url stringlengths 62 66 | repository_url stringclasses 1
value | labels_url stringlengths 76 80 | comments_url stringlengths 71 75 | events_url stringlengths 69 73 | html_url stringlengths 50 56 | id int64 377M 2.15B | node_id stringlengths 18 32 | number int64 1 29.2k | title stringlengths 1 487 | user dict | labels list | state stringclasses 2
values | locked bool 2
classes | assignee dict | assignees list | comments list | created_at int64 1.54k 1.71k | updated_at int64 1.54k 1.71k | closed_at int64 1.54k 1.71k ⌀ | author_association stringclasses 4
values | active_lock_reason stringclasses 2
values | body stringlengths 0 234k ⌀ | reactions dict | timeline_url stringlengths 71 75 | state_reason stringclasses 3
values | draft bool 2
classes | pull_request dict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/huggingface/transformers/issues/27748 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27748/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27748/comments | https://api.github.com/repos/huggingface/transformers/issues/27748/events | https://github.com/huggingface/transformers/pull/27748 | 2,014,816,598 | PR_kwDOCUB6oc5gkxlV | 27,748 | Deprecate `LegacyIndex` used in `RagRetriever` | {
"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/... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27748). All of your documentation changes will be reflected on that endpoint.",
"Hi @LysandreJik \r\n\r\nAfter #27776, this PR doesn't add much value. Let me know if we would like to close this or still keep this deprecatio... | 1,701 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
A first step to reduce the security issue of pickle for `RAG`.
In general, we can also ask users to set some environment variable to enable using the code block that contains `pickle.load` to have even better security. But that could be done in a follow up PR (which also applies to `Transf... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27748/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/27748/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27748",
"html_url": "https://github.com/huggingface/transformers/pull/27748",
"diff_url": "https://github.com/huggingface/transformers/pull/27748.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27748.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27747 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27747/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27747/comments | https://api.github.com/repos/huggingface/transformers/issues/27747/events | https://github.com/huggingface/transformers/issues/27747 | 2,014,620,199 | I_kwDOCUB6oc54FKon | 27,747 | List index out of range | {
"login": "MinnTrit",
"id": 151976884,
"node_id": "U_kgDOCQ77tA",
"avatar_url": "https://avatars.githubusercontent.com/u/151976884?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MinnTrit",
"html_url": "https://github.com/MinnTrit",
"followers_url": "https://api.github.com/users/MinnTrit/... | [] | closed | false | null | [] | [
"Hello 🤗 \r\nI would recommend you to try putting some breakpoints here and there to check what you are feeding to the model. \r\nAlso truncation and padding should be done with the ` truncation` and `padding` arguments of the ` __call__` method of the tokenizers. You should also set ` return_tensors = \"pt\" ` t... | 1,701 | 1,704 | 1,704 | NONE | null | ### System Info
Hi everyone, I'm currently having the issue with my code. This is my code:
```python
import transformers
from transformers import AutoTokenizer, RobertaForSequenceClassification
import pandas as pd
import numpy as np
import torch
df = pd.read_excel(r'D:\Data Intense\Level 2\Day9\comme... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27747/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/27747/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27746 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27746/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27746/comments | https://api.github.com/repos/huggingface/transformers/issues/27746/events | https://github.com/huggingface/transformers/pull/27746 | 2,014,522,949 | PR_kwDOCUB6oc5gjwsS | 27,746 | Move tensors to same device to enable IDEFICS naive MP training | {
"login": "willemsenbram",
"id": 22574774,
"node_id": "MDQ6VXNlcjIyNTc0Nzc0",
"avatar_url": "https://avatars.githubusercontent.com/u/22574774?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/willemsenbram",
"html_url": "https://github.com/willemsenbram",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"For the test I suggest to do it in a follow-up PR to test that for all architectures, through a test that runs on the daily CI"
] | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Fixes #27736
`logits`, `labels`, and `attention_mask` are expected to be on the same device.
This PR moves these tensors to the same device, making IDEFICS training compatible with naive model parallelism.
See also #22561
@amyeroberts @ArthurZucker @younesbelkada | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27746/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/27746/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27746",
"html_url": "https://github.com/huggingface/transformers/pull/27746",
"diff_url": "https://github.com/huggingface/transformers/pull/27746.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27746.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27745 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27745/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27745/comments | https://api.github.com/repos/huggingface/transformers/issues/27745/events | https://github.com/huggingface/transformers/pull/27745 | 2,014,464,762 | PR_kwDOCUB6oc5gjj6S | 27,745 | Enforce pin memory disabling when using cpu only | {
"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/... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27745). All of your documentation changes will be reflected on that endpoint."
] | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Fixes #26556
## 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#start-contributing-pull-requests)... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27745/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/27745/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27745",
"html_url": "https://github.com/huggingface/transformers/pull/27745",
"diff_url": "https://github.com/huggingface/transformers/pull/27745.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27745.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27744 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27744/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27744/comments | https://api.github.com/repos/huggingface/transformers/issues/27744/events | https://github.com/huggingface/transformers/issues/27744 | 2,014,404,231 | I_kwDOCUB6oc54EV6H | 27,744 | Trainer fails when using torchrun for distributed run of transformer model wrapped with PEFT | {
"login": "Ahmed-Roushdy",
"id": 68569076,
"node_id": "MDQ6VXNlcjY4NTY5MDc2",
"avatar_url": "https://avatars.githubusercontent.com/u/68569076?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Ahmed-Roushdy",
"html_url": "https://github.com/Ahmed-Roushdy",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md) a... | 1,701 | 1,707 | 1,707 | NONE | null | ### System Info
- `transformers` version: 4.33.3 [1/1854]
- Platform: Linux-5.8.0-59-generic-x86_64-with-glibc2.31
- Python version: 3.9.18
- Huggingface_hub version: 0.19.2
- Safetensors vers... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27744/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/27744/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27743 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27743/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27743/comments | https://api.github.com/repos/huggingface/transformers/issues/27743/events | https://github.com/huggingface/transformers/pull/27743 | 2,014,400,635 | PR_kwDOCUB6oc5gjVvs | 27,743 | restructure AMD scheduled CI | {
"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/... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27743). All of your documentation changes will be reflected on that endpoint."
] | 1,701 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
So far, the AMD scheduled CI is run as a single workflow with `mi210` and `mi250` both in it (each has ~500 jobs): see [here](https://github.com/huggingface/transformers/actions/runs/6999845840)
<img width="791" alt="Screenshot 2023-11-28 140000" src="https://github.com/huggingface/transfo... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27743/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/27743/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27743",
"html_url": "https://github.com/huggingface/transformers/pull/27743",
"diff_url": "https://github.com/huggingface/transformers/pull/27743.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27743.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27742 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27742/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27742/comments | https://api.github.com/repos/huggingface/transformers/issues/27742/events | https://github.com/huggingface/transformers/pull/27742 | 2,014,334,077 | PR_kwDOCUB6oc5gjHJV | 27,742 | Add Swinv2 backbone | {
"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/use... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27742). All of your documentation changes will be reflected on that endpoint."
] | 1,701 | 1,703 | 1,703 | CONTRIBUTOR | null | # What does this PR do?
This PR is a continuation of #25799. It adds `Swinv2Backbone`, such that it can be combined with the DPT framework (known as [MiDaS 3.1](https://github.com/isl-org/MiDaS/releases)).
It also required some updates to `modeling_swinv2.py` which are fully backwards compatible (all slow integra... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27742/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/27742/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27742",
"html_url": "https://github.com/huggingface/transformers/pull/27742",
"diff_url": "https://github.com/huggingface/transformers/pull/27742.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27742.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27741 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27741/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27741/comments | https://api.github.com/repos/huggingface/transformers/issues/27741/events | https://github.com/huggingface/transformers/issues/27741 | 2,014,328,953 | I_kwDOCUB6oc54EDh5 | 27,741 | FuyuForCausalLM.forward() got an unexpected keyword argument 'labels' | {
"login": "Nyandwi",
"id": 52796597,
"node_id": "MDQ6VXNlcjUyNzk2NTk3",
"avatar_url": "https://avatars.githubusercontent.com/u/52796597?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Nyandwi",
"html_url": "https://github.com/Nyandwi",
"followers_url": "https://api.github.com/users/Nyandw... | [] | closed | false | null | [] | [
"Hey! Yes this was updated recently, could you print\r\n```python \r\nimport transformers\r\nprint(transformers.__versions__)\r\n```\r\n",
"Thanks for quick reply. I have `4.36.0.dev0`. ",
"It is also the version I get when I install from the source again(now).",
"That's strange see [here](https://github.com/... | 1,701 | 1,701 | 1,701 | NONE | null | ### System Info
- `transformers` version: 4.36.0.dev0
- Platform: Linux-4.18.0-372.32.1.el8_6.x86_64-x86_64-with-glibc2.28
- Python version: 3.10.13
- Huggingface_hub version: 0.17.3
- Safetensors version: 0.4.0
- Accelerate version: 0.24.1
- Accelerate config: not found
- PyTorch version (GPU?): 2.1.0+cu118... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27741/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/27741/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27740 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27740/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27740/comments | https://api.github.com/repos/huggingface/transformers/issues/27740/events | https://github.com/huggingface/transformers/pull/27740 | 2,014,199,375 | PR_kwDOCUB6oc5gipx0 | 27,740 | Docs: Fix broken cross-references, i.e. `~transformer.` -> `~transformers.` | {
"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/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._",
"Gladly! Thank you for your work on the docs! It's always an underappreciated part of any OS project."
] | 1,701 | 1,701 | 1,701 | MEMBER | null | # What does this PR do?
I noticed that some of the docs use incorrect cross-references, e.g. here: https://huggingface.co/docs/transformers/v4.35.2/en/main_classes/trainer#transformers.Trainer.add_callback
Then I noticed these use `~transformer.foo` rather than `~transformers.foo`. I've replaced all `~transformer... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27740/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/27740/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27740",
"html_url": "https://github.com/huggingface/transformers/pull/27740",
"diff_url": "https://github.com/huggingface/transformers/pull/27740.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27740.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27739 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27739/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27739/comments | https://api.github.com/repos/huggingface/transformers/issues/27739/events | https://github.com/huggingface/transformers/issues/27739 | 2,014,160,509 | I_kwDOCUB6oc54DaZ9 | 27,739 | RuntimeError: Could not infer dtype of JpegImageFile | {
"login": "realbigi",
"id": 96737615,
"node_id": "U_kgDOBcQZTw",
"avatar_url": "https://avatars.githubusercontent.com/u/96737615?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/realbigi",
"html_url": "https://github.com/realbigi",
"followers_url": "https://api.github.com/users/realbigi/fo... | [] | closed | false | null | [] | [
"Hi @realbigi, thanks for raising an issue! \r\n\r\nThe error itself is arising because torch can't decode jpeg files. From the notebook, it's not immediately obvious what's happening, as the images should be opened as PIL.Image.Image. I can see there was an error reading in the dataset and some of the cells have b... | 1,701 | 1,704 | 1,704 | NONE | null | ### System Info
transformers, windows 11, python 3.11.6
I have been trying to replicate this https://huggingface.co/docs/transformers/tasks/image_classification image classification with PyTorch. But when I run the skript I get this error:
RuntimeError: Could not infer dtype of JpegImageFile
Does anyone have ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27739/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/27739/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27738 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27738/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27738/comments | https://api.github.com/repos/huggingface/transformers/issues/27738/events | https://github.com/huggingface/transformers/pull/27738 | 2,014,053,752 | PR_kwDOCUB6oc5giKMW | 27,738 | single word should be set to False | {
"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.c... | [] | closed | false | null | [] | [] | 1,701 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
Fixes qn issue reported on the hub:
```python
from transformers import AutoTokenizer, MT5Tokenizer
tokenizer = MT5Tokenizer.from_pretrained("google/mt5-base", extra_ids = 100)
```
(this is rarely tested) trigers this
```diff
# for legacy purpose, we keep this. Will be removed ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27738/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/27738/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27738",
"html_url": "https://github.com/huggingface/transformers/pull/27738",
"diff_url": "https://github.com/huggingface/transformers/pull/27738.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27738.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27737 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27737/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27737/comments | https://api.github.com/repos/huggingface/transformers/issues/27737/events | https://github.com/huggingface/transformers/issues/27737 | 2,013,567,783 | I_kwDOCUB6oc54BJsn | 27,737 | How to save the generated output of BarkModel to an npz file? | {
"login": "chet-chen",
"id": 16471378,
"node_id": "MDQ6VXNlcjE2NDcxMzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/16471378?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/chet-chen",
"html_url": "https://github.com/chet-chen",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"cc @ylacombe ",
"Hey @chet-chen, let me take a look!\r\nFrom the back of my mind, I think it requires to add something similar to `output_full = True` to BarkModel. It could be great that you do this if you add time, WDYT ?",
"@ylacombe I'm relatively new to Python, and I'm facing a challenge in converting the... | 1,701 | 1,704 | 1,704 | NONE | null | Hello there!
I'm using the BarkModel from Hugging Face Transformers and I'm wondering how to save the generated results to an npz file. I'd like to use these saved results as history prompts for the next generation.
In the [suno-ai/bark](https://github.com/suno-ai/bark) , when using the [`semantic_to_waveform`](h... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27737/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/27737/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27736 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27736/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27736/comments | https://api.github.com/repos/huggingface/transformers/issues/27736/events | https://github.com/huggingface/transformers/issues/27736 | 2,013,402,394 | I_kwDOCUB6oc54AhUa | 27,736 | RuntimeError(s) when attempting multi-GPU fine-tuning of IDEFICS with naive model parallelism | {
"login": "willemsenbram",
"id": 22574774,
"node_id": "MDQ6VXNlcjIyNTc0Nzc0",
"avatar_url": "https://avatars.githubusercontent.com/u/22574774?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/willemsenbram",
"html_url": "https://github.com/willemsenbram",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"Hi @willemsenbram - thanks for raising this issue! \r\n\r\nThe proposed solution sounds good to me - it's consistent with other model implementations. If you open a PR we'll be happy to review :) ",
"Hi @amyeroberts, thanks for checking! I've opened the [PR](https://github.com/huggingface/transformers/pull/27746... | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | ### System Info
- `transformers` version: 4.34.0
- Platform: Linux-5.4.0-148-generic-x86_64-with-glibc2.31
- Python version: 3.10.13
- Huggingface_hub version: 0.17.3
- Safetensors version: 0.4.0
- Accelerate version: 0.24.1
- Accelerate config: not found
- PyTorch version (GPU?): 2.1.0 (True)
- Tensorflow ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27736/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/27736/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27735 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27735/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27735/comments | https://api.github.com/repos/huggingface/transformers/issues/27735/events | https://github.com/huggingface/transformers/pull/27735 | 2,013,261,833 | PR_kwDOCUB6oc5gfdp6 | 27,735 | Adding SegGPT | {
"login": "EduardoPach",
"id": 69953243,
"node_id": "MDQ6VXNlcjY5OTUzMjQz",
"avatar_url": "https://avatars.githubusercontent.com/u/69953243?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/EduardoPach",
"html_url": "https://github.com/EduardoPach",
"followers_url": "https://api.github.com/... | [] | open | false | null | [] | [
"@NielsRogge just pinging you here",
"Hey @NielsRogge, could you take a look at the PR? Code quality is failing, but is unrelated and the documentation tests I don't know what is going on because it's failing on an example that it's not even in the `modeling_seggpt.py`",
"Looks like we're nearly there. Just on... | 1,701 | 1,708 | null | NONE | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27735/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/27735/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27735",
"html_url": "https://github.com/huggingface/transformers/pull/27735",
"diff_url": "https://github.com/huggingface/transformers/pull/27735.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27735.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27734 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27734/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27734/comments | https://api.github.com/repos/huggingface/transformers/issues/27734/events | https://github.com/huggingface/transformers/pull/27734 | 2,013,072,983 | PR_kwDOCUB6oc5gezhS | 27,734 | Deberta can now be exported to TorchScript | {
"login": "Szustarol",
"id": 61427290,
"node_id": "MDQ6VXNlcjYxNDI3Mjkw",
"avatar_url": "https://avatars.githubusercontent.com/u/61427290?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Szustarol",
"html_url": "https://github.com/Szustarol",
"followers_url": "https://api.github.com/users/... | [] | open | false | null | [] | [
"Great work @Szustarol ! I think this solution still pins the device for the traced model. One solution is to create an ONNX model instead. Perhaps you might have a better solution? ",
"Thanks for the response @demq, I would be more than happy to expand on my solution, however I am not quite sure if I get your s... | 1,701 | 1,707 | null | NONE | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27734/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/27734/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27734",
"html_url": "https://github.com/huggingface/transformers/pull/27734",
"diff_url": "https://github.com/huggingface/transformers/pull/27734.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27734.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27733 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27733/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27733/comments | https://api.github.com/repos/huggingface/transformers/issues/27733/events | https://github.com/huggingface/transformers/issues/27733 | 2,012,763,875 | I_kwDOCUB6oc53-Fbj | 27,733 | ZERO loss while finetuning Llama2 usin SFT trainer and the use of collator | {
"login": "Sosycs",
"id": 6597399,
"node_id": "MDQ6VXNlcjY1OTczOTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6597399?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Sosycs",
"html_url": "https://github.com/Sosycs",
"followers_url": "https://api.github.com/users/Sosycs/foll... | [] | closed | false | null | [] | [
"Hey 🤗 thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discuss.huggingface.co/) instead? I'm sure the community will be of help!\r\n\r\nThanks!",
"Sure!\r\nand Sorry for the inconvenience",
"This issue has been auto... | 1,701 | 1,704 | 1,704 | NONE | null | Hello everyone,
my code is:
```
response_template = "Answer: [/INST]"
collator = DataCollatorForCompletionOnlyLM(response_template=response_template_tokenized , tokenizer=tokenizer)
example = """<s>[INST] <<SYS>> Please select the correct answer from the given multiple Options based on the given Context: <</SY... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27733/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/27733/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27732 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27732/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27732/comments | https://api.github.com/repos/huggingface/transformers/issues/27732/events | https://github.com/huggingface/transformers/pull/27732 | 2,012,732,723 | PR_kwDOCUB6oc5gdo2g | 27,732 | Fix AMD Push CI not triggered | {
"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/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,701 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
#26940 introduced a bug, and AMD Push CI is not triggered in the past week.
See https://github.com/huggingface/transformers/actions/workflows/self-push-amd-mi210-caller.yml | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27732/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/27732/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27732",
"html_url": "https://github.com/huggingface/transformers/pull/27732",
"diff_url": "https://github.com/huggingface/transformers/pull/27732.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27732.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27731 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27731/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27731/comments | https://api.github.com/repos/huggingface/transformers/issues/27731/events | https://github.com/huggingface/transformers/pull/27731 | 2,012,630,069 | PR_kwDOCUB6oc5gdShK | 27,731 | Use proper weights name when logging in `PreTrainedModel.save_pretrained` | {
"login": "michaelbenayoun",
"id": 25418079,
"node_id": "MDQ6VXNlcjI1NDE4MDc5",
"avatar_url": "https://avatars.githubusercontent.com/u/25418079?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/michaelbenayoun",
"html_url": "https://github.com/michaelbenayoun",
"followers_url": "https://api... | [] | closed | false | null | [] | [
"Could you rebase and make sure CIs are green? 🤗 ",
"I think the issue is fixed in the current `main` branch now.",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27731). All of your documentation changes will be reflected on that endpoint. The docs are available unt... | 1,701 | 1,705 | 1,705 | MEMBER | null | # What does this PR do?
It seems that `WEIGHTS_NAME` even when `safe_serialization=True`. This PR fixes that by using `weights_name` which is defined as follows:
```
if not _hf_peft_config_loaded:
weights_name = SAFE_WEIGHTS_NAME if safe_serialization else WEIGHTS_NAME
weights_nam... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27731/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/27731/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27731",
"html_url": "https://github.com/huggingface/transformers/pull/27731",
"diff_url": "https://github.com/huggingface/transformers/pull/27731.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27731.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27730 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27730/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27730/comments | https://api.github.com/repos/huggingface/transformers/issues/27730/events | https://github.com/huggingface/transformers/issues/27730 | 2,012,579,385 | I_kwDOCUB6oc539YY5 | 27,730 | Add flag for easily finetuning heads / linear probing to AutoModelforSequenceClassification | {
"login": "0amp",
"id": 28636996,
"node_id": "MDQ6VXNlcjI4NjM2OTk2",
"avatar_url": "https://avatars.githubusercontent.com/u/28636996?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/0amp",
"html_url": "https://github.com/0amp",
"followers_url": "https://api.github.com/users/0amp/followers"... | [
{
"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 | [] | [] | 1,701 | 1,704 | null | NONE | null | ### Feature request
Previous work has shown that last layer linear probing is cheaper and often generalizes better than normal finetuning (see [1](https://arxiv.org/pdf/2202.10054.pdf)). I imagine this could be implemented as a flag to AutoModelforSequenceClassification so that only the last layer classification head ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27730/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/27730/timeline | reopened | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27729 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27729/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27729/comments | https://api.github.com/repos/huggingface/transformers/issues/27729/events | https://github.com/huggingface/transformers/pull/27729 | 2,012,556,720 | PR_kwDOCUB6oc5gdCbI | 27,729 | Improve forward signature test | {
"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/use... | [] | closed | false | null | [] | [] | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
As a follow-up of #27681, I'd like to make sure that every new backbone that is added to the library follows the same API. Hence, this PR extends the `test_forward_signature` test to make sure this is tested.
Rather than only checking the first keyword argument, it checks all of them. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27729/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/27729/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27729",
"html_url": "https://github.com/huggingface/transformers/pull/27729",
"diff_url": "https://github.com/huggingface/transformers/pull/27729.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27729.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27728 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27728/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27728/comments | https://api.github.com/repos/huggingface/transformers/issues/27728/events | https://github.com/huggingface/transformers/issues/27728 | 2,012,537,779 | I_kwDOCUB6oc539OOz | 27,728 | `hub_strategy`'s documentation for `checkpoint` option is wrong and misleading | {
"login": "omermazig",
"id": 95534441,
"node_id": "U_kgDOBbG9aQ",
"avatar_url": "https://avatars.githubusercontent.com/u/95534441?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/omermazig",
"html_url": "https://github.com/omermazig",
"followers_url": "https://api.github.com/users/omermazi... | [] | closed | false | null | [] | [
"A PR to the docs would be great, I also believe that if we use `save_total_limit`, it should take into account *not* deleting the one that has the best metric, and looking at the oldest scale to delete before that. I can look into this if it's a bit too complex on your end 😄 ",
"This issue has been automaticall... | 1,701 | 1,704 | 1,704 | NONE | null | ### System Info
The documentation for [hub_strategy](https://huggingface.co/docs/transformers/main_classes/trainer#transformers.TrainingArguments.hub_strategy) seems to be mistaken, because it stated:
> "checkpoint": like "every_save" but the latest checkpoint is also pushed in a subfolder named last-checkpoint, ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27728/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/27728/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27727 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27727/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27727/comments | https://api.github.com/repos/huggingface/transformers/issues/27727/events | https://github.com/huggingface/transformers/issues/27727 | 2,012,212,416 | I_kwDOCUB6oc537-zA | 27,727 | Streaming support in automatic-speech-recognition pipeline | {
"login": "CoderHam",
"id": 11223643,
"node_id": "MDQ6VXNlcjExMjIzNjQz",
"avatar_url": "https://avatars.githubusercontent.com/u/11223643?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/CoderHam",
"html_url": "https://github.com/CoderHam",
"followers_url": "https://api.github.com/users/Cod... | [] | closed | false | null | [] | [
"Hey! This seems to already be supported see this: #21196 does this not work for you problem? ",
"Shoot I wasn't aware this is already a feature. Yes I can make it work for me with the changes from that PR\r\n\r\nClosing this ticket but if there are official docs for the above I'd appreciate a pointer to it. Than... | 1,701 | 1,701 | 1,701 | NONE | null | ### Feature request
I'd like to request for the ability to stream back chunks of audio transcripts instead of having to wait for the entire audio to be processed. For real time use cases, it helps to be able to have small chunks of audio (10 seconds) and have their transcriptions returned as soon as they are available... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27727/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/27727/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27726 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27726/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27726/comments | https://api.github.com/repos/huggingface/transformers/issues/27726/events | https://github.com/huggingface/transformers/issues/27726 | 2,012,080,747 | I_kwDOCUB6oc537epr | 27,726 | How to load PixArtAlphaPipeline in 8bit? | {
"login": "FurkanGozukara",
"id": 19240467,
"node_id": "MDQ6VXNlcjE5MjQwNDY3",
"avatar_url": "https://avatars.githubusercontent.com/u/19240467?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/FurkanGozukara",
"html_url": "https://github.com/FurkanGozukara",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | [
"Hello, @sayakpaul and @SunMarc should know more about this. Untagging myself. Thank you.",
"You need to point to the exact line of code that breaks stuff and I am unable to see what that is. \r\n\r\n```python\r\ntext_encoder = T5EncoderModel.from_pretrained(\r\n \"PixArt-alpha/PixArt-XL-2-1024-MS\",\r\n su... | 1,701 | 1,704 | 1,704 | NONE | null | I know there is example but I couldn't make it work. I am trying to make an auto installer and gradio interface for Pix Art Alpha Pipeline so common people can install and use on their Windows PCs
Currently my below code working and I want to make it load in 8 bit is that possible?
```
if torch.cuda.is_available... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27726/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/27726/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27725 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27725/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27725/comments | https://api.github.com/repos/huggingface/transformers/issues/27725/events | https://github.com/huggingface/transformers/pull/27725 | 2,012,055,894 | PR_kwDOCUB6oc5gbVD8 | 27,725 | Fix oneformer instance segmentation RuntimeError | {
"login": "yhshin11",
"id": 5031800,
"node_id": "MDQ6VXNlcjUwMzE4MDA=",
"avatar_url": "https://avatars.githubusercontent.com/u/5031800?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yhshin11",
"html_url": "https://github.com/yhshin11",
"followers_url": "https://api.github.com/users/yhshi... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27725). All of your documentation changes will be reflected on that endpoint."
] | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Fixes RuntimeError when running on GPU due to inconsistent devices in OneFormerImageProcessor.
Same issue was fixed in [image_processing_mask2former.py](https://github.com/huggingface/transformers/blame/514de24abfd4416aeba6a6455ad5920f57f3567d/src/transformers/models/mask2former/image_proc... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27725/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/27725/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27725",
"html_url": "https://github.com/huggingface/transformers/pull/27725",
"diff_url": "https://github.com/huggingface/transformers/pull/27725.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27725.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27723 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27723/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27723/comments | https://api.github.com/repos/huggingface/transformers/issues/27723/events | https://github.com/huggingface/transformers/pull/27723 | 2,011,968,032 | PR_kwDOCUB6oc5gbBta | 27,723 | [Table Transformer] Convert more 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/use... | [] | open | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27723). All of your documentation changes will be reflected on that endpoint.",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please... | 1,701 | 1,707 | null | CONTRIBUTOR | null | # What does this PR do?
Microsoft released some new Table Transformer (TATR) checkpoints, I've converted them and pushed them to the hub:
* https://huggingface.co/microsoft/table-transformer-structure-recognition-v1.1-fin
* https://huggingface.co/microsoft/table-transformer-structure-recognition-v1.1-pub
* http... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27723/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/27723/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27723",
"html_url": "https://github.com/huggingface/transformers/pull/27723",
"diff_url": "https://github.com/huggingface/transformers/pull/27723.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27723.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27722 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27722/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27722/comments | https://api.github.com/repos/huggingface/transformers/issues/27722/events | https://github.com/huggingface/transformers/issues/27722 | 2,011,952,478 | I_kwDOCUB6oc536_Ve | 27,722 | Adding support for prompt lookup decoding (variant of assisted generation) | {
"login": "apoorvumang",
"id": 1957903,
"node_id": "MDQ6VXNlcjE5NTc5MDM=",
"avatar_url": "https://avatars.githubusercontent.com/u/1957903?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/apoorvumang",
"html_url": "https://github.com/apoorvumang",
"followers_url": "https://api.github.com/us... | [] | closed | false | null | [] | [
"tagging @gante since you have recently worked on lookahead decoding",
"Hi @apoorvumang 👋 \r\n\r\nFirst of all, thank you for creating this clever strategy and for sharing it openly! It's simple and elegant, which makes it really great.\r\n\r\nI've been thinking about it from a software point of view. The core f... | 1,701 | 1,705 | 1,705 | CONTRIBUTOR | null | ### Feature request
Recently proposed method prompt lookup decoding, which replaces the draft model with string matching in prompt
Code: https://github.com/apoorvumang/prompt-lookup-decoding
### Motivation
- The method gives significant speedups in input grounded tasks (2x-4x)
- Applicable to all decoder models,... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27722/reactions",
"total_count": 7,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 7,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27722/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27721 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27721/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27721/comments | https://api.github.com/repos/huggingface/transformers/issues/27721/events | https://github.com/huggingface/transformers/pull/27721 | 2,011,939,210 | PR_kwDOCUB6oc5ga7du | 27,721 | Log a warning in `TransfoXLTokenizer.__init__` | {
"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/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,701 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
I missed this part before merging #27607
https://github.com/huggingface/transformers/pull/27607#pullrequestreview-1747452403 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27721/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/27721/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27721",
"html_url": "https://github.com/huggingface/transformers/pull/27721",
"diff_url": "https://github.com/huggingface/transformers/pull/27721.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27721.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27720 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27720/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27720/comments | https://api.github.com/repos/huggingface/transformers/issues/27720/events | https://github.com/huggingface/transformers/pull/27720 | 2,011,907,053 | PR_kwDOCUB6oc5ga0hp | 27,720 | Add common processor tests | {
"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/use... | [] | open | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27720). All of your documentation changes will be reflected on that endpoint.",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please... | 1,701 | 1,706 | null | CONTRIBUTOR | null | # What does this PR do?
Multimodal processors currently don't have common tests. This PR aims to work towards having a common API for our multimodal processors, making sure they all have the same inputs and outputs (e.g. making sure text+vision processors accept `text` as first kwarg, then `images`, etc.).
As a f... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27720/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/27720/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27720",
"html_url": "https://github.com/huggingface/transformers/pull/27720",
"diff_url": "https://github.com/huggingface/transformers/pull/27720.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27720.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27719 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27719/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27719/comments | https://api.github.com/repos/huggingface/transformers/issues/27719/events | https://github.com/huggingface/transformers/issues/27719 | 2,011,845,097 | I_kwDOCUB6oc536lHp | 27,719 | Batch QuestionAnsweringPipeline prediction with different postprocess_params (e.g. max_answer_len) | {
"login": "KatHaruto",
"id": 74958594,
"node_id": "MDQ6VXNlcjc0OTU4NTk0",
"avatar_url": "https://avatars.githubusercontent.com/u/74958594?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/KatHaruto",
"html_url": "https://github.com/KatHaruto",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"Hey! 🤗 I don't think this is possible out of the box as it's a custom usage! Feel free to share your solution here or ask on the [forum](https://discuss.huggingface.co/)! "
] | 1,701 | 1,701 | 1,701 | NONE | null | Is it possible to change `QuestionAnsweringPipeline.postprocess` parameters such as `max_answer_len` for each question in batch prediction?
In my use case, `max_answer_len` is an important parameter for the output I want. And I want to use batch prediction to improve performance, but currently I can only set a singl... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27719/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/27719/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27718 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27718/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27718/comments | https://api.github.com/repos/huggingface/transformers/issues/27718/events | https://github.com/huggingface/transformers/pull/27718 | 2,011,737,709 | PR_kwDOCUB6oc5gaQWx | 27,718 | Add CogVLM | {
"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/use... | [] | closed | false | null | [] | [
"A cleaner implementation I'm working on is here: https://github.com/NielsRogge/transformers/tree/add_cogvlm_cleaner. It implements the model like llava, by adding the image tokens inside the model, rather than creating them in the processor class.",
"Closing this one in favor of the PR above."
] | 1,701 | 1,703 | 1,703 | CONTRIBUTOR | null | # What does this PR do?
This PR adds CogVLM natively into the Transformers library (it's already usable with `trust_remote_code=True`, but with this PR one can run it without the xformers, einops and triton dependencies).
To do:
- [x] remove triton dependency for rotary embeddings (or make `FastRotaryEmbedding... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27718/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/27718/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27718",
"html_url": "https://github.com/huggingface/transformers/pull/27718",
"diff_url": "https://github.com/huggingface/transformers/pull/27718.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27718.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27717 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27717/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27717/comments | https://api.github.com/repos/huggingface/transformers/issues/27717/events | https://github.com/huggingface/transformers/pull/27717 | 2,011,646,997 | PR_kwDOCUB6oc5gZ82R | 27,717 | [`NllbTokenizer`] refactor with added tokens decoder | {
"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.c... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27717). All of your documentation changes will be reflected on that endpoint.",
"@amyeroberts failing test is flaky, no idea why it's behaving that way will check main in the mean time",
"Regarding failing tests - there's... | 1,701 | 1,707 | 1,707 | COLLABORATOR | null | # What does this PR do?
Fixes #26497 by make the list of languages optional. By default these languages will be added, but otherwise by setting `additional_special_tokens = None` can be suppressed in both the fast and slow tokenizer. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27717/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/27717/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27717",
"html_url": "https://github.com/huggingface/transformers/pull/27717",
"diff_url": "https://github.com/huggingface/transformers/pull/27717.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27717.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27716 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27716/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27716/comments | https://api.github.com/repos/huggingface/transformers/issues/27716/events | https://github.com/huggingface/transformers/pull/27716 | 2,011,528,153 | PR_kwDOCUB6oc5gZiju | 27,716 | Fix unsupported setting of self._n_gpu in training_args on XPU devices | {
"login": "Liangliang-Ma",
"id": 17159645,
"node_id": "MDQ6VXNlcjE3MTU5NjQ1",
"avatar_url": "https://avatars.githubusercontent.com/u/17159645?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Liangliang-Ma",
"html_url": "https://github.com/Liangliang-Ma",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"> Hey! This seems to have been introduce in #25714, and thus I am not convinced that the fix is as simple as that! Would you mind also sharing a reproducer of the issue? Might be related to specific hard / soft versions\r\n\r\n Hi, @ArthurZucker! I have discussed with @abhilash1910 about the issue before this PR a... | 1,701 | 1,701 | 1,701 | CONTRIBUTOR | null | In current training_args, self._n_gpu is set to device_count on XPU device, which will cause crash on XPU devices.
In Trainer, if self.args.n_gpu greater than one, it will utilize torch.nn.DataParallel to wrap the model. But Ipex(intel_extension_for_pytorch) don't support DataParallel, while it suggests using DDP inst... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27716/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/27716/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27716",
"html_url": "https://github.com/huggingface/transformers/pull/27716",
"diff_url": "https://github.com/huggingface/transformers/pull/27716.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27716.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27715 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27715/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27715/comments | https://api.github.com/repos/huggingface/transformers/issues/27715/events | https://github.com/huggingface/transformers/pull/27715 | 2,011,342,832 | PR_kwDOCUB6oc5gY7QJ | 27,715 | Enhancing Code Readability and Maintainability with Simplified Activation Function Selection. | {
"login": "hi-sushanta",
"id": 93595990,
"node_id": "U_kgDOBZQpVg",
"avatar_url": "https://avatars.githubusercontent.com/u/93595990?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hi-sushanta",
"html_url": "https://github.com/hi-sushanta",
"followers_url": "https://api.github.com/users/hi... | [] | closed | false | null | [] | [
"Yes, because it is defined once globally and can be used throughout the entire codebase, improving code clarity and reducing redundancy.\r\n\r\n\r\n",
"Sorry 🤗 I mean that these are choices and we chose to have `nn.functional` rather than `F` which for us is less readable ",
"Thank you for your feedback. I un... | 1,701 | 1,704 | 1,704 | CONTRIBUTOR | null | This code optimization enhances code readability and maintainability by utilizing aliases, simplified activation function selection, and consistent function definitions.
## 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 ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27715/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/27715/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27715",
"html_url": "https://github.com/huggingface/transformers/pull/27715",
"diff_url": "https://github.com/huggingface/transformers/pull/27715.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27715.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27714 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27714/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27714/comments | https://api.github.com/repos/huggingface/transformers/issues/27714/events | https://github.com/huggingface/transformers/issues/27714 | 2,011,276,844 | I_kwDOCUB6oc534aYs | 27,714 | Overflow error: Can't convert negative int to unsigned [finetuning Bart] | {
"login": "matsuobasho",
"id": 13874772,
"node_id": "MDQ6VXNlcjEzODc0Nzcy",
"avatar_url": "https://avatars.githubusercontent.com/u/13874772?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/matsuobasho",
"html_url": "https://github.com/matsuobasho",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"Not a bug, merely has to do with `compute_metrics` not having a step to convert the -100 padded label values to `pad_token_id`. Adding that line gets rid of the error.\r\n```\r\ndef compute_metrics(eval_preds):\r\n preds, labels = eval_preds \r\n if isinstance(preds, tuple):\r\n preds = preds[0]\r\n... | 1,701 | 1,701 | 1,701 | NONE | null | ### System Info
Platform: Windows 10
Device: cpu
Python: 3.9.6
Transformers: 4.35.2
Datasets: 2.15.0
Torch: 2.1.1
### Who can help?
@muellerzr @pacman100
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [X] An officially supported task in the `ex... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27714/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/27714/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27713 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27713/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27713/comments | https://api.github.com/repos/huggingface/transformers/issues/27713/events | https://github.com/huggingface/transformers/pull/27713 | 2,011,261,063 | PR_kwDOCUB6oc5gYqML | 27,713 | Added Guarani Language Code to the Whisper Model | {
"login": "mfidabel",
"id": 18636378,
"node_id": "MDQ6VXNlcjE4NjM2Mzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/18636378?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mfidabel",
"html_url": "https://github.com/mfidabel",
"followers_url": "https://api.github.com/users/mfi... | [] | closed | false | null | [] | [
"Hey @mfidabel! The dictionary [`LANGUAGES`](https://github.com/huggingface/transformers/blob/9270ab082740a55344a851049a0b69673b6cbdc5/src/transformers/models/whisper/tokenization_whisper.py#L94) is defined as the languages that the pre-trained Whisper model supports. Since Guarani is not supported by the pre-train... | 1,701 | 1,707 | 1,704 | NONE | null | # What does this PR do?
This Pull Request adds the Guarani Language Code to the whisper tokenizer so it can be fine-tuned for the Guaraní Language.
The change is so small that it shouldn't break anything.
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27713/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/27713/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27713",
"html_url": "https://github.com/huggingface/transformers/pull/27713",
"diff_url": "https://github.com/huggingface/transformers/pull/27713.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27713.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27712 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27712/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27712/comments | https://api.github.com/repos/huggingface/transformers/issues/27712/events | https://github.com/huggingface/transformers/issues/27712 | 2,011,204,646 | I_kwDOCUB6oc534Iwm | 27,712 | Add support for llama.cpp | {
"login": "oobabooga",
"id": 112222186,
"node_id": "U_kgDOBrBf6g",
"avatar_url": "https://avatars.githubusercontent.com/u/112222186?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/oobabooga",
"html_url": "https://github.com/oobabooga",
"followers_url": "https://api.github.com/users/oobabo... | [] | open | false | null | [] | [
"Hi @oobabooga !\r\nApologies for my late reply \r\nIn general we are very interested in adding new quantization schemes in HF transformers. Currently, we're waiting to merge https://github.com/huggingface/transformers/pull/26610 in order to make the support for new quantization methods easier for anyone in the fut... | 1,701 | 1,706 | null | CONTRIBUTOR | null | ### Feature request
I would like to request [llama.cpp](https://github.com/ggerganov/llama.cpp) as a new model backend in the transformers library.
### Motivation
llama.cpp offers:
1) Excellent performance in scenarios where memory bandwidth is an issue, namely CPU inference and GPU + CPU inference.
2) Sup... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27712/reactions",
"total_count": 9,
"+1": 9,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27712/timeline | null | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27711 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27711/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27711/comments | https://api.github.com/repos/huggingface/transformers/issues/27711/events | https://github.com/huggingface/transformers/issues/27711 | 2,011,195,770 | I_kwDOCUB6oc534Gl6 | 27,711 | Inquiry about the difference between two Approaches to Mask Infilling using BART Model in the official document | {
"login": "Hyfred",
"id": 38806779,
"node_id": "MDQ6VXNlcjM4ODA2Nzc5",
"avatar_url": "https://avatars.githubusercontent.com/u/38806779?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Hyfred",
"html_url": "https://github.com/Hyfred",
"followers_url": "https://api.github.com/users/Hyfred/fo... | [] | closed | false | null | [] | [
"Hey! Sorry I didn't really have time to dig and must say that I didn't know the generate pipeline could fill up the hole for you! \r\nThe `beam-search` algorithm is used for that, which for me would mean that you explore more with this technique. Otherwise it should be similar 😉 ",
"By default the generate meth... | 1,701 | 1,705 | 1,705 | NONE | null | I came across two different methods of using the BART model for mask infilling at this URL: https://huggingface.co/docs/transformers/main/en/model_doc/bart#bart
The first method involves taking the logit corresponding to the mask word in the decoding last hidden representation and using softmax to directly predict w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27711/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/27711/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27710 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27710/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27710/comments | https://api.github.com/repos/huggingface/transformers/issues/27710/events | https://github.com/huggingface/transformers/issues/27710 | 2,011,127,681 | I_kwDOCUB6oc5331-B | 27,710 | Resume training with deepspeed resets learning rate | {
"login": "BiEchi",
"id": 60613238,
"node_id": "MDQ6VXNlcjYwNjEzMjM4",
"avatar_url": "https://avatars.githubusercontent.com/u/60613238?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BiEchi",
"html_url": "https://github.com/BiEchi",
"followers_url": "https://api.github.com/users/BiEchi/fo... | [] | closed | false | null | [] | [
"Solved, we should add the specification for scheduler in the config json for deepspeed as well."
] | 1,701 | 1,701 | 1,701 | NONE | null | ### System Info
- `transformers` version: 4.32.0.dev0
- Platform: Linux-3.10.0-1160.31.1.el7.x86_64-x86_64-with-glibc2.17
- Python version: 3.9.0
- Huggingface_hub version: 0.16.4
- Safetensors version: 0.3.1
- Accelerate version: 0.21.0
- Accelerate config: not found
- PyTorch version (GPU?): 2.0.1+cu117 (F... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27710/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/27710/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27709 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27709/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27709/comments | https://api.github.com/repos/huggingface/transformers/issues/27709/events | https://github.com/huggingface/transformers/pull/27709 | 2,011,104,926 | PR_kwDOCUB6oc5gYLFL | 27,709 | [`from_pretrained`] Make from_pretrained fast again | {
"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.c... | [] | closed | false | null | [] | [
"- explicit overwrite breaks fx and is not that faster \r\n- Non initialized is not always zeros (failing tests) just make sure it’s not initialized ",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27709). All of your documentation changes will be reflected on that end... | 1,701 | 1,702 | 1,702 | COLLABORATOR | null | # what does this PR do
Skips all layer initialization when loading from pretrained without accelerate.
From ~20seconds to 5 seconds for a 7B model like llama.
The Weights are effectively initialized in « init_weights_ » of the pretrained method. All internal calls are skipped
- [x] Check that if a linear la... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27709/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/27709/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27709",
"html_url": "https://github.com/huggingface/transformers/pull/27709",
"diff_url": "https://github.com/huggingface/transformers/pull/27709.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27709.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27708 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27708/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27708/comments | https://api.github.com/repos/huggingface/transformers/issues/27708/events | https://github.com/huggingface/transformers/issues/27708 | 2,010,989,582 | I_kwDOCUB6oc533UQO | 27,708 | `load_in_4bit=True` works only with models in `safetensors` format | {
"login": "danielkorat",
"id": 32893314,
"node_id": "MDQ6VXNlcjMyODkzMzE0",
"avatar_url": "https://avatars.githubusercontent.com/u/32893314?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/danielkorat",
"html_url": "https://github.com/danielkorat",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"Hey! Thanks for reporting. I would not be surprised that `safetensors` supports loading each parameters indivudally, casting them which is not supported by torch. I think that using smaller shards is usually what we do in this case, feel free to use: https://huggingface.co/ybelkada/Mistral-7B-v0.1-bf16-sharded",
... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | ### System Info
Hi,
When testing on `Google Colab (Free Tier T4 GPU)`, this code crashes with RAM OOM [(Notebook)](https://colab.research.google.com/drive/1zAzdcH_KRQuc_0zWBEzYuaV1h4ERgzPy?usp=sharing):
```python
AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", load_in_4bit=True, device_map="... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27708/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/27708/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27707 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27707/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27707/comments | https://api.github.com/repos/huggingface/transformers/issues/27707/events | https://github.com/huggingface/transformers/issues/27707 | 2,010,981,157 | I_kwDOCUB6oc533SMl | 27,707 | When labels is integer instead of float, the training crashes in MPS / CUDA. Could this be validated and fail with a better error higher up the stack ? | {
"login": "przem8k",
"id": 1824302,
"node_id": "MDQ6VXNlcjE4MjQzMDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1824302?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/przem8k",
"html_url": "https://github.com/przem8k",
"followers_url": "https://api.github.com/users/przem8k/... | [
{
"id": 5616426447,
"node_id": "LA_kwDOCUB6oc8AAAABTsPdzw",
"url": "https://api.github.com/repos/huggingface/transformers/labels/solved",
"name": "solved",
"color": "B1D6DC",
"default": false,
"description": ""
}
] | closed | false | null | [] | [
"Hey! Thanks for raising the issue! 🤗 \r\nIt's kind of hard to have architecture specific errors for something that works very well anywhere else! specifically given that it makes more sense for labels to be int rather than floats 😅 So not really in favor of this\r\n",
"Hi @ArthurZucker , thanks for taking the ... | 1,700 | 1,705 | 1,704 | NONE | null | ### System Info
- `transformers` version: 4.31.0
- Platform: macOS-14.1.1-arm64-arm-64bit
- Python version: 3.11.4
- Huggingface_hub version: 0.16.4
- Safetensors version: 0.3.1
- Accelerate version: 0.21.0
- Accelerate config: not found
- PyTorch version (GPU?): 2.1.0.dev20230804 (False)
- Tensorflow ver... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27707/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/27707/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27706 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27706/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27706/comments | https://api.github.com/repos/huggingface/transformers/issues/27706/events | https://github.com/huggingface/transformers/issues/27706 | 2,010,978,026 | I_kwDOCUB6oc533Rbq | 27,706 | implement TemplateConstraints in class transformers.Constraint | {
"login": "MrzEsma",
"id": 55921249,
"node_id": "MDQ6VXNlcjU1OTIxMjQ5",
"avatar_url": "https://avatars.githubusercontent.com/u/55921249?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MrzEsma",
"html_url": "https://github.com/MrzEsma",
"followers_url": "https://api.github.com/users/MrzEsm... | [
{
"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 | [] | [
"Hey! This might be implemented if the community is found needing this feature, and if it makes sens in the overall design of the API so I can't give you any estimate of the time this might take!\r\nLeaving this issue open and pinging @gante ! ",
"Hello @gante,\r\n\r\nI saw that @ArthurZucker brought you into the... | 1,700 | 1,704 | null | NONE | null | ### Feature request
Hello,
I recently came across your [blog post](https://huggingface.co/blog/constrained-beam-search) on constrained beam search and I am thrilled to see such implementations being made available. They are quite impressive!
Currently, I find myself in need of the feature `TemplateConstraints` I... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27706/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27706/timeline | null | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27705 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27705/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27705/comments | https://api.github.com/repos/huggingface/transformers/issues/27705/events | https://github.com/huggingface/transformers/issues/27705 | 2,010,941,935 | I_kwDOCUB6oc533Inv | 27,705 | Bug in the code of owlv2 algorithm | {
"login": "zhongwenkun886",
"id": 42397285,
"node_id": "MDQ6VXNlcjQyMzk3Mjg1",
"avatar_url": "https://avatars.githubusercontent.com/u/42397285?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhongwenkun886",
"html_url": "https://github.com/zhongwenkun886",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | [
"See #27205, which will be fixed by #27698. See also my [demo notebook](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/OWLv2/Zero_and_one_shot_object_detection_with_OWLv2.ipynb), you need to provide the `target_sizes` of the padded image rather than the original one for visualization.",
"thanks"... | 1,700 | 1,701 | 1,701 | NONE | null | ### System Info
python, linux
### 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
The p... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27705/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/27705/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27704 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27704/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27704/comments | https://api.github.com/repos/huggingface/transformers/issues/27704/events | https://github.com/huggingface/transformers/issues/27704 | 2,010,854,927 | I_kwDOCUB6oc532zYP | 27,704 | Stopping criteria does not work for Llama-2-13B | {
"login": "Eichhof",
"id": 6844011,
"node_id": "MDQ6VXNlcjY4NDQwMTE=",
"avatar_url": "https://avatars.githubusercontent.com/u/6844011?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Eichhof",
"html_url": "https://github.com/Eichhof",
"followers_url": "https://api.github.com/users/Eichhof/... | [] | closed | false | null | [] | [
"Hey! 🤗 \r\nI don't have access to `StoppingCriteriaSub` (missing form the reproducer) but this is very similar to #23852, and #26959, which most probably has the answers you are looking for. \r\nNow what you need to check thoroughly is not the strings that are decoder, but the token ids that you feed to logit pr... | 1,700 | 1,701 | 1,701 | NONE | null | ### System Info
- `transformers` version: 4.35.0
- Platform: Linux-5.15.0-89-generic-x86_64-with-glibc2.35
- Python version: 3.9.0
- Huggingface_hub version: 0.16.4
- Safetensors version: 0.4.0
- Accelerate version: 0.24.1
- Accelerate config: not found
- PyTorch version (GPU?): 2.1.0+cu121 (True)
- Tensorflo... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27704/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/27704/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27703 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27703/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27703/comments | https://api.github.com/repos/huggingface/transformers/issues/27703/events | https://github.com/huggingface/transformers/issues/27703 | 2,010,735,754 | I_kwDOCUB6oc532WSK | 27,703 | speech recognition with speecht5 | {
"login": "poojitharamachandra",
"id": 39840406,
"node_id": "MDQ6VXNlcjM5ODQwNDA2",
"avatar_url": "https://avatars.githubusercontent.com/u/39840406?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/poojitharamachandra",
"html_url": "https://github.com/poojitharamachandra",
"followers_url": ... | [] | closed | false | null | [] | [
"Hey! Seems like you are not using an officially shared snippet / an external library (call to `sd.rec`) to make sure we can help you, would you mind sharing the full snippet? \r\n",
"```python\r\nimport sounddevice as sd\r\nfrom transformers import SpeechT5Processor, SpeechT5ForTextToSpeech, SpeechT5ForSpeechToT... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
```python
processor = SpeechT5Processor.from_pretrained("microsoftt5_tts")
model = SpeechT5ForSpeechToText.from_pretrained("microsoftt5_tts")
duration = 10
sampling_rate = 16000
audio = sd.rec(int(sampling_rate * duration), samplerate=sampling_rate, channels=1)
input_features = processor(audio=... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27703/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/27703/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27702 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27702/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27702/comments | https://api.github.com/repos/huggingface/transformers/issues/27702/events | https://github.com/huggingface/transformers/issues/27702 | 2,010,524,705 | I_kwDOCUB6oc531iwh | 27,702 | Issue with Fine-tuning LLM for Classification | {
"login": "NickL77",
"id": 8673939,
"node_id": "MDQ6VXNlcjg2NzM5Mzk=",
"avatar_url": "https://avatars.githubusercontent.com/u/8673939?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/NickL77",
"html_url": "https://github.com/NickL77",
"followers_url": "https://api.github.com/users/NickL77/... | [] | closed | false | null | [] | [
"Hey 🤗 thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discuss.huggingface.co/) instead?\r\n\r\nRegarding quantization, you cannot train a fully quantized model it's just not supported (not supported by any library AFAI... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
```
- `transformers` version: 4.34.0
- Platform: Linux-6.5.4-76060504-generic-x86_64-with-glibc2.35
- Python version: 3.10.12
- Huggingface_hub version: 0.19.4
- Safetensors version: 0.4.0
- Accelerate version: 0.23.0
- Accelerate config: not found
- PyTorch version (GPU?): 2.1.0+cu121 (True... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27702/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/27702/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27701 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27701/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27701/comments | https://api.github.com/repos/huggingface/transformers/issues/27701/events | https://github.com/huggingface/transformers/issues/27701 | 2,010,509,759 | I_kwDOCUB6oc531fG_ | 27,701 | type object 'OPTDecoder' has no attribute '_prepare_decoder_attention_mask'. | {
"login": "muzi0111",
"id": 151991120,
"node_id": "U_kgDOCQ8zUA",
"avatar_url": "https://avatars.githubusercontent.com/u/151991120?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/muzi0111",
"html_url": "https://github.com/muzi0111",
"followers_url": "https://api.github.com/users/muzi0111/... | [] | closed | false | null | [] | [
"Hey! I think that version 4.34.0 should work. This attribute was removed in #27086 and it was a private method so not a breaking change",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlea... | 1,700 | 1,704 | 1,704 | NONE | null |
I'm currently facing an issue as follows: which version of Transformers should I install? The latest version does not seem to have this configuration option. I'm getting an AttributeError: type object 'OPTDecoder' has no attribute '_prepare_decoder_attention_mask'. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27701/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/27701/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27700 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27700/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27700/comments | https://api.github.com/repos/huggingface/transformers/issues/27700/events | https://github.com/huggingface/transformers/pull/27700 | 2,010,450,877 | PR_kwDOCUB6oc5gWIvf | 27,700 | Fix precision errors from casting rotary parameters to FP16 with AMP | {
"login": "kevinhu",
"id": 6051736,
"node_id": "MDQ6VXNlcjYwNTE3MzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/6051736?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kevinhu",
"html_url": "https://github.com/kevinhu",
"followers_url": "https://api.github.com/users/kevinhu/... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27700). All of your documentation changes will be reflected on that endpoint.",
"FYI @gante and @Rocketknight1 if we see anything failing. I ran slow tests locally and it was all good"
] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
When training with AMP, using `einsum` to multiply `t` and `self.inv_freq` will introduce precision errors because it casts the result to FP16. This can be avoided by using `torch.outer` instead, as originally mentioned here: https://github.com/Dao-AILab/flash-attention/blob/2c3baba4a63c4007c8a... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27700/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/27700/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27700",
"html_url": "https://github.com/huggingface/transformers/pull/27700",
"diff_url": "https://github.com/huggingface/transformers/pull/27700.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27700.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27699 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27699/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27699/comments | https://api.github.com/repos/huggingface/transformers/issues/27699/events | https://github.com/huggingface/transformers/issues/27699 | 2,010,325,647 | I_kwDOCUB6oc530yKP | 27,699 | TypeError: Llama.create_completion() got an unexpected keyword argument 'min_p' | {
"login": "mclassen",
"id": 711016,
"node_id": "MDQ6VXNlcjcxMTAxNg==",
"avatar_url": "https://avatars.githubusercontent.com/u/711016?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mclassen",
"html_url": "https://github.com/mclassen",
"followers_url": "https://api.github.com/users/mclasse... | [] | closed | false | null | [] | [
"Me to, just getting started will follow ",
"Run the set up again. when you updated there was a new requirements.txt. I had the same issue and running the setup again fixed it right up.",
"I've also been getting this error for several days now. Tried resetting the pod, creating new pods with different gpus or d... | 1,700 | 1,705 | 1,705 | NONE | null | ### System Info
I should have the latest version.
When executing a query against a GGUF model, I get this error:
Traceback (most recent call last):
File "/workspace/text-generation-webui/modules/callbacks.py", line 57, in gentask
ret = self.mfunc(callback=_callback, *args, **self.kwargs)
File "/workspac... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27699/reactions",
"total_count": 2,
"+1": 1,
"-1": 0,
"laugh": 1,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27699/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27698 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27698/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27698/comments | https://api.github.com/repos/huggingface/transformers/issues/27698/events | https://github.com/huggingface/transformers/pull/27698 | 2,010,168,139 | PR_kwDOCUB6oc5gVRQv | 27,698 | Fix owlv2 code snippet | {
"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/use... | [] | closed | false | null | [] | [] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Fixes #27205 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27698/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/27698/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27698",
"html_url": "https://github.com/huggingface/transformers/pull/27698",
"diff_url": "https://github.com/huggingface/transformers/pull/27698.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27698.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27697 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27697/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27697/comments | https://api.github.com/repos/huggingface/transformers/issues/27697/events | https://github.com/huggingface/transformers/issues/27697 | 2,010,076,389 | I_kwDOCUB6oc53z1Tl | 27,697 | using SFT for finetuning Llama2, TypeError: TextEncodeInput must be Union[TextInputSequence, Tuple[InputSequence, InputSequence]] | {
"login": "Sosycs",
"id": 6597399,
"node_id": "MDQ6VXNlcjY1OTczOTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6597399?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Sosycs",
"html_url": "https://github.com/Sosycs",
"followers_url": "https://api.github.com/users/Sosycs/foll... | [] | closed | false | null | [] | [
"Hi @Sosycs, I'm thoroughly impressed with your code's performance on my local machine. I strongly recommend reviewing the package version to ensure compatibility with the latest updates \r\n\r\n**Package-Version:**\r\n> *Transformers:* 4.36.0.dev0\r\n> *Trl :* 0.7.4\r\n\r\n**My-Code:**\r\n```\r\nfrom transformers... | 1,700 | 1,704 | 1,704 | NONE | null | Hello,
I am experiencing the following error:
```
You're using a LlamaTokenizerFast tokenizer. Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the text followed by a call to the `pad` method to get a padded encoding.
-------------------------------------... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27697/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/27697/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27696 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27696/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27696/comments | https://api.github.com/repos/huggingface/transformers/issues/27696/events | https://github.com/huggingface/transformers/pull/27696 | 2,010,011,896 | PR_kwDOCUB6oc5gUuoL | 27,696 | Fix Past CI | {
"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/... | [] | closed | false | null | [] | [] | 1,700 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
Since mid-November, we have hundreds of following failures in each past CI
> (line 1108) NotImplementedError: Loading a dataset cached in a LocalFileSystem is not supported.
It is caused by `fsspec==2023.10.0` with an old `datasets`.
This PR just updates `datasets` (at CI runtime) t... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27696/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/27696/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27696",
"html_url": "https://github.com/huggingface/transformers/pull/27696",
"diff_url": "https://github.com/huggingface/transformers/pull/27696.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27696.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27695 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27695/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27695/comments | https://api.github.com/repos/huggingface/transformers/issues/27695/events | https://github.com/huggingface/transformers/pull/27695 | 2,009,945,094 | PR_kwDOCUB6oc5gUgDr | 27,695 | Fix `TVPModelTest` | {
"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/... | [] | closed | false | null | [] | [] | 1,700 | 1,700 | 1,700 | COLLABORATOR | null | # What does this PR do?
Just device issue. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27695/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/27695/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27695",
"html_url": "https://github.com/huggingface/transformers/pull/27695",
"diff_url": "https://github.com/huggingface/transformers/pull/27695.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27695.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27694 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27694/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27694/comments | https://api.github.com/repos/huggingface/transformers/issues/27694/events | https://github.com/huggingface/transformers/pull/27694 | 2,009,916,536 | PR_kwDOCUB6oc5gUZ6O | 27,694 | Introduce SegGPT model | {
"login": "raghavanone",
"id": 115454562,
"node_id": "U_kgDOBuGyYg",
"avatar_url": "https://avatars.githubusercontent.com/u/115454562?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/raghavanone",
"html_url": "https://github.com/raghavanone",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"cc @amyeroberts and @niels",
"Hi @raghavanone - thanks for opening this PR! Let us know when it's ready for review ",
"@raghavanone Hey, I also was doing the implementation of `SegGPT` to add to the library (you can find my work here https://github.com/huggingface/transformers/pull/27735) I believe my current ... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | Introduce SegGPT model.
#27514 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27694/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/27694/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27694",
"html_url": "https://github.com/huggingface/transformers/pull/27694",
"diff_url": "https://github.com/huggingface/transformers/pull/27694.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27694.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27693 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27693/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27693/comments | https://api.github.com/repos/huggingface/transformers/issues/27693/events | https://github.com/huggingface/transformers/pull/27693 | 2,009,819,879 | PR_kwDOCUB6oc5gUE04 | 27,693 | Trigger corresponding pipeline tests if `tests/utils/tiny_model_summary.json` is modified | {
"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/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,700 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
Trigger corresponding pipeline tests if `tests/utils/tiny_model_summary.json` is modified.
It happened several times that we merged a PR and pipeline testing failed, because the CI triggered on those PRs didn't cover all files (as `tests/utils/tiny_model_summary.json` is not a python file) | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27693/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/27693/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27693",
"html_url": "https://github.com/huggingface/transformers/pull/27693",
"diff_url": "https://github.com/huggingface/transformers/pull/27693.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27693.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27692 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27692/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27692/comments | https://api.github.com/repos/huggingface/transformers/issues/27692/events | https://github.com/huggingface/transformers/issues/27692 | 2,009,677,525 | I_kwDOCUB6oc53yT7V | 27,692 | RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:1 and cuda:0! (when checking argument for argument weight in method wrapper_CUDA__native_layer_norm) | {
"login": "nirdoshrawal009",
"id": 102213412,
"node_id": "U_kgDOBhenJA",
"avatar_url": "https://avatars.githubusercontent.com/u/102213412?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nirdoshrawal009",
"html_url": "https://github.com/nirdoshrawal009",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"Hi @nirdoshrawal009, can you provide a minimal reproducer ? Thanks ",
"Meet the same problem, here is a minimal reproducer of mine:\r\n\r\n```python\r\nfrom transformers import pipeline\r\n\r\n# In my case, audio is passed through HTTP requests\r\ndef process(audio):\r\n\tpipe = pipeline(\"automatic-speech-recog... | 1,700 | 1,703 | 1,703 | NONE | null | ### System Info
I want to finetune Falcon 7b models using SFTTrainer from Transformers library.
I have set the device_map = 'auto' while loading the model and cuda_visible_devices = '0,1'
But getting this error.
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modi... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27692/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/27692/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27691 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27691/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27691/comments | https://api.github.com/repos/huggingface/transformers/issues/27691/events | https://github.com/huggingface/transformers/pull/27691 | 2,009,497,162 | PR_kwDOCUB6oc5gS-NX | 27,691 | Reorder the code on the Hub to explicit that sharing on the Hub isn't a requirement | {
"login": "LysandreJik",
"id": 30755778,
"node_id": "MDQ6VXNlcjMwNzU1Nzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LysandreJik",
"html_url": "https://github.com/LysandreJik",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,700 | 1,701 | 1,701 | MEMBER | null | cc @julien-c @gary149 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27691/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27691/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27691",
"html_url": "https://github.com/huggingface/transformers/pull/27691",
"diff_url": "https://github.com/huggingface/transformers/pull/27691.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27691.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27690 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27690/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27690/comments | https://api.github.com/repos/huggingface/transformers/issues/27690/events | https://github.com/huggingface/transformers/pull/27690 | 2,009,488,459 | PR_kwDOCUB6oc5gS8Ut | 27,690 | Make image processors more general | {
"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/use... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27690). All of your documentation changes will be reflected on that endpoint.",
"Thanks for opening this PR! \r\n\r\n>Ideally it should also allow for the following (cc @amyeroberts)\r\n> \r\n> size = {\"longest_edge\": ...... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
This PR undoes https://github.com/huggingface/transformers/pull/26965 and instead makes image processors more general, by removing the hardcoded "shortest_side" arguments, allowing to pass the following to image processors:
```
size = {"height": ..., "width": ...}
size = {"shortest_edge"... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27690/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/27690/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27690",
"html_url": "https://github.com/huggingface/transformers/pull/27690",
"diff_url": "https://github.com/huggingface/transformers/pull/27690.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27690.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27689 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27689/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27689/comments | https://api.github.com/repos/huggingface/transformers/issues/27689/events | https://github.com/huggingface/transformers/pull/27689 | 2,009,430,217 | PR_kwDOCUB6oc5gSvrw | 27,689 | fix warning | {
"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.c... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,700 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
Reverts a change from #27519 to fix #27678
cc @Rocketknight1 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27689/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/27689/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27689",
"html_url": "https://github.com/huggingface/transformers/pull/27689",
"diff_url": "https://github.com/huggingface/transformers/pull/27689.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27689.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27688 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27688/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27688/comments | https://api.github.com/repos/huggingface/transformers/issues/27688/events | https://github.com/huggingface/transformers/issues/27688 | 2,009,422,069 | I_kwDOCUB6oc53xVj1 | 27,688 | Remote code improvements | {
"login": "LysandreJik",
"id": 30755778,
"node_id": "MDQ6VXNlcjMwNzU1Nzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LysandreJik",
"html_url": "https://github.com/LysandreJik",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"1. was resolved by @MKhalusova in https://github.com/huggingface/transformers/pull/27213, it will be in the next release :hugs: ",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note... | 1,700 | 1,703 | 1,703 | MEMBER | null | Originally posted by @Jackmin801 on the awesome [_jinaai/jina-embeddings-v2-base-en_](https://huggingface.co/jinaai/jina-embeddings-v2-base-en/discussions/5#654102670a2101c338f3d737) repository.
> 1. Documentation
> The clarity of the transformers [documentation on custom models](https://huggingface.co/docs/transfo... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27688/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/27688/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27687 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27687/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27687/comments | https://api.github.com/repos/huggingface/transformers/issues/27687/events | https://github.com/huggingface/transformers/pull/27687 | 2,009,285,159 | PR_kwDOCUB6oc5gSQYH | 27,687 | Skip pipeline tests for 2 models for now | {
"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/... | [] | closed | false | null | [] | [] | 1,700 | 1,700 | 1,700 | COLLABORATOR | null | # What does this PR do?
Skip 2 pipeline tests for 2 moels for now. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27687/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/27687/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27687",
"html_url": "https://github.com/huggingface/transformers/pull/27687",
"diff_url": "https://github.com/huggingface/transformers/pull/27687.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27687.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27686 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27686/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27686/comments | https://api.github.com/repos/huggingface/transformers/issues/27686/events | https://github.com/huggingface/transformers/issues/27686 | 2,009,178,905 | I_kwDOCUB6oc53waMZ | 27,686 | 'LlamaTokenizerFast' object has no attribute 'prefix_id' | {
"login": "Kushalamummigatti",
"id": 62338340,
"node_id": "MDQ6VXNlcjYyMzM4MzQw",
"avatar_url": "https://avatars.githubusercontent.com/u/62338340?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Kushalamummigatti",
"html_url": "https://github.com/Kushalamummigatti",
"followers_url": "https... | [] | closed | false | null | [] | [
"Hello, 🤗 from the look of it, you are not using the `CodeLlamaTokenizer` but the `LlamaTokenizer`. You code does not include a reproducer so I cannot really help you here. ",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed ... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
Using transformer version `transformers` version: 4.34.1
- Platform: Linux-5.10.0-26-cloud-amd64-x86_64-with-glibc2.31
- Python version: 3.10.10
- Huggingface_hub version: 0.19.4
- Safetensors version: 0.3.3
- Accelerate version: 0.23.0
- Accelerate config: not found
- PyTorch version (GPU?): ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27686/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/27686/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27685 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27685/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27685/comments | https://api.github.com/repos/huggingface/transformers/issues/27685/events | https://github.com/huggingface/transformers/pull/27685 | 2,008,971,092 | PR_kwDOCUB6oc5gROJF | 27,685 | Add Flash Attention 2 to Persimmon | {
"login": "jeromeku",
"id": 2455711,
"node_id": "MDQ6VXNlcjI0NTU3MTE=",
"avatar_url": "https://avatars.githubusercontent.com/u/2455711?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jeromeku",
"html_url": "https://github.com/jeromeku",
"followers_url": "https://api.github.com/users/jerom... | [] | open | false | null | [] | [
"@younesbelkada \r\n\r\nRe-based, installed `ruff==0.1.5`, and re-ran `make style`, still getting test failure for `PhiModelTest.test_pipeline_text_generation`.",
"cc @molbap as younes is offline"
] | 1,700 | 1,706 | null | NONE | null | # What does this PR do?
Integrates FA2 to Persimmon per #26350, #27052 (former branch was messed up after trying to rebase, so PR'ing a new branch).
## 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 gui... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27685/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/27685/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27685",
"html_url": "https://github.com/huggingface/transformers/pull/27685",
"diff_url": "https://github.com/huggingface/transformers/pull/27685.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27685.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27684 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27684/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27684/comments | https://api.github.com/repos/huggingface/transformers/issues/27684/events | https://github.com/huggingface/transformers/issues/27684 | 2,008,953,546 | I_kwDOCUB6oc53vjLK | 27,684 | transformers/utils/generic.py contains function deprecated in latest PyTorch | {
"login": "rationalism",
"id": 813306,
"node_id": "MDQ6VXNlcjgxMzMwNg==",
"avatar_url": "https://avatars.githubusercontent.com/u/813306?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rationalism",
"html_url": "https://github.com/rationalism",
"followers_url": "https://api.github.com/user... | [] | closed | false | null | [] | [
"Hey! Thanks for opening this 🤗 \r\nWould be down to update this but we need to make sure we support other versions of `torch` (you are using the latest). \r\nWould you like to open a PR and makes sure this works on previous versions as well? ",
"+1",
"+1",
"same issue ",
"#27803 fixes this! Make sure to u... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
- `transformers` version: 4.35.2
- Platform: Linux-5.19.0-50-generic-x86_64-with-glibc2.35
- Python version: 3.10.13
- Huggingface_hub version: 0.17.3
- Safetensors version: 0.3.1
- Accelerate version: 0.24.1
- Accelerate config: - compute_environment: LOCAL_MACHINE
- distributed_type: NO
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27684/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27684/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27683 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27683/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27683/comments | https://api.github.com/repos/huggingface/transformers/issues/27683/events | https://github.com/huggingface/transformers/pull/27683 | 2,008,857,050 | PR_kwDOCUB6oc5gQ2Pu | 27,683 | tokenizer_kwargs in text-generation pipeline __call__() | {
"login": "thedamnedrhino",
"id": 8396998,
"node_id": "MDQ6VXNlcjgzOTY5OTg=",
"avatar_url": "https://avatars.githubusercontent.com/u/8396998?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thedamnedrhino",
"html_url": "https://github.com/thedamnedrhino",
"followers_url": "https://api.gith... | [] | closed | false | null | [] | [
"@ArthurZucker what do you think about adding a `truncation` (and `max_length`) arg to the pipeline call/constructor instead of this? For all text pipelines. ",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment o... | 1,700 | 1,707 | 1,704 | CONTRIBUTOR | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27683/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/27683/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27683",
"html_url": "https://github.com/huggingface/transformers/pull/27683",
"diff_url": "https://github.com/huggingface/transformers/pull/27683.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27683.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27682 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27682/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27682/comments | https://api.github.com/repos/huggingface/transformers/issues/27682/events | https://github.com/huggingface/transformers/issues/27682 | 2,008,854,973 | I_kwDOCUB6oc53vLG9 | 27,682 | Mistral with Flash atteniton v2 give error on long sequence input and max_new_tokens | {
"login": "binarycrayon",
"id": 10211,
"node_id": "MDQ6VXNlcjEwMjEx",
"avatar_url": "https://avatars.githubusercontent.com/u/10211?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/binarycrayon",
"html_url": "https://github.com/binarycrayon",
"followers_url": "https://api.github.com/users/b... | [] | closed | false | null | [] | [
"Hi @binarycrayon \r\nThis issue should be fixed with https://github.com/huggingface/transformers/pull/27548",
"Thank you, subscribed. I will close this once it's merged"
] | 1,700 | 1,701 | 1,701 | NONE | null | ### System Info
## system information
WSL on windows 11
## Hardware
RTX3090Ti
## Software
Python: 3.10.13
Transformers: 4.35
## Input Sequence Length
15901 throws error
13266 was fine
## Model Initialization
```
model_id = "mistralai/Mistral-7B-v0.1"
model = AutoModelForCausalLM.from_pretrained(
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27682/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/27682/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27681 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27681/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27681/comments | https://api.github.com/repos/huggingface/transformers/issues/27681/events | https://github.com/huggingface/transformers/pull/27681 | 2,008,640,581 | PR_kwDOCUB6oc5gQHM9 | 27,681 | Update forward signature test for vision models | {
"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/use... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._",
"`main_input_name` is defined in `PreTrainedModel` (and the corresponding TF/Flax counterpart), so\r\n\r\n> all (NLP) models have this properly set\r\n\r\nis True"
] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
This PR makes sure that vision-only models don't need to overwrite `test_forward_signature`. Instead, `model.main_input_name` is leveraged. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27681/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/27681/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27681",
"html_url": "https://github.com/huggingface/transformers/pull/27681",
"diff_url": "https://github.com/huggingface/transformers/pull/27681.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27681.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27680 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27680/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27680/comments | https://api.github.com/repos/huggingface/transformers/issues/27680/events | https://github.com/huggingface/transformers/pull/27680 | 2,008,640,225 | PR_kwDOCUB6oc5gQHIE | 27,680 | Fix sampling method to handle all -inf scores in next_token_scores | {
"login": "Saibo-creator",
"id": 53392976,
"node_id": "MDQ6VXNlcjUzMzkyOTc2",
"avatar_url": "https://avatars.githubusercontent.com/u/53392976?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Saibo-creator",
"html_url": "https://github.com/Saibo-creator",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"@Saibo-creator see related comment [here](https://github.com/huggingface/transformers/issues/27676#issuecomment-1831735403) :)",
"check PR #27797 which fixes the relevant issues"
] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | This commit addresses a bug in the token generation process where next_token_scores are all -inf, leading to NaN probabilities after softmax. This typically happens in the constrained generation process where no tokens are allowed to be generated anymore. The fix involves adjusting probabilities in such scenarios to en... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27680/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/27680/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27680",
"html_url": "https://github.com/huggingface/transformers/pull/27680",
"diff_url": "https://github.com/huggingface/transformers/pull/27680.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27680.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27679 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27679/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27679/comments | https://api.github.com/repos/huggingface/transformers/issues/27679/events | https://github.com/huggingface/transformers/pull/27679 | 2,008,532,360 | PR_kwDOCUB6oc5gPvoi | 27,679 | using env var to skip check_imports err | {
"login": "wqh17101",
"id": 26429138,
"node_id": "MDQ6VXNlcjI2NDI5MTM4",
"avatar_url": "https://avatars.githubusercontent.com/u/26429138?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/wqh17101",
"html_url": "https://github.com/wqh17101",
"followers_url": "https://api.github.com/users/wqh... | [] | closed | false | null | [] | [
"@amyeroberts",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27679). All of your documentation changes will be reflected on that endpoint.",
"Hi @wqh17101 - thanks for opening this PR with a solution! As this is a highly specific fix (adding in an environment variab... | 1,700 | 1,704 | 1,704 | NONE | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27679/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/27679/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27679",
"html_url": "https://github.com/huggingface/transformers/pull/27679",
"diff_url": "https://github.com/huggingface/transformers/pull/27679.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27679.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27678 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27678/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27678/comments | https://api.github.com/repos/huggingface/transformers/issues/27678/events | https://github.com/huggingface/transformers/issues/27678 | 2,008,463,663 | I_kwDOCUB6oc53trkv | 27,678 | AttributeError: module 'transformers.utils.logging' has no attribute 'warning' | {
"login": "benzom",
"id": 31667178,
"node_id": "MDQ6VXNlcjMxNjY3MTc4",
"avatar_url": "https://avatars.githubusercontent.com/u/31667178?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/benzom",
"html_url": "https://github.com/benzom",
"followers_url": "https://api.github.com/users/benzom/fo... | [] | closed | false | null | [] | [
"Hey! Indeed, opening a PR for a fix! 🤗 thanks for reporting"
] | 1,700 | 1,701 | 1,701 | NONE | null | ### System Info
Hi everyone. I'm trying to run example from here https://github.com/huggingface/transformers/tree/main/examples/pytorch
Transformers library was installed from the source as it was requested during the first run
accelerate==0.24.1
torch==1.13.0a0+936e930
The running command:
```
accelerate lau... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27678/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/27678/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27677 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27677/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27677/comments | https://api.github.com/repos/huggingface/transformers/issues/27677/events | https://github.com/huggingface/transformers/pull/27677 | 2,008,433,180 | PR_kwDOCUB6oc5gPZzA | 27,677 | Add ChatGLM model. | {
"login": "xunkai55",
"id": 4828553,
"node_id": "MDQ6VXNlcjQ4Mjg1NTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/4828553?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/xunkai55",
"html_url": "https://github.com/xunkai55",
"followers_url": "https://api.github.com/users/xunka... | [] | closed | false | null | [] | [
"Hey! Thanks for opening a PR 🤗 I think the implementation of chatGLM should follow the comment I posted [here](https://github.com/huggingface/transformers/pull/27267#issuecomment-1802207501). See #27267 which also wanted to add support for ChatGM 😉 ",
"This issue has been automatically marked as stale because ... | 1,700 | 1,704 | 1,704 | NONE | null | # What does this PR do?
Add ChatGLM model support in HF Transformers repo.
## 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... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27677/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/27677/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27677",
"html_url": "https://github.com/huggingface/transformers/pull/27677",
"diff_url": "https://github.com/huggingface/transformers/pull/27677.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27677.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27676 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27676/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27676/comments | https://api.github.com/repos/huggingface/transformers/issues/27676/events | https://github.com/huggingface/transformers/issues/27676 | 2,008,428,965 | I_kwDOCUB6oc53tjGl | 27,676 | RuntimeError with prefix_allowed_tokens_fn and do_sample=True When Allowed Tokens List is Empty | {
"login": "Saibo-creator",
"id": 53392976,
"node_id": "MDQ6VXNlcjUzMzkyOTc2",
"avatar_url": "https://avatars.githubusercontent.com/u/53392976?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Saibo-creator",
"html_url": "https://github.com/Saibo-creator",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"Hey! Thanks for reporting, this would require use to check that the output of `self._prefix_allowed_tokens_fn(batch_id, sent)` on each token is not `[]` before applying the mask. It does make sense because we don't specify that the list cannot be empty. \r\nWould you like to open a PR for a fix? (meaning something... | 1,700 | 1,702 | 1,702 | CONTRIBUTOR | null | ### System Info
- `transformers` version: 4.36.0.dev0
- Platform: macOS-13.4.1-arm64-arm-64bit
- Python version: 3.10.13
- Huggingface_hub version: 0.19.3
- Safetensors version: 0.3.3
- Accelerate version: 0.23.0
- Accelerate config: not found
- PyTorch version (GPU?): 2.0.1 (False)
- Tensorflow version (GP... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27676/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/27676/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27675 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27675/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27675/comments | https://api.github.com/repos/huggingface/transformers/issues/27675/events | https://github.com/huggingface/transformers/pull/27675 | 2,008,356,275 | PR_kwDOCUB6oc5gPIwM | 27,675 | Fix semantic error in evaluation section | {
"login": "anihm136",
"id": 49116134,
"node_id": "MDQ6VXNlcjQ5MTE2MTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/49116134?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anihm136",
"html_url": "https://github.com/anihm136",
"followers_url": "https://api.github.com/users/ani... | [] | closed | false | null | [] | [] | 1,700 | 1,700 | 1,700 | CONTRIBUTOR | null | # What does this PR do?
Change "convert predictions to logits" to "convert logits to
predictions" to fix semantic error in the evaluation section. Logits
need to be converted to predictions to evaluate the accuracy, not the
other way round
## Before submitting
- [x] This PR fixes a typo or improves the docs (you can ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27675/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/27675/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27675",
"html_url": "https://github.com/huggingface/transformers/pull/27675",
"diff_url": "https://github.com/huggingface/transformers/pull/27675.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27675.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27674 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27674/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27674/comments | https://api.github.com/repos/huggingface/transformers/issues/27674/events | https://github.com/huggingface/transformers/pull/27674 | 2,008,350,483 | PR_kwDOCUB6oc5gPHeh | 27,674 | Update tiny model creation script | {
"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/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,700 | 1,701 | 1,701 | COLLABORATOR | null | # What does this PR do?
A few fixes or improvements I found necessary while working on #27388.
See comments along the changes | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27674/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/27674/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27674",
"html_url": "https://github.com/huggingface/transformers/pull/27674",
"diff_url": "https://github.com/huggingface/transformers/pull/27674.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27674.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27673 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27673/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27673/comments | https://api.github.com/repos/huggingface/transformers/issues/27673/events | https://github.com/huggingface/transformers/pull/27673 | 2,008,285,315 | PR_kwDOCUB6oc5gO5Mr | 27,673 | Translating en/model_doc folder docs to Japanese(from `blip` to `clap`) 🇯🇵 | {
"login": "rajveer43",
"id": 64583161,
"node_id": "MDQ6VXNlcjY0NTgzMTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/64583161?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rajveer43",
"html_url": "https://github.com/rajveer43",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27673). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.",
"> LGTM! Can you try running `make fixup` and `pip uninstall black && pip install -... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27673/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/27673/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27673",
"html_url": "https://github.com/huggingface/transformers/pull/27673",
"diff_url": "https://github.com/huggingface/transformers/pull/27673.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27673.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27672 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27672/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27672/comments | https://api.github.com/repos/huggingface/transformers/issues/27672/events | https://github.com/huggingface/transformers/pull/27672 | 2,008,284,054 | PR_kwDOCUB6oc5gO460 | 27,672 | Update TVP arxiv link | {
"login": "amyeroberts",
"id": 22614925,
"node_id": "MDQ6VXNlcjIyNjE0OTI1",
"avatar_url": "https://avatars.githubusercontent.com/u/22614925?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/amyeroberts",
"html_url": "https://github.com/amyeroberts",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._"
] | 1,700 | 1,700 | 1,700 | COLLABORATOR | null | # What does this PR do?
TVP on the doc page wasn't rendering the architecture image which highlighted two issues:
* Link in HTML tag had been copied from TVLT. This PR updates to point to the TVP paper
* TVP architecture pic hadn't been added ([resolved here](https://huggingface.co/datasets/huggingface/document... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27672/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/27672/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27672",
"html_url": "https://github.com/huggingface/transformers/pull/27672",
"diff_url": "https://github.com/huggingface/transformers/pull/27672.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27672.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27671 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27671/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27671/comments | https://api.github.com/repos/huggingface/transformers/issues/27671/events | https://github.com/huggingface/transformers/pull/27671 | 2,008,210,958 | PR_kwDOCUB6oc5gOo4T | 27,671 | [WIP][Splinter] Fixes #16627 by implementing the test cases for splinter | {
"login": "nileshkokane01",
"id": 8201108,
"node_id": "MDQ6VXNlcjgyMDExMDg=",
"avatar_url": "https://avatars.githubusercontent.com/u/8201108?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nileshkokane01",
"html_url": "https://github.com/nileshkokane01",
"followers_url": "https://api.gith... | [
{
"id": 2796628563,
"node_id": "MDU6TGFiZWwyNzk2NjI4NTYz",
"url": "https://api.github.com/repos/huggingface/transformers/labels/WIP",
"name": "WIP",
"color": "234C99",
"default": false,
"description": "Label your PR/Issue with WIP for some long outstanding Issues/PRs that are work in pro... | open | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md) a... | 1,700 | 1,706 | null | CONTRIBUTOR | null | # What does this PR do?
Fixes #16627 by implementing test cases for splinter
<!--
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 y... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27671/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/27671/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27671",
"html_url": "https://github.com/huggingface/transformers/pull/27671",
"diff_url": "https://github.com/huggingface/transformers/pull/27671.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27671.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27670 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27670/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27670/comments | https://api.github.com/repos/huggingface/transformers/issues/27670/events | https://github.com/huggingface/transformers/issues/27670 | 2,008,200,663 | I_kwDOCUB6oc53srXX | 27,670 | Min P style sampling - an alternative to Top P/TopK | {
"login": "kalomaze",
"id": 66376113,
"node_id": "MDQ6VXNlcjY2Mzc2MTEz",
"avatar_url": "https://avatars.githubusercontent.com/u/66376113?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kalomaze",
"html_url": "https://github.com/kalomaze",
"followers_url": "https://api.github.com/users/kal... | [
{
"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 | [] | [
"fyi @gante 🤗 ",
"Hi @kalomaze 👋 Thank you for opening this issue!\r\n\r\nIn addition to Temperature, Top p, and Top k, which apply distribution-agnostic transformations, we have three other distribution-aware transformations:\r\n1. [Typical P Decoding](https://huggingface.co/docs/transformers/internal/generati... | 1,700 | 1,701 | null | NONE | null | ### Feature request
This is a sampler method already present in other LLM inference backends that aims to simplify the truncation process & help accomodate for the flaws/failings of Top P & Top K.
**Min P**.
![image](https://github.com/huggingface/transformers/assets/66376113/53113071-20ed-43bf-a8ae-c9f083840d9... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27670/reactions",
"total_count": 6,
"+1": 5,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27670/timeline | null | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27669 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27669/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27669/comments | https://api.github.com/repos/huggingface/transformers/issues/27669/events | https://github.com/huggingface/transformers/issues/27669 | 2,007,992,511 | I_kwDOCUB6oc53r4i_ | 27,669 | [i18n-jp] Translating `en/model_doc` folder docs to Japanese 🇯🇵 | {
"login": "rajveer43",
"id": 64583161,
"node_id": "MDQ6VXNlcjY0NTgzMTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/64583161?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rajveer43",
"html_url": "https://github.com/rajveer43",
"followers_url": "https://api.github.com/users/... | [
{
"id": 2796628563,
"node_id": "MDU6TGFiZWwyNzk2NjI4NTYz",
"url": "https://api.github.com/repos/huggingface/transformers/labels/WIP",
"name": "WIP",
"color": "234C99",
"default": false,
"description": "Label your PR/Issue with WIP for some long outstanding Issues/PRs that are work in pro... | closed | false | null | [] | [] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | <!--
Note: Please search to see if an issue already exists for the language you are trying to translate.
-->
Hi!
Let's bring the documentation to all the japanese-speaking community 🌐 (currently 0 out of 267 complete)
Who would want to translate? Please follow the 🤗 [TRANSLATING guide](https://github.com/h... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27669/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/27669/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27668 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27668/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27668/comments | https://api.github.com/repos/huggingface/transformers/issues/27668/events | https://github.com/huggingface/transformers/pull/27668 | 2,007,905,340 | PR_kwDOCUB6oc5gNr5I | 27,668 | memory efficient attention (Flash V2) initial support - encoder-only and not combined with relative attention | {
"login": "YoelShoshan",
"id": 7043815,
"node_id": "MDQ6VXNlcjcwNDM4MTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/7043815?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YoelShoshan",
"html_url": "https://github.com/YoelShoshan",
"followers_url": "https://api.github.com/us... | [] | closed | false | null | [] | [] | 1,700 | 1,700 | 1,700 | NONE | null | # What does this PR do?
adds support for memory efficient attention, based on using xformers library memory_efficient_attention() op.
it only supports the following:
1. Encoder-only (no causal attention)
2. Relative attention (which gets injected into the "scores" (the results of Q@K.transpose()) is NOT support... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27668/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/27668/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27668",
"html_url": "https://github.com/huggingface/transformers/pull/27668",
"diff_url": "https://github.com/huggingface/transformers/pull/27668.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27668.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27667 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27667/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27667/comments | https://api.github.com/repos/huggingface/transformers/issues/27667/events | https://github.com/huggingface/transformers/pull/27667 | 2,007,663,534 | PR_kwDOCUB6oc5gM2sB | 27,667 | fix: fix gradient accumulate step for learning rate | {
"login": "pphuc25",
"id": 81808312,
"node_id": "MDQ6VXNlcjgxODA4MzEy",
"avatar_url": "https://avatars.githubusercontent.com/u/81808312?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pphuc25",
"html_url": "https://github.com/pphuc25",
"followers_url": "https://api.github.com/users/pphuc2... | [] | closed | false | null | [] | [] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | Hi,
I think this is the mistake seen the learning rate should take the param `num_train_steps` as `total_train_steps`, instead of `len(vectorized_datasets["train"])`
So I fix them to the right suitable that learning rate will go down to zero as init
I would like to cc @sanchit-gandhi to review my PR | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27667/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/27667/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27667",
"html_url": "https://github.com/huggingface/transformers/pull/27667",
"diff_url": "https://github.com/huggingface/transformers/pull/27667.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27667.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27666 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27666/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27666/comments | https://api.github.com/repos/huggingface/transformers/issues/27666/events | https://github.com/huggingface/transformers/issues/27666 | 2,007,613,877 | I_kwDOCUB6oc53qcG1 | 27,666 | how to remove punctuation marks. | {
"login": "chanyong-owl",
"id": 57178312,
"node_id": "MDQ6VXNlcjU3MTc4MzEy",
"avatar_url": "https://avatars.githubusercontent.com/u/57178312?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/chanyong-owl",
"html_url": "https://github.com/chanyong-owl",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | [
"Hey 🤗 thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discuss.huggingface.co/) instead? I'm sure the community will be of help!\r\n\r\nThanks!",
"This issue has been automatically marked as stale because it has not h... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
i trained t5-large for translation.
the result of train was good
But when i input some sentence, the result is like that "What are you doing now?.??....."
[?.??......] <- how to delete that punctuation marks.
i put some parameter like max_length. But i can not solve that situation
### Who ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27666/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/27666/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27665 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27665/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27665/comments | https://api.github.com/repos/huggingface/transformers/issues/27665/events | https://github.com/huggingface/transformers/pull/27665 | 2,007,502,981 | PR_kwDOCUB6oc5gMTvp | 27,665 | Docs/Add conversion code to the musicgen docs | {
"login": "yoinked-h",
"id": 63889420,
"node_id": "MDQ6VXNlcjYzODg5NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/63889420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yoinked-h",
"html_url": "https://github.com/yoinked-h",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"Hey! Updated the people you pinged, let's minimize it 😉 ",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27665). All of your documentation changes will be reflected on that endpoint."
] | 1,700 | 1,700 | 1,700 | CONTRIBUTOR | null |
# What does this PR do?
Adds conversion code to the musicgen docs (the conversion code is quite hidden and needs some changes to be run on other model dirs)
## 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 [contr... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27665/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/27665/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27665",
"html_url": "https://github.com/huggingface/transformers/pull/27665",
"diff_url": "https://github.com/huggingface/transformers/pull/27665.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27665.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27663 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27663/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27663/comments | https://api.github.com/repos/huggingface/transformers/issues/27663/events | https://github.com/huggingface/transformers/pull/27663 | 2,007,090,060 | PR_kwDOCUB6oc5gK93b | 27,663 | Fix yolos resizing | {
"login": "amyeroberts",
"id": 22614925,
"node_id": "MDQ6VXNlcjIyNjE0OTI1",
"avatar_url": "https://avatars.githubusercontent.com/u/22614925?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/amyeroberts",
"html_url": "https://github.com/amyeroberts",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [] | 1,700 | 1,703 | 1,703 | COLLABORATOR | null | # What does this PR do?
On main - running the following produces an image outside of the limits set by "longest_edge" in an images config:
```py
from transformers import AutoProcessor
from PIL import Image
import requests
processor = AutoProcessor.from_pretrained("Xenova/yolos-small-300") # or hustvl/yolos... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27663/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/27663/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27663",
"html_url": "https://github.com/huggingface/transformers/pull/27663",
"diff_url": "https://github.com/huggingface/transformers/pull/27663.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27663.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27662 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27662/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27662/comments | https://api.github.com/repos/huggingface/transformers/issues/27662/events | https://github.com/huggingface/transformers/pull/27662 | 2,006,973,324 | PR_kwDOCUB6oc5gKkJ5 | 27,662 | [`Llava`] Add Llava to transformers | {
"login": "younesbelkada",
"id": 49240599,
"node_id": "MDQ6VXNlcjQ5MjQwNTk5",
"avatar_url": "https://avatars.githubusercontent.com/u/49240599?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/younesbelkada",
"html_url": "https://github.com/younesbelkada",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27662). All of your documentation changes will be reflected on that endpoint.",
"Bakllava weights successfully converted here: https://huggingface.co/ybelkada/BakLlava-v1-hf",
"All converted checkpoints are now under this... | 1,700 | 1,706 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Adds Llava - a multimodal model in transformers library.
Llava is a multi-modal model that claims to have competitive performance than GPT-4 for multi-modal tasks. There are currently 3 main variants of this architecture:
- Llava - Llama
- Llava - MPT
- Llava - Mistral (known as Bakl... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27662/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/27662/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27662",
"html_url": "https://github.com/huggingface/transformers/pull/27662",
"diff_url": "https://github.com/huggingface/transformers/pull/27662.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27662.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27661 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27661/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27661/comments | https://api.github.com/repos/huggingface/transformers/issues/27661/events | https://github.com/huggingface/transformers/pull/27661 | 2,006,949,847 | PR_kwDOCUB6oc5gKe8I | 27,661 | [`FA-2`] Add Flash Attention to `Phi` | {
"login": "susnato",
"id": 56069179,
"node_id": "MDQ6VXNlcjU2MDY5MTc5",
"avatar_url": "https://avatars.githubusercontent.com/u/56069179?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/susnato",
"html_url": "https://github.com/susnato",
"followers_url": "https://api.github.com/users/susnat... | [] | closed | false | null | [] | [
"All the FA tests pass except the `test_flash_attn_2_generate_padding_right`.\r\n\r\n\r\n\r\nThis is odd given that the `flash_attn_2_inference_padding_right` test is passing as do... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | 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 w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27661/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/27661/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27661",
"html_url": "https://github.com/huggingface/transformers/pull/27661",
"diff_url": "https://github.com/huggingface/transformers/pull/27661.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27661.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27660 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27660/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27660/comments | https://api.github.com/repos/huggingface/transformers/issues/27660/events | https://github.com/huggingface/transformers/pull/27660 | 2,006,906,570 | PR_kwDOCUB6oc5gKVaC | 27,660 | Log learning rate | {
"login": "muellerzr",
"id": 7831895,
"node_id": "MDQ6VXNlcjc4MzE4OTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/7831895?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/muellerzr",
"html_url": "https://github.com/muellerzr",
"followers_url": "https://api.github.com/users/mu... | [] | open | false | null | [] | [
"Is it a draft or should i review? ",
"@ArthurZucker still draft, hit one failure when I was trying it so should be good next week hopefully 🤞 ",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this threa... | 1,700 | 1,707 | null | CONTRIBUTOR | null | # What does this PR do?
Logs the learning rate on each logging instance when training, and specifically uses scientific notation to save on space
Fulfills https://github.com/huggingface/transformers/issues/27631
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27660/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/27660/timeline | null | true | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27660",
"html_url": "https://github.com/huggingface/transformers/pull/27660",
"diff_url": "https://github.com/huggingface/transformers/pull/27660.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27660.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27659 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27659/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27659/comments | https://api.github.com/repos/huggingface/transformers/issues/27659/events | https://github.com/huggingface/transformers/pull/27659 | 2,006,855,559 | PR_kwDOCUB6oc5gKJ-r | 27,659 | [i18n-fr] Translate autoclass tutorial to French | {
"login": "NoB0",
"id": 28621493,
"node_id": "MDQ6VXNlcjI4NjIxNDkz",
"avatar_url": "https://avatars.githubusercontent.com/u/28621493?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/NoB0",
"html_url": "https://github.com/NoB0",
"followers_url": "https://api.github.com/users/NoB0/followers"... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27659). All of your documentation changes will be reflected on that endpoint.",
"Thanks for the translation! "
] | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Translated the autoclass_tutorial.md file of the documentation to French.
Part of #21456
Thank you in advance for your review.
## 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 [contrib... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27659/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/27659/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27659",
"html_url": "https://github.com/huggingface/transformers/pull/27659",
"diff_url": "https://github.com/huggingface/transformers/pull/27659.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27659.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27658 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27658/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27658/comments | https://api.github.com/repos/huggingface/transformers/issues/27658/events | https://github.com/huggingface/transformers/pull/27658 | 2,006,844,016 | PR_kwDOCUB6oc5gKHZp | 27,658 | [Whisper] Finalize batched SOTA long-form generation | {
"login": "patrickvonplaten",
"id": 23423619,
"node_id": "MDQ6VXNlcjIzNDIzNjE5",
"avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/patrickvonplaten",
"html_url": "https://github.com/patrickvonplaten",
"followers_url": "https://... | [] | closed | false | null | [] | [
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_27658). All of your documentation changes will be reflected on that endpoint.",
"Follow-up of: https://github.com/huggingface/transformers/pull/27492",
"Failing test is a timeout which should not happen when running again... | 1,700 | 1,708 | 1,705 | MEMBER | null | 🚨🚨 ***Disclaimer***: **All the credit of this PR goes to the original codebase: https://github.com/openai/whisper/blob/main/whisper/transcribe.py (this PR is to 90% copied from there). The reason it's added here (instead of contributing to https://github.com/openai/whisper) is because the original repo fundamentally ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27658/reactions",
"total_count": 3,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 3,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/27658/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27658",
"html_url": "https://github.com/huggingface/transformers/pull/27658",
"diff_url": "https://github.com/huggingface/transformers/pull/27658.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27658.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27657 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27657/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27657/comments | https://api.github.com/repos/huggingface/transformers/issues/27657/events | https://github.com/huggingface/transformers/pull/27657 | 2,006,766,771 | PR_kwDOCUB6oc5gJ2Wz | 27,657 | [i18n-fr] Translate installation to French | {
"login": "NoB0",
"id": 28621493,
"node_id": "MDQ6VXNlcjI4NjIxNDkz",
"avatar_url": "https://avatars.githubusercontent.com/u/28621493?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/NoB0",
"html_url": "https://github.com/NoB0",
"followers_url": "https://api.github.com/users/NoB0/followers"... | [] | closed | false | null | [] | [
"Heads up, a few typos I was able to spot:\r\n\r\n- line 31: missing 's' -> environnements\r\n- line 57: en seule ligne -> en une seule ligne\r\n- line 74 : suivant -> suivants\r\n- line 110: utilie -> utile",
"> Heads up, a few typos I was able to spot:\r\n> \r\n> * line 31: missing 's' -> environnements\r\n> ... | 1,700 | 1,701 | 1,701 | CONTRIBUTOR | null | # What does this PR do?
Translated the quicktour.mdx file of the documentation to French.
Part of #21456
Thank you in advance for your review.
## 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 g... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27657/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/27657/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27657",
"html_url": "https://github.com/huggingface/transformers/pull/27657",
"diff_url": "https://github.com/huggingface/transformers/pull/27657.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27657.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27656 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27656/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27656/comments | https://api.github.com/repos/huggingface/transformers/issues/27656/events | https://github.com/huggingface/transformers/pull/27656 | 2,006,508,658 | PR_kwDOCUB6oc5gI9ek | 27,656 | [Auto Safetensors] Websocket -> SSE | {
"login": "LysandreJik",
"id": 30755778,
"node_id": "MDQ6VXNlcjMwNzU1Nzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LysandreJik",
"html_url": "https://github.com/LysandreJik",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._",
"Let's merge and iterate on the other PR"
] | 1,700 | 1,700 | 1,700 | MEMBER | null | cc @Narsil | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27656/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/27656/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27656",
"html_url": "https://github.com/huggingface/transformers/pull/27656",
"diff_url": "https://github.com/huggingface/transformers/pull/27656.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27656.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27655 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27655/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27655/comments | https://api.github.com/repos/huggingface/transformers/issues/27655/events | https://github.com/huggingface/transformers/pull/27655 | 2,006,369,439 | PR_kwDOCUB6oc5gIeqV | 27,655 | [DPT, Dinov2] Add resources | {
"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/use... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._",
"@NielsRogge Shall I merge?",
"Yes this can be merged :)"
] | 1,700 | 1,700 | 1,700 | CONTRIBUTOR | null | # What does this PR do?
To make more people aware that DPT is now compatible with `AutoBackbone` (https://github.com/huggingface/transformers/pull/26092), I've added some usage tips and resources to the docs. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27655/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/27655/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27655",
"html_url": "https://github.com/huggingface/transformers/pull/27655",
"diff_url": "https://github.com/huggingface/transformers/pull/27655.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27655.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27653 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27653/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27653/comments | https://api.github.com/repos/huggingface/transformers/issues/27653/events | https://github.com/huggingface/transformers/issues/27653 | 2,006,313,091 | I_kwDOCUB6oc53leiD | 27,653 | Silent failure when using max_length value that is too low | {
"login": "Kroshtan",
"id": 31923442,
"node_id": "MDQ6VXNlcjMxOTIzNDQy",
"avatar_url": "https://avatars.githubusercontent.com/u/31923442?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Kroshtan",
"html_url": "https://github.com/Kroshtan",
"followers_url": "https://api.github.com/users/Kro... | [] | closed | false | null | [] | [
"Hey! Feel free to open a PR for a fix, have seen this a few times but don't think it's very important and fast tokenizers also don't have the same outputs as slow for this.\r\nThough not sure if there are a lot of relavant usages for this? 😅 ",
"This issue has been automatically marked as stale because it has n... | 1,700 | 1,704 | 1,704 | NONE | null | ### System Info
- `transformers` version: 4.34.1
- Platform: Linux-5.15.0-89-generic-x86_64-with-glibc2.35
- Python version: 3.10.12
- Huggingface_hub version: 0.17.3
- Safetensors version: 0.4.0
- Accelerate version: not installed
- Accelerate config: not found
- PyTorch version (GPU?): 2.0.0+cu117 (True)
- T... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27653/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/27653/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27652 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27652/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27652/comments | https://api.github.com/repos/huggingface/transformers/issues/27652/events | https://github.com/huggingface/transformers/pull/27652 | 2,006,193,793 | PR_kwDOCUB6oc5gH30w | 27,652 | Refactoring Trainer, adds `save_only_model` arg and simplifying FSDP integration | {
"login": "pacman100",
"id": 13534540,
"node_id": "MDQ6VXNlcjEzNTM0NTQw",
"avatar_url": "https://avatars.githubusercontent.com/u/13534540?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pacman100",
"html_url": "https://github.com/pacman100",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"_The documentation is not available anymore as the PR was closed or merged._",
"Hi, \r\n\r\nThe (Circle)CI did run all the tests: for example, in `torch_job`, we can see\r\n\r\n```\r\npython -m pytest --junitxml=test-results/junit.xml -n 6 --max-worker-restart=0 --dist=loadfile --make-reports=tests_torch tests/b... | 1,700 | 1,701 | 1,700 | CONTRIBUTOR | null | # What does this PR do?
1. Bumps up the minimum Accelerate version to 0.21.0
2. Add `save_only_model` arg - This enables the feature request https://github.com/huggingface/transformers/issues/26706
3. Simplifies a lot of logic in FSDP:
a. Currently, FSDP-XLA logic is custom in Trainer and normal FSDP is using ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27652/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/27652/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27652",
"html_url": "https://github.com/huggingface/transformers/pull/27652",
"diff_url": "https://github.com/huggingface/transformers/pull/27652.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27652.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27651 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27651/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27651/comments | https://api.github.com/repos/huggingface/transformers/issues/27651/events | https://github.com/huggingface/transformers/issues/27651 | 2,006,176,100 | I_kwDOCUB6oc53k9Fk | 27,651 | In which function it is best way to use the temperature parameter .from_pretrained() or .generate() | {
"login": "pradeepdev-1995",
"id": 41164884,
"node_id": "MDQ6VXNlcjQxMTY0ODg0",
"avatar_url": "https://avatars.githubusercontent.com/u/41164884?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pradeepdev-1995",
"html_url": "https://github.com/pradeepdev-1995",
"followers_url": "https://api... | [] | closed | false | null | [] | [
"Hey 🤗 thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discuss.huggingface.co/) instead? \r\nI think the answer depends on the use case 😉 \r\nThanks!",
"This issue has been automatically marked as stale because it ha... | 1,700 | 1,704 | 1,704 | NONE | null | ### Feature request
in which function it is the best way to use the temperature parameter .from_pretrained or .generate()
It seems that in both functions we can pass the temperature parameter. Which is the best way to do this?
```
Model = AutoModelForCausalLM.from_pretrained(model_name,
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27651/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/27651/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27650 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27650/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27650/comments | https://api.github.com/repos/huggingface/transformers/issues/27650/events | https://github.com/huggingface/transformers/issues/27650 | 2,006,116,599 | I_kwDOCUB6oc53kuj3 | 27,650 | Mask2Former slowdown starting from version 4.32.0 | {
"login": "matteot11",
"id": 15927868,
"node_id": "MDQ6VXNlcjE1OTI3ODY4",
"avatar_url": "https://avatars.githubusercontent.com/u/15927868?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/matteot11",
"html_url": "https://github.com/matteot11",
"followers_url": "https://api.github.com/users/... | [] | open | false | null | [] | [
"cc @amyeroberts ",
"Hi @matteot11 - thanks for raising! Looking into it 🕵️ ",
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing ... | 1,700 | 1,707 | null | NONE | null | ### System Info
- System information: x86_64 GNU/Linux (with Titan RTX GPU)
- Ubuntu version: 18.04
- Python version: 3.8.12
- CUDA version: 11.1
- PyTorch version: 2.0.1
- transformers version: 4.32.0
### Who can help?
@amyeroberts
### Information
- [ ] The official example scripts
- [ ] My own modified scri... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27650/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/27650/timeline | reopened | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27649 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27649/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27649/comments | https://api.github.com/repos/huggingface/transformers/issues/27649/events | https://github.com/huggingface/transformers/issues/27649 | 2,006,086,108 | I_kwDOCUB6oc53knHc | 27,649 | Adding support for lookahead decoding for autoregressive (decoder + encoder-decoder) models | {
"login": "shermansiu",
"id": 12627125,
"node_id": "MDQ6VXNlcjEyNjI3MTI1",
"avatar_url": "https://avatars.githubusercontent.com/u/12627125?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/shermansiu",
"html_url": "https://github.com/shermansiu",
"followers_url": "https://api.github.com/use... | [
{
"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 | [] | [
"FYI @gante so we keep track of this. Shared offline but might be good after the cache refactoring. ",
"The current reference implementation builds directly on top of Huggingface transformers, but the authors have mentioned that they plan to release a custom CUDA kernel to speed up the method.\r\n\r\nShould we wa... | 1,700 | 1,701 | null | CONTRIBUTOR | null | ### Feature request
Fu et al. propose a novel decoding technique that accelerates greedy decoding on Llama 2 and Code-Llama by 1.5-2x across various parameters sizes, without a draft model. This method can be extended to work on beam search decoding.
Blog post: https://lmsys.org/blog/2023-11-21-lookahead-decoding/
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27649/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/27649/timeline | null | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27648 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27648/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27648/comments | https://api.github.com/repos/huggingface/transformers/issues/27648/events | https://github.com/huggingface/transformers/issues/27648 | 2,005,914,513 | I_kwDOCUB6oc53j9OR | 27,648 | kosmos_processor `sorted_length =sorted([(idx, len(x)) for idx, x in enumerate(text_encoding.input_ids)])` wrong? | {
"login": "YamingZhang",
"id": 50822118,
"node_id": "MDQ6VXNlcjUwODIyMTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/50822118?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YamingZhang",
"html_url": "https://github.com/YamingZhang",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"Mmmm yeah seems like it. Would you like to open a PR? \r\nfyi @ydshieh ",
"Sure, I'd be happy to help out",
"Hi @YamingZhang \r\n\r\nthe latest version is\r\n\r\n```\r\n sorted_length = sorted(\r\n [(idx, len(x)) for idx, x in enumerate(text_encoding.input_ids)], key=lambda x:... | 1,700 | 1,700 | 1,700 | NONE | null | https://github.com/huggingface/transformers/blob/514de24abfd4416aeba6a6455ad5920f57f3567d/src/transformers/models/kosmos2/processing_kosmos2.py#L214C17-L214C17
The sorted code lacks a sorting function and should be changed to
`sorted_length =sorted([(idx, len(x)) for idx, x in enumerate(text_encoding.input_ids)], ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27648/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/27648/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/27647 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27647/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27647/comments | https://api.github.com/repos/huggingface/transformers/issues/27647/events | https://github.com/huggingface/transformers/pull/27647 | 2,005,859,302 | PR_kwDOCUB6oc5gGuT_ | 27,647 | Add push_to_hub also for image_processor | {
"login": "correll",
"id": 399192,
"node_id": "MDQ6VXNlcjM5OTE5Mg==",
"avatar_url": "https://avatars.githubusercontent.com/u/399192?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/correll",
"html_url": "https://github.com/correll",
"followers_url": "https://api.github.com/users/correll/fo... | [] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md) a... | 1,700 | 1,704 | 1,704 | NONE | null |
# What does this PR do?
Without this the image-segmentation pipeline will fail with a user's own model.
<!--
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 f... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27647/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/27647/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27647",
"html_url": "https://github.com/huggingface/transformers/pull/27647",
"diff_url": "https://github.com/huggingface/transformers/pull/27647.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27647.patch",
"merged_at... |
https://api.github.com/repos/huggingface/transformers/issues/27646 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/27646/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/27646/comments | https://api.github.com/repos/huggingface/transformers/issues/27646/events | https://github.com/huggingface/transformers/pull/27646 | 2,005,853,813 | PR_kwDOCUB6oc5gGtHa | 27,646 | Add "accelerate" to the list of required pacakges | {
"login": "correll",
"id": 399192,
"node_id": "MDQ6VXNlcjM5OTE5Mg==",
"avatar_url": "https://avatars.githubusercontent.com/u/399192?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/correll",
"html_url": "https://github.com/correll",
"followers_url": "https://api.github.com/users/correll/fo... | [] | closed | false | null | [] | [
"I include the error message when running \"train()\". Apparently it is only needed when using torch. \r\n\r\n---------------------------------------------------------------------------\r\nImportError Traceback (most recent call last)\r\n[<ipython-input-11-6385862da3e9>](https://localh... | 1,700 | 1,700 | 1,700 | NONE | null |
# What does this PR do?
Added "accelerate" to the list of required packages
<!--
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 o... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/27646/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/27646/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/27646",
"html_url": "https://github.com/huggingface/transformers/pull/27646",
"diff_url": "https://github.com/huggingface/transformers/pull/27646.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/27646.patch",
"merged_at... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.