50it/s]
Loading checkpoint shards: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:05<00:00, 1.58it/s]
['Once upon a time, in a kingdom far away\n\n**Title: "The Last King"**\n\n**Title** \n']
['Once upon a time in the bustling city of Neo-Tokyo, a city that was the epitome where high-tech']
['Once upon a time, a kind and helpful friend named Yoonhee became "Sangha," the friend of all']
``` | {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40283/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40283/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40282 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40282/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40282/comments | https://api.github.com/repos/huggingface/transformers/issues/40282/events | https://github.com/huggingface/transformers/pull/40282 | 3,334,525,303 | PR_kwDOCUB6oc6kToGs | 40,282 | Remove deprecated functions | {
"login": "cyyever",
"id": 17618148,
"node_id": "MDQ6VXNlcjE3NjE4MTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/17618148?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cyyever",
"html_url": "https://github.com/cyyever",
"followers_url": "https://api.github.com/users/cyyever/followers",
"following_url": "https://api.github.com/users/cyyever/following{/other_user}",
"gists_url": "https://api.github.com/users/cyyever/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cyyever/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cyyever/subscriptions",
"organizations_url": "https://api.github.com/users/cyyever/orgs",
"repos_url": "https://api.github.com/users/cyyever/repos",
"events_url": "https://api.github.com/users/cyyever/events{/privacy}",
"received_events_url": "https://api.github.com/users/cyyever/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T14:04:44 | 2025-08-31T14:49:46 | 2025-08-31T14:49:42 | CONTRIBUTOR | null | null | true | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40282",
"html_url": "https://github.com/huggingface/transformers/pull/40282",
"diff_url": "https://github.com/huggingface/transformers/pull/40282.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40282.patch",
"merged_at": null
} | # What does this PR do?
The test utility functions for FT&JAX have been deprecated and not used. | {
"login": "cyyever",
"id": 17618148,
"node_id": "MDQ6VXNlcjE3NjE4MTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/17618148?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cyyever",
"html_url": "https://github.com/cyyever",
"followers_url": "https://api.github.com/users/cyyever/followers",
"following_url": "https://api.github.com/users/cyyever/following{/other_user}",
"gists_url": "https://api.github.com/users/cyyever/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cyyever/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cyyever/subscriptions",
"organizations_url": "https://api.github.com/users/cyyever/orgs",
"repos_url": "https://api.github.com/users/cyyever/repos",
"events_url": "https://api.github.com/users/cyyever/events{/privacy}",
"received_events_url": "https://api.github.com/users/cyyever/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40282/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40282/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40281 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40281/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40281/comments | https://api.github.com/repos/huggingface/transformers/issues/40281/events | https://github.com/huggingface/transformers/pull/40281 | 3,334,460,631 | PR_kwDOCUB6oc6kTaH2 | 40,281 | Fix AttributeError in AwqQuantizer when version attribute is missing | {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T13:46:48 | 2025-08-20T12:59:19 | 2025-08-20T12:59:19 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40281",
"html_url": "https://github.com/huggingface/transformers/pull/40281",
"diff_url": "https://github.com/huggingface/transformers/pull/40281.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40281.patch",
"merged_at": null
} | ## What does this PR do?
This PR adds an explicit version check for **AutoAWQ** when using `modules_to_not_convert` (module quantization skipping).
- Module skipping requires functionality added in **AutoAWQ 0.1.8**.
- Older versions or missing installs previously caused confusing runtime errors.
- This update ensures clear, actionable errors and avoids unnecessary checks when skipping is not requested.
---
## Issue link
Fixes #39798
---
## Motivation
**Before:**
Using `modules_to_not_convert` with old or missing AutoAWQ led to runtime failures and unclear error messages.
**After:**
- If AutoAWQ <0.1.8 → users see:
*"Your current version of autoawq (<version>) does not support module quantization skipping. Please upgrade autoawq to >=0.1.8."*
- If AutoAWQ is not installed → users see:
*"AutoAWQ >= 0.1.8 is required for module quantization skipping. Install with: `pip install autoawq>=0.1.8`"*
- If not using `modules_to_not_convert` → no check is triggered, no behavioral changes.
---
## Changes
### `quantizers/quantizer_awq.py`
- Added `_check_autoawq_version_for_module_skipping()` helper
- Detects AutoAWQ version (`__version__` or `version`)
- Raises clear ImportError if version <0.1.8 or AutoAWQ not installed
- Invoked only when `modules_to_not_convert` is non-empty
- Preserves CUDA/XPU/IPEX logic
### `tests/test_quantizer_awq.py`
- `test_old_autoawq_raises`: AutoAWQ 0.1.7 → raises upgrade error
- `test_new_autoawq_passes`: AutoAWQ >=0.1.8 → passes
- `test_autoawq_not_installed`: Raises install guidance
- `test_no_modules_to_not_convert_no_check`: Skips check when `None`
- `test_empty_modules_list_no_check`: Skips check when `[]`
---
## Backward compatibility
- ✅ No breaking changes
- ✅ Workflows without module skipping unaffected
- ✅ Preserves CUDA/XPU/IPEX gating
---
## Testing
- Added 5 new unit tests covering:
1. Old AutoAWQ (<0.1.8)
2. New AutoAWQ (>=0.1.8)
3. AutoAWQ not installed
4. `modules_to_not_convert=None`
5. `modules_to_not_convert=[]`
- All tests pass locally
---
## Checklist
- [x] Fixes #39798
- [x] Improves error messages and robustness
- [x] Adds new tests
- [x] No BC breaks
- [x] Minimal scope (two files only)
---
⚡ Maintainers: This is a focused improvement gated behind AutoAWQ>=0.1.8.
Happy to adjust helper naming or placement if preferred.
| {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40281/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40281/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40280 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40280/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40280/comments | https://api.github.com/repos/huggingface/transformers/issues/40280/events | https://github.com/huggingface/transformers/pull/40280 | 3,334,318,324 | PR_kwDOCUB6oc6kS7Mj | 40,280 | Fix: Apply `get_placeholder_mask` in Ovis2 | {
"login": "thisisiron",
"id": 23303033,
"node_id": "MDQ6VXNlcjIzMzAzMDMz",
"avatar_url": "https://avatars.githubusercontent.com/u/23303033?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thisisiron",
"html_url": "https://github.com/thisisiron",
"followers_url": "https://api.github.com/users/thisisiron/followers",
"following_url": "https://api.github.com/users/thisisiron/following{/other_user}",
"gists_url": "https://api.github.com/users/thisisiron/gists{/gist_id}",
"starred_url": "https://api.github.com/users/thisisiron/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/thisisiron/subscriptions",
"organizations_url": "https://api.github.com/users/thisisiron/orgs",
"repos_url": "https://api.github.com/users/thisisiron/repos",
"events_url": "https://api.github.com/users/thisisiron/events{/privacy}",
"received_events_url": "https://api.github.com/users/thisisiron/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T13:06:21 | 2025-08-20T15:12:11 | 2025-08-20T15:12:11 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40280",
"html_url": "https://github.com/huggingface/transformers/pull/40280",
"diff_url": "https://github.com/huggingface/transformers/pull/40280.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40280.patch",
"merged_at": "2025-08-20T15:12:11"
} | # What does this PR do?
The `get_placeholder_mask` function is currently unused. I’ve updated the code so that this function is now properly utilized.
## Who can review?
@Cyrilvallez
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40280/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40280/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40279 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40279/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40279/comments | https://api.github.com/repos/huggingface/transformers/issues/40279/events | https://github.com/huggingface/transformers/pull/40279 | 3,334,201,997 | PR_kwDOCUB6oc6kShyW | 40,279 | Adjust ROCm test output expectations | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T12:30:20 | 2025-08-19T19:21:45 | 2025-08-19T19:21:45 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40279",
"html_url": "https://github.com/huggingface/transformers/pull/40279",
"diff_url": "https://github.com/huggingface/transformers/pull/40279.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40279.patch",
"merged_at": "2025-08-19T19:21:45"
} | # What does this PR do?
This PR corrects the expected output for ROCm devices.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40279/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40279/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40278 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40278/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40278/comments | https://api.github.com/repos/huggingface/transformers/issues/40278/events | https://github.com/huggingface/transformers/issues/40278 | 3,334,069,222 | I_kwDOCUB6oc7Gudvm | 40,278 | [BUG] GPT-OSS weight dequantization discrepancy for online and offline methods | {
"login": "TheTinyTeddy",
"id": 171109504,
"node_id": "U_kgDOCjLsgA",
"avatar_url": "https://avatars.githubusercontent.com/u/171109504?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/TheTinyTeddy",
"html_url": "https://github.com/TheTinyTeddy",
"followers_url": "https://api.github.com/users/TheTinyTeddy/followers",
"following_url": "https://api.github.com/users/TheTinyTeddy/following{/other_user}",
"gists_url": "https://api.github.com/users/TheTinyTeddy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/TheTinyTeddy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/TheTinyTeddy/subscriptions",
"organizations_url": "https://api.github.com/users/TheTinyTeddy/orgs",
"repos_url": "https://api.github.com/users/TheTinyTeddy/repos",
"events_url": "https://api.github.com/users/TheTinyTeddy/events{/privacy}",
"received_events_url": "https://api.github.com/users/TheTinyTeddy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T11:47:29 | 2025-08-28T06:39:15 | 2025-08-28T06:39:15 | NONE | null | null | null | null | For offline dequantization using quantization_config=Mxfp4Config(dequantize=True) when converting, the dequantization function is `convert_moe_packed_tensors`.
For online dequantization using triton kernel, the weight is dequantized on-the-fly in the `triton _matmul_ogs` function.
Theoretically, they should produce the same output since they both dequantize the weight from mxfp4 to bf16, but I've tested the actual outputs are quite different, so maybe implementation issue in either function? | {
"login": "TheTinyTeddy",
"id": 171109504,
"node_id": "U_kgDOCjLsgA",
"avatar_url": "https://avatars.githubusercontent.com/u/171109504?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/TheTinyTeddy",
"html_url": "https://github.com/TheTinyTeddy",
"followers_url": "https://api.github.com/users/TheTinyTeddy/followers",
"following_url": "https://api.github.com/users/TheTinyTeddy/following{/other_user}",
"gists_url": "https://api.github.com/users/TheTinyTeddy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/TheTinyTeddy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/TheTinyTeddy/subscriptions",
"organizations_url": "https://api.github.com/users/TheTinyTeddy/orgs",
"repos_url": "https://api.github.com/users/TheTinyTeddy/repos",
"events_url": "https://api.github.com/users/TheTinyTeddy/events{/privacy}",
"received_events_url": "https://api.github.com/users/TheTinyTeddy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40278/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40278/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40277 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40277/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40277/comments | https://api.github.com/repos/huggingface/transformers/issues/40277/events | https://github.com/huggingface/transformers/pull/40277 | 3,334,016,607 | PR_kwDOCUB6oc6kR5QG | 40,277 | [generate] handle support for cache classes when num enc layers != num dec layers | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T11:29:55 | 2025-08-21T16:35:21 | 2025-08-21T16:35:18 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40277",
"html_url": "https://github.com/huggingface/transformers/pull/40277",
"diff_url": "https://github.com/huggingface/transformers/pull/40277.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40277.patch",
"merged_at": "2025-08-21T16:35:18"
} | # What does this PR do?
I think this bug has been in since `EncoderDecoderCache` was added: we naively assumed that # encoder layers == # decoder layers, as we used `config.num_hidden_layers` (# encoder layers in encoder-decoder models) to instantiate both caches 👀
This PR adds the missing logic:
- we can specify which part of the config we want to pull with `get_text_config()` (previously we could only isolate the decoder). We also update `num_hidden_layers` accordingly, so that the caches see the right number of layers.
- we pull the right config (encoder vs decoder) before parameterizing `EncoderDecoderCache` in `generate`
- (adds tests)
Fixes #40120
_________________________
Example of failing checkpoint, taken from #40120 (needs num encoder layers != num decoder layers)
```py
import torch
from transformers import AutoTokenizer, pipeline
torch_dtype = torch.float16
device_map = "cpu"
model_kwargs = dict(torch_dtype=torch_dtype, device_map=device_map)
model_id = "sshleifer/distilbart-cnn-12-6"
tokenizer = AutoTokenizer.from_pretrained(model_id)
generator = pipeline(
"summarization",
model=model_id,
tokenizer=tokenizer,
model_kwargs=model_kwargs,
)
generation_config = generator.model.generation_config
generation_config.do_sample = True
generation_config.use_cache = True
generation_config.temperature = 1.0
generation_config.num_beams = 1
generation_config.max_new_tokens = 100
generation_config.min_new_tokens = 100
generation_config.top_p = 1.0
generation_config.cache_implementation="static"
prompt = "I like math"
output = generator(prompt , batch_size=1, generation_config=generation_config) # the cache will have an incorrect number of layers on `main`, but it runs
output = generator(prompt , batch_size=1, generation_config=generation_config) # crashes on `main`
``` | {
"login": "gante",
"id": 12240844,
"node_id": "MDQ6VXNlcjEyMjQwODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gante",
"html_url": "https://github.com/gante",
"followers_url": "https://api.github.com/users/gante/followers",
"following_url": "https://api.github.com/users/gante/following{/other_user}",
"gists_url": "https://api.github.com/users/gante/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gante/subscriptions",
"organizations_url": "https://api.github.com/users/gante/orgs",
"repos_url": "https://api.github.com/users/gante/repos",
"events_url": "https://api.github.com/users/gante/events{/privacy}",
"received_events_url": "https://api.github.com/users/gante/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40277/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40277/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40276 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40276/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40276/comments | https://api.github.com/repos/huggingface/transformers/issues/40276/events | https://github.com/huggingface/transformers/pull/40276 | 3,333,923,906 | PR_kwDOCUB6oc6kRktL | 40,276 | One cache class to rule them all | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T11:01:36 | 2025-08-20T17:36:13 | 2025-08-20T17:36:11 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40276",
"html_url": "https://github.com/huggingface/transformers/pull/40276",
"diff_url": "https://github.com/huggingface/transformers/pull/40276.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40276.patch",
"merged_at": "2025-08-20T17:36:11"
} | # What does this PR do?
Now that we moved the cache logic to a per-layer approach and added DynamicSliding layer, duplicating the number of Cache classes is kind of useless, and brings a lot of unnecessary complexity.
This PR deprecates all Cache classes to keep only 3:
- DynamicCache -> each layer type is inferred from config and correctly dispatched
- StaticCache -> each layer type is inferred from config and correctly dispatched
- QuantizedCache -> only has 1 layer type anyway
This simplifies quite a lot as we had created a lot of classes before, which are always the same up to the layer type.
Also, offloading is now ALWAYS possible, independently of the layer combination, which was not the case before: `OffloadedCache` would work with only full layers, and we had 2 different classes for `StaticOffloaded`, depending on hybrid structure or only full layer, but fully sliding models such as Mistral could not be offloaded with static implementation.
Finally, `cache_implementation="static"` will correctly instantiate the correct layer types, independently of the model, which is not the case currently (it would use only full layers all the time, even if model uses sliding layers, wasting memory) | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40276/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40276/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40275 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40275/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40275/comments | https://api.github.com/repos/huggingface/transformers/issues/40275/events | https://github.com/huggingface/transformers/pull/40275 | 3,333,907,863 | PR_kwDOCUB6oc6kRhMi | 40,275 | SmolVLM test fixes | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T10:56:36 | 2025-08-19T19:22:06 | 2025-08-19T19:22:06 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40275",
"html_url": "https://github.com/huggingface/transformers/pull/40275",
"diff_url": "https://github.com/huggingface/transformers/pull/40275.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40275.patch",
"merged_at": "2025-08-19T19:22:06"
} | # What does this PR do?
This PR makes the CI for SmolVLM again: fixes the export test + adjusts the expected values for AMD.
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40275/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40275/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40274 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40274/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40274/comments | https://api.github.com/repos/huggingface/transformers/issues/40274/events | https://github.com/huggingface/transformers/pull/40274 | 3,333,866,325 | PR_kwDOCUB6oc6kRYCb | 40,274 | Fix AttributeError in AwqQuantizer when version attribute is missing | {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T10:43:30 | 2025-08-19T12:41:25 | 2025-08-19T12:41:25 | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40274",
"html_url": "https://github.com/huggingface/transformers/pull/40274",
"diff_url": "https://github.com/huggingface/transformers/pull/40274.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40274.patch",
"merged_at": null
} | ## What does this PR do?
This PR adds an explicit version check for **AutoAWQ** when using `modules_to_not_convert` (module quantization skipping).
- Module skipping requires functionality added in **AutoAWQ 0.1.8**.
- Older versions or missing installs previously caused confusing runtime errors.
- This update ensures clear, actionable errors and avoids unnecessary checks when skipping is not requested.
---
## Issue link
Fixes #39798
---
## Motivation / Rationale
**Before:**
Using `modules_to_not_convert` with old or missing AutoAWQ led to runtime failures and unclear error messages.
**After:**
- If AutoAWQ <0.1.8 → users see:
*"Your current version of autoawq (<version>) does not support module quantization skipping. Please upgrade autoawq to >=0.1.8."*
- If AutoAWQ is not installed → users see:
*"AutoAWQ >= 0.1.8 is required for module quantization skipping. Install with: `pip install autoawq>=0.1.8`"*
- If not using `modules_to_not_convert` → no check is triggered, no behavioral changes.
---
## Changes
### `quantizers/quantizer_awq.py`
- Added `_check_autoawq_version_for_module_skipping()` helper
- Detects AutoAWQ version (`__version__` or `version`)
- Raises clear ImportError if version <0.1.8 or AutoAWQ not installed
- Invoked only when `modules_to_not_convert` is non-empty
- Preserves CUDA/XPU/IPEX logic
### `tests/test_quantizer_awq.py`
- `test_old_autoawq_raises`: AutoAWQ 0.1.7 → raises upgrade error
- `test_new_autoawq_passes`: AutoAWQ >=0.1.8 → passes
- `test_autoawq_not_installed`: Raises install guidance
- `test_no_modules_to_not_convert_no_check`: Skips check when `None`
- `test_empty_modules_list_no_check`: Skips check when `[]`
---
## Backward compatibility
- ✅ No breaking changes
- ✅ Workflows without module skipping unaffected
- ✅ Preserves CUDA/XPU/IPEX gating
---
## Testing
- Added 5 new unit tests covering:
1. Old AutoAWQ (<0.1.8)
2. New AutoAWQ (>=0.1.8)
3. AutoAWQ not installed
4. `modules_to_not_convert=None`
5. `modules_to_not_convert=[]`
- All tests pass locally
---
## Checklist
- [x] Fixes #39798
- [x] Improves error messages and robustness
- [x] Adds new tests
- [x] No BC breaks
- [x] Minimal scope (two files only)
---
⚡ Maintainers: This is a focused improvement gated behind AutoAWQ>=0.1.8.
Happy to adjust helper naming or placement if preferred.
| {
"login": "SnehaDeshmukh28",
"id": 91473383,
"node_id": "MDQ6VXNlcjkxNDczMzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/91473383?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SnehaDeshmukh28",
"html_url": "https://github.com/SnehaDeshmukh28",
"followers_url": "https://api.github.com/users/SnehaDeshmukh28/followers",
"following_url": "https://api.github.com/users/SnehaDeshmukh28/following{/other_user}",
"gists_url": "https://api.github.com/users/SnehaDeshmukh28/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SnehaDeshmukh28/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SnehaDeshmukh28/subscriptions",
"organizations_url": "https://api.github.com/users/SnehaDeshmukh28/orgs",
"repos_url": "https://api.github.com/users/SnehaDeshmukh28/repos",
"events_url": "https://api.github.com/users/SnehaDeshmukh28/events{/privacy}",
"received_events_url": "https://api.github.com/users/SnehaDeshmukh28/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40274/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40274/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40273 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40273/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40273/comments | https://api.github.com/repos/huggingface/transformers/issues/40273/events | https://github.com/huggingface/transformers/pull/40273 | 3,333,623,715 | PR_kwDOCUB6oc6kQknC | 40,273 | docs(layoutlm): add missing `id=usage` to `<hfoptions>` tag in LayoutLM model card | {
"login": "Jin-HoMLee",
"id": 123657753,
"node_id": "U_kgDOB17eGQ",
"avatar_url": "https://avatars.githubusercontent.com/u/123657753?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Jin-HoMLee",
"html_url": "https://github.com/Jin-HoMLee",
"followers_url": "https://api.github.com/users/Jin-HoMLee/followers",
"following_url": "https://api.github.com/users/Jin-HoMLee/following{/other_user}",
"gists_url": "https://api.github.com/users/Jin-HoMLee/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Jin-HoMLee/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Jin-HoMLee/subscriptions",
"organizations_url": "https://api.github.com/users/Jin-HoMLee/orgs",
"repos_url": "https://api.github.com/users/Jin-HoMLee/repos",
"events_url": "https://api.github.com/users/Jin-HoMLee/events{/privacy}",
"received_events_url": "https://api.github.com/users/Jin-HoMLee/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T09:27:22 | 2025-08-19T16:14:44 | 2025-08-19T16:14:44 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40273",
"html_url": "https://github.com/huggingface/transformers/pull/40273",
"diff_url": "https://github.com/huggingface/transformers/pull/40273.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40273.patch",
"merged_at": "2025-08-19T16:14:44"
} | LayoutLM model card
# Add missing `id=usage` to `<hfoptions>` tag
This PR fixes a documentation issue in the LayoutLM model card by adding the missing `id=usage` option to the `<hfoptions>` tag.
This ensures correct rendering via Hugging Face doc-builder.
No code changes.
No dependencies.
Related to #36979 and discussed in https://github.com/huggingface/transformers/pull/40129#issuecomment-3197541436
## Before submitting
- [x] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [x] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [x] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests? (not applicable)
## Who can review?
Anyone in the community is free to review the PR once the tests have passed.
For documentation fixes: @stevhliu
| {
"login": "stevhliu",
"id": 59462357,
"node_id": "MDQ6VXNlcjU5NDYyMzU3",
"avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevhliu",
"html_url": "https://github.com/stevhliu",
"followers_url": "https://api.github.com/users/stevhliu/followers",
"following_url": "https://api.github.com/users/stevhliu/following{/other_user}",
"gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions",
"organizations_url": "https://api.github.com/users/stevhliu/orgs",
"repos_url": "https://api.github.com/users/stevhliu/repos",
"events_url": "https://api.github.com/users/stevhliu/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevhliu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40273/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40273/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40272 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40272/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40272/comments | https://api.github.com/repos/huggingface/transformers/issues/40272/events | https://github.com/huggingface/transformers/pull/40272 | 3,333,568,074 | PR_kwDOCUB6oc6kQY2- | 40,272 | Addiing ByteDance Seed Seed-OSS | {
"login": "Fazziekey",
"id": 55798671,
"node_id": "MDQ6VXNlcjU1Nzk4Njcx",
"avatar_url": "https://avatars.githubusercontent.com/u/55798671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Fazziekey",
"html_url": "https://github.com/Fazziekey",
"followers_url": "https://api.github.com/users/Fazziekey/followers",
"following_url": "https://api.github.com/users/Fazziekey/following{/other_user}",
"gists_url": "https://api.github.com/users/Fazziekey/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Fazziekey/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Fazziekey/subscriptions",
"organizations_url": "https://api.github.com/users/Fazziekey/orgs",
"repos_url": "https://api.github.com/users/Fazziekey/repos",
"events_url": "https://api.github.com/users/Fazziekey/events{/privacy}",
"received_events_url": "https://api.github.com/users/Fazziekey/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1843244711,
"node_id": "MDU6TGFiZWwxODQzMjQ0NzEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model",
"name": "New model",
"color": "fbca04",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2025-08-19T09:10:58 | 2025-08-22T12:54:34 | 2025-08-22T12:54:28 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40272",
"html_url": "https://github.com/huggingface/transformers/pull/40272",
"diff_url": "https://github.com/huggingface/transformers/pull/40272.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40272.patch",
"merged_at": "2025-08-22T12:54:28"
} | # What does this PR do?
This PR adds the support of codes for the coming Seed Oss models
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40272/reactions",
"total_count": 14,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 12,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 2
} | https://api.github.com/repos/huggingface/transformers/issues/40272/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40271 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40271/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40271/comments | https://api.github.com/repos/huggingface/transformers/issues/40271/events | https://github.com/huggingface/transformers/pull/40271 | 3,333,429,404 | PR_kwDOCUB6oc6kP79k | 40,271 | Clean up xcodec addition. | {
"login": "ebezzam",
"id": 4757445,
"node_id": "MDQ6VXNlcjQ3NTc0NDU=",
"avatar_url": "https://avatars.githubusercontent.com/u/4757445?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ebezzam",
"html_url": "https://github.com/ebezzam",
"followers_url": "https://api.github.com/users/ebezzam/followers",
"following_url": "https://api.github.com/users/ebezzam/following{/other_user}",
"gists_url": "https://api.github.com/users/ebezzam/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ebezzam/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ebezzam/subscriptions",
"organizations_url": "https://api.github.com/users/ebezzam/orgs",
"repos_url": "https://api.github.com/users/ebezzam/repos",
"events_url": "https://api.github.com/users/ebezzam/events{/privacy}",
"received_events_url": "https://api.github.com/users/ebezzam/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T08:26:35 | 2025-08-20T10:16:34 | 2025-08-20T10:16:28 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40271",
"html_url": "https://github.com/huggingface/transformers/pull/40271",
"diff_url": "https://github.com/huggingface/transformers/pull/40271.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40271.patch",
"merged_at": "2025-08-20T10:16:28"
} | # What does this PR do?
Clean up recent model addition for X-Codec #38248
- Remove unnecessary inputs to config (many are actually derived from other parameters
- Make more consistent with other codecs (DAC and Encodec), as discussed in [X-Codec2 PR](#37868) with @eustlb
- Add all codec models (5) from original authors (see [here](https://github.com/zhenye234/xcodec?tab=readme-ov-file#available-models), previously just speech codec was added). All model cards are here: https://huggingface.co/collections/hf-audio/x-codec-68a434ba4cc4d363eb813dce
| {
"login": "ebezzam",
"id": 4757445,
"node_id": "MDQ6VXNlcjQ3NTc0NDU=",
"avatar_url": "https://avatars.githubusercontent.com/u/4757445?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ebezzam",
"html_url": "https://github.com/ebezzam",
"followers_url": "https://api.github.com/users/ebezzam/followers",
"following_url": "https://api.github.com/users/ebezzam/following{/other_user}",
"gists_url": "https://api.github.com/users/ebezzam/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ebezzam/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ebezzam/subscriptions",
"organizations_url": "https://api.github.com/users/ebezzam/orgs",
"repos_url": "https://api.github.com/users/ebezzam/repos",
"events_url": "https://api.github.com/users/ebezzam/events{/privacy}",
"received_events_url": "https://api.github.com/users/ebezzam/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40271/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40271/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40270 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40270/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40270/comments | https://api.github.com/repos/huggingface/transformers/issues/40270/events | https://github.com/huggingface/transformers/pull/40270 | 3,333,384,972 | PR_kwDOCUB6oc6kPykq | 40,270 | Remove MI300 CI | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T08:12:51 | 2025-08-19T08:24:13 | 2025-08-19T08:23:39 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40270",
"html_url": "https://github.com/huggingface/transformers/pull/40270",
"diff_url": "https://github.com/huggingface/transformers/pull/40270.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40270.patch",
"merged_at": "2025-08-19T08:23:39"
} | Cleaning up as we no longer have runners for these types. In commit history if we need it back :) | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40270/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40270/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40269 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40269/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40269/comments | https://api.github.com/repos/huggingface/transformers/issues/40269/events | https://github.com/huggingface/transformers/pull/40269 | 3,333,203,385 | PR_kwDOCUB6oc6kPMpr | 40,269 | Fix attention mask for RT-DETR-based models | {
"login": "materight",
"id": 26522190,
"node_id": "MDQ6VXNlcjI2NTIyMTkw",
"avatar_url": "https://avatars.githubusercontent.com/u/26522190?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/materight",
"html_url": "https://github.com/materight",
"followers_url": "https://api.github.com/users/materight/followers",
"following_url": "https://api.github.com/users/materight/following{/other_user}",
"gists_url": "https://api.github.com/users/materight/gists{/gist_id}",
"starred_url": "https://api.github.com/users/materight/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/materight/subscriptions",
"organizations_url": "https://api.github.com/users/materight/orgs",
"repos_url": "https://api.github.com/users/materight/repos",
"events_url": "https://api.github.com/users/materight/events{/privacy}",
"received_events_url": "https://api.github.com/users/materight/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T07:14:35 | 2025-08-19T09:16:28 | 2025-08-19T09:15:56 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40269",
"html_url": "https://github.com/huggingface/transformers/pull/40269",
"diff_url": "https://github.com/huggingface/transformers/pull/40269.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40269.patch",
"merged_at": "2025-08-19T09:15:56"
} | # What does this PR do?
Currently the attention mask is not handled correctly for RT-DETR-based models (`rt_detr`, `rt_detr_v2`, `d_fine`). The docs for `get_contrastive_denoising_training_group` say that it should return a `torch.FloatTensor` for the attention mask, but it actually returns `torch.bool`. The mask is then just summed up with the attention weights, so it just adds 0 or 1.
This PR changes `get_contrastive_denoising_training_group` to return a `torch.FloatTensor` with `-inf` on the masked weights, and adapts all the custom `MultiHeadAttention` to correctly do the conversion if a boolean mask is given in input, as done in [`nn.MultiheadAttention`](https://github.com/pytorch/pytorch/blob/ba56102387ef21a3b04b357e5b183d48f0afefc7/torch/nn/modules/activation.py#L1233), where the boolean mask is [converted to a float tensor](https://github.com/pytorch/pytorch/blob/8f31aa97a3e1e17bed29b6cedf9884f0c6b145e9/torch/nn/functional.py#L6050).
Related to [this issue](https://github.com/Peterande/D-FINE/issues/276#issuecomment-3052369924).
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [x] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@qubvel @VladOS95-cyber
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40269/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40269/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40268 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40268/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40268/comments | https://api.github.com/repos/huggingface/transformers/issues/40268/events | https://github.com/huggingface/transformers/issues/40268 | 3,333,180,731 | I_kwDOCUB6oc7GrE07 | 40,268 | Some csm testcases fail because of incompatible with torch_comiple | {
"login": "yuanwu2017",
"id": 34643241,
"node_id": "MDQ6VXNlcjM0NjQzMjQx",
"avatar_url": "https://avatars.githubusercontent.com/u/34643241?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuanwu2017",
"html_url": "https://github.com/yuanwu2017",
"followers_url": "https://api.github.com/users/yuanwu2017/followers",
"following_url": "https://api.github.com/users/yuanwu2017/following{/other_user}",
"gists_url": "https://api.github.com/users/yuanwu2017/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yuanwu2017/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yuanwu2017/subscriptions",
"organizations_url": "https://api.github.com/users/yuanwu2017/orgs",
"repos_url": "https://api.github.com/users/yuanwu2017/repos",
"events_url": "https://api.github.com/users/yuanwu2017/events{/privacy}",
"received_events_url": "https://api.github.com/users/yuanwu2017/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T07:06:21 | 2025-08-20T05:25:23 | 2025-08-20T05:25:23 | CONTRIBUTOR | null | null | null | null | ### System Info
root@d1d4a7d93896:/workspace/upstream/transformers# transformers env
2025-08-19 16:10:36.294548: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2025-08-19 16:10:36.329950: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2025-08-19 16:10:36.329980: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2025-08-19 16:10:36.330834: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2025-08-19 16:10:36.335881: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2025-08-19 16:10:37.005075: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
WARNING:tensorflow:From /workspace/upstream/transformers/src/transformers/commands/env.py:120: is_gpu_available (from tensorflow.python.framework.test_util) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.config.list_physical_devices('GPU')` instead.
2025-08-19 16:10:40.370177: W tensorflow/core/common_runtime/gpu/gpu_device.cc:2256] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
No GPU/TPU found, falling back to CPU. (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.)
Copy-and-paste the text below in your GitHub issue and FILL OUT the two last points.
- `transformers` version: 4.56.0.dev0
- Platform: Linux-5.4.292-1.el8.elrepo.x86_64-x86_64-with-glibc2.35
- Python version: 3.10.12
- Huggingface_hub version: 0.34.2
- Safetensors version: 0.5.3
- Accelerate version: 1.3.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.6.0+cu124 (CUDA)
- Tensorflow version (GPU?): 2.15.1 (False)
- Flax version (CPU?/GPU?/TPU?): 0.7.0 (cpu)
- Jax version: 0.4.13
- JaxLib version: 0.4.13
- Using distributed or parallel set-up in script?: <fill in>
- Using GPU in script?: <fill in>
- GPU type: NVIDIA A100 80GB PCIe
### Who can help?
_No response_
### Information
- [x] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
Reproduce steps：
`pytest tests/models/csm`
<img width="789" height="171" alt="Image" src="https://github.com/user-attachments/assets/2793e35d-78bb-4d16-9425-c78315e8fb2d" />
### Expected behavior
Pytest pass | {
"login": "yuanwu2017",
"id": 34643241,
"node_id": "MDQ6VXNlcjM0NjQzMjQx",
"avatar_url": "https://avatars.githubusercontent.com/u/34643241?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuanwu2017",
"html_url": "https://github.com/yuanwu2017",
"followers_url": "https://api.github.com/users/yuanwu2017/followers",
"following_url": "https://api.github.com/users/yuanwu2017/following{/other_user}",
"gists_url": "https://api.github.com/users/yuanwu2017/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yuanwu2017/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yuanwu2017/subscriptions",
"organizations_url": "https://api.github.com/users/yuanwu2017/orgs",
"repos_url": "https://api.github.com/users/yuanwu2017/repos",
"events_url": "https://api.github.com/users/yuanwu2017/events{/privacy}",
"received_events_url": "https://api.github.com/users/yuanwu2017/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40268/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40268/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40267 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40267/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40267/comments | https://api.github.com/repos/huggingface/transformers/issues/40267/events | https://github.com/huggingface/transformers/pull/40267 | 3,333,116,486 | PR_kwDOCUB6oc6kO6ct | 40,267 | Skip some csm testcases because of incompatible with torch_comiple | {
"login": "yuanwu2017",
"id": 34643241,
"node_id": "MDQ6VXNlcjM0NjQzMjQx",
"avatar_url": "https://avatars.githubusercontent.com/u/34643241?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuanwu2017",
"html_url": "https://github.com/yuanwu2017",
"followers_url": "https://api.github.com/users/yuanwu2017/followers",
"following_url": "https://api.github.com/users/yuanwu2017/following{/other_user}",
"gists_url": "https://api.github.com/users/yuanwu2017/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yuanwu2017/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yuanwu2017/subscriptions",
"organizations_url": "https://api.github.com/users/yuanwu2017/orgs",
"repos_url": "https://api.github.com/users/yuanwu2017/repos",
"events_url": "https://api.github.com/users/yuanwu2017/events{/privacy}",
"received_events_url": "https://api.github.com/users/yuanwu2017/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-19T06:42:55 | 2025-08-20T05:24:53 | 2025-08-20T05:24:52 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40267",
"html_url": "https://github.com/huggingface/transformers/pull/40267",
"diff_url": "https://github.com/huggingface/transformers/pull/40267.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40267.patch",
"merged_at": null
} | # What does this PR do?
Skip some csm testcases because of CSM output recording incompatible with torch_comiple
Before:
<img width="790" height="149" alt="image" src="https://github.com/user-attachments/assets/30a66625-f35c-4a43-84eb-e31ff96d8609" />
After:
<img width="777" height="217" alt="image" src="https://github.com/user-attachments/assets/19de3e08-2565-4147-b3f6-698cd4a3e9a8" />
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # ([issue 40268](https://github.com/huggingface/transformers/issues/40268))
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "yuanwu2017",
"id": 34643241,
"node_id": "MDQ6VXNlcjM0NjQzMjQx",
"avatar_url": "https://avatars.githubusercontent.com/u/34643241?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuanwu2017",
"html_url": "https://github.com/yuanwu2017",
"followers_url": "https://api.github.com/users/yuanwu2017/followers",
"following_url": "https://api.github.com/users/yuanwu2017/following{/other_user}",
"gists_url": "https://api.github.com/users/yuanwu2017/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yuanwu2017/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yuanwu2017/subscriptions",
"organizations_url": "https://api.github.com/users/yuanwu2017/orgs",
"repos_url": "https://api.github.com/users/yuanwu2017/repos",
"events_url": "https://api.github.com/users/yuanwu2017/events{/privacy}",
"received_events_url": "https://api.github.com/users/yuanwu2017/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40267/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40267/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40266 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40266/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40266/comments | https://api.github.com/repos/huggingface/transformers/issues/40266/events | https://github.com/huggingface/transformers/issues/40266 | 3,333,071,970 | I_kwDOCUB6oc7GqqRi | 40,266 | PretrainedConfig `from_pretrained` does not maintain `save_pretrained` values with `sub_configs` | {
"login": "AmitMY",
"id": 5757359,
"node_id": "MDQ6VXNlcjU3NTczNTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/5757359?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AmitMY",
"html_url": "https://github.com/AmitMY",
"followers_url": "https://api.github.com/users/AmitMY/followers",
"following_url": "https://api.github.com/users/AmitMY/following{/other_user}",
"gists_url": "https://api.github.com/users/AmitMY/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AmitMY/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AmitMY/subscriptions",
"organizations_url": "https://api.github.com/users/AmitMY/orgs",
"repos_url": "https://api.github.com/users/AmitMY/repos",
"events_url": "https://api.github.com/users/AmitMY/events{/privacy}",
"received_events_url": "https://api.github.com/users/AmitMY/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T06:27:35 | 2025-08-19T14:12:58 | 2025-08-19T14:12:58 | NONE | null | null | null | null | ### System Info
- `transformers` version: 4.55.0.dev0
- Platform: macOS-15.6-arm64-arm-64bit
- Python version: 3.12.2
- Huggingface_hub version: 0.34.3
- Safetensors version: 0.5.3
- Accelerate version: 1.9.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.7.1 (NA)
- Tensorflow version (GPU?): not installed (NA)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
### Who can help?
Don't know.
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
You create a config class with two `sub_configs`
```py
class CustomModelConfig(PretrainedConfig):
model_type = "custom_model"
sub_configs = {
"image_encoder": AutoConfig,
"bytes_encoder": AutoConfig,
}
def __init__(self,
image_encoder: Optional[Union[AutoConfig, dict]] = None,
bytes_encoder: Optional[Union[AutoConfig, dict]] = None,
**kwargs):
# Configuration defaults
super().__init__(**kwargs)
self.image_encoder = image_encoder
self.bytes_encoder = bytes_encoder
```
You imagine that `save_pretrained` would work, but it fails:
> AttributeError: 'dict' object has no attribute 'to_dict'
So you add a few fixes:
```py
self.fix_sub_config("image_encoder")
self.fix_sub_config("bytes_encoder")
def fix_sub_config(self, name: str):
config = getattr(self, name, None)
if isinstance(config, dict):
model_type = getattr(config, "model_type", None)
config_cls = CONFIG_MAPPING[model_type] if model_type else PretrainedConfig
config = config_cls(**config)
setattr(self, name, config)
```
and now `save_pretrained` works, and the config looks correct.
But if you run `CustomModelConfig.from_pretrained(temp_dir)`, you find a different config, with `model_type` missing, leading to not using the correct sub-config class.
### Expected behavior
`save_pretrained` then `load_pretrained` should result in the same config.
```py
temp_dir = "output"
config = CustomModelConfig(
bytes_encoder=AutoConfig.from_pretrained("EleutherAI/pythia-70m"),
image_encoder=AutoConfig.from_pretrained("WinKawaks/vit-tiny-patch16-224"),
)
config.save_pretrained(save_directory=temp_dir, push_to_hub=False)
new_config = CustomModelConfig.from_pretrained(temp_dir)
old_config_text = config.to_json_string()
new_config_text = new_config.to_json_string()
with open("output/config1.json", "w") as f:
f.write(old_config_text)
with open("output/config2.json", "w") as f:
f.write(new_config_text)
assert new_config_text == old_config_text
``` | {
"login": "AmitMY",
"id": 5757359,
"node_id": "MDQ6VXNlcjU3NTczNTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/5757359?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AmitMY",
"html_url": "https://github.com/AmitMY",
"followers_url": "https://api.github.com/users/AmitMY/followers",
"following_url": "https://api.github.com/users/AmitMY/following{/other_user}",
"gists_url": "https://api.github.com/users/AmitMY/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AmitMY/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AmitMY/subscriptions",
"organizations_url": "https://api.github.com/users/AmitMY/orgs",
"repos_url": "https://api.github.com/users/AmitMY/repos",
"events_url": "https://api.github.com/users/AmitMY/events{/privacy}",
"received_events_url": "https://api.github.com/users/AmitMY/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40266/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40266/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40265 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40265/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40265/comments | https://api.github.com/repos/huggingface/transformers/issues/40265/events | https://github.com/huggingface/transformers/pull/40265 | 3,333,038,215 | PR_kwDOCUB6oc6kOpqB | 40,265 | Enable PLW and PLE rules | {
"login": "cyyever",
"id": 17618148,
"node_id": "MDQ6VXNlcjE3NjE4MTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/17618148?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cyyever",
"html_url": "https://github.com/cyyever",
"followers_url": "https://api.github.com/users/cyyever/followers",
"following_url": "https://api.github.com/users/cyyever/following{/other_user}",
"gists_url": "https://api.github.com/users/cyyever/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cyyever/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cyyever/subscriptions",
"organizations_url": "https://api.github.com/users/cyyever/orgs",
"repos_url": "https://api.github.com/users/cyyever/repos",
"events_url": "https://api.github.com/users/cyyever/events{/privacy}",
"received_events_url": "https://api.github.com/users/cyyever/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | open | false | null | [] | null | [] | 2025-08-19T06:16:32 | 2025-10-19T02:44:50 | null | CONTRIBUTOR | null | null | true | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40265",
"html_url": "https://github.com/huggingface/transformers/pull/40265",
"diff_url": "https://github.com/huggingface/transformers/pull/40265.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40265.patch",
"merged_at": null
} | # What does this PR do?
Enable ruff `PLW` and `PLE` rules, which implement warnings and errors of Pylint. | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40265/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40265/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40264 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40264/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40264/comments | https://api.github.com/repos/huggingface/transformers/issues/40264/events | https://github.com/huggingface/transformers/issues/40264 | 3,333,008,893 | I_kwDOCUB6oc7Gqa39 | 40,264 | Model loads with inconsistent `dtype` based on `from_pretrained` and `from_config` | {
"login": "AmitMY",
"id": 5757359,
"node_id": "MDQ6VXNlcjU3NTczNTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/5757359?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AmitMY",
"html_url": "https://github.com/AmitMY",
"followers_url": "https://api.github.com/users/AmitMY/followers",
"following_url": "https://api.github.com/users/AmitMY/following{/other_user}",
"gists_url": "https://api.github.com/users/AmitMY/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AmitMY/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AmitMY/subscriptions",
"organizations_url": "https://api.github.com/users/AmitMY/orgs",
"repos_url": "https://api.github.com/users/AmitMY/repos",
"events_url": "https://api.github.com/users/AmitMY/events{/privacy}",
"received_events_url": "https://api.github.com/users/AmitMY/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T06:04:46 | 2025-08-19T13:42:15 | 2025-08-19T13:42:15 | NONE | null | null | null | null | ### System Info
- `transformers` version: 4.55.0.dev0
- Platform: macOS-15.6-arm64-arm-64bit
- Python version: 3.12.2
- Huggingface_hub version: 0.34.3
- Safetensors version: 0.5.3
- Accelerate version: 1.9.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.7.1 (NA)
- Tensorflow version (GPU?): not installed (NA)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
### Who can help?
I think maybe @Rocketknight1
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
Loading the model once `from_pretrained` and once `from_config` yields different `dtype`.
```py
from transformers import AutoConfig, AutoModelForCausalLM
config = AutoConfig.from_pretrained("EleutherAI/pythia-70m")
model1 = AutoModelForCausalLM.from_config(config)
print(model1.dtype)
model2 = AutoModelForCausalLM.from_pretrained("EleutherAI/pythia-70m")
print(model2.dtype)
assert model1.dtype == model2.dtype, "Models should have the same dtype "
```
### Expected behavior
`dtype` should be consistent. | {
"login": "AmitMY",
"id": 5757359,
"node_id": "MDQ6VXNlcjU3NTczNTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/5757359?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AmitMY",
"html_url": "https://github.com/AmitMY",
"followers_url": "https://api.github.com/users/AmitMY/followers",
"following_url": "https://api.github.com/users/AmitMY/following{/other_user}",
"gists_url": "https://api.github.com/users/AmitMY/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AmitMY/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AmitMY/subscriptions",
"organizations_url": "https://api.github.com/users/AmitMY/orgs",
"repos_url": "https://api.github.com/users/AmitMY/repos",
"events_url": "https://api.github.com/users/AmitMY/events{/privacy}",
"received_events_url": "https://api.github.com/users/AmitMY/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40264/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40264/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40263 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40263/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40263/comments | https://api.github.com/repos/huggingface/transformers/issues/40263/events | https://github.com/huggingface/transformers/issues/40263 | 3,332,905,191 | I_kwDOCUB6oc7GqBjn | 40,263 | [VLMs] How to process a batch that contains samples with and without images? | {
"login": "qgallouedec",
"id": 45557362,
"node_id": "MDQ6VXNlcjQ1NTU3MzYy",
"avatar_url": "https://avatars.githubusercontent.com/u/45557362?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qgallouedec",
"html_url": "https://github.com/qgallouedec",
"followers_url": "https://api.github.com/users/qgallouedec/followers",
"following_url": "https://api.github.com/users/qgallouedec/following{/other_user}",
"gists_url": "https://api.github.com/users/qgallouedec/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qgallouedec/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qgallouedec/subscriptions",
"organizations_url": "https://api.github.com/users/qgallouedec/orgs",
"repos_url": "https://api.github.com/users/qgallouedec/repos",
"events_url": "https://api.github.com/users/qgallouedec/events{/privacy}",
"received_events_url": "https://api.github.com/users/qgallouedec/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [] | 2025-08-19T05:09:36 | 2025-09-18T08:08:51 | 2025-09-18T08:08:51 | MEMBER | null | null | null | null | Is there a **standard** way to process a batch that contains samples with and without images?
For example:
```python
from transformers import AutoProcessor
from PIL import Image
import numpy as np
model_id = ... # tested are "google/gemma-3-4b-it", "HuggingFaceM4/idefics2-8b", "HuggingFaceM4/Idefics3-8B-Llama3", "HuggingFaceTB/SmolVLM2-2.2B-Instruct", "llava-hf/llava-1.5-7b-hf", "llava-hf/llava-v1.6-mistral-7b-hf", "OpenGVLab/InternVL3-8B-hf", "Qwen/Qwen2-VL-2B-Instruct","Qwen/Qwen2.5-VL-3B-Instruct"]
processor = AutoProcessor.from_pretrained(model_id)
messages = [
[{"role": "user", "content": [{"type": "text", "text": "What's the capital of France?"}]}],
[{"role": "user", "content": [{"type": "image"}, {"type": "text", "text": "What is it?"}]}],
]
texts = processor.apply_chat_template(messages)
image = Image.fromarray(
np.random.uniform(low=0.0, high=255.0, size=(32, 48, 3)).astype(np.uint8)
)
images = [[], [image]]
processor(images=images, text=texts)
```
This fails for all models I tested.
```python
images=[image] # The only syntax I found that works for some models: llava-hf/llava-1.5-7b-hf, llava-hf/llava-v1.6-mistral-7b-hf, OpenGVLab/InternVL3-8B-hf, Qwen/Qwen2-VL-2B-Instruct, Qwen/Qwen2.5-VL-3B-Instruct
images = [None, [image]] # always fails
images = [None, image] # always fails
images = [[], [image]] # always fails
```
### Expected behavior
There should be a standard / documented way to batch process mixed inputs (some samples with images, some without).
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40263/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40263/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40262 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40262/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40262/comments | https://api.github.com/repos/huggingface/transformers/issues/40262/events | https://github.com/huggingface/transformers/issues/40262 | 3,332,724,748 | I_kwDOCUB6oc7GpVgM | 40,262 | v4.55.3 tag miss some codes for patch, affect FA2 on Ascend NPU | {
"login": "FightingZhen",
"id": 26176607,
"node_id": "MDQ6VXNlcjI2MTc2NjA3",
"avatar_url": "https://avatars.githubusercontent.com/u/26176607?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/FightingZhen",
"html_url": "https://github.com/FightingZhen",
"followers_url": "https://api.github.com/users/FightingZhen/followers",
"following_url": "https://api.github.com/users/FightingZhen/following{/other_user}",
"gists_url": "https://api.github.com/users/FightingZhen/gists{/gist_id}",
"starred_url": "https://api.github.com/users/FightingZhen/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/FightingZhen/subscriptions",
"organizations_url": "https://api.github.com/users/FightingZhen/orgs",
"repos_url": "https://api.github.com/users/FightingZhen/repos",
"events_url": "https://api.github.com/users/FightingZhen/events{/privacy}",
"received_events_url": "https://api.github.com/users/FightingZhen/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-19T03:05:33 | 2025-08-21T09:50:39 | 2025-08-21T09:50:39 | CONTRIBUTOR | null | null | null | null | ### System Info
codes related to Flash-Attention2 judgment logic in `4.55.3` is different from `main`, `4.55.3` branch code is unavailable on Ascend NPU, `main` branch is OK
For Ascend NPU, `4.55.3` branch will go into else logic branch, and search `flash-attn` package, which will cause `ImportError` on Ascend NPU.
in `4.55.3`
https://github.com/huggingface/transformers/blob/7dbc054e2a0c3cafd3ea22db0566db700b3a8cbf/src/transformers/modeling_utils.py#L2483-L2492
in `main`
https://github.com/huggingface/transformers/blob/6b5bd117231f969713ed79fd4870903ab3c93edf/src/transformers/modeling_utils.py#L2479-L2487
### Who can help?
@ArthurZucker
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
Tested through `LLaMA-Factory`, which is not strong related with above error.
### Expected behavior
`4.55.3` branch keep the code logic in `main` branch | {
"login": "ArthurZucker",
"id": 48595927,
"node_id": "MDQ6VXNlcjQ4NTk1OTI3",
"avatar_url": "https://avatars.githubusercontent.com/u/48595927?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ArthurZucker",
"html_url": "https://github.com/ArthurZucker",
"followers_url": "https://api.github.com/users/ArthurZucker/followers",
"following_url": "https://api.github.com/users/ArthurZucker/following{/other_user}",
"gists_url": "https://api.github.com/users/ArthurZucker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ArthurZucker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ArthurZucker/subscriptions",
"organizations_url": "https://api.github.com/users/ArthurZucker/orgs",
"repos_url": "https://api.github.com/users/ArthurZucker/repos",
"events_url": "https://api.github.com/users/ArthurZucker/events{/privacy}",
"received_events_url": "https://api.github.com/users/ArthurZucker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40262/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40262/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40261 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40261/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40261/comments | https://api.github.com/repos/huggingface/transformers/issues/40261/events | https://github.com/huggingface/transformers/pull/40261 | 3,332,327,141 | PR_kwDOCUB6oc6kMYez | 40,261 | Fix slow static cache export tests | {
"login": "jackzhxng",
"id": 32371937,
"node_id": "MDQ6VXNlcjMyMzcxOTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/32371937?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jackzhxng",
"html_url": "https://github.com/jackzhxng",
"followers_url": "https://api.github.com/users/jackzhxng/followers",
"following_url": "https://api.github.com/users/jackzhxng/following{/other_user}",
"gists_url": "https://api.github.com/users/jackzhxng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jackzhxng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jackzhxng/subscriptions",
"organizations_url": "https://api.github.com/users/jackzhxng/orgs",
"repos_url": "https://api.github.com/users/jackzhxng/repos",
"events_url": "https://api.github.com/users/jackzhxng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jackzhxng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 6886428489,
"node_id": "LA_kwDOCUB6oc8AAAABmnaPSQ",
"url": "https://api.github.com/repos/huggingface/transformers/labels/run-slow",
"name": "run-slow",
"color": "E1D519",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2025-08-18T22:36:18 | 2025-08-19T09:24:07 | 2025-08-19T09:24:07 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40261",
"html_url": "https://github.com/huggingface/transformers/pull/40261",
"diff_url": "https://github.com/huggingface/transformers/pull/40261.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40261.patch",
"merged_at": "2025-08-19T09:24:07"
} | # What does this PR do?
Fix failures on RUN_SLOW `export_static_cache` tests introduced by https://github.com/huggingface/transformers/pull/39836. Reverts to export on token-by-token since:
1. The [generate function used for testing](https://github.com/huggingface/transformers/blob/main/src/transformers/integrations/executorch.py#L349) does prefill sequentially atm.
2. Dynamic shapes were not previously included in the export
Would like to follow up to include dynamic shapes in export - https://github.com/huggingface/transformers/issues/40260
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [x] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@zucchini-nlp | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40261/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40261/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40260 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40260/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40260/comments | https://api.github.com/repos/huggingface/transformers/issues/40260/events | https://github.com/huggingface/transformers/issues/40260 | 3,332,323,865 | I_kwDOCUB6oc7GnzoZ | 40,260 | Add dynamic shapes to ExecuTorch static cache export tests | {
"login": "jackzhxng",
"id": 32371937,
"node_id": "MDQ6VXNlcjMyMzcxOTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/32371937?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jackzhxng",
"html_url": "https://github.com/jackzhxng",
"followers_url": "https://api.github.com/users/jackzhxng/followers",
"following_url": "https://api.github.com/users/jackzhxng/following{/other_user}",
"gists_url": "https://api.github.com/users/jackzhxng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jackzhxng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jackzhxng/subscriptions",
"organizations_url": "https://api.github.com/users/jackzhxng/orgs",
"repos_url": "https://api.github.com/users/jackzhxng/repos",
"events_url": "https://api.github.com/users/jackzhxng/events{/privacy}",
"received_events_url": "https://api.github.com/users/jackzhxng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 2648621985,
"node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1",
"url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request",
"name": "Feature request",
"color": "FBCA04",
"default": false,
"description": "Request for a new feature"
}
] | open | false | null | [] | null | [] | 2025-08-18T22:34:13 | 2025-08-28T17:12:18 | null | CONTRIBUTOR | null | null | null | null | ### Feature request
Add dynamic shapes to the ExecuTorch `text_export_static_cache` tests on [causalLM models such as Qwen](https://github.com/huggingface/transformers/blob/main/tests/models/qwen3/test_modeling_qwen3.py#L297).
Then modify the test generation function to prefill all at once [instead of sequentially](https://github.com/huggingface/transformers/blob/main/src/transformers/integrations/executorch.py#L386).
### Motivation
Tests the export more accurately since production use cases will almost always export with dynamic cache to prefill all at once during inference.
### Your contribution
I submit the PR @jackzhxng | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40260/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40260/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | false |
https://api.github.com/repos/huggingface/transformers/issues/40259 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40259/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40259/comments | https://api.github.com/repos/huggingface/transformers/issues/40259/events | https://github.com/huggingface/transformers/pull/40259 | 3,332,301,263 | PR_kwDOCUB6oc6kMTGz | 40,259 | Add LFM2-VL support | {
"login": "ankke",
"id": 48625325,
"node_id": "MDQ6VXNlcjQ4NjI1MzI1",
"avatar_url": "https://avatars.githubusercontent.com/u/48625325?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ankke",
"html_url": "https://github.com/ankke",
"followers_url": "https://api.github.com/users/ankke/followers",
"following_url": "https://api.github.com/users/ankke/following{/other_user}",
"gists_url": "https://api.github.com/users/ankke/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ankke/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ankke/subscriptions",
"organizations_url": "https://api.github.com/users/ankke/orgs",
"repos_url": "https://api.github.com/users/ankke/repos",
"events_url": "https://api.github.com/users/ankke/events{/privacy}",
"received_events_url": "https://api.github.com/users/ankke/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T22:19:14 | 2025-09-19T10:48:18 | 2025-09-19T10:48:18 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40259",
"html_url": "https://github.com/huggingface/transformers/pull/40259",
"diff_url": "https://github.com/huggingface/transformers/pull/40259.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40259.patch",
"merged_at": null
} | Add support for LFM2-VL models.
LFM2‑VL is [Liquid AI](https://www.liquid.ai/)'s first series of multimodal models, designed to process text and images with variable resolutions. Built on the [LFM2](https://huggingface.co/collections/LiquidAI/lfm2-686d721927015b2ad73eaa38) backbone, it is optimized for low-latency and edge AI applications.
Checkpoints available [here](https://huggingface.co/collections/LiquidAI/lfm2-vl-68963bbc84a610f7638d5ffa).
| {
"login": "ankke",
"id": 48625325,
"node_id": "MDQ6VXNlcjQ4NjI1MzI1",
"avatar_url": "https://avatars.githubusercontent.com/u/48625325?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ankke",
"html_url": "https://github.com/ankke",
"followers_url": "https://api.github.com/users/ankke/followers",
"following_url": "https://api.github.com/users/ankke/following{/other_user}",
"gists_url": "https://api.github.com/users/ankke/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ankke/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ankke/subscriptions",
"organizations_url": "https://api.github.com/users/ankke/orgs",
"repos_url": "https://api.github.com/users/ankke/repos",
"events_url": "https://api.github.com/users/ankke/events{/privacy}",
"received_events_url": "https://api.github.com/users/ankke/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40259/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40259/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40258 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40258/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40258/comments | https://api.github.com/repos/huggingface/transformers/issues/40258/events | https://github.com/huggingface/transformers/issues/40258 | 3,331,813,260 | I_kwDOCUB6oc7Gl2-M | 40,258 | RuntimeError in loss computation with label_smoothing_factor > 0 for multi-label classification | {
"login": "andreabosisio",
"id": 79710398,
"node_id": "MDQ6VXNlcjc5NzEwMzk4",
"avatar_url": "https://avatars.githubusercontent.com/u/79710398?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/andreabosisio",
"html_url": "https://github.com/andreabosisio",
"followers_url": "https://api.github.com/users/andreabosisio/followers",
"following_url": "https://api.github.com/users/andreabosisio/following{/other_user}",
"gists_url": "https://api.github.com/users/andreabosisio/gists{/gist_id}",
"starred_url": "https://api.github.com/users/andreabosisio/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/andreabosisio/subscriptions",
"organizations_url": "https://api.github.com/users/andreabosisio/orgs",
"repos_url": "https://api.github.com/users/andreabosisio/repos",
"events_url": "https://api.github.com/users/andreabosisio/events{/privacy}",
"received_events_url": "https://api.github.com/users/andreabosisio/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T19:16:18 | 2025-08-25T14:23:32 | 2025-08-25T14:23:32 | NONE | null | null | null | null | ### System Info
```
OS: Linux-5.15.0-1092-azure-x86_64-with-glibc2.35
Python: 3.10.12 (main, May 27 2025, 17:12:29) [GCC 11.4.0]
PyTorch: 2.0.1+cu118
TensorFlow: 2.14.1
transformers: 4.36.1
```
### Who can help?
@zach-huggingface @SunMarc
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
```
from transformers import (
AutoTokenizer,
AutoModelForSequenceClassification,
TrainingArguments,
Trainer
)
import torch
from torch.utils.data import Dataset
# ----- Minimal fake dataset -----
class DummyDataset(Dataset):
def __init__(self, tokenizer):
self.texts = ["hello world", "foo bar", "transformers test"]
# 3 examples, 3 labels (multi-label one-hot)
self.labels = torch.tensor([[1, 0, 1], [0, 1, 0], [1, 1, 0]], dtype=torch.float)
self.encodings = tokenizer(self.texts, truncation=True, padding=True)
def __len__(self):
return len(self.texts)
def __getitem__(self, idx):
item = {k: torch.tensor(v[idx]) for k, v in self.encodings.items()}
item["labels"] = self.labels[idx]
return item
# ----- Load tokenizer and model -----
model_name = "intfloat/multilingual-e5-large-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_name)
# Define model for 3 labels, multi-label classification
model = AutoModelForSequenceClassification.from_pretrained(
model_name,
num_labels=3,
problem_type="multi_label_classification"
)
# ----- Prepare dataset -----
train_dataset = DummyDataset(tokenizer)
# ----- Training arguments -----
training_args = TrainingArguments(
output_dir="./test-trainer",
per_device_train_batch_size=2,
num_train_epochs=1,
learning_rate=5e-5,
label_smoothing_factor=0.1,
logging_steps=1,
remove_unused_columns=False,
report_to="none"
)
# ----- Trainer -----
trainer = Trainer(
model=model,
args=training_args,
train_dataset=train_dataset,
)
# ----- Run training -----
trainer.train()
```
Got
```
/databricks/python/lib/python3.10/site-packages/huggingface_hub/file_download.py:795: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
warnings.warn(
Some weights of XLMRobertaForSequenceClassification were not initialized from the model checkpoint at intfloat/multilingual-e5-large-instruct and are newly initialized: ['classifier.dense.bias', 'classifier.dense.weight', 'classifier.out_proj.bias', 'classifier.out_proj.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
[W reducer.cpp:1300] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
```
```
RuntimeError: gather(): Expected dtype int64 for index
File <command-6212259188830850>, line 60
53 trainer = Trainer(
54 model=model,
55 args=training_args,
56 train_dataset=train_dataset,
57 )
59 # ----- Run training -----
---> 60 trainer.train()
File <command-6212259188830850>, line 60
53 trainer = Trainer(
54 model=model,
55 args=training_args,
56 train_dataset=train_dataset,
57 )
59 # ----- Run training -----
---> 60 trainer.train()
File /databricks/python/lib/python3.10/site-packages/mlflow/utils/autologging_utils/safety.py:451, in safe_patch.<locals>.safe_patch_function(*args, **kwargs)
436 if (
437 active_session_failed
438 or autologging_is_disabled(autologging_integration)
(...)
445 # warning behavior during original function execution, since autologging is being
446 # skipped
447 with set_non_mlflow_warnings_behavior_for_current_thread(
448 disable_warnings=False,
449 reroute_warnings=False,
450 ):
--> 451 return original(*args, **kwargs)
453 # Whether or not the original / underlying function has been called during the
454 # execution of patched code
455 original_has_been_called = False
File /databricks/python_shell/lib/dbruntime/huggingface_patches/transformers.py:54, in _create_patch_function.<locals>.patched_fit_function(self, *args, **kwargs)
52 call_succeeded = False
53 try:
---> 54 model = original_method(self, *args, **kwargs)
55 call_succeeded = True
56 return model
File /databricks/python/lib/python3.10/site-packages/transformers/trainer.py:1537, in Trainer.train(self, resume_from_checkpoint, trial, ignore_keys_for_eval, **kwargs)
1535 hf_hub_utils.enable_progress_bars()
1536 else:
-> 1537 return inner_training_loop(
1538 args=args,
1539 resume_from_checkpoint=resume_from_checkpoint,
1540 trial=trial,
1541 ignore_keys_for_eval=ignore_keys_for_eval,
1542 )
File /databricks/python/lib/python3.10/site-packages/transformers/trainer.py:1854, in Trainer._inner_training_loop(self, batch_size, args, resume_from_checkpoint, trial, ignore_keys_for_eval)
1851 self.control = self.callback_handler.on_step_begin(args, self.state, self.control)
1853 with self.accelerator.accumulate(model):
-> 1854 tr_loss_step = self.training_step(model, inputs)
1856 if (
1857 args.logging_nan_inf_filter
1858 and not is_torch_tpu_available()
1859 and (torch.isnan(tr_loss_step) or torch.isinf(tr_loss_step))
1860 ):
1861 # if loss is nan or inf simply add the average of previous logged losses
1862 tr_loss += tr_loss / (1 + self.state.global_step - self._globalstep_last_logged)
File /databricks/python/lib/python3.10/site-packages/transformers/trainer.py:2728, in Trainer.training_step(self, model, inputs)
2725 return loss_mb.reduce_mean().detach().to(self.args.device)
2727 with self.compute_loss_context_manager():
-> 2728 loss = self.compute_loss(model, inputs)
2730 if self.args.n_gpu > 1:
2731 loss = loss.mean() # mean() to average on multi-gpu parallel training
File /databricks/python/lib/python3.10/site-packages/transformers/trainer.py:2766, in Trainer.compute_loss(self, model, inputs, return_outputs)
2764 loss = self.label_smoother(outputs, labels, shift_labels=True)
2765 else:
-> 2766 loss = self.label_smoother(outputs, labels)
2767 else:
2768 if isinstance(outputs, dict) and "loss" not in outputs:
File /databricks/python/lib/python3.10/site-packages/transformers/trainer_pt_utils.py:508, in LabelSmoother.__call__(self, model_output, labels, shift_labels)
505 # In case the ignore_index is -100, the gather will fail, so we replace labels by 0. The padding_mask
506 # will ignore them in any case.
507 labels = torch.clamp(labels, min=0)
--> 508 nll_loss = log_probs.gather(dim=-1, index=labels)
509 # works for fp16 input tensor too, by internally upcasting it to fp32
510 smoothed_loss = log_probs.sum(dim=-1, keepdim=True, dtype=torch.float32)
```
### Expected behavior
A warning or error if `label_smoothing_factor>0` is not compatible with `problem_type="multi_label_classification"` | {
"login": "SunMarc",
"id": 57196510,
"node_id": "MDQ6VXNlcjU3MTk2NTEw",
"avatar_url": "https://avatars.githubusercontent.com/u/57196510?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SunMarc",
"html_url": "https://github.com/SunMarc",
"followers_url": "https://api.github.com/users/SunMarc/followers",
"following_url": "https://api.github.com/users/SunMarc/following{/other_user}",
"gists_url": "https://api.github.com/users/SunMarc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SunMarc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SunMarc/subscriptions",
"organizations_url": "https://api.github.com/users/SunMarc/orgs",
"repos_url": "https://api.github.com/users/SunMarc/repos",
"events_url": "https://api.github.com/users/SunMarc/events{/privacy}",
"received_events_url": "https://api.github.com/users/SunMarc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40258/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40258/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40257 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40257/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40257/comments | https://api.github.com/repos/huggingface/transformers/issues/40257/events | https://github.com/huggingface/transformers/pull/40257 | 3,331,801,274 | PR_kwDOCUB6oc6kKmyh | 40,257 | Fix: correct `k_proj` weight and bias slicing in D-FINE | {
"login": "notkisk",
"id": 107971634,
"node_id": "U_kgDOBm-EMg",
"avatar_url": "https://avatars.githubusercontent.com/u/107971634?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/notkisk",
"html_url": "https://github.com/notkisk",
"followers_url": "https://api.github.com/users/notkisk/followers",
"following_url": "https://api.github.com/users/notkisk/following{/other_user}",
"gists_url": "https://api.github.com/users/notkisk/gists{/gist_id}",
"starred_url": "https://api.github.com/users/notkisk/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/notkisk/subscriptions",
"organizations_url": "https://api.github.com/users/notkisk/orgs",
"repos_url": "https://api.github.com/users/notkisk/repos",
"events_url": "https://api.github.com/users/notkisk/events{/privacy}",
"received_events_url": "https://api.github.com/users/notkisk/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T19:12:11 | 2025-08-19T09:45:21 | 2025-08-19T09:44:37 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40257",
"html_url": "https://github.com/huggingface/transformers/pull/40257",
"diff_url": "https://github.com/huggingface/transformers/pull/40257.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40257.patch",
"merged_at": "2025-08-19T09:44:37"
} | This PR fixes the slicing of the `k_proj` weight and bias during checkpoint conversion for D-FINE.
Fixes #40253
**Changes:**
* Corrected `k_proj.weight` from `in_proj_weight[256:384, :]` → `in_proj_weight[128:256, :]`.
* Corrected `k_proj.bias` from `in_proj_bias[256:384]` → `in_proj_bias[128:256]`.
**Impact:**
Ensures the attention mechanism in D-FINE correctly loads the original PyTorch checkpoint without misaligned projections. | {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40257/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40257/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40256 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40256/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40256/comments | https://api.github.com/repos/huggingface/transformers/issues/40256/events | https://github.com/huggingface/transformers/pull/40256 | 3,331,735,086 | PR_kwDOCUB6oc6kKY2C | 40,256 | make model docs device agnostic (2) | {
"login": "yao-matrix",
"id": 7245027,
"node_id": "MDQ6VXNlcjcyNDUwMjc=",
"avatar_url": "https://avatars.githubusercontent.com/u/7245027?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yao-matrix",
"html_url": "https://github.com/yao-matrix",
"followers_url": "https://api.github.com/users/yao-matrix/followers",
"following_url": "https://api.github.com/users/yao-matrix/following{/other_user}",
"gists_url": "https://api.github.com/users/yao-matrix/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yao-matrix/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yao-matrix/subscriptions",
"organizations_url": "https://api.github.com/users/yao-matrix/orgs",
"repos_url": "https://api.github.com/users/yao-matrix/repos",
"events_url": "https://api.github.com/users/yao-matrix/events{/privacy}",
"received_events_url": "https://api.github.com/users/yao-matrix/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T18:46:21 | 2025-08-19T20:26:47 | 2025-08-19T20:10:03 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40256",
"html_url": "https://github.com/huggingface/transformers/pull/40256",
"diff_url": "https://github.com/huggingface/transformers/pull/40256.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40256.patch",
"merged_at": "2025-08-19T20:10:03"
} | @stevhliu , following https://github.com/huggingface/transformers/pull/40143, i am updating more model docs to make them device agnostic, pls help review, thx very much. | {
"login": "stevhliu",
"id": 59462357,
"node_id": "MDQ6VXNlcjU5NDYyMzU3",
"avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevhliu",
"html_url": "https://github.com/stevhliu",
"followers_url": "https://api.github.com/users/stevhliu/followers",
"following_url": "https://api.github.com/users/stevhliu/following{/other_user}",
"gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions",
"organizations_url": "https://api.github.com/users/stevhliu/orgs",
"repos_url": "https://api.github.com/users/stevhliu/repos",
"events_url": "https://api.github.com/users/stevhliu/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevhliu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40256/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40256/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40255 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40255/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40255/comments | https://api.github.com/repos/huggingface/transformers/issues/40255/events | https://github.com/huggingface/transformers/issues/40255 | 3,331,657,077 | I_kwDOCUB6oc7GlQ11 | 40,255 | SAM decoder training error | {
"login": "alexdzm",
"id": 63930909,
"node_id": "MDQ6VXNlcjYzOTMwOTA5",
"avatar_url": "https://avatars.githubusercontent.com/u/63930909?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/alexdzm",
"html_url": "https://github.com/alexdzm",
"followers_url": "https://api.github.com/users/alexdzm/followers",
"following_url": "https://api.github.com/users/alexdzm/following{/other_user}",
"gists_url": "https://api.github.com/users/alexdzm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/alexdzm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/alexdzm/subscriptions",
"organizations_url": "https://api.github.com/users/alexdzm/orgs",
"repos_url": "https://api.github.com/users/alexdzm/repos",
"events_url": "https://api.github.com/users/alexdzm/events{/privacy}",
"received_events_url": "https://api.github.com/users/alexdzm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T18:14:53 | 2025-09-04T17:32:41 | 2025-09-04T17:32:41 | NONE | null | null | null | null | ### System Info
- `transformers` version: 4.55.2
- Platform: Linux-4.18.0-553.16.1.el8_10.cloud.0.1.x86_64-x86_64-with-glibc2.28
- Python version: 3.12.10
- Huggingface_hub version: 0.34.4
- Safetensors version: 0.6.2
- Accelerate version: 1.10.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.8.0+cu128 (CUDA)
- Tensorflow version (GPU?): not installed (NA)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
- Using distributed or parallel set-up in script?: No
- Using GPU in script?: Yes
- GPU type: NVIDIA A100-SXM4-80GB
### Who can help?
@amyeroberts, @qubvel, @vasqu
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
Training SAM with a transformers trainer causes this error. It seems that `dropout_p` is not defined anywhere for this object and this logic was updated in [a recent PR](https://github.com/huggingface/transformers/commit/433d2a23d734914e0a9903c6b7039d28daa6cbc1). When I manually replace this if else with a set dropout value it works. It would be nice to remove this please!
This can be reproduced by running the `sam_trainer.py` script in [this repo ](https://github.com/i-dot-ai/extract-finetuning)
```
Exception has occurred: AttributeError (note: full exception trace is shown but execution is paused at: _run_module_as_main)
'SamAttention' object has no attribute 'dropout_p'
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1962, in __getattr__
raise AttributeError(
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/models/sam/modeling_sam.py", line 269, in forward
dropout=0.0 if not self.training else self.dropout_p,
^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/models/sam/modeling_sam.py", line 325, in forward
queries, _ = self.self_attn(query=queries, key=queries, value=queries)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/models/sam/modeling_sam.py", line 397, in forward
queries, keys, _ = layer(
^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/models/sam/modeling_sam.py", line 512, in forward
point_embedding, image_embeddings = self.transformer(
^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/models/sam/modeling_sam.py", line 1353, in forward
low_res_masks, iou_predictions = self.mask_decoder(
^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/utils/generic.py", line 1083, in wrapper
outputs = func(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/amp/autocast_mode.py", line 44, in decorate_autocast
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/accelerate/utils/operations.py", line 806, in __call__
return convert_to_fp32(self.model_forward(*args, **kwargs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/accelerate/utils/operations.py", line 818, in forward
return model_forward(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/sam_trainer.py", line 119, in compute_loss
outputs = model(**inputs, multimask_output=False)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/trainer.py", line 3796, in training_step
loss = self.compute_loss(model, inputs, num_items_in_batch=num_items_in_batch)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/trainer.py", line 2582, in _inner_training_loop
tr_loss_step = self.training_step(model, inputs, num_items_in_batch)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/.venv/lib/python3.12/site-packages/transformers/trainer.py", line 2238, in train
return inner_training_loop(
^^^^^^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/sam_trainer.py", line 234, in main
train_result = trainer.train()
^^^^^^^^^^^^^^^
File "/mnt/data1/extract-finetuning/sam_trainer.py", line 240, in <module>
main()
File "/mnt/data1/.local/share/uv/python/cpython-3.12.10-linux-x86_64-gnu/lib/python3.12/runpy.py", line 88, in _run_code
exec(code, run_globals)
File "/mnt/data1/.local/share/uv/python/cpython-3.12.10-linux-x86_64-gnu/lib/python3.12/runpy.py", line 198, in _run_module_as_main (Current frame)
return _run_code(code, main_globals, None,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'SamAttention' object has no attribute 'dropout_p'
```
### Expected behavior
The model should train without throwing this error | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40255/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40255/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40254 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40254/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40254/comments | https://api.github.com/repos/huggingface/transformers/issues/40254/events | https://github.com/huggingface/transformers/pull/40254 | 3,331,628,188 | PR_kwDOCUB6oc6kKCSd | 40,254 | Fix benchmark workflow | {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T18:03:34 | 2025-08-18T18:14:50 | 2025-08-18T18:14:16 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40254",
"html_url": "https://github.com/huggingface/transformers/pull/40254",
"diff_url": "https://github.com/huggingface/transformers/pull/40254.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40254.patch",
"merged_at": "2025-08-18T18:14:16"
} | # What does this PR do?
This PR fixes a path issue introduced in https://github.com/huggingface/transformers/pull/39768.
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "ahadnagy",
"id": 21314428,
"node_id": "MDQ6VXNlcjIxMzE0NDI4",
"avatar_url": "https://avatars.githubusercontent.com/u/21314428?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ahadnagy",
"html_url": "https://github.com/ahadnagy",
"followers_url": "https://api.github.com/users/ahadnagy/followers",
"following_url": "https://api.github.com/users/ahadnagy/following{/other_user}",
"gists_url": "https://api.github.com/users/ahadnagy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ahadnagy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ahadnagy/subscriptions",
"organizations_url": "https://api.github.com/users/ahadnagy/orgs",
"repos_url": "https://api.github.com/users/ahadnagy/repos",
"events_url": "https://api.github.com/users/ahadnagy/events{/privacy}",
"received_events_url": "https://api.github.com/users/ahadnagy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40254/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40254/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40253 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40253/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40253/comments | https://api.github.com/repos/huggingface/transformers/issues/40253/events | https://github.com/huggingface/transformers/issues/40253 | 3,331,494,526 | I_kwDOCUB6oc7GkpJ- | 40,253 | Wrong attention weights conversion for D-FINE | {
"login": "materight",
"id": 26522190,
"node_id": "MDQ6VXNlcjI2NTIyMTkw",
"avatar_url": "https://avatars.githubusercontent.com/u/26522190?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/materight",
"html_url": "https://github.com/materight",
"followers_url": "https://api.github.com/users/materight/followers",
"following_url": "https://api.github.com/users/materight/following{/other_user}",
"gists_url": "https://api.github.com/users/materight/gists{/gist_id}",
"starred_url": "https://api.github.com/users/materight/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/materight/subscriptions",
"organizations_url": "https://api.github.com/users/materight/orgs",
"repos_url": "https://api.github.com/users/materight/repos",
"events_url": "https://api.github.com/users/materight/events{/privacy}",
"received_events_url": "https://api.github.com/users/materight/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T17:11:34 | 2025-08-20T10:05:24 | 2025-08-19T09:44:38 | CONTRIBUTOR | null | null | null | null | ### System Info
I was trying to train D-FINE on a custom dataset, but for some reason I could never get it to work with huggingface's implementation, while with the original repo it was training just fine. I went over the two implementation layer by layer, and the self-attention weights in the decoder layer seem to be different.
[Here](https://github.com/huggingface/transformers/blob/01c03bf4eea3f9a37031fc6dd30a01a6ba130fda/src/transformers/models/d_fine/modeling_d_fine.py#L245-L246) `k_proj` and `v_proj` have the same weights. I think there's an issue in the [weights conversion script](https://github.com/huggingface/transformers/blob/2bcf9f6c7ef7f84b3c3aeec51e117429ecdc1254/src/transformers/models/d_fine/convert_d_fine_original_pytorch_checkpoint_to_hf.py#L367C1-L370C96). `[256:384, :]` and `[-128:, :]` refer to the same slice, one of those should be `[128:256, :]` instead.
### Who can help?
@VladOS95-cyber @qubvel
### Information
- [x] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [x] My own task or dataset (give details below)
### Reproduction
```python
from transformers import DFineForObjectDetection
model = DFineForObjectDetection.from_pretrained("ustc-community/dfine-nano-coco")
eq = model.model.decoder.layers[0].self_attn.k_proj.weight == model.model.decoder.layers[0].self_attn.v_proj.weight
print(eq.all())
```
### Expected behavior
`k_proj` and `v_proj` should have different weights, as in the original model. | {
"login": "qubvel",
"id": 31920396,
"node_id": "MDQ6VXNlcjMxOTIwMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31920396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/qubvel",
"html_url": "https://github.com/qubvel",
"followers_url": "https://api.github.com/users/qubvel/followers",
"following_url": "https://api.github.com/users/qubvel/following{/other_user}",
"gists_url": "https://api.github.com/users/qubvel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/qubvel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/qubvel/subscriptions",
"organizations_url": "https://api.github.com/users/qubvel/orgs",
"repos_url": "https://api.github.com/users/qubvel/repos",
"events_url": "https://api.github.com/users/qubvel/events{/privacy}",
"received_events_url": "https://api.github.com/users/qubvel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40253/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40253/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40252 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40252/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40252/comments | https://api.github.com/repos/huggingface/transformers/issues/40252/events | https://github.com/huggingface/transformers/pull/40252 | 3,331,408,308 | PR_kwDOCUB6oc6kJTHz | 40,252 | Update image_processing_perception_lm_fast.py to allow for proper override of vision_input_type | {
"login": "tyleryzhu",
"id": 22010655,
"node_id": "MDQ6VXNlcjIyMDEwNjU1",
"avatar_url": "https://avatars.githubusercontent.com/u/22010655?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tyleryzhu",
"html_url": "https://github.com/tyleryzhu",
"followers_url": "https://api.github.com/users/tyleryzhu/followers",
"following_url": "https://api.github.com/users/tyleryzhu/following{/other_user}",
"gists_url": "https://api.github.com/users/tyleryzhu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tyleryzhu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tyleryzhu/subscriptions",
"organizations_url": "https://api.github.com/users/tyleryzhu/orgs",
"repos_url": "https://api.github.com/users/tyleryzhu/repos",
"events_url": "https://api.github.com/users/tyleryzhu/events{/privacy}",
"received_events_url": "https://api.github.com/users/tyleryzhu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T16:40:48 | 2025-08-19T12:25:20 | 2025-08-19T11:41:27 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40252",
"html_url": "https://github.com/huggingface/transformers/pull/40252",
"diff_url": "https://github.com/huggingface/transformers/pull/40252.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40252.patch",
"merged_at": "2025-08-19T11:41:27"
} | Allow for a proper override of vision_input_type in hf fast image processor, otherwise we need to resort to manually setting the attribute.
# What does this PR do?
Allow for a proper override of vision_input_type in hf fast image processor, otherwise we need to resort to manually setting the attribute.
Fixes #40251
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [x] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40252/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40252/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40251 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40251/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40251/comments | https://api.github.com/repos/huggingface/transformers/issues/40251/events | https://github.com/huggingface/transformers/issues/40251 | 3,331,404,490 | I_kwDOCUB6oc7GkTLK | 40,251 | PerceptionLM Image Processor doesn't properly allow for vision_input_type override | {
"login": "tyleryzhu",
"id": 22010655,
"node_id": "MDQ6VXNlcjIyMDEwNjU1",
"avatar_url": "https://avatars.githubusercontent.com/u/22010655?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tyleryzhu",
"html_url": "https://github.com/tyleryzhu",
"followers_url": "https://api.github.com/users/tyleryzhu/followers",
"following_url": "https://api.github.com/users/tyleryzhu/following{/other_user}",
"gists_url": "https://api.github.com/users/tyleryzhu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tyleryzhu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tyleryzhu/subscriptions",
"organizations_url": "https://api.github.com/users/tyleryzhu/orgs",
"repos_url": "https://api.github.com/users/tyleryzhu/repos",
"events_url": "https://api.github.com/users/tyleryzhu/events{/privacy}",
"received_events_url": "https://api.github.com/users/tyleryzhu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T16:39:34 | 2025-08-19T11:41:28 | 2025-08-19T11:41:28 | CONTRIBUTOR | null | null | null | null | ### System Info
- `transformers` version: 4.55.2
- Platform: Linux-5.10.0-35-cloud-amd64-x86_64-with-glibc2.31
- Python version: 3.13.5
- Huggingface_hub version: 0.34.4
- Safetensors version: 0.6.2
- Accelerate version: 1.10.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.8.0+cu128 (CUDA)
- Tensorflow version (GPU?): not installed (NA)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
- Using distributed or parallel set-up in script?: no
- Using GPU in script?: doesn't matter, but yes
- GPU type: NVIDIA A100-SXM4-80GB
### Who can help?
Currently the PerceptionLM image processor does not properly allow us to override the vision_input_type from thumb+tile to vanilla if we'd like when applying the processor, as it only takes it from initialization which does not offer an override capability. I'd propose to making a change to [this line](https://github.com/huggingface/transformers/blob/main/src/transformers/models/perception_lm/image_processing_perception_lm_fast.py#L286), which would still successfully inherit the default value but also be overridden by kwargs when manually creating the processor.
@shuminghu is likely the best POC for this.
### Information
- [x] The official example scripts
- [ ] My own modified scripts
### Tasks
- [x] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
```
from transformers import AutoProcessor, AutoModelForImageTextToText
from huggingface_hub import hf_hub_download
MODEL_PATH = "facebook/Perception-LM-1B"
processor = AutoProcessor.from_pretrained(MODEL_PATH, use_fast=True)
model = AutoModelForImageTextToText.from_pretrained(MODEL_PATH).to("cuda")
test_image_file = hf_hub_download(
repo_id="shumingh/perception_lm_test_images",
filename="14496_0.PNG",
repo_type="dataset",
)
conversation = [
{
"role": "user",
"content": [
{
"type": "image",
"url": test_image_file,
},
{"type": "text", "text": "Describe the bar plot in the image."},
],
}
]
inputs = processor.apply_chat_template(
[conversation],
add_generation_prompt=True,
tokenize=True,
return_dict=True,
return_tensors="pt",
images_kwargs={
"vision_input_type": "vanilla",
"tile_size": 448,
"max_num_tiles": 1,
}
)
```
### Expected behavior
Allow the override to happen like in the non-huggingface code. | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40251/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40251/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40250 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40250/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40250/comments | https://api.github.com/repos/huggingface/transformers/issues/40250/events | https://github.com/huggingface/transformers/pull/40250 | 3,331,200,315 | PR_kwDOCUB6oc6kIm0F | 40,250 | Standardize BertGeneration model card | {
"login": "nemitha2005",
"id": 139318679,
"node_id": "U_kgDOCE3Vlw",
"avatar_url": "https://avatars.githubusercontent.com/u/139318679?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nemitha2005",
"html_url": "https://github.com/nemitha2005",
"followers_url": "https://api.github.com/users/nemitha2005/followers",
"following_url": "https://api.github.com/users/nemitha2005/following{/other_user}",
"gists_url": "https://api.github.com/users/nemitha2005/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nemitha2005/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nemitha2005/subscriptions",
"organizations_url": "https://api.github.com/users/nemitha2005/orgs",
"repos_url": "https://api.github.com/users/nemitha2005/repos",
"events_url": "https://api.github.com/users/nemitha2005/events{/privacy}",
"received_events_url": "https://api.github.com/users/nemitha2005/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T15:34:50 | 2025-08-19T18:22:13 | 2025-08-19T18:22:13 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40250",
"html_url": "https://github.com/huggingface/transformers/pull/40250",
"diff_url": "https://github.com/huggingface/transformers/pull/40250.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40250.patch",
"merged_at": "2025-08-19T18:22:13"
} | # What does this PR do?
#36979
Updated the BertGeneration model card to follow the new standardized format including:
- New consistent layout with badges
- Friendly description written for accessibility
- Usage examples with Pipeline, AutoModel, and transformers-cli
- Quantization example with BitsAndBytesConfig
- Updated resources section with proper links
This standardizes the BertGeneration documentation to match the new template format requested in issue #36979.
## Before submitting
- [x] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you read the [[contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request)](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [[forum](https://discuss.huggingface.co/)](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs)](https://github.com/huggingface/transformers/tree/main/docs), and
[[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation)](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@stevhliu - Documentation lead who is managing the model card standardization project.
Anyone in the community is free to review the PR once the tests have passed. | {
"login": "stevhliu",
"id": 59462357,
"node_id": "MDQ6VXNlcjU5NDYyMzU3",
"avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevhliu",
"html_url": "https://github.com/stevhliu",
"followers_url": "https://api.github.com/users/stevhliu/followers",
"following_url": "https://api.github.com/users/stevhliu/following{/other_user}",
"gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions",
"organizations_url": "https://api.github.com/users/stevhliu/orgs",
"repos_url": "https://api.github.com/users/stevhliu/repos",
"events_url": "https://api.github.com/users/stevhliu/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevhliu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40250/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40250/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40249 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40249/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40249/comments | https://api.github.com/repos/huggingface/transformers/issues/40249/events | https://github.com/huggingface/transformers/pull/40249 | 3,331,173,583 | PR_kwDOCUB6oc6kIg9F | 40,249 | [`CI`] Fix repo consistency and docs | {
"login": "vasqu",
"id": 73884904,
"node_id": "MDQ6VXNlcjczODg0OTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/73884904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vasqu",
"html_url": "https://github.com/vasqu",
"followers_url": "https://api.github.com/users/vasqu/followers",
"following_url": "https://api.github.com/users/vasqu/following{/other_user}",
"gists_url": "https://api.github.com/users/vasqu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vasqu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vasqu/subscriptions",
"organizations_url": "https://api.github.com/users/vasqu/orgs",
"repos_url": "https://api.github.com/users/vasqu/repos",
"events_url": "https://api.github.com/users/vasqu/events{/privacy}",
"received_events_url": "https://api.github.com/users/vasqu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T15:27:26 | 2025-09-22T16:56:58 | 2025-08-18T15:32:17 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40249",
"html_url": "https://github.com/huggingface/transformers/pull/40249",
"diff_url": "https://github.com/huggingface/transformers/pull/40249.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40249.patch",
"merged_at": "2025-08-18T15:32:17"
} | Missed in #37088, just a typo to be passed from modular :eyes: | {
"login": "Cyrilvallez",
"id": 71554963,
"node_id": "MDQ6VXNlcjcxNTU0OTYz",
"avatar_url": "https://avatars.githubusercontent.com/u/71554963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Cyrilvallez",
"html_url": "https://github.com/Cyrilvallez",
"followers_url": "https://api.github.com/users/Cyrilvallez/followers",
"following_url": "https://api.github.com/users/Cyrilvallez/following{/other_user}",
"gists_url": "https://api.github.com/users/Cyrilvallez/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Cyrilvallez/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Cyrilvallez/subscriptions",
"organizations_url": "https://api.github.com/users/Cyrilvallez/orgs",
"repos_url": "https://api.github.com/users/Cyrilvallez/repos",
"events_url": "https://api.github.com/users/Cyrilvallez/events{/privacy}",
"received_events_url": "https://api.github.com/users/Cyrilvallez/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40249/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40249/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40248 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40248/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40248/comments | https://api.github.com/repos/huggingface/transformers/issues/40248/events | https://github.com/huggingface/transformers/pull/40248 | 3,330,454,122 | PR_kwDOCUB6oc6kGFEd | 40,248 | set inputs_embeds to None while generate to avoid audio encoder forward in generation process | {
"login": "BakerBunker",
"id": 17872844,
"node_id": "MDQ6VXNlcjE3ODcyODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/17872844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BakerBunker",
"html_url": "https://github.com/BakerBunker",
"followers_url": "https://api.github.com/users/BakerBunker/followers",
"following_url": "https://api.github.com/users/BakerBunker/following{/other_user}",
"gists_url": "https://api.github.com/users/BakerBunker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/BakerBunker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/BakerBunker/subscriptions",
"organizations_url": "https://api.github.com/users/BakerBunker/orgs",
"repos_url": "https://api.github.com/users/BakerBunker/repos",
"events_url": "https://api.github.com/users/BakerBunker/events{/privacy}",
"received_events_url": "https://api.github.com/users/BakerBunker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T12:13:04 | 2025-08-19T08:46:26 | 2025-08-19T08:45:57 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40248",
"html_url": "https://github.com/huggingface/transformers/pull/40248",
"diff_url": "https://github.com/huggingface/transformers/pull/40248.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40248.patch",
"merged_at": "2025-08-19T08:45:57"
} |
# What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change.
Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Avoid audio encoder forward in generation process.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
@vasqu @zucchini-nlp
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @
If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
Please tag fewer than 3 people.
Models:
- text models: @ArthurZucker
- vision models: @amyeroberts, @qubvel
- speech models: @eustlb
- graph models: @clefourrier
Library:
- flax: @gante and @Rocketknight1
- generate: @zucchini-nlp (visual-language models) or @gante (all others)
- pipelines: @Rocketknight1
- tensorflow: @gante and @Rocketknight1
- tokenizers: @ArthurZucker
- trainer: @zach-huggingface, @SunMarc and @qgallouedec
- chat templates: @Rocketknight1
Integrations:
- deepspeed: HF Trainer/Accelerate: @SunMarc @zach-huggingface
- ray/raytune: @richardliaw, @amogkam
- Big Model Inference: @SunMarc
- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
Documentation: @stevhliu
HF projects:
- accelerate: [different repo](https://github.com/huggingface/accelerate)
- datasets: [different repo](https://github.com/huggingface/datasets)
- diffusers: [different repo](https://github.com/huggingface/diffusers)
- rust tokenizers: [different repo](https://github.com/huggingface/tokenizers)
Maintained examples (not research project or legacy):
- Flax: @Rocketknight1
- PyTorch: See Models above and tag the person corresponding to the modality of the example.
- TensorFlow: @Rocketknight1
-->
| {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40248/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40248/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40247 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40247/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40247/comments | https://api.github.com/repos/huggingface/transformers/issues/40247/events | https://github.com/huggingface/transformers/pull/40247 | 3,330,378,290 | PR_kwDOCUB6oc6kF0_c | 40,247 | ⚠️ 🔴 Add ministral model | {
"login": "manueldeprada",
"id": 6536835,
"node_id": "MDQ6VXNlcjY1MzY4MzU=",
"avatar_url": "https://avatars.githubusercontent.com/u/6536835?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/manueldeprada",
"html_url": "https://github.com/manueldeprada",
"followers_url": "https://api.github.com/users/manueldeprada/followers",
"following_url": "https://api.github.com/users/manueldeprada/following{/other_user}",
"gists_url": "https://api.github.com/users/manueldeprada/gists{/gist_id}",
"starred_url": "https://api.github.com/users/manueldeprada/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/manueldeprada/subscriptions",
"organizations_url": "https://api.github.com/users/manueldeprada/orgs",
"repos_url": "https://api.github.com/users/manueldeprada/repos",
"events_url": "https://api.github.com/users/manueldeprada/events{/privacy}",
"received_events_url": "https://api.github.com/users/manueldeprada/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1843244711,
"node_id": "MDU6TGFiZWwxODQzMjQ0NzEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model",
"name": "New model",
"color": "fbca04",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2025-08-18T11:49:20 | 2025-09-11T08:30:39 | 2025-09-11T08:30:39 | CONTRIBUTOR | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40247",
"html_url": "https://github.com/huggingface/transformers/pull/40247",
"diff_url": "https://github.com/huggingface/transformers/pull/40247.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40247.patch",
"merged_at": "2025-09-11T08:30:39"
} | Coming from #39799, adding Ministral model to support its interleaved attention.
| {
"login": "manueldeprada",
"id": 6536835,
"node_id": "MDQ6VXNlcjY1MzY4MzU=",
"avatar_url": "https://avatars.githubusercontent.com/u/6536835?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/manueldeprada",
"html_url": "https://github.com/manueldeprada",
"followers_url": "https://api.github.com/users/manueldeprada/followers",
"following_url": "https://api.github.com/users/manueldeprada/following{/other_user}",
"gists_url": "https://api.github.com/users/manueldeprada/gists{/gist_id}",
"starred_url": "https://api.github.com/users/manueldeprada/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/manueldeprada/subscriptions",
"organizations_url": "https://api.github.com/users/manueldeprada/orgs",
"repos_url": "https://api.github.com/users/manueldeprada/repos",
"events_url": "https://api.github.com/users/manueldeprada/events{/privacy}",
"received_events_url": "https://api.github.com/users/manueldeprada/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40247/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40247/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40246 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40246/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40246/comments | https://api.github.com/repos/huggingface/transformers/issues/40246/events | https://github.com/huggingface/transformers/issues/40246 | 3,330,362,088 | I_kwDOCUB6oc7GgUro | 40,246 | MT5ForConditionalGeneration no longer importable after torch and transformers upgrade | {
"login": "MarcoDiMarek",
"id": 75703661,
"node_id": "MDQ6VXNlcjc1NzAzNjYx",
"avatar_url": "https://avatars.githubusercontent.com/u/75703661?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MarcoDiMarek",
"html_url": "https://github.com/MarcoDiMarek",
"followers_url": "https://api.github.com/users/MarcoDiMarek/followers",
"following_url": "https://api.github.com/users/MarcoDiMarek/following{/other_user}",
"gists_url": "https://api.github.com/users/MarcoDiMarek/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MarcoDiMarek/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MarcoDiMarek/subscriptions",
"organizations_url": "https://api.github.com/users/MarcoDiMarek/orgs",
"repos_url": "https://api.github.com/users/MarcoDiMarek/repos",
"events_url": "https://api.github.com/users/MarcoDiMarek/events{/privacy}",
"received_events_url": "https://api.github.com/users/MarcoDiMarek/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T11:44:17 | 2025-08-18T11:57:21 | 2025-08-18T11:56:58 | NONE | null | null | null | null | ### System Info
ModuleNotFoundError: Could not import module 'MT5ForConditionalGeneration'. Are this object's requirements defined correctly?
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
from transformers import MT5ForConditionalGeneration
Name: transformers
Version: 4.55.2
Name: torch
Version: 2.8.0
Platform tested: Windows
### Expected behavior
Import of the module that has existed for years.... | {
"login": "MarcoDiMarek",
"id": 75703661,
"node_id": "MDQ6VXNlcjc1NzAzNjYx",
"avatar_url": "https://avatars.githubusercontent.com/u/75703661?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MarcoDiMarek",
"html_url": "https://github.com/MarcoDiMarek",
"followers_url": "https://api.github.com/users/MarcoDiMarek/followers",
"following_url": "https://api.github.com/users/MarcoDiMarek/following{/other_user}",
"gists_url": "https://api.github.com/users/MarcoDiMarek/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MarcoDiMarek/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MarcoDiMarek/subscriptions",
"organizations_url": "https://api.github.com/users/MarcoDiMarek/orgs",
"repos_url": "https://api.github.com/users/MarcoDiMarek/repos",
"events_url": "https://api.github.com/users/MarcoDiMarek/events{/privacy}",
"received_events_url": "https://api.github.com/users/MarcoDiMarek/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40246/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40246/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40245 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40245/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40245/comments | https://api.github.com/repos/huggingface/transformers/issues/40245/events | https://github.com/huggingface/transformers/issues/40245 | 3,330,281,091 | I_kwDOCUB6oc7GgA6D | 40,245 | When elastic training is enabled, data may be lost or over-read. | {
"login": "zhengchenyu",
"id": 10381583,
"node_id": "MDQ6VXNlcjEwMzgxNTgz",
"avatar_url": "https://avatars.githubusercontent.com/u/10381583?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhengchenyu",
"html_url": "https://github.com/zhengchenyu",
"followers_url": "https://api.github.com/users/zhengchenyu/followers",
"following_url": "https://api.github.com/users/zhengchenyu/following{/other_user}",
"gists_url": "https://api.github.com/users/zhengchenyu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zhengchenyu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zhengchenyu/subscriptions",
"organizations_url": "https://api.github.com/users/zhengchenyu/orgs",
"repos_url": "https://api.github.com/users/zhengchenyu/repos",
"events_url": "https://api.github.com/users/zhengchenyu/events{/privacy}",
"received_events_url": "https://api.github.com/users/zhengchenyu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T11:17:41 | 2025-10-05T08:02:52 | 2025-10-05T08:02:52 | NONE | null | null | null | null | ### System Info
master
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder (such as GLUE/SQuAD, ...)
- [ ] My own task or dataset (give details below)
### Reproduction
When elastic training is enabled, data may be lost or over-read.
If one workers is trainning, and then save checkpoint, and then two workers continue to run from this checkpoint, data may be lost.
Here is a unit test to reproduce this. Firstly, one work read one batch, get [0, 1, 2, 3], then save checkpoint. Secondly, the two worker will continue from this checkpoint. The index 0 worker read one batch, will get [8, 9, 10,11], the index 1 worker read one batch, will get [12, 13, 14, 15]. We found that [4, 5, 6, 7] is missing.
```
dataset = list(range(16))
# 1 Read samples by two processes
## 1.1 Read samples by 0 index
batch_sampler = BatchSampler(dataset, batch_size=4, drop_last=False)
batch_sampler_0 = BatchSamplerShard(batch_sampler=batch_sampler, num_processes=1, process_index=0)
shard_dataloader_0 = DataLoaderShard(dataset=dataset, batch_sampler=batch_sampler_0)
shard_dataloader_0 = skip_first_batches(shard_dataloader_0, num_samples=0)
shard_dataloader_iter_0 = iter(shard_dataloader_0)
assert (next(shard_dataloader_iter_0).tolist() == [0, 1, 2, 3])
# 2 Continue to read samples
# 2.1 Read samples by 0 index with skip
batch_sampler = BatchSampler(dataset, batch_size=4, drop_last=False)
batch_sampler_0 = BatchSamplerShard(batch_sampler=batch_sampler, num_processes=2, process_index=0)
shard_dataloader_0 = DataLoaderShard(dataset=dataset, batch_sampler=batch_sampler_0)
shard_dataloader_0 = skip_first_batches(shard_dataloader_0, num_batches=1)
shard_dataloader_iter_0 = iter(shard_dataloader_0)
# Here [4, 5, 6, 7] is missing
self.assertEqual(next(shard_dataloader_iter_0).tolist(), [8, 9, 10, 11])
```
This is because we use global_steps to resume training. So sample is lost.
### Expected behavior
Maybe we should use samples_trained_in_current_epoch to resume. | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40245/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40245/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40244 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40244/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40244/comments | https://api.github.com/repos/huggingface/transformers/issues/40244/events | https://github.com/huggingface/transformers/pull/40244 | 3,330,245,564 | PR_kwDOCUB6oc6kFYhi | 40,244 | add-loftr-keypoints-to-map | {
"login": "alexdzm",
"id": 63930909,
"node_id": "MDQ6VXNlcjYzOTMwOTA5",
"avatar_url": "https://avatars.githubusercontent.com/u/63930909?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/alexdzm",
"html_url": "https://github.com/alexdzm",
"followers_url": "https://api.github.com/users/alexdzm/followers",
"following_url": "https://api.github.com/users/alexdzm/following{/other_user}",
"gists_url": "https://api.github.com/users/alexdzm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/alexdzm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/alexdzm/subscriptions",
"organizations_url": "https://api.github.com/users/alexdzm/orgs",
"repos_url": "https://api.github.com/users/alexdzm/repos",
"events_url": "https://api.github.com/users/alexdzm/events{/privacy}",
"received_events_url": "https://api.github.com/users/alexdzm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | open | false | null | [] | null | [] | 2025-08-18T11:05:20 | 2025-08-18T11:14:54 | null | NONE | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40244",
"html_url": "https://github.com/huggingface/transformers/pull/40244",
"diff_url": "https://github.com/huggingface/transformers/pull/40244.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40244.patch",
"merged_at": null
} | # What does this PR do?
The[ official demo for eloftr](https://huggingface.co/zju-community/matchanything_eloftr) does not work with the current transformers because `EfficientLoFTRForKeypointMatching` is not instantiated when the model is called. This fixes the problem by adding it to the model mapping.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#create-a-pull-request),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes? Here are the
[documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and
[here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@qubvel tagging you as you reviewed the pr that added this model
| null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40244/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40244/timeline | null | null | null | null | true | false |
https://api.github.com/repos/huggingface/transformers/issues/40243 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40243/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40243/comments | https://api.github.com/repos/huggingface/transformers/issues/40243/events | https://github.com/huggingface/transformers/pull/40243 | 3,330,187,534 | PR_kwDOCUB6oc6kFL6d | 40,243 | AMD scheduled CI ref env file | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T10:45:51 | 2025-08-18T20:23:31 | 2025-08-18T13:23:27 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40243",
"html_url": "https://github.com/huggingface/transformers/pull/40243",
"diff_url": "https://github.com/huggingface/transformers/pull/40243.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40243.patch",
"merged_at": "2025-08-18T13:23:27"
} | Some of our AMD runners require us to reference an env-file, some do not.
Added new input to the workflow to enable this functionality [here](https://github.com/huggingface/hf-workflows/pull/43). | {
"login": "ivarflakstad",
"id": 69173633,
"node_id": "MDQ6VXNlcjY5MTczNjMz",
"avatar_url": "https://avatars.githubusercontent.com/u/69173633?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ivarflakstad",
"html_url": "https://github.com/ivarflakstad",
"followers_url": "https://api.github.com/users/ivarflakstad/followers",
"following_url": "https://api.github.com/users/ivarflakstad/following{/other_user}",
"gists_url": "https://api.github.com/users/ivarflakstad/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ivarflakstad/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ivarflakstad/subscriptions",
"organizations_url": "https://api.github.com/users/ivarflakstad/orgs",
"repos_url": "https://api.github.com/users/ivarflakstad/repos",
"events_url": "https://api.github.com/users/ivarflakstad/events{/privacy}",
"received_events_url": "https://api.github.com/users/ivarflakstad/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40243/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40243/timeline | null | null | null | null | true | true |
https://api.github.com/repos/huggingface/transformers/issues/40242 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40242/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40242/comments | https://api.github.com/repos/huggingface/transformers/issues/40242/events | https://github.com/huggingface/transformers/issues/40242 | 3,329,967,711 | I_kwDOCUB6oc7Ge0Zf | 40,242 | All versions of Transformers after 4.52.4 break batch inference on MPS when padding_side = left with SmolVLM | {
"login": "Matteo-Omenetti",
"id": 43417658,
"node_id": "MDQ6VXNlcjQzNDE3NjU4",
"avatar_url": "https://avatars.githubusercontent.com/u/43417658?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Matteo-Omenetti",
"html_url": "https://github.com/Matteo-Omenetti",
"followers_url": "https://api.github.com/users/Matteo-Omenetti/followers",
"following_url": "https://api.github.com/users/Matteo-Omenetti/following{/other_user}",
"gists_url": "https://api.github.com/users/Matteo-Omenetti/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Matteo-Omenetti/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Matteo-Omenetti/subscriptions",
"organizations_url": "https://api.github.com/users/Matteo-Omenetti/orgs",
"repos_url": "https://api.github.com/users/Matteo-Omenetti/repos",
"events_url": "https://api.github.com/users/Matteo-Omenetti/events{/privacy}",
"received_events_url": "https://api.github.com/users/Matteo-Omenetti/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 3817266200,
"node_id": "MDU6TGFiZWwzODE3MjY2MjAw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": null
}
] | closed | false | null | [] | null | [] | 2025-08-18T09:39:16 | 2025-09-29T08:02:33 | 2025-09-29T08:02:33 | NONE | null | null | null | null | ### System Info
- `transformers` version: 4.55.2
- Platform: macOS-15.6-arm64-arm-64bit-Mach-O
- Python version: 3.13.3
- Huggingface_hub version: 0.34.4
- Safetensors version: 0.6.2
- Accelerate version: 1.9.0
- Accelerate config: not found
- DeepSpeed version: not installed
- PyTorch version (accelerator?): 2.7.1 (NA)
- Tensorflow version (GPU?): not installed (NA)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
- Using distributed or parallel set-up in script?: No
### Information
All versions of Transformers after 4.52.4 break batch inference on MPS when padding_side = left.
When padding_side = "left" and inputs of difference length are given such that padding is introduced, then the output for the shortest padded input is token 0 repeated for ever.
This seems to be happening only on MPS, setting device = CPU solves the issue.
### Reproduction
Minimal reproducer:
```python
from transformers import AutoModelForImageTextToText, AutoProcessor
from transformers.image_utils import load_image
DEVICE = "mps" # if device == "cpu", it works
model_name = "HuggingFaceTB/SmolVLM-256M-Instruct"
processor = AutoProcessor.from_pretrained(model_name)
processor.tokenizer.padding_side = "left" # if set to right, you get "wrong" output but not infinite loop of 0 token
model = AutoModelForImageTextToText.from_pretrained(model_name, device_map=DEVICE)
model.eval()
image_urls = [
"https://upload.wikimedia.org/wikipedia/commons/7/76/GazettedeFrance.jpg",
"https://upload.wikimedia.org/wikipedia/commons/1/15/Cat_August_2010-4.jpg",
]
images = [load_image(u) for u in image_urls]
def get_prompt(query):
messages = [
{
"role": "user",
"content": [{"type": "image"}, {"type": "text", "text": query}],
},
]
prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
return prompt
prompts = [get_prompt("OCR this image"), get_prompt("Describe in details the content of this picture.")]
inputs = processor(
text=prompts,
images=images,
return_tensors="pt",
padding=True
)
inputs = inputs.to(DEVICE)
gen_kwargs = dict(
max_new_tokens=1000 - inputs.input_ids.shape[1],
use_cache=True,
do_sample=False,
)
generated_ids = model.generate(**inputs, **gen_kwargs)
outputs = processor.batch_decode(
generated_ids[:, inputs.input_ids.shape[1] :], skip_special_tokens=False
)
for i, output in enumerate(outputs):
print(f"=========== Output {i} ===========")
print(output)
print("\n\n")
```
Output:
> =========== Output 0 ===========
<|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|>
> =========== Output 1 ===========
> The image depicts a cat lying on a white surface, likely a ledge or a wall. The cat is positioned in a relaxed manner, with its front paws slightly outstretched and its hind legs bent. The cat has a striped pattern on its fur, which is a common feature in many cat breeds. The stripes are horizontal and run from the tip of its tail to the base of its tail, creating a pattern that is both eye-catching and visually appealing.
>
> The cat's fur is light brown or black, and its eyes are a deep amber color, which is a common feature in
### Expected behavior
The model should produce the same output as on CPU. | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40242/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40242/timeline | null | completed | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | false | true |
https://api.github.com/repos/huggingface/transformers/issues/40241 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/40241/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/40241/comments | https://api.github.com/repos/huggingface/transformers/issues/40241/events | https://github.com/huggingface/transformers/pull/40241 | 3,329,932,354 | PR_kwDOCUB6oc6kEWz6 | 40,241 | [configuration] allow to overwrite kwargs from subconfigs | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2025-08-18T09:28:45 | 2025-08-22T11:31:26 | 2025-08-22T11:31:26 | MEMBER | null | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/40241",
"html_url": "https://github.com/huggingface/transformers/pull/40241",
"diff_url": "https://github.com/huggingface/transformers/pull/40241.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/40241.patch",
"merged_at": "2025-08-22T11:31:26"
} | # What does this PR do?
Fixes https://github.com/huggingface/transformers/issues/38258 and allows users to overwrite kwargs from subconfigs in the same way as we allow to overwrite in the main config. See below code for what has changed
```python
# Create a config that is same as the LLaVa-NeXT Mistral with smaller vocab size
# Before
config = AutoConfig.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf", text_config={"vocab_size": 900})
config.text_config.sliding_window
>>> 4096 # default value for sliding_window in Mistral config
config.text_config.vocab_size
>>> 900 # Vocab size was changed to user-defined kwarg
# After
config = AutoConfig.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf", text_config={"vocab_size": 900})
config.text_config.sliding_window
>>> None # sliding_window from loaded checkpoint
config.text_config.vocab_size
>>> 900 # Vocab size was changed to user-defined kwarg
``` | {
"login": "zucchini-nlp",
"id": 100715397,
"node_id": "U_kgDOBgDLhQ",
"avatar_url": "https://avatars.githubusercontent.com/u/100715397?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zucchini-nlp",
"html_url": "https://github.com/zucchini-nlp",
"followers_url": "https://api.github.com/users/zucchini-nlp/followers",
"following_url": "https://api.github.com/users/zucchini-nlp/following{/other_user}",
"gists_url": "https://api.github.com/users/zucchini-nlp/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zucchini-nlp/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zucchini-nlp/subscriptions",
"organizations_url": "https://api.github.com/users/zucchini-nlp/orgs",
"repos_url": "https://api.github.com/users/zucchini-nlp/repos",
"events_url": "https://api.github.com/users/zucchini-nlp/events{/privacy}",
"received_events_url": "https://api.github.com/users/zucchini-nlp/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/40241/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/40241/timeline | null | null | null | null | true | true |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.