Datasets:

DanielPFlorian

/

Transformers-Github-Issues

like 0

{ "login": "gante", "id": 12240844, "node_id": "MDQ6VXNlcjEyMjQwODQ0", "avatar_url": "https://avatars.githubusercontent.com/u/12240844?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gante", "html_url": "https://github.com/gante", "followers_url": "https://api.github.com/users/gante/followers", "following_url": "https://api.github.com/users/gante/following{/other_user}", "gists_url": "https://api.github.com/users/gante/gists{/gist_id}", "starred_url": "https://api.github.com/users/gante/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gante/subscriptions", "organizations_url": "https://api.github.com/users/gante/orgs", "repos_url": "https://api.github.com/users/gante/repos", "events_url": "https://api.github.com/users/gante/events{/privacy}", "received_events_url": "https://api.github.com/users/gante/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28760). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "@ArthurZucker I've added a test case that shows how it was failing before 👍 It only covers one model, though, as the test has to be tailored to each model.\r\n\r\nThe test will also be useful to show to users the importance of this ugly pattern :D " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28760/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28760", "html_url": "https://github.com/huggingface/transformers/pull/28760", "diff_url": "https://github.com/huggingface/transformers/pull/28760.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28760.patch", "merged_at": 1706711948000 }

{ "login": "jmamou", "id": 19263306, "node_id": "MDQ6VXNlcjE5MjYzMzA2", "avatar_url": "https://avatars.githubusercontent.com/u/19263306?v=4", "gravatar_id": "", "url": "https://api.github.com/users/jmamou", "html_url": "https://github.com/jmamou", "followers_url": "https://api.github.com/users/jmamou/followers", "following_url": "https://api.github.com/users/jmamou/following{/other_user}", "gists_url": "https://api.github.com/users/jmamou/gists{/gist_id}", "starred_url": "https://api.github.com/users/jmamou/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/jmamou/subscriptions", "organizations_url": "https://api.github.com/users/jmamou/orgs", "repos_url": "https://api.github.com/users/jmamou/repos", "events_url": "https://api.github.com/users/jmamou/events{/privacy}", "received_events_url": "https://api.github.com/users/jmamou/received_events", "type": "User", "site_admin": false }

[]

[]

[ "> Thanks for adding this!\r\n> \r\n> For the code quality checks, you'll need to run `make fixup` and push the changes.\r\n> \r\n> From the docstring, it mentions behaviour about `\"heuristic_transient\" being reset, but I don't see logic relating to it in this diff. Does this already happen?\r\n\r\nI just fixed docstring", "@jmamou Let's try and make the CI green :) You'll need to resolve the quality checks by running `make fixup` and pushing the changes. \r\n\r\nFor the other failing tests, you'll need to try rebasing on main.", "@amyeroberts \r\ntests_torch failed. \r\nAre you familiar with the error?", "@jmamou Tbh, I'm not sure what the cause of these failures are. I would first suggest rebasing on main to make sure you have all of the most recent commits. This will trigger a re-run of the CI too.", "@amyeroberts \r\nsame test still fails 👎 ", "@jmamou Our apologies, [this PR](https://github.com/huggingface/transformers/pull/29027) fixed it. Could you try rebasing again please? 🤗 ", "@amyeroberts both CI failures seem unrelated to this PR, including `check_repository_consistency` 👀 do you have an idea of what might be causing it?", "@jmamou Apologies for all of the current issues you've been experiencing with unrelated failures on this PR. \r\n\r\nThe two current batches of failing tests should have been resolved with #29037, #29043\r\n\r\nCould you try one (final 🤞) rebase to get this CI green 🟢 ? ", "@amyeroberts \r\ntests passed on CI 👍 ", "@jmamou Thanks for this contribution and your patience with our misbehaving CI! ", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28759). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28759/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28759", "html_url": "https://github.com/huggingface/transformers/pull/28759", "diff_url": "https://github.com/huggingface/transformers/pull/28759.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28759.patch", "merged_at": 1708083898000 }

{ "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/users/amyeroberts/followers", "following_url": "https://api.github.com/users/amyeroberts/following{/other_user}", "gists_url": "https://api.github.com/users/amyeroberts/gists{/gist_id}", "starred_url": "https://api.github.com/users/amyeroberts/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/amyeroberts/subscriptions", "organizations_url": "https://api.github.com/users/amyeroberts/orgs", "repos_url": "https://api.github.com/users/amyeroberts/repos", "events_url": "https://api.github.com/users/amyeroberts/events{/privacy}", "received_events_url": "https://api.github.com/users/amyeroberts/received_events", "type": "User", "site_admin": false }

[]

[]

[ "@amyeroberts you are too fast to address my first comment, so I am not sure if you see my 2nd one\r\n\r\n\r\n\r\nsrc/transformers/dependency_versions_table.py need to be updated by running make deps_table_update\r\n", "> @amyeroberts you are too fast to address my first comment, so I am not sure if you see my 2nd one\r\n> \r\n> src/transformers/dependency_versions_table.py need to be updated by running make deps_table_update\r\n\r\nI didn't - but I've run and pushed the change now :) ", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28758). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28758/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28758", "html_url": "https://github.com/huggingface/transformers/pull/28758", "diff_url": "https://github.com/huggingface/transformers/pull/28758.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28758.patch", "merged_at": 1706541734000 }

{ "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/users/amyeroberts/followers", "following_url": "https://api.github.com/users/amyeroberts/following{/other_user}", "gists_url": "https://api.github.com/users/amyeroberts/gists{/gist_id}", "starred_url": "https://api.github.com/users/amyeroberts/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/amyeroberts/subscriptions", "organizations_url": "https://api.github.com/users/amyeroberts/orgs", "repos_url": "https://api.github.com/users/amyeroberts/repos", "events_url": "https://api.github.com/users/amyeroberts/events{/privacy}", "received_events_url": "https://api.github.com/users/amyeroberts/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28757). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28757/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28757", "html_url": "https://github.com/huggingface/transformers/pull/28757", "diff_url": "https://github.com/huggingface/transformers/pull/28757.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28757.patch", "merged_at": 1706541745000 }

{ "login": "tleyden", "id": 296876, "node_id": "MDQ6VXNlcjI5Njg3Ng==", "avatar_url": "https://avatars.githubusercontent.com/u/296876?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tleyden", "html_url": "https://github.com/tleyden", "followers_url": "https://api.github.com/users/tleyden/followers", "following_url": "https://api.github.com/users/tleyden/following{/other_user}", "gists_url": "https://api.github.com/users/tleyden/gists{/gist_id}", "starred_url": "https://api.github.com/users/tleyden/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tleyden/subscriptions", "organizations_url": "https://api.github.com/users/tleyden/orgs", "repos_url": "https://api.github.com/users/tleyden/repos", "events_url": "https://api.github.com/users/tleyden/events{/privacy}", "received_events_url": "https://api.github.com/users/tleyden/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28756). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "@ArthurZucker sounds good, I'll look into adding a test. If you know of any existing tests that I should look at, LMK!", "this one seems relevant https://github.com/huggingface/transformers/blob/main/tests/trainer/test_trainer.py#L835" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28756/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28756", "html_url": "https://github.com/huggingface/transformers/pull/28756", "diff_url": "https://github.com/huggingface/transformers/pull/28756.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28756.patch", "merged_at": null }

{ "login": "notsyncing", "id": 2649806, "node_id": "MDQ6VXNlcjI2NDk4MDY=", "avatar_url": "https://avatars.githubusercontent.com/u/2649806?v=4", "gravatar_id": "", "url": "https://api.github.com/users/notsyncing", "html_url": "https://github.com/notsyncing", "followers_url": "https://api.github.com/users/notsyncing/followers", "following_url": "https://api.github.com/users/notsyncing/following{/other_user}", "gists_url": "https://api.github.com/users/notsyncing/gists{/gist_id}", "starred_url": "https://api.github.com/users/notsyncing/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/notsyncing/subscriptions", "organizations_url": "https://api.github.com/users/notsyncing/orgs", "repos_url": "https://api.github.com/users/notsyncing/repos", "events_url": "https://api.github.com/users/notsyncing/events{/privacy}", "received_events_url": "https://api.github.com/users/notsyncing/received_events", "type": "User", "site_admin": false }

[]

[]

[ "@muellerzr would you mind taking a look? thanks!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28755/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28755", "html_url": "https://github.com/huggingface/transformers/pull/28755", "diff_url": "https://github.com/huggingface/transformers/pull/28755.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28755.patch", "merged_at": null }

{ "login": "karl-hajjar", "id": 24575019, "node_id": "MDQ6VXNlcjI0NTc1MDE5", "avatar_url": "https://avatars.githubusercontent.com/u/24575019?v=4", "gravatar_id": "", "url": "https://api.github.com/users/karl-hajjar", "html_url": "https://github.com/karl-hajjar", "followers_url": "https://api.github.com/users/karl-hajjar/followers", "following_url": "https://api.github.com/users/karl-hajjar/following{/other_user}", "gists_url": "https://api.github.com/users/karl-hajjar/gists{/gist_id}", "starred_url": "https://api.github.com/users/karl-hajjar/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/karl-hajjar/subscriptions", "organizations_url": "https://api.github.com/users/karl-hajjar/orgs", "repos_url": "https://api.github.com/users/karl-hajjar/repos", "events_url": "https://api.github.com/users/karl-hajjar/events{/privacy}", "received_events_url": "https://api.github.com/users/karl-hajjar/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The code quality check fails, saying `src/transformers/models/llama/convert_llama_weights_to_hf.py` needs to be reformatted, but I'm not sure which part is problematic.", "`make style` should automatically reformat it! Also could the history is a bit messed up, recommending you to merge and force push! ", "Yes, sorry things got messed up when I tried to merge and I didn't pay attention. I'll fix the formatting with `make style` and push.", "@ArthurZucker, @amyeroberts, I think this PR does the same as [this other PR](https://github.com/huggingface/transformers/pull/28767) which was opened 13 hours ago. I think only one of them should be kept, I'll let you decide what is the way forward considering there are duplicate PRs.", "Usually the first opened is the one we try to merge so yours in this case 😉 ", "Hi @amyeroberts, @ArthurZucker, the PR is still awaiting your review before it can be merged ! Is there anything else you would want me to add ? For example, should I also add 'code' as one of the versions for codeLLama ? ", "Hi @karl-hajjar. At the moment, it seems there are many files with diffs unrelated to this PR. From the commit history, it looks like you might have rebased and pushed, but not force pushed. When rebasing it's necessary to do `git push -f`, as it's essentially rewriting the history. ", "Yes I have rebased when it wasn't necessary unfortunately. If I force push on the next commit, do you think this should solve the issue ? ", "@karl-hajjar It shouldn't matter when you rebase - it's just going the put the commits in the PR on top of the head of the `main` branch. Yes, force pushing should resolve this. ", "Two tests are failing because of an error when importing `transformers.models.nat.modeling_nat` and `transformers.models.detr.modeling_detr`. I don't know why these fail considering the very minor changes of the previous commit ...", "The issue is still that we can't review the PR with such a huge diff 😓 once you rebase, force push to erase the history of the commits ", "@ArthurZucker do I understand correctly that you suggest I rebase again and then force push ? (I have forced push after the last commit [42744f2](https://github.com/huggingface/transformers/pull/28754/commits/42744f2c315850bd2b38f4791551476a288f8239) but I don't think it has erased the commit history)", "I would do this:\r\n- `git reset HEAD~0dfd39ebb317707c6e3186b08737337b6f34aa11`\r\n- `git add . `\r\n- `git commit -m\"force push\"`\r\n- `git push -f`\r\n", "Two tests are failing because of an error when importing `transformers.models.nat.modeling_nat` and `transformers.models.detr.modeling_detr`. I don't know why these fail considering the very minor changes of the last commit", "@amyeroberts @ArthurZucker two tests are failing but I'm not sure why. an error when importing `transformers.models.nat.modeling_nat` and `transformers.models.detr.modeling_detr`. I don't understand how the small changes I made to the llama files could impact those imports...", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28754). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28754/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28754", "html_url": "https://github.com/huggingface/transformers/pull/28754", "diff_url": "https://github.com/huggingface/transformers/pull/28754.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28754.patch", "merged_at": 1707474241000 }

{ "login": "johko", "id": 2843485, "node_id": "MDQ6VXNlcjI4NDM0ODU=", "avatar_url": "https://avatars.githubusercontent.com/u/2843485?v=4", "gravatar_id": "", "url": "https://api.github.com/users/johko", "html_url": "https://github.com/johko", "followers_url": "https://api.github.com/users/johko/followers", "following_url": "https://api.github.com/users/johko/following{/other_user}", "gists_url": "https://api.github.com/users/johko/gists{/gist_id}", "starred_url": "https://api.github.com/users/johko/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/johko/subscriptions", "organizations_url": "https://api.github.com/users/johko/orgs", "repos_url": "https://api.github.com/users/johko/repos", "events_url": "https://api.github.com/users/johko/events{/privacy}", "received_events_url": "https://api.github.com/users/johko/received_events", "type": "User", "site_admin": false }

[ { "id": 1843244711, "node_id": "MDU6TGFiZWwxODQzMjQ0NzEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model", "name": "New model", "color": "fbca04", "default": false, "description": "" } ]

[]

[ "Opened a PR which leverages [PyTorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) on the original repo: https://github.com/TonyLianLong/CrossMAE/pull/2" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28753/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "nic-olo", "id": 89006260, "node_id": "MDQ6VXNlcjg5MDA2MjYw", "avatar_url": "https://avatars.githubusercontent.com/u/89006260?v=4", "gravatar_id": "", "url": "https://api.github.com/users/nic-olo", "html_url": "https://github.com/nic-olo", "followers_url": "https://api.github.com/users/nic-olo/followers", "following_url": "https://api.github.com/users/nic-olo/following{/other_user}", "gists_url": "https://api.github.com/users/nic-olo/gists{/gist_id}", "starred_url": "https://api.github.com/users/nic-olo/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/nic-olo/subscriptions", "organizations_url": "https://api.github.com/users/nic-olo/orgs", "repos_url": "https://api.github.com/users/nic-olo/repos", "events_url": "https://api.github.com/users/nic-olo/events{/privacy}", "received_events_url": "https://api.github.com/users/nic-olo/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28752/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "Vinyzu", "id": 50874994, "node_id": "MDQ6VXNlcjUwODc0OTk0", "avatar_url": "https://avatars.githubusercontent.com/u/50874994?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Vinyzu", "html_url": "https://github.com/Vinyzu", "followers_url": "https://api.github.com/users/Vinyzu/followers", "following_url": "https://api.github.com/users/Vinyzu/following{/other_user}", "gists_url": "https://api.github.com/users/Vinyzu/gists{/gist_id}", "starred_url": "https://api.github.com/users/Vinyzu/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Vinyzu/subscriptions", "organizations_url": "https://api.github.com/users/Vinyzu/orgs", "repos_url": "https://api.github.com/users/Vinyzu/repos", "events_url": "https://api.github.com/users/Vinyzu/events{/privacy}", "received_events_url": "https://api.github.com/users/Vinyzu/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28751). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28751/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28751", "html_url": "https://github.com/huggingface/transformers/pull/28751", "diff_url": "https://github.com/huggingface/transformers/pull/28751.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28751.patch", "merged_at": 1706522812000 }

{ "login": "dashapetr", "id": 54349415, "node_id": "MDQ6VXNlcjU0MzQ5NDE1", "avatar_url": "https://avatars.githubusercontent.com/u/54349415?v=4", "gravatar_id": "", "url": "https://api.github.com/users/dashapetr", "html_url": "https://github.com/dashapetr", "followers_url": "https://api.github.com/users/dashapetr/followers", "following_url": "https://api.github.com/users/dashapetr/following{/other_user}", "gists_url": "https://api.github.com/users/dashapetr/gists{/gist_id}", "starred_url": "https://api.github.com/users/dashapetr/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/dashapetr/subscriptions", "organizations_url": "https://api.github.com/users/dashapetr/orgs", "repos_url": "https://api.github.com/users/dashapetr/repos", "events_url": "https://api.github.com/users/dashapetr/events{/privacy}", "received_events_url": "https://api.github.com/users/dashapetr/received_events", "type": "User", "site_admin": false }

[ { "id": 2648621985, "node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1", "url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request", "name": "Feature request", "color": "FBCA04", "default": false, "description": "Request for a new feature" } ]

[]

[ "> Thanks! Won't this default to `max_new_tokens=20` that is set in generate if the arg is not passed?\r\n\r\nI am not sure about this.\r\n\r\nI didn't manage to find what exact place the limitation of `max_new_token `<= 192 comes from. If we look at the tool level, each separate tool (model) has its own limitation of `max_new_tokens`. If we refer to [API's documentation](https://huggingface.co/docs/api-inference/detailed_parameters#detailed-parameters), there is no default value for `max_new_tokens`.\r\n\r\nOne more possible solution I see here is to set `max_new_tokens`=192 so the validation error won't appear.", "So the HFAgent and the Local agent who both use huggingface tools will be limited by the `max_new_tokens` that you give them / by the default `max_new_token` / `max_length` of the model on the hub. I think the goal of the agents where to have simple APIs so 200 tokens default seemed alright for me. Otherwise it can never stop for a lot of model. \r\nI'll add a feature request for setting this but think we should keep it that way for now!", "> I think the goal of the agents where to have simple APIs so 200 tokens default seemed alright for me. Otherwise it can never stop for a lot of model.\r\n\r\nAgree with this. But the issue is still here: \r\n\r\n\r\n\r\nI believe it is because the HFAgent limits `max_new_tokens` to **200**, but the `starcoder` API needs **192** tokens or less.\r\n", "Would you suggest adding a controllable argument? " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28750/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28750", "html_url": "https://github.com/huggingface/transformers/pull/28750", "diff_url": "https://github.com/huggingface/transformers/pull/28750.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28750.patch", "merged_at": null }

{ "login": "Arwa491", "id": 117582570, "node_id": "U_kgDOBwIq6g", "avatar_url": "https://avatars.githubusercontent.com/u/117582570?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Arwa491", "html_url": "https://github.com/Arwa491", "followers_url": "https://api.github.com/users/Arwa491/followers", "following_url": "https://api.github.com/users/Arwa491/following{/other_user}", "gists_url": "https://api.github.com/users/Arwa491/gists{/gist_id}", "starred_url": "https://api.github.com/users/Arwa491/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Arwa491/subscriptions", "organizations_url": "https://api.github.com/users/Arwa491/orgs", "repos_url": "https://api.github.com/users/Arwa491/repos", "events_url": "https://api.github.com/users/Arwa491/events{/privacy}", "received_events_url": "https://api.github.com/users/Arwa491/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @Arwa491, thanks for raising an issue! \r\n\r\nThis is a question best placed in our [forums](https://discuss.huggingface.co/). We try to reserve the github issues for feature requests and bug reports." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28749/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "KaifAhmad1", "id": 98801504, "node_id": "U_kgDOBeOXYA", "avatar_url": "https://avatars.githubusercontent.com/u/98801504?v=4", "gravatar_id": "", "url": "https://api.github.com/users/KaifAhmad1", "html_url": "https://github.com/KaifAhmad1", "followers_url": "https://api.github.com/users/KaifAhmad1/followers", "following_url": "https://api.github.com/users/KaifAhmad1/following{/other_user}", "gists_url": "https://api.github.com/users/KaifAhmad1/gists{/gist_id}", "starred_url": "https://api.github.com/users/KaifAhmad1/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/KaifAhmad1/subscriptions", "organizations_url": "https://api.github.com/users/KaifAhmad1/orgs", "repos_url": "https://api.github.com/users/KaifAhmad1/repos", "events_url": "https://api.github.com/users/KaifAhmad1/events{/privacy}", "received_events_url": "https://api.github.com/users/KaifAhmad1/received_events", "type": "User", "site_admin": false }

[]

[]

[ "I would recommend you to set `CUDA_LAUNCH_BLOCKING=1` and add a breakpoint to see where this is happening. Usually at the embedding layer, when the tokenizer has an extra token and the vocab was not resized 🤗 " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28748/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "wendlerc", "id": 8095362, "node_id": "MDQ6VXNlcjgwOTUzNjI=", "avatar_url": "https://avatars.githubusercontent.com/u/8095362?v=4", "gravatar_id": "", "url": "https://api.github.com/users/wendlerc", "html_url": "https://github.com/wendlerc", "followers_url": "https://api.github.com/users/wendlerc/followers", "following_url": "https://api.github.com/users/wendlerc/following{/other_user}", "gists_url": "https://api.github.com/users/wendlerc/gists{/gist_id}", "starred_url": "https://api.github.com/users/wendlerc/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/wendlerc/subscriptions", "organizations_url": "https://api.github.com/users/wendlerc/orgs", "repos_url": "https://api.github.com/users/wendlerc/repos", "events_url": "https://api.github.com/users/wendlerc/events{/privacy}", "received_events_url": "https://api.github.com/users/wendlerc/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Ok I guess that's because they implement a linear attention." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28747/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "lh0x00", "id": 9839768, "node_id": "MDQ6VXNlcjk4Mzk3Njg=", "avatar_url": "https://avatars.githubusercontent.com/u/9839768?v=4", "gravatar_id": "", "url": "https://api.github.com/users/lh0x00", "html_url": "https://github.com/lh0x00", "followers_url": "https://api.github.com/users/lh0x00/followers", "following_url": "https://api.github.com/users/lh0x00/following{/other_user}", "gists_url": "https://api.github.com/users/lh0x00/gists{/gist_id}", "starred_url": "https://api.github.com/users/lh0x00/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/lh0x00/subscriptions", "organizations_url": "https://api.github.com/users/lh0x00/orgs", "repos_url": "https://api.github.com/users/lh0x00/repos", "events_url": "https://api.github.com/users/lh0x00/events{/privacy}", "received_events_url": "https://api.github.com/users/lh0x00/received_events", "type": "User", "site_admin": false }

[]

[]

[ "hiii, any feedback are welcome, I waiting to a new version to use instead of a fork version or give me feedback if my pr having any issues. many thanks @pacman100 @muellerzr ", "Nice! Assuming this works, really excited for it to get merged in, as I've had this issue with checkpoints.\r\n\r\nQuick one - it looks like your local formatter may have different line length settings to the default for `transformers`, since the diff shows a bunch of changes that aren't relevant to your updates. Up to you if you want to revert those, but I'd hate for them to hold up the PR :)\r\n\r\n(I don't expect it to be a problem since the style and quality checks passed, but it would make the PR look smaller and keep the file consistent)", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28746). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "cc @younesbelkada", "yeah, @nathan-az I was able to continue my fine tuning, and it worked well. You can use it through my branch while waiting for the PR to be merged.", "waiting for this merge ", "me too, waiting for this merge -.-", "Hi everyone! Sure yes, please give us some time to review properly this PR and we'll make sure the fix is going to be landed! ", "thanks for reviewed, @pacman100 @younesbelkada " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28746/reactions", "total_count": 3, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 1, "rocket": 2, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28746", "html_url": "https://github.com/huggingface/transformers/pull/28746", "diff_url": "https://github.com/huggingface/transformers/pull/28746.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28746.patch", "merged_at": 1706705907000 }

{ "login": "FanaticPythoner", "id": 45826736, "node_id": "MDQ6VXNlcjQ1ODI2NzM2", "avatar_url": "https://avatars.githubusercontent.com/u/45826736?v=4", "gravatar_id": "", "url": "https://api.github.com/users/FanaticPythoner", "html_url": "https://github.com/FanaticPythoner", "followers_url": "https://api.github.com/users/FanaticPythoner/followers", "following_url": "https://api.github.com/users/FanaticPythoner/following{/other_user}", "gists_url": "https://api.github.com/users/FanaticPythoner/gists{/gist_id}", "starred_url": "https://api.github.com/users/FanaticPythoner/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/FanaticPythoner/subscriptions", "organizations_url": "https://api.github.com/users/FanaticPythoner/orgs", "repos_url": "https://api.github.com/users/FanaticPythoner/repos", "events_url": "https://api.github.com/users/FanaticPythoner/events{/privacy}", "received_events_url": "https://api.github.com/users/FanaticPythoner/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @FanaticPythoner, thanks for opening this feature request! \r\n\r\nPydantic v2 is supported on the development branch c.f. #28728 #27933 \r\n\r\nTo use, install from source: `pip install git+https://github.com/huggingface/transformers`\r\n", "Great thank you. Is there an ETA for the next release?", "The schedule for releases is around once a month. The latest release was last week, so probably around 4 weeks or so. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28745/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "YourSaDady", "id": 99607923, "node_id": "U_kgDOBe_lcw", "avatar_url": "https://avatars.githubusercontent.com/u/99607923?v=4", "gravatar_id": "", "url": "https://api.github.com/users/YourSaDady", "html_url": "https://github.com/YourSaDady", "followers_url": "https://api.github.com/users/YourSaDady/followers", "following_url": "https://api.github.com/users/YourSaDady/following{/other_user}", "gists_url": "https://api.github.com/users/YourSaDady/gists{/gist_id}", "starred_url": "https://api.github.com/users/YourSaDady/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/YourSaDady/subscriptions", "organizations_url": "https://api.github.com/users/YourSaDady/orgs", "repos_url": "https://api.github.com/users/YourSaDady/repos", "events_url": "https://api.github.com/users/YourSaDady/events{/privacy}", "received_events_url": "https://api.github.com/users/YourSaDady/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @YourSaDady, thanks for raising this issue! \r\n\r\nIf the program is killed when using `offload_folder` it would indicate that you're hitting OOM. If you look at your GPU memory utilization - `watch -n1 nvidia-smi` - and your CPU utilization e.g. with`top` are you able to see these topping out when loading the model?" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28744/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "kiehls90", "id": 101498700, "node_id": "U_kgDOBgy_TA", "avatar_url": "https://avatars.githubusercontent.com/u/101498700?v=4", "gravatar_id": "", "url": "https://api.github.com/users/kiehls90", "html_url": "https://github.com/kiehls90", "followers_url": "https://api.github.com/users/kiehls90/followers", "following_url": "https://api.github.com/users/kiehls90/following{/other_user}", "gists_url": "https://api.github.com/users/kiehls90/gists{/gist_id}", "starred_url": "https://api.github.com/users/kiehls90/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/kiehls90/subscriptions", "organizations_url": "https://api.github.com/users/kiehls90/orgs", "repos_url": "https://api.github.com/users/kiehls90/repos", "events_url": "https://api.github.com/users/kiehls90/events{/privacy}", "received_events_url": "https://api.github.com/users/kiehls90/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @kiehls90, thanks for raising an issue! \r\n\r\nThe scripts under `research_examples` aren't actively maintained. If a dataset is very large, you can try [streaming it](https://huggingface.co/docs/datasets/stream) instead of downloading the whole thing. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28743/reactions", "total_count": 1, "+1": 1, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "tamanna-mostafa", "id": 156403336, "node_id": "U_kgDOCVKGiA", "avatar_url": "https://avatars.githubusercontent.com/u/156403336?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tamanna-mostafa", "html_url": "https://github.com/tamanna-mostafa", "followers_url": "https://api.github.com/users/tamanna-mostafa/followers", "following_url": "https://api.github.com/users/tamanna-mostafa/following{/other_user}", "gists_url": "https://api.github.com/users/tamanna-mostafa/gists{/gist_id}", "starred_url": "https://api.github.com/users/tamanna-mostafa/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tamanna-mostafa/subscriptions", "organizations_url": "https://api.github.com/users/tamanna-mostafa/orgs", "repos_url": "https://api.github.com/users/tamanna-mostafa/repos", "events_url": "https://api.github.com/users/tamanna-mostafa/events{/privacy}", "received_events_url": "https://api.github.com/users/tamanna-mostafa/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @tamanna-mostafa 👋 \r\n\r\nYou'll have to help us figure out what's wrong: can you get us a short and reproducible script that showcases the issue on the `transformers` size? I see two exceptions in your pasted code, one about `text-generation-inference` and another about `safetensors`", "@gante \r\nThanks for your comments. Here are the codes I ran (please let me know if you need any further details):\r\n\r\n```\r\n#Config for SFT\r\nmistral-7b-sft-MM-RLAIF:\r\n dtype: bf16\r\n log_dir: \"mistral-7b-sft-MM-PS\"\r\n learning_rate: 2e-5\r\n model_name: /mnt/efs/workspace/sakhaki/models/Mistral-7B-v0.1\r\n deepspeed_config: configs/zero_config_sft_65b.json #configs/zero_config_pretrain.json\r\n output_dir: /mnt/efs/data/tammosta/files_t/output_sft_32k\r\n weight_decay: 0.01\r\n max_length: 4096\r\n warmup_steps: 100\r\n gradient_checkpointing: true\r\n gradient_accumulation_steps: 8\r\n per_device_train_batch_size: 1\r\n per_device_eval_batch_size: 1\r\n eval_steps: 500000\r\n save_steps: 100\r\n num_train_epochs: 2\r\n save_total_limit: 4\r\n use_flash_attention: false\r\n residual_dropout: 0.0\r\n residual_dropout_lima: true\r\n save_strategy: steps\r\n peft_model: false\r\n only_last_turn_loss: false\r\n use_custom_sampler: true\r\n datasets:\r\n - sft-custom:\r\n data_files: /mnt/efs/data/tammosta/files_t/SFT_inp_26787_RBS_plus_Optima.json\r\n #fraction : 0.75\r\n max_val_set: 300\r\n val_split: 0.0001\r\n - oasst_export:\r\n lang: \"bg,ca,cs,da,de,en,es,fr,hr,hu,it,nl,pl,pt,ro,ru,sl,sr,sv,uk\" # sft-8.0\r\n hf_dataset_name: OpenAssistant/oasst1\r\n fraction : 0.5\r\n val_split: 0.0001\r\n max_val_set: 300\r\n top_k: 1\r\n#run SFT on mistral 7b model\r\ndeepspeed trainer_sft_d.py --configs mistral-7b-sft-MM-RLAIF --wandb-entity tammosta --show_dataset_stats --deepspeed\r\n\r\n#Run DPO on the SFT model\r\naccelerate launch --config_file ./accelerate_configs/ds_zero3.yaml rlhf_dpo.py \\\r\n--model_name_or_path=\"/mnt/efs/data/tammosta/files_t/output_sft_32k\" \\\r\n--output_dir=\"/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k\" \\\r\n--data_path=\"/mnt/efs/data/tammosta/files_t/DPO_data_rbs_clean_AIF.json\" \\\r\n--use_lamma2_peft_config False \\\r\n--beta 0.1 \\\r\n--optimizer_type adamw_hf \\\r\n--learning_rate 1e-6 \\\r\n--warmup_steps 50 \\\r\n--per_device_train_batch_size 1 \\\r\n--per_device_eval_batch_size 1 \\\r\n--gradient_accumulation_steps 8 \\\r\n--lora_alpha 16 \\\r\n--lora_dropout 0.05 \\\r\n--lora_r 8 \\\r\n--max_prompt_length 2048 \\\r\n--max_length 4096 \\\r\n--num_train_epochs 4 \\\r\n--logging_steps 20 \\\r\n--save_steps 100 \\\r\n--save_total_limit 8 \\\r\n--eval_steps 50 \\\r\n--gradient_checkpointing True \\\r\n--report_to \"wandb\"\r\n\r\n```\r\n#Contents of the DPO output folder\r\n```\r\nubuntu@ip-172-31-8-218:/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k$ ls \r\nREADME.md adapter_model.safetensors checkpoint-100 checkpoint-300 checkpoint-500 checkpoint-700 global_step736 special_tokens_map.json tokenizer.model training_args.bin\r\nadapter_config.json added_tokens.json checkpoint-200 checkpoint-400 checkpoint-600 final_checkpoint latest tokenizer.json tokenizer_config.json zero_to_fp32.py\r\n\r\n```\r\n```\r\n# merge the lora adaptors\r\npython merge_peft_adaptors_gpu.py --base_model_name_or_path /mnt/efs/data/tammosta/files_t/output_sft_32k --peft_model_path /mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k --output_dir /mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k_merged --safe_serialization\r\n```\r\n\r\n```\r\n#Content of merge_peft_adators_gpu.py\r\n\r\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\r\nfrom peft import PeftModel\r\nimport torch\r\n\r\nimport os\r\nimport argparse\r\n\r\ndef get_args():\r\n parser = argparse.ArgumentParser()\r\n parser.add_argument(\"--base_model_name_or_path\", type=str)\r\n parser.add_argument(\"--peft_model_path\", type=str)\r\n parser.add_argument(\"--output_dir\", type=str)\r\n parser.add_argument(\"--device\", type=str, default=\"auto\")\r\n parser.add_argument(\"--safe_serialization\", action=\"store_true\")\r\n\r\n return parser.parse_args()\r\n####\r\ndef main():\r\n args = get_args()\r\n\r\n if args.device == 'auto':\r\n device_arg = { 'device_map': 'auto' }\r\n else:\r\n device_arg = { 'device_map': { \"\": args.device} }\r\n\r\n print(f\"Loading base model: {args.base_model_name_or_path}\")\r\n base_model = AutoModelForCausalLM.from_pretrained(\r\n args.base_model_name_or_path,\r\n return_dict=True,\r\n torch_dtype=torch.float16,\r\n trust_remote_code=True,\r\n **device_arg\r\n )\r\n #device = torch.device('cpu')\r\n #base_model.to(device)\r\n\r\n print(f\"Loading PEFT: {args.peft_model_path}\")\r\n model = PeftModel.from_pretrained(base_model, args.peft_model_path)\r\n print(\"Peft Model : \", model.device)\r\n print(f\"Running merge_and_unload\")\r\n model = model.merge_and_unload()\r\n\r\n tokenizer = AutoTokenizer.from_pretrained(args.base_model_name_or_path)\r\n\r\n\r\n model.save_pretrained(f\"{args.output_dir}\",max_shard_size='9GB',safe_serialization=args.safe_serialization)\r\n tokenizer.save_pretrained(f\"{args.output_dir}\",max_shard_size='9GB',safe_serialization=args.safe_serialization)\r\n print(f\"Model saved to {args.output_dir}\")\r\n####\r\nif __name__ == \"__main__\" :\r\n main()\r\n\r\n```\r\n```\r\n#The error I get while running the code above \r\n\r\nLoading base model: /mnt/efs/data/tammosta/files_t/output_sft_32k\r\nLoading checkpoint shards: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:04<00:00, 1.40s/it]\r\nLoading PEFT: /mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/scripts_hb/merge_peft_adaptors_gpu.py\", line 51, in <module>\r\n main()\r\n File \"/mnt/efs/data/tammosta/scripts_hb/merge_peft_adaptors_gpu.py\", line 38, in main\r\n model = PeftModel.from_pretrained(base_model, args.peft_model_path)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 352, in from_pretrained\r\n model.load_adapter(model_id, adapter_name, is_trainable=is_trainable, **kwargs)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 689, in load_adapter\r\n adapters_weights = load_peft_weights(model_id, device=torch_device, **hf_hub_download_kwargs)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/utils/save_and_load.py\", line 270, in load_peft_weights\r\n adapters_weights = safe_load_file(filename, device=device)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/safetensors/torch.py\", line 308, in load_file\r\n with safe_open(filename, framework=\"pt\", device=device) as f:\r\nsafetensors_rust.SafetensorError: Error while deserializing header: InvalidHeaderDeserialization\r\n\r\n\r\n```", "Hi @tamanna-mostafa 👋 looking at your stack trace, it looks like a `peft` error, you should open an issue there :)", "@gante\r\nIssue opened here: https://github.com/huggingface/peft/issues/1443", "Closing this ticket since the issue is reported in the ticket above. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28742/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "khipp", "id": 9824526, "node_id": "MDQ6VXNlcjk4MjQ1MjY=", "avatar_url": "https://avatars.githubusercontent.com/u/9824526?v=4", "gravatar_id": "", "url": "https://api.github.com/users/khipp", "html_url": "https://github.com/khipp", "followers_url": "https://api.github.com/users/khipp/followers", "following_url": "https://api.github.com/users/khipp/following{/other_user}", "gists_url": "https://api.github.com/users/khipp/gists{/gist_id}", "starred_url": "https://api.github.com/users/khipp/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/khipp/subscriptions", "organizations_url": "https://api.github.com/users/khipp/orgs", "repos_url": "https://api.github.com/users/khipp/repos", "events_url": "https://api.github.com/users/khipp/events{/privacy}", "received_events_url": "https://api.github.com/users/khipp/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28741). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28741/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28741", "html_url": "https://github.com/huggingface/transformers/pull/28741", "diff_url": "https://github.com/huggingface/transformers/pull/28741.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28741.patch", "merged_at": 1706522791000 }

{ "login": "michaelgruner", "id": 717880, "node_id": "MDQ6VXNlcjcxNzg4MA==", "avatar_url": "https://avatars.githubusercontent.com/u/717880?v=4", "gravatar_id": "", "url": "https://api.github.com/users/michaelgruner", "html_url": "https://github.com/michaelgruner", "followers_url": "https://api.github.com/users/michaelgruner/followers", "following_url": "https://api.github.com/users/michaelgruner/following{/other_user}", "gists_url": "https://api.github.com/users/michaelgruner/gists{/gist_id}", "starred_url": "https://api.github.com/users/michaelgruner/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/michaelgruner/subscriptions", "organizations_url": "https://api.github.com/users/michaelgruner/orgs", "repos_url": "https://api.github.com/users/michaelgruner/repos", "events_url": "https://api.github.com/users/michaelgruner/events{/privacy}", "received_events_url": "https://api.github.com/users/michaelgruner/received_events", "type": "User", "site_admin": false }

[]

{ "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/users/amyeroberts/followers", "following_url": "https://api.github.com/users/amyeroberts/following{/other_user}", "gists_url": "https://api.github.com/users/amyeroberts/gists{/gist_id}", "starred_url": "https://api.github.com/users/amyeroberts/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/amyeroberts/subscriptions", "organizations_url": "https://api.github.com/users/amyeroberts/orgs", "repos_url": "https://api.github.com/users/amyeroberts/repos", "events_url": "https://api.github.com/users/amyeroberts/events{/privacy}", "received_events_url": "https://api.github.com/users/amyeroberts/received_events", "type": "User", "site_admin": false }

[ { "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/users/amyeroberts/followers", "following_url": "https://api.github.com/users/amyeroberts/following{/other_user}", "gists_url": "https://api.github.com/users/amyeroberts/gists{/gist_id}", "starred_url": "https://api.github.com/users/amyeroberts/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/amyeroberts/subscriptions", "organizations_url": "https://api.github.com/users/amyeroberts/orgs", "repos_url": "https://api.github.com/users/amyeroberts/repos", "events_url": "https://api.github.com/users/amyeroberts/events{/privacy}", "received_events_url": "https://api.github.com/users/amyeroberts/received_events", "type": "User", "site_admin": false } ]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28740/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "stevhliu", "id": 59462357, "node_id": "MDQ6VXNlcjU5NDYyMzU3", "avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4", "gravatar_id": "", "url": "https://api.github.com/users/stevhliu", "html_url": "https://github.com/stevhliu", "followers_url": "https://api.github.com/users/stevhliu/followers", "following_url": "https://api.github.com/users/stevhliu/following{/other_user}", "gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}", "starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions", "organizations_url": "https://api.github.com/users/stevhliu/orgs", "repos_url": "https://api.github.com/users/stevhliu/repos", "events_url": "https://api.github.com/users/stevhliu/events{/privacy}", "received_events_url": "https://api.github.com/users/stevhliu/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28739). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Hmm not sure why the test is failing with `ERROR dummy.py - AttributeError: 'HfDoctestModule' object has no attribute 'getini'`. Any idea how to fix or if I can go ahead and merge?", "we can merge but let's rebase on main it should fix it!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28739/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28739", "html_url": "https://github.com/huggingface/transformers/pull/28739", "diff_url": "https://github.com/huggingface/transformers/pull/28739.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28739.patch", "merged_at": 1706807777000 }

{ "login": "goelayu", "id": 31916840, "node_id": "MDQ6VXNlcjMxOTE2ODQw", "avatar_url": "https://avatars.githubusercontent.com/u/31916840?v=4", "gravatar_id": "", "url": "https://api.github.com/users/goelayu", "html_url": "https://github.com/goelayu", "followers_url": "https://api.github.com/users/goelayu/followers", "following_url": "https://api.github.com/users/goelayu/following{/other_user}", "gists_url": "https://api.github.com/users/goelayu/gists{/gist_id}", "starred_url": "https://api.github.com/users/goelayu/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/goelayu/subscriptions", "organizations_url": "https://api.github.com/users/goelayu/orgs", "repos_url": "https://api.github.com/users/goelayu/repos", "events_url": "https://api.github.com/users/goelayu/events{/privacy}", "received_events_url": "https://api.github.com/users/goelayu/received_events", "type": "User", "site_admin": false }

[ { "id": 2648621985, "node_id": "MDU6TGFiZWwyNjQ4NjIxOTg1", "url": "https://api.github.com/repos/huggingface/transformers/labels/Feature%20request", "name": "Feature request", "color": "FBCA04", "default": false, "description": "Request for a new feature" } ]

[]

[ "cc @ArthurZucker " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28738/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "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/users/amyeroberts/followers", "following_url": "https://api.github.com/users/amyeroberts/following{/other_user}", "gists_url": "https://api.github.com/users/amyeroberts/gists{/gist_id}", "starred_url": "https://api.github.com/users/amyeroberts/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/amyeroberts/subscriptions", "organizations_url": "https://api.github.com/users/amyeroberts/orgs", "repos_url": "https://api.github.com/users/amyeroberts/repos", "events_url": "https://api.github.com/users/amyeroberts/events{/privacy}", "received_events_url": "https://api.github.com/users/amyeroberts/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28737). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28737/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28737", "html_url": "https://github.com/huggingface/transformers/pull/28737", "diff_url": "https://github.com/huggingface/transformers/pull/28737.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28737.patch", "merged_at": 1706454614000 }

{ "login": "RezaYazdaniAminabadi", "id": 44502768, "node_id": "MDQ6VXNlcjQ0NTAyNzY4", "avatar_url": "https://avatars.githubusercontent.com/u/44502768?v=4", "gravatar_id": "", "url": "https://api.github.com/users/RezaYazdaniAminabadi", "html_url": "https://github.com/RezaYazdaniAminabadi", "followers_url": "https://api.github.com/users/RezaYazdaniAminabadi/followers", "following_url": "https://api.github.com/users/RezaYazdaniAminabadi/following{/other_user}", "gists_url": "https://api.github.com/users/RezaYazdaniAminabadi/gists{/gist_id}", "starred_url": "https://api.github.com/users/RezaYazdaniAminabadi/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/RezaYazdaniAminabadi/subscriptions", "organizations_url": "https://api.github.com/users/RezaYazdaniAminabadi/orgs", "repos_url": "https://api.github.com/users/RezaYazdaniAminabadi/repos", "events_url": "https://api.github.com/users/RezaYazdaniAminabadi/events{/privacy}", "received_events_url": "https://api.github.com/users/RezaYazdaniAminabadi/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28736/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28736", "html_url": "https://github.com/huggingface/transformers/pull/28736", "diff_url": "https://github.com/huggingface/transformers/pull/28736.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28736.patch", "merged_at": null }

{ "login": "sanchit-gandhi", "id": 93869735, "node_id": "U_kgDOBZhWpw", "avatar_url": "https://avatars.githubusercontent.com/u/93869735?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sanchit-gandhi", "html_url": "https://github.com/sanchit-gandhi", "followers_url": "https://api.github.com/users/sanchit-gandhi/followers", "following_url": "https://api.github.com/users/sanchit-gandhi/following{/other_user}", "gists_url": "https://api.github.com/users/sanchit-gandhi/gists{/gist_id}", "starred_url": "https://api.github.com/users/sanchit-gandhi/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sanchit-gandhi/subscriptions", "organizations_url": "https://api.github.com/users/sanchit-gandhi/orgs", "repos_url": "https://api.github.com/users/sanchit-gandhi/repos", "events_url": "https://api.github.com/users/sanchit-gandhi/events{/privacy}", "received_events_url": "https://api.github.com/users/sanchit-gandhi/received_events", "type": "User", "site_admin": false }

[]

[]

[ "cc @ArthurZucker for possible up-coming Flax model additions", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28735). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28735/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28735", "html_url": "https://github.com/huggingface/transformers/pull/28735", "diff_url": "https://github.com/huggingface/transformers/pull/28735.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28735.patch", "merged_at": 1706293240000 }

{ "login": "Rocketknight1", "id": 12866554, "node_id": "MDQ6VXNlcjEyODY2NTU0", "avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Rocketknight1", "html_url": "https://github.com/Rocketknight1", "followers_url": "https://api.github.com/users/Rocketknight1/followers", "following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}", "gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}", "starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions", "organizations_url": "https://api.github.com/users/Rocketknight1/orgs", "repos_url": "https://api.github.com/users/Rocketknight1/repos", "events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}", "received_events_url": "https://api.github.com/users/Rocketknight1/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28734). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "This has now been rebased following the tf-keras PR, will merge once the CI is green." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28734/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28734", "html_url": "https://github.com/huggingface/transformers/pull/28734", "diff_url": "https://github.com/huggingface/transformers/pull/28734.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28734.patch", "merged_at": 1706707122000 }

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28733). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Thanks @Wauplin Will update it (along with `top_k` too maybe!)", "FYI:\r\n\r\nAlso changed\r\n\r\n> Assign labels to the image(s) passed as inputs.\r\n\r\nto \r\n\r\n> Predict the depth(s) of the image(s) passed as inputs.\r\n\r\nand removed `top_k` part\r\n\r\nhttps://github.com/huggingface/transformers/pull/28733/commits/22ff25c6cec0eba56ea78646c3b88ec9f728e86a" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28733/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28733", "html_url": "https://github.com/huggingface/transformers/pull/28733", "diff_url": "https://github.com/huggingface/transformers/pull/28733.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28733.patch", "merged_at": 1706521376000 }

{ "login": "zhukpm", "id": 52332744, "node_id": "MDQ6VXNlcjUyMzMyNzQ0", "avatar_url": "https://avatars.githubusercontent.com/u/52332744?v=4", "gravatar_id": "", "url": "https://api.github.com/users/zhukpm", "html_url": "https://github.com/zhukpm", "followers_url": "https://api.github.com/users/zhukpm/followers", "following_url": "https://api.github.com/users/zhukpm/following{/other_user}", "gists_url": "https://api.github.com/users/zhukpm/gists{/gist_id}", "starred_url": "https://api.github.com/users/zhukpm/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/zhukpm/subscriptions", "organizations_url": "https://api.github.com/users/zhukpm/orgs", "repos_url": "https://api.github.com/users/zhukpm/repos", "events_url": "https://api.github.com/users/zhukpm/events{/privacy}", "received_events_url": "https://api.github.com/users/zhukpm/received_events", "type": "User", "site_admin": false }

[]

[]

[ "This seems like a duplicate of https://github.com/huggingface/transformers/issues/25420#issuecomment-1775317535 ", "Yeah, it is. Thanks" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28732/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "VINUK0", "id": 58259367, "node_id": "MDQ6VXNlcjU4MjU5MzY3", "avatar_url": "https://avatars.githubusercontent.com/u/58259367?v=4", "gravatar_id": "", "url": "https://api.github.com/users/VINUK0", "html_url": "https://github.com/VINUK0", "followers_url": "https://api.github.com/users/VINUK0/followers", "following_url": "https://api.github.com/users/VINUK0/following{/other_user}", "gists_url": "https://api.github.com/users/VINUK0/gists{/gist_id}", "starred_url": "https://api.github.com/users/VINUK0/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/VINUK0/subscriptions", "organizations_url": "https://api.github.com/users/VINUK0/orgs", "repos_url": "https://api.github.com/users/VINUK0/repos", "events_url": "https://api.github.com/users/VINUK0/events{/privacy}", "received_events_url": "https://api.github.com/users/VINUK0/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @VINUK0, thanks for raising this issue! \r\n\r\nI believe this is a torch and hardware issue, rather than something in the transformers library. \r\n\r\nOther people have reported issues between torch and sdp, when loading in bf16 on T4. See [this issue](https://github.com/Lightning-AI/lit-gpt/issues/327) and [comment](https://github.com/Lightning-AI/lit-gpt/issues/327#issuecomment-1664674460), which others said helped resolve their issue. \r\n\r\nOther related issues:\r\n* https://github.com/pytorch/pytorch/issues/102029#issuecomment-1560071148\r\n* https://github.com/fishaudio/fish-speech/issues/7\r\n\r\nI'd suggest trying to load your model in fp16 instead of bf16. \r\n\r\ncc @fxmarty for reference", "Thank you, I can reproduce the issue on a T4:\r\n\r\n```python\r\nimport torch\r\n\r\nquery = torch.rand(32, 8, 128, 64, dtype=torch.bfloat16, device=\"cuda\")\r\nkey = torch.rand(32, 8, 128, 64, dtype=torch.bfloat16, device=\"cuda\")\r\nvalue = torch.rand(32, 8, 128, 64, dtype=torch.bfloat16, device=\"cuda\")\r\n\r\nres = torch.nn.functional.scaled_dot_product_attention(query,key,value)\r\n```\r\n\r\nUsing `torch.backends.cuda.enable_mem_efficient_sdp(False)` fixes the issue (https://pytorch.org/docs/master/backends.html#torch.backends.cuda.enable_mem_efficient_sdp) and can be used as a temporary workaround. Alternatively, you can load the model with the `attn_implementation=\"eager\"` argument to avoid using SDPA: `model = AutoModel.from_pretrained(..., attn_implementation=\"eager\")`.\r\n\r\n@drisspg Do you think this is a bug? Should the auto-dispatch of SDPA automatically fall back on the math path in case it is used with a device not supporting bf16?", "This definitely was a bug, but should have been fixed with this PR: https://github.com/pytorch/pytorch/pull/116272\r\nDo you know which version you are on for PyTorch when you repoed above?", "This was with torch==2.1.2, and appears to be fixed in torch==2.2.0. @VINUK0 Please update to 2.2.0 or use `torch.backends.cuda.enable_mem_efficient_sdp(False)`. Thank you!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28731/reactions", "total_count": 2, "+1": 2, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "akbayt", "id": 11097700, "node_id": "MDQ6VXNlcjExMDk3NzAw", "avatar_url": "https://avatars.githubusercontent.com/u/11097700?v=4", "gravatar_id": "", "url": "https://api.github.com/users/akbayt", "html_url": "https://github.com/akbayt", "followers_url": "https://api.github.com/users/akbayt/followers", "following_url": "https://api.github.com/users/akbayt/following{/other_user}", "gists_url": "https://api.github.com/users/akbayt/gists{/gist_id}", "starred_url": "https://api.github.com/users/akbayt/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/akbayt/subscriptions", "organizations_url": "https://api.github.com/users/akbayt/orgs", "repos_url": "https://api.github.com/users/akbayt/repos", "events_url": "https://api.github.com/users/akbayt/events{/privacy}", "received_events_url": "https://api.github.com/users/akbayt/received_events", "type": "User", "site_admin": false }

[ { "id": 1843244711, "node_id": "MDU6TGFiZWwxODQzMjQ0NzEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model", "name": "New model", "color": "fbca04", "default": false, "description": "" } ]

[]

[ "Hey @akbayt. Thanks a lot for opening this new model request and contributing to the HF ecosystem! 🤗\r\nWe have recently been trying to push for `model on the hub` and have as much support as we can there. It will also be easier to integrate it. Here is a [tutorial](https://huggingface.co/docs/transformers/custom_models) if that sound good to you!", "Hi @amyeroberts. Thanks for your quick response and guidance. As far as I can see, sharing the models on hub will make it easier for me and will add a slight work when someone wants to use it. This is still acceptable. However, the models I have trained so far are very modest in scale (you can see the details in the paper), and I am actively seeking support and resources to train larger models. I believe that having my model integrated into the 🤗 transformers library, rather than sharing it on the hub, could increase its visibility within the community. In turn, it might help me find some support easier. Thanks again.", "The best way to help make your model visible is by making sure the model is easy to find (be on the hub), easy to use (has a detailed model card with code snippet users showing how to get started) and a space demo showing the model's capabilities. Many models have become popular that were just on the hub. For example, falcon and phi-2 started on the hub. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28730/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "osanseviero", "id": 7246357, "node_id": "MDQ6VXNlcjcyNDYzNTc=", "avatar_url": "https://avatars.githubusercontent.com/u/7246357?v=4", "gravatar_id": "", "url": "https://api.github.com/users/osanseviero", "html_url": "https://github.com/osanseviero", "followers_url": "https://api.github.com/users/osanseviero/followers", "following_url": "https://api.github.com/users/osanseviero/following{/other_user}", "gists_url": "https://api.github.com/users/osanseviero/gists{/gist_id}", "starred_url": "https://api.github.com/users/osanseviero/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/osanseviero/subscriptions", "organizations_url": "https://api.github.com/users/osanseviero/orgs", "repos_url": "https://api.github.com/users/osanseviero/repos", "events_url": "https://api.github.com/users/osanseviero/events{/privacy}", "received_events_url": "https://api.github.com/users/osanseviero/received_events", "type": "User", "site_admin": false }

[]

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[ { "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false } ]

[ "Thanks @osanseviero ! Indeed, will open a PR to fix this" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28729/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28728). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28728/reactions", "total_count": 1, "+1": 0, "-1": 0, "laugh": 0, "hooray": 1, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28728", "html_url": "https://github.com/huggingface/transformers/pull/28728", "diff_url": "https://github.com/huggingface/transformers/pull/28728.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28728.patch", "merged_at": 1706287173000 }

{ "login": "xkszltl", "id": 5203025, "node_id": "MDQ6VXNlcjUyMDMwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/5203025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/xkszltl", "html_url": "https://github.com/xkszltl", "followers_url": "https://api.github.com/users/xkszltl/followers", "following_url": "https://api.github.com/users/xkszltl/following{/other_user}", "gists_url": "https://api.github.com/users/xkszltl/gists{/gist_id}", "starred_url": "https://api.github.com/users/xkszltl/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/xkszltl/subscriptions", "organizations_url": "https://api.github.com/users/xkszltl/orgs", "repos_url": "https://api.github.com/users/xkszltl/repos", "events_url": "https://api.github.com/users/xkszltl/events{/privacy}", "received_events_url": "https://api.github.com/users/xkszltl/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28727). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Nice catch on checking on GH !", "Sounds good to me, added.", "BTW why 4.40?\r\nAny convention and consideration regarding the deprecation cycle?", "> BTW why 4.40? Any convention and consideration regarding the deprecation cycle?\r\n\r\nWe typically deprecate in 2 release cycles and this commit will be part of 4.38. ", "@xkszltl For the failing tests on CI, running `make fixup` and pushing the changes should resolve the issue and trigger a new CI run", "> @xkszltl For the failing tests on CI, running `make fixup` and pushing the changes should resolve the issue and trigger a new CI run\r\n\r\nWould be great if we can actually lint in CI (not just check) and print out the git diff.\r\nThis way we don't have to run anything locally and can make whatever change needed directly from the log.\r\n\r\nAnd that's also how I usually setup linter CI, based on observation that not everyone likes the extra step.", "> This way we don't have to run anything locally and can make whatever change needed directly from the log.\r\n\r\nIsn't having to look at a diff in the CI run and manually apply the changes more work than just having the linter run locally? ", "> Isn't having to look at a diff in the CI run and manually apply the changes more work than just having the linter run locally?\r\n\r\nGood question.\r\nMy opinion is: not always.\r\n- For people continuously developing on the project, yes, a simple linter command is the easiest.\r\n- For people occasionally joining, or currently focus on other projects, spending extra effort to setup another environment locally with possibly conflict dependencies is a lot more than copy-pasting diff from CI log.", "Fixed.\r\nAlso change to log only once or it may be too annoying." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28727/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28727", "html_url": "https://github.com/huggingface/transformers/pull/28727", "diff_url": "https://github.com/huggingface/transformers/pull/28727.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28727.patch", "merged_at": 1706541900000 }

{ "login": "hungdinhxuan", "id": 79694464, "node_id": "MDQ6VXNlcjc5Njk0NDY0", "avatar_url": "https://avatars.githubusercontent.com/u/79694464?v=4", "gravatar_id": "", "url": "https://api.github.com/users/hungdinhxuan", "html_url": "https://github.com/hungdinhxuan", "followers_url": "https://api.github.com/users/hungdinhxuan/followers", "following_url": "https://api.github.com/users/hungdinhxuan/following{/other_user}", "gists_url": "https://api.github.com/users/hungdinhxuan/gists{/gist_id}", "starred_url": "https://api.github.com/users/hungdinhxuan/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/hungdinhxuan/subscriptions", "organizations_url": "https://api.github.com/users/hungdinhxuan/orgs", "repos_url": "https://api.github.com/users/hungdinhxuan/repos", "events_url": "https://api.github.com/users/hungdinhxuan/events{/privacy}", "received_events_url": "https://api.github.com/users/hungdinhxuan/received_events", "type": "User", "site_admin": false }

[]

[]

[ "cc @sanchit-gandhi @ylacombe " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28726/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28725). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28725/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28725", "html_url": "https://github.com/huggingface/transformers/pull/28725", "diff_url": "https://github.com/huggingface/transformers/pull/28725.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28725.patch", "merged_at": 1706270449000 }

{ "login": "fxmarty", "id": 9808326, "node_id": "MDQ6VXNlcjk4MDgzMjY=", "avatar_url": "https://avatars.githubusercontent.com/u/9808326?v=4", "gravatar_id": "", "url": "https://api.github.com/users/fxmarty", "html_url": "https://github.com/fxmarty", "followers_url": "https://api.github.com/users/fxmarty/followers", "following_url": "https://api.github.com/users/fxmarty/following{/other_user}", "gists_url": "https://api.github.com/users/fxmarty/gists{/gist_id}", "starred_url": "https://api.github.com/users/fxmarty/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/fxmarty/subscriptions", "organizations_url": "https://api.github.com/users/fxmarty/orgs", "repos_url": "https://api.github.com/users/fxmarty/repos", "events_url": "https://api.github.com/users/fxmarty/events{/privacy}", "received_events_url": "https://api.github.com/users/fxmarty/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28724). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Thank you @amyeroberts, added a test that fails on the current `main` due to the controlflow not being correctly captured." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28724/reactions", "total_count": 1, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 1 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28724", "html_url": "https://github.com/huggingface/transformers/pull/28724", "diff_url": "https://github.com/huggingface/transformers/pull/28724.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28724.patch", "merged_at": 1706777102000 }

{ "login": "tomaarsen", "id": 37621491, "node_id": "MDQ6VXNlcjM3NjIxNDkx", "avatar_url": "https://avatars.githubusercontent.com/u/37621491?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tomaarsen", "html_url": "https://github.com/tomaarsen", "followers_url": "https://api.github.com/users/tomaarsen/followers", "following_url": "https://api.github.com/users/tomaarsen/following{/other_user}", "gists_url": "https://api.github.com/users/tomaarsen/gists{/gist_id}", "starred_url": "https://api.github.com/users/tomaarsen/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tomaarsen/subscriptions", "organizations_url": "https://api.github.com/users/tomaarsen/orgs", "repos_url": "https://api.github.com/users/tomaarsen/repos", "events_url": "https://api.github.com/users/tomaarsen/events{/privacy}", "received_events_url": "https://api.github.com/users/tomaarsen/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @tomaarsen, thanks for raising this! \r\n\r\nIndeed, this shouldn't be happening. Looking into it", "Looking into it, this a consequence of https://github.com/huggingface/transformers/pull/27282 - when `weights_only=True`, then [_legacy_load is used by torch to load the state dict](https://github.com/pytorch/pytorch/blob/a72190fd51f19cbfb5c09ae3088729f94aef7141/torch/serialization.py#L1036). The error is thrown if the `storage_dtype` is type e.g. `torch.FloatStorage`.\r\n\r\ncc'ing in @Narsil who knows more about the tensors and how they're stored. In particular, in the warning message we have: \r\n\r\n```\r\n/data/ml/lib/python3.10/site-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly. To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()\r\n```\r\n\r\nWill this cause issue for use? I'm unsure what is meant by \"using storages directly\". ", "If it's not relevant for us (e.g. if we do not \"use storages directly\", then we could adopt this to avoid the warning: https://github.com/pytorch/pytorch/blob/a72190fd51f19cbfb5c09ae3088729f94aef7141/torch/storage.py#L475C9-L475C36\r\n", "No it's just using a legacy load for this particular file.\r\nNewer files can load perfectly with weights_only=True: https://github.com/pytorch/pytorch/blob/a72190fd51f19cbfb5c09ae3088729f94aef7141/torch/serialization.py#L1016\r\n\r\nThis should be fixed upstream, as I doubt it's our calls that trigger the warning." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28723/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "kevin3314", "id": 37268015, "node_id": "MDQ6VXNlcjM3MjY4MDE1", "avatar_url": "https://avatars.githubusercontent.com/u/37268015?v=4", "gravatar_id": "", "url": "https://api.github.com/users/kevin3314", "html_url": "https://github.com/kevin3314", "followers_url": "https://api.github.com/users/kevin3314/followers", "following_url": "https://api.github.com/users/kevin3314/following{/other_user}", "gists_url": "https://api.github.com/users/kevin3314/gists{/gist_id}", "starred_url": "https://api.github.com/users/kevin3314/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/kevin3314/subscriptions", "organizations_url": "https://api.github.com/users/kevin3314/orgs", "repos_url": "https://api.github.com/users/kevin3314/repos", "events_url": "https://api.github.com/users/kevin3314/events{/privacy}", "received_events_url": "https://api.github.com/users/kevin3314/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28722/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "Bachstelze", "id": 19904888, "node_id": "MDQ6VXNlcjE5OTA0ODg4", "avatar_url": "https://avatars.githubusercontent.com/u/19904888?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Bachstelze", "html_url": "https://github.com/Bachstelze", "followers_url": "https://api.github.com/users/Bachstelze/followers", "following_url": "https://api.github.com/users/Bachstelze/following{/other_user}", "gists_url": "https://api.github.com/users/Bachstelze/gists{/gist_id}", "starred_url": "https://api.github.com/users/Bachstelze/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Bachstelze/subscriptions", "organizations_url": "https://api.github.com/users/Bachstelze/orgs", "repos_url": "https://api.github.com/users/Bachstelze/repos", "events_url": "https://api.github.com/users/Bachstelze/events{/privacy}", "received_events_url": "https://api.github.com/users/Bachstelze/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @Bachstelze, thanks for raising an issue! \r\n\r\nThe `EncoderDecoder` models are composite models which use `AutoModel` to load the encoder and decoder respectively. As per [the BertGeneration docs](https://github.com/huggingface/transformers/blob/bbe30c6968c66bc77f7f1f246e64743d74419770/docs/source/en/model_doc/bert-generation.md?plain=1#L89), you can load the model using:\r\n\r\n```py\r\nfrom transformers import EncoderDecoderModel\r\nmodel = EncoderDecoderModel.from_pretrained(\"Bachstelze/instructionRoberta-base\", output_attentions=True)\r\n```", "@amyeroberts Yes it is possible to load it as EncoderDecoderModel, though many libraries load generic models just with the Automodel, so EncoderDecoderModels yield an error." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28721/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "StrivedTye", "id": 19620650, "node_id": "MDQ6VXNlcjE5NjIwNjUw", "avatar_url": "https://avatars.githubusercontent.com/u/19620650?v=4", "gravatar_id": "", "url": "https://api.github.com/users/StrivedTye", "html_url": "https://github.com/StrivedTye", "followers_url": "https://api.github.com/users/StrivedTye/followers", "following_url": "https://api.github.com/users/StrivedTye/following{/other_user}", "gists_url": "https://api.github.com/users/StrivedTye/gists{/gist_id}", "starred_url": "https://api.github.com/users/StrivedTye/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/StrivedTye/subscriptions", "organizations_url": "https://api.github.com/users/StrivedTye/orgs", "repos_url": "https://api.github.com/users/StrivedTye/repos", "events_url": "https://api.github.com/users/StrivedTye/events{/privacy}", "received_events_url": "https://api.github.com/users/StrivedTye/received_events", "type": "User", "site_admin": false }

[]

[]

[ "cc @ydshieh ", "Sorry, indeed, it's my bad\r\n\r\n#28506 didn't fix #27282 properly, I will open a PR", "@StrivedTye Fix is merged. I tried it and it works. Let me know if you have further issue with this. Thanks for reporting the issue." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28720/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "velaia", "id": 1515904, "node_id": "MDQ6VXNlcjE1MTU5MDQ=", "avatar_url": "https://avatars.githubusercontent.com/u/1515904?v=4", "gravatar_id": "", "url": "https://api.github.com/users/velaia", "html_url": "https://github.com/velaia", "followers_url": "https://api.github.com/users/velaia/followers", "following_url": "https://api.github.com/users/velaia/following{/other_user}", "gists_url": "https://api.github.com/users/velaia/gists{/gist_id}", "starred_url": "https://api.github.com/users/velaia/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/velaia/subscriptions", "organizations_url": "https://api.github.com/users/velaia/orgs", "repos_url": "https://api.github.com/users/velaia/repos", "events_url": "https://api.github.com/users/velaia/events{/privacy}", "received_events_url": "https://api.github.com/users/velaia/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @velaia, thanks for contributing to improving our docs! \r\n\r\nCould you combine the changes in the PR with the others in #28718? ", "Hi @amyeroberts , combining was what I originally had in mind but didn't know how. I think I did it somehow by making the changes in my fork of the repo. Can you check pls?", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28719). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28719/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28719", "html_url": "https://github.com/huggingface/transformers/pull/28719", "diff_url": "https://github.com/huggingface/transformers/pull/28719.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28719.patch", "merged_at": 1706270337000 }

{ "login": "velaia", "id": 1515904, "node_id": "MDQ6VXNlcjE1MTU5MDQ=", "avatar_url": "https://avatars.githubusercontent.com/u/1515904?v=4", "gravatar_id": "", "url": "https://api.github.com/users/velaia", "html_url": "https://github.com/velaia", "followers_url": "https://api.github.com/users/velaia/followers", "following_url": "https://api.github.com/users/velaia/following{/other_user}", "gists_url": "https://api.github.com/users/velaia/gists{/gist_id}", "starred_url": "https://api.github.com/users/velaia/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/velaia/subscriptions", "organizations_url": "https://api.github.com/users/velaia/orgs", "repos_url": "https://api.github.com/users/velaia/repos", "events_url": "https://api.github.com/users/velaia/events{/privacy}", "received_events_url": "https://api.github.com/users/velaia/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28718/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28718", "html_url": "https://github.com/huggingface/transformers/pull/28718", "diff_url": "https://github.com/huggingface/transformers/pull/28718.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28718.patch", "merged_at": null }

{ "login": "ShukantPal", "id": 22450567, "node_id": "MDQ6VXNlcjIyNDUwNTY3", "avatar_url": "https://avatars.githubusercontent.com/u/22450567?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ShukantPal", "html_url": "https://github.com/ShukantPal", "followers_url": "https://api.github.com/users/ShukantPal/followers", "following_url": "https://api.github.com/users/ShukantPal/following{/other_user}", "gists_url": "https://api.github.com/users/ShukantPal/gists{/gist_id}", "starred_url": "https://api.github.com/users/ShukantPal/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ShukantPal/subscriptions", "organizations_url": "https://api.github.com/users/ShukantPal/orgs", "repos_url": "https://api.github.com/users/ShukantPal/repos", "events_url": "https://api.github.com/users/ShukantPal/events{/privacy}", "received_events_url": "https://api.github.com/users/ShukantPal/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28717). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28717/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28717", "html_url": "https://github.com/huggingface/transformers/pull/28717", "diff_url": "https://github.com/huggingface/transformers/pull/28717.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28717.patch", "merged_at": 1706270374000 }

{ "login": "Mickls", "id": 41884581, "node_id": "MDQ6VXNlcjQxODg0NTgx", "avatar_url": "https://avatars.githubusercontent.com/u/41884581?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Mickls", "html_url": "https://github.com/Mickls", "followers_url": "https://api.github.com/users/Mickls/followers", "following_url": "https://api.github.com/users/Mickls/following{/other_user}", "gists_url": "https://api.github.com/users/Mickls/gists{/gist_id}", "starred_url": "https://api.github.com/users/Mickls/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Mickls/subscriptions", "organizations_url": "https://api.github.com/users/Mickls/orgs", "repos_url": "https://api.github.com/users/Mickls/repos", "events_url": "https://api.github.com/users/Mickls/events{/privacy}", "received_events_url": "https://api.github.com/users/Mickls/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @Mickls, thanks for raising this issue! \r\n\r\nA fix was merged in #28637. Could you try installing transformers from source and see if this resolves the issue for you? \r\n\r\n`pip install git+https://github.com/huggingface/transformers`", "> Hi @Mickls, thanks for raising this issue!\r\n> \r\n> A fix was merged in #28637. Could you try installing transformers from source and see if this resolves the issue for you?\r\n> \r\n> `pip install git+https://github.com/huggingface/transformers`\r\n\r\nYes it works fine now, thanks" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28716/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "stevhliu", "id": 59462357, "node_id": "MDQ6VXNlcjU5NDYyMzU3", "avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4", "gravatar_id": "", "url": "https://api.github.com/users/stevhliu", "html_url": "https://github.com/stevhliu", "followers_url": "https://api.github.com/users/stevhliu/followers", "following_url": "https://api.github.com/users/stevhliu/following{/other_user}", "gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}", "starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions", "organizations_url": "https://api.github.com/users/stevhliu/orgs", "repos_url": "https://api.github.com/users/stevhliu/repos", "events_url": "https://api.github.com/users/stevhliu/events{/privacy}", "received_events_url": "https://api.github.com/users/stevhliu/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28715/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28715", "html_url": "https://github.com/huggingface/transformers/pull/28715", "diff_url": "https://github.com/huggingface/transformers/pull/28715.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28715.patch", "merged_at": 1706290147000 }

{ "login": "ekgren", "id": 1921821, "node_id": "MDQ6VXNlcjE5MjE4MjE=", "avatar_url": "https://avatars.githubusercontent.com/u/1921821?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ekgren", "html_url": "https://github.com/ekgren", "followers_url": "https://api.github.com/users/ekgren/followers", "following_url": "https://api.github.com/users/ekgren/following{/other_user}", "gists_url": "https://api.github.com/users/ekgren/gists{/gist_id}", "starred_url": "https://api.github.com/users/ekgren/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ekgren/subscriptions", "organizations_url": "https://api.github.com/users/ekgren/orgs", "repos_url": "https://api.github.com/users/ekgren/repos", "events_url": "https://api.github.com/users/ekgren/events{/privacy}", "received_events_url": "https://api.github.com/users/ekgren/received_events", "type": "User", "site_admin": false }

[]

[]

[ "#26678 fixed this, I can't push everything to the hub but Llama tokenizer will have a fix soon. \r\nThis is a duplicate of #26455", "Thank you for all the hard work @ArthurZucker, closing this issue then!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28714/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "DavidAfonsoValente", "id": 74915610, "node_id": "MDQ6VXNlcjc0OTE1NjEw", "avatar_url": "https://avatars.githubusercontent.com/u/74915610?v=4", "gravatar_id": "", "url": "https://api.github.com/users/DavidAfonsoValente", "html_url": "https://github.com/DavidAfonsoValente", "followers_url": "https://api.github.com/users/DavidAfonsoValente/followers", "following_url": "https://api.github.com/users/DavidAfonsoValente/following{/other_user}", "gists_url": "https://api.github.com/users/DavidAfonsoValente/gists{/gist_id}", "starred_url": "https://api.github.com/users/DavidAfonsoValente/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/DavidAfonsoValente/subscriptions", "organizations_url": "https://api.github.com/users/DavidAfonsoValente/orgs", "repos_url": "https://api.github.com/users/DavidAfonsoValente/repos", "events_url": "https://api.github.com/users/DavidAfonsoValente/events{/privacy}", "received_events_url": "https://api.github.com/users/DavidAfonsoValente/received_events", "type": "User", "site_admin": false }

[]

[]

[ "@DavidAfonsoValente Thanks for opening a PR and adding FA2 for this model! Let us know when the PR is ready to be reviewed 🤗 ", "> @DavidAfonsoValente Thanks for opening a PR and adding FA2 for this model! Let us know when the PR is ready to be reviewed 🤗\r\n\r\nThanks! I believe it's ready to be reviewed.", "The tests that are failing are due to the changes made in the functions that are labeled as:\r\n Copied from transformers.models.roberta.modeling_roberta.RobertaSelfAttention with Roberta->XLMRoberta\r\n\r\nWhat is the standard procedure when I have made changes to these functions in order to accommodate the new feature?\r\n", "@DavidAfonsoValente If you click on the CI runs, you'll see the error messages relating to the code quality failures. These detail how to resolve the issue: run `make fix-copies` and push the changes. \r\n\r\nIf another model is copying this model's attention classes, then it's great because you're adding FA2 for two models 🥳 You just need to apply the equivalent changes to the model that's copying. ", "Hello there! \r\n\r\nI'm working on integrating scaled_dot_product_attention to BERT #28802, and there might be some merge conflicts with this change. \r\n\r\nMost of the changes I'm making will propagate through to XML-RoBERTa (through fix-copies). From what I can, our approaches are a bit different from this in 2 areas:\r\n1. I've kept the `#Copies from` line at the class level, so any changes to BERT's init() will be propagated through to the fix-copies classes. I've modified the `#Copies from` line to use the corresponding ATTENTION_CLASS as necessary.\r\n2. I've created a BertSdpaSelfAttention to go with BertSelfAttention, instead of creating a BertSdpaAttention. The rationale behind this is that the code in BertSdpaSelfAttention is adapted from the code in BertSelfAttention, so I figured it's better to name it as such.\r\n\r\nLet me know if you have any questions about these. I think we should discuss about which approach is better and adopt it. Would you mind looking through #28802 and let me know what you think?", "> Hello there!\r\n> \r\n> I'm working on integrating scaled_dot_product_attention to BERT #28802, and there might be some merge conflicts with this change.\r\n> \r\n> Most of the changes I'm making will propagate through to XML-RoBERTa (through fix-copies). From what I can, our approaches are a bit different from this in 2 areas:\r\n> \r\n> 1. I've kept the `#Copies from` line at the class level, so any changes to BERT's init() will be propagated through to the fix-copies classes. I've modified the `#Copies from` line to use the corresponding ATTENTION_CLASS as necessary.\r\n> 2. I've created a BertSdpaSelfAttention to go with BertSelfAttention, instead of creating a BertSdpaAttention. The rationale behind this is that the code in BertSdpaSelfAttention is adapted from the code in BertSelfAttention, so I figured it's better to name it as such.\r\n> \r\n> Let me know if you have any questions about these. I think we should discuss about which approach is better and adopt it. Would you mind looking through #28802 and let me know what you think?\r\n\r\nIt seems like your implementation doesnt create complicated conflitcts, it should be simple to merge, what do you think?", "It should be simple to merge. I think the main question is whether or not we want to use the XLM_ROBERTA_SELF_ATTENTION_CLASSES or XLM_ROBERTA_ATTENTION_CLASSES approach?", "I think in order to be consistent with other models we should keep the name XLM_ROBERTA_ATTENTION_CLASSES since this is how other models have their attention classes named.", "Thanks for your input.\r\n\r\nI am leaning towards using the \"SelfAttention\" convention in this case, because XLMRobertaFlashAttention2 actually tries to copy the logic of both XLMRobertaAttention and XLMRobertaSelfAttention into one. I think it's cleaner to have a XLMRobertaSelfFlashAttention2 that mirrors XLMRobertaSelfAttention, and then reuse the logic inside XLMRobertaAttention for both types of self attentions.\r\n\r\nReusing the code in XLMRobertaAttention should help avoid some bugs. Judging from the existing code, I think there's already a bit of an issue with the way self.output() is called in XLMRobertaFlashAttention2 (I think you need more than the call to self.output.dense(), otherwise you're missing the dropout and LayerNorm).\r\n", "Okk, maybe it'll be easier if you merge your PR and I work on top of your version.", "Thanks :thumbsup: I'll iterate as fast as possible to get it merged, and will let you know as soon as it's done!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28713/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28713", "html_url": "https://github.com/huggingface/transformers/pull/28713", "diff_url": "https://github.com/huggingface/transformers/pull/28713.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28713.patch", "merged_at": null }

{ "login": "Rocketknight1", "id": 12866554, "node_id": "MDQ6VXNlcjEyODY2NTU0", "avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Rocketknight1", "html_url": "https://github.com/Rocketknight1", "followers_url": "https://api.github.com/users/Rocketknight1/followers", "following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}", "gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}", "starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions", "organizations_url": "https://api.github.com/users/Rocketknight1/orgs", "repos_url": "https://api.github.com/users/Rocketknight1/repos", "events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}", "received_events_url": "https://api.github.com/users/Rocketknight1/received_events", "type": "User", "site_admin": false }

[]

[]

[ "@amyeroberts sorry, let me explain what I meant with the model output being 'used'!\r\n\r\nThe issue occurs when the main `BlipModel` calls the `BlipText` sub-model, and the sub-model returns a `ModelOutput` which the main `BlipModel` reads values from. As far as I can tell, the issue does **not** arise if the `BlipText` model is called directly. So this seems to be some issue with the TF compiler not really understanding what kind of object a `ModelOutput` is when it has to compile through one inside a model `call()`.", "I ran the slow tests and all passed!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28712/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28712", "html_url": "https://github.com/huggingface/transformers/pull/28712", "diff_url": "https://github.com/huggingface/transformers/pull/28712.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28712.patch", "merged_at": 1706271750000 }

{ "login": "molbap", "id": 39954772, "node_id": "MDQ6VXNlcjM5OTU0Nzcy", "avatar_url": "https://avatars.githubusercontent.com/u/39954772?v=4", "gravatar_id": "", "url": "https://api.github.com/users/molbap", "html_url": "https://github.com/molbap", "followers_url": "https://api.github.com/users/molbap/followers", "following_url": "https://api.github.com/users/molbap/following{/other_user}", "gists_url": "https://api.github.com/users/molbap/gists{/gist_id}", "starred_url": "https://api.github.com/users/molbap/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/molbap/subscriptions", "organizations_url": "https://api.github.com/users/molbap/orgs", "repos_url": "https://api.github.com/users/molbap/repos", "events_url": "https://api.github.com/users/molbap/events{/privacy}", "received_events_url": "https://api.github.com/users/molbap/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28711). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28711/reactions", "total_count": 2, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 2, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28711", "html_url": "https://github.com/huggingface/transformers/pull/28711", "diff_url": "https://github.com/huggingface/transformers/pull/28711.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28711.patch", "merged_at": null }

{ "login": "DavidAfonsoValente", "id": 74915610, "node_id": "MDQ6VXNlcjc0OTE1NjEw", "avatar_url": "https://avatars.githubusercontent.com/u/74915610?v=4", "gravatar_id": "", "url": "https://api.github.com/users/DavidAfonsoValente", "html_url": "https://github.com/DavidAfonsoValente", "followers_url": "https://api.github.com/users/DavidAfonsoValente/followers", "following_url": "https://api.github.com/users/DavidAfonsoValente/following{/other_user}", "gists_url": "https://api.github.com/users/DavidAfonsoValente/gists{/gist_id}", "starred_url": "https://api.github.com/users/DavidAfonsoValente/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/DavidAfonsoValente/subscriptions", "organizations_url": "https://api.github.com/users/DavidAfonsoValente/orgs", "repos_url": "https://api.github.com/users/DavidAfonsoValente/repos", "events_url": "https://api.github.com/users/DavidAfonsoValente/events{/privacy}", "received_events_url": "https://api.github.com/users/DavidAfonsoValente/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28710/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28710", "html_url": "https://github.com/huggingface/transformers/pull/28710", "diff_url": "https://github.com/huggingface/transformers/pull/28710.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28710.patch", "merged_at": null }

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28709). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28709/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28709", "html_url": "https://github.com/huggingface/transformers/pull/28709", "diff_url": "https://github.com/huggingface/transformers/pull/28709.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28709.patch", "merged_at": 1706256153000 }

{ "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.github.com/users/nileshkokane01/followers", "following_url": "https://api.github.com/users/nileshkokane01/following{/other_user}", "gists_url": "https://api.github.com/users/nileshkokane01/gists{/gist_id}", "starred_url": "https://api.github.com/users/nileshkokane01/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/nileshkokane01/subscriptions", "organizations_url": "https://api.github.com/users/nileshkokane01/orgs", "repos_url": "https://api.github.com/users/nileshkokane01/repos", "events_url": "https://api.github.com/users/nileshkokane01/events{/privacy}", "received_events_url": "https://api.github.com/users/nileshkokane01/received_events", "type": "User", "site_admin": false }

[]

[]

[ "@younesbelkada ,\r\nCan you please review the changes? ", "@younesbelkada ,\r\nI rebased and resolved the conflict. I hope its a right way, or else let me know.", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28708). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "cc @younesbelkada ", "@ArthurZucker IMO this is not really breaking, it is even the opposite as it fixes some subtle bugs with respect to training with BLIP- see @NielsRogge ' comment here: https://github.com/huggingface/transformers/issues/28167#issuecomment-1867398976", "thinking a bit about it, indeed maybe we should make that configurable through a variable in the config so that potentially users could revert to original behaviour if needed. \r\n@nileshkokane01 would be happy to adjust the PR accordingly? You just need to add a new variable `label_smoothing` in the blip config class and set it to 0", "@younesbelkada sure! I'll do that.", "thank you @nileshkokane01 !", "@younesbelkada ,\r\n\r\nDo I have to change the nll loss to the following as well: \r\n\r\n`\r\nloss_fct = CrossEntropyLoss(reduction=reduction, label_smoothing=self.config.lable_smoothing)`", "hi @nileshkokane01 \r\nYes please, this sounds great", "@younesbelkada ,\r\ncan you have a look ? " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28708/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28708", "html_url": "https://github.com/huggingface/transformers/pull/28708", "diff_url": "https://github.com/huggingface/transformers/pull/28708.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28708.patch", "merged_at": 1708390335000 }

{ "login": "oroojlooy", "id": 20797260, "node_id": "MDQ6VXNlcjIwNzk3MjYw", "avatar_url": "https://avatars.githubusercontent.com/u/20797260?v=4", "gravatar_id": "", "url": "https://api.github.com/users/oroojlooy", "html_url": "https://github.com/oroojlooy", "followers_url": "https://api.github.com/users/oroojlooy/followers", "following_url": "https://api.github.com/users/oroojlooy/following{/other_user}", "gists_url": "https://api.github.com/users/oroojlooy/gists{/gist_id}", "starred_url": "https://api.github.com/users/oroojlooy/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/oroojlooy/subscriptions", "organizations_url": "https://api.github.com/users/oroojlooy/orgs", "repos_url": "https://api.github.com/users/oroojlooy/repos", "events_url": "https://api.github.com/users/oroojlooy/events{/privacy}", "received_events_url": "https://api.github.com/users/oroojlooy/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @oroojlooy, thanks for raising this issue! \r\n\r\nI believe there's an accumulation of gradients happening due to the multiple forward passes on the model. \r\n\r\nPutting the pipeline calls in the `torch.no_grad` context should help:\r\n\r\n```py\r\nwith torch.no_grad():\r\n outputs = self.pipeline(prompt, max_new_tokens=self.max_new_tokens, do_sample=self.do_sample, temperature=self.temperature, top_k=self.top_k, top_p=self.top_p)\r\n```\r\n\r\nAs a side note - if do_sample=False, then parameters like top_p won't have any effect on the generation. ", "Hi @amyeroberts \r\nThanks for your reply. I tried adding `torch.no_grad`, it does not help and the memory keeps increasing. I also tried running the model via TGI which is supposed to manage the process efficiently, still the memory increases with that; although, no CUDA memory error when I use that. The CUDA memory usage with TGI increases up to around 39Gb on all cores and stays there. \r\n\r\nCould not be this related to the structure of MixTral 8*7B, where it has eight expert models? i.e., I observe a memory jump when one of the expert models gets loaded? \r\n\r\nAlso thanks for the tip about `do_sample`!", "Hi @oroojlooy, \r\n\r\nYou'll need to provide some more details about how the memory increases, in particular for the non-TGI case: is there a sudden spike? Does it go up and down? After how many calls do you see this increase? \r\n\r\nNote: in the case of generation, you're making autoregressive calls to the model i.e. the model is being repeatedly called with an increasing input length. If options such as `use_cache` aren't selected, then you would expect a memory increase, even after the model has been loaded. \r\n\r\nI'd suggest not using the pipeline, and using the modeling code directly. This will give you more control and enable you to monitor better what is causing increases in memory. \r\n\r\n```py\r\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\r\nimport torch\r\n\r\n\r\nclass GenerationModel:\r\n def __init__(self, model_id, temperature=0.0, max_new_tokens=356, do_sample=False, top_k=50, top_p=0.7):\r\n self.model = AutoModelForCausalLM.from_pretrained(model_id, device_map=\"auto\", torch_dtype=torch.float16)\r\n self.tokenizer = AutoTokenizer.from_pretrained(model_id)\r\n self.temperature = temperature\r\n self.max_new_tokens = max_new_tokens\r\n self.do_sample = do_sample\r\n self.top_k = top_k\r\n self.top_p = top_p\r\n\r\n if do_sample and temperature == 0.0:\r\n raise ValueError(\r\n \"`temperature` (=0.0) has to be a strictly positive float, otherwise your next token scores will be \"\r\n \"invalid. If you're looking for greedy decoding strategies, set `do_sample=False`\")\r\n\r\n def __call__(self, raw_messages: str) -> str:\r\n \"\"\"\r\n An example of message is:\r\n messages = [{\"role\": \"user\", \"content\": \"Explain what a Mixture of Experts is in less than 100 words.\"}]\r\n \"\"\"\r\n try:\r\n messages = [{\"role\": \"user\", \"content\": raw_messages}]\r\n prompt = self.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)\r\n inputs = self.tokenizer(prompt, return_tensors=\"pt\").to(self.model.device)\r\n with torch.no_grad():\r\n outputs = self.model.generate(**inputs, max_length=len(prompt[0]) + self.max_new_tokens, use_cache=True)\r\n\r\n generated_text = self.tokenizer.decode(outputs[0], skip_special_tokens=True)\r\n\r\n return generated_text\r\n except Exception as e:\r\n print(e)\r\n\r\n\r\nif __name__ == \"__main__\":\r\n model = GenerationModel(\r\n model_id=\"mistralai/Mixtral-8x7B-Instruct-v0.1\",\r\n temperature=0.0,\r\n max_new_tokens=356,\r\n do_sample=False,\r\n )\r\n messages = \"Explain what a Mixture of Experts is in less than 100 words.\"\r\n out = model(messages)\r\n print(out)\r\n```\r\n\r\ncc @gante The generation wizard who will know more about getting this to run well\r\n\r\n> Could not be this related to the structure of MixTral 8*7B, where it has eight expert models? i.e., I observe a memory jump when one of the expert models gets loaded?\r\n\r\nYou can run an experiment with a non-MoE model and see :) \r\n\r\n\r\n", "Hi @oroojlooy :wave:\r\n\r\nAs @amyeroberts wrote, the memory consumption in `transformers` is expected to grow throughout generation (i.e. the pipeline call in your script), as the input/output grows longer. This is because we don't pre-allocate memory, contrarily to TGI (that's why you see a fixed memory footprint after the model gets loaded). It is also independent of being a MoE model, it's how text generation works.\r\n\r\nTo confirm that there is no memory leak, you can try a simple test: call your pipeline repeatedly with the same input and with `do_sample=False`. You should not see memory increases as you repeat the calls.", "Thanks @amyeroberts and @gante for the replies. \r\n@amyeroberts I tried non-MoE model as you suggested and I. can confirm what you pointed out! \r\n@gante I actually have tried what you suggested and did not see any memory jump. \r\n\r\nBut, I am still confused about how the inference works within language models. I understanding was, for a fixed batch-size, the memory usage of the network should be fixed. Because, regardless of how big is an input+output, the max-len of input+output is always capped by the context size of the LLM, which a memory is allocated for that we load the model. So, for batch-size `b`, the memory utilization of the model is equal to `dtype_size*b*num_params` + `dtype_size*num_operations`. Should not this memory-size utilization be fixed through the inference time? \r\n\r\n\r\n", "@oroojlooy to understand why the memory grows (and what you can do about it), have a look at [this guide](https://huggingface.co/docs/transformers/llm_tutorial_optimization) -- especially section 2, which covers the self-attention layer :)", "@gante So, if I understand correctly, the matrix `QK^T` with different size of `N^2` stays at cache, and that cause the surge of memory usage? \r\n", "@oroojlooy There are sources of memory requirements increase as the sequence length (`N`) increases when caches are used:\r\n1. The materialization of the `QK^T` multiplication, which may grow as quickly as `N^2` (flash attention decreases this), as you wrote;\r\n2. The cached key and values, which grow linearly with `N`.\r\n\r\n`transformers` is eager, it only allocates when needed. TGI checks the maximum possible memory usage at startup time. Both will have roughly the same peak memory usage, for a given model/maximum input length/attention implementation." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28707/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ylacombe", "id": 52246514, "node_id": "MDQ6VXNlcjUyMjQ2NTE0", "avatar_url": "https://avatars.githubusercontent.com/u/52246514?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ylacombe", "html_url": "https://github.com/ylacombe", "followers_url": "https://api.github.com/users/ylacombe/followers", "following_url": "https://api.github.com/users/ylacombe/following{/other_user}", "gists_url": "https://api.github.com/users/ylacombe/gists{/gist_id}", "starred_url": "https://api.github.com/users/ylacombe/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ylacombe/subscriptions", "organizations_url": "https://api.github.com/users/ylacombe/orgs", "repos_url": "https://api.github.com/users/ylacombe/repos", "events_url": "https://api.github.com/users/ylacombe/events{/privacy}", "received_events_url": "https://api.github.com/users/ylacombe/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28706). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28706/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28706", "html_url": "https://github.com/huggingface/transformers/pull/28706", "diff_url": "https://github.com/huggingface/transformers/pull/28706.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28706.patch", "merged_at": null }

{ "login": "NielsRogge", "id": 48327001, "node_id": "MDQ6VXNlcjQ4MzI3MDAx", "avatar_url": "https://avatars.githubusercontent.com/u/48327001?v=4", "gravatar_id": "", "url": "https://api.github.com/users/NielsRogge", "html_url": "https://github.com/NielsRogge", "followers_url": "https://api.github.com/users/NielsRogge/followers", "following_url": "https://api.github.com/users/NielsRogge/following{/other_user}", "gists_url": "https://api.github.com/users/NielsRogge/gists{/gist_id}", "starred_url": "https://api.github.com/users/NielsRogge/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/NielsRogge/subscriptions", "organizations_url": "https://api.github.com/users/NielsRogge/orgs", "repos_url": "https://api.github.com/users/NielsRogge/repos", "events_url": "https://api.github.com/users/NielsRogge/events{/privacy}", "received_events_url": "https://api.github.com/users/NielsRogge/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28705). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28705/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28705", "html_url": "https://github.com/huggingface/transformers/pull/28705", "diff_url": "https://github.com/huggingface/transformers/pull/28705.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28705.patch", "merged_at": 1708352549000 }

{ "login": "menouarazib", "id": 99955425, "node_id": "U_kgDOBfUy4Q", "avatar_url": "https://avatars.githubusercontent.com/u/99955425?v=4", "gravatar_id": "", "url": "https://api.github.com/users/menouarazib", "html_url": "https://github.com/menouarazib", "followers_url": "https://api.github.com/users/menouarazib/followers", "following_url": "https://api.github.com/users/menouarazib/following{/other_user}", "gists_url": "https://api.github.com/users/menouarazib/gists{/gist_id}", "starred_url": "https://api.github.com/users/menouarazib/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/menouarazib/subscriptions", "organizations_url": "https://api.github.com/users/menouarazib/orgs", "repos_url": "https://api.github.com/users/menouarazib/repos", "events_url": "https://api.github.com/users/menouarazib/events{/privacy}", "received_events_url": "https://api.github.com/users/menouarazib/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @menouarazib, thanks for raising this issue! \r\n\r\nThis is because the model being loading with this checkpoint, is from [code on the hub](https://huggingface.co/tiiuae/falcon-7b/blob/main/modeling_falcon.py) -- [mapping here](https://huggingface.co/tiiuae/falcon-7b/blob/898df1396f35e447d5fe44e0a3ccaaaa69f30d36/config.json#L14). \r\n\r\nWhereas the code in the library [does support FA2](https://github.com/huggingface/transformers/blob/f40b87de0ca234df61f76928956c4a2118c0b548/src/transformers/models/falcon/modeling_falcon.py#L946). \r\n\r\nThis divergence is happening because the model was originally on the hub and then was ported into this library. The good news is, [FA has already been implemented for this model](https://github.com/huggingface/transformers/blob/f40b87de0ca234df61f76928956c4a2118c0b548/src/transformers/models/falcon/modeling_falcon.py#L540) and so it should be pretty easy to add to the hub model. I'd suggest opening a discussion there and requesting the addition with the model repo owners. \r\n\r\ncc @Rocketknight1 who know more about the intended hub vs transformers code usage and the recommended way to map between the two. \r\n\r\nIf you want to use FA2 directly with Falcon, you can import the model from transformers directly instead of using the auto model: \r\n\r\n```py\r\nimport torch\r\nfrom transformers import AutoTokenizer, BitsAndBytesConfig, FalconForCausalLM\r\n\r\n# Hugging Face Falcon-7B model ID\r\nmodel_id = \"tiiuae/falcon-7b\"\r\n\r\n# BitsAndBytesConfig for 4-bit integers\r\nbnb_config = BitsAndBytesConfig(\r\n load_in_4bit=True,\r\n bnb_4bit_use_double_quant=True,\r\n bnb_4bit_quant_type=\"nf4\",\r\n bnb_4bit_compute_dtype=torch.bfloat16\r\n)\r\n\r\n# Load the model and tokenizer\r\nmodel = FalconForCausalLM.from_pretrained(\r\n model_id,\r\n trust_remote_code=True,\r\n device_map=\"auto\",\r\n attn_implementation=\"flash_attention_2\",\r\n torch_dtype=torch.bfloat16,\r\n quantization_config=bnb_config\r\n)\r\n```", "Thanks, @amyeroberts, for this clarification. It works when using `FalconForCausalLM from transformers `. \r\nI have created a discussion to add Flash Attention to the model hosted on the hub:\r\nhttps://huggingface.co/tiiuae/falcon-7b/discussions/98\r\n\r\nThanks" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28704/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "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.github.com/users/younesbelkada/followers", "following_url": "https://api.github.com/users/younesbelkada/following{/other_user}", "gists_url": "https://api.github.com/users/younesbelkada/gists{/gist_id}", "starred_url": "https://api.github.com/users/younesbelkada/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/younesbelkada/subscriptions", "organizations_url": "https://api.github.com/users/younesbelkada/orgs", "repos_url": "https://api.github.com/users/younesbelkada/repos", "events_url": "https://api.github.com/users/younesbelkada/events{/privacy}", "received_events_url": "https://api.github.com/users/younesbelkada/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Do you see possibility of people mixing and matching different quantization params or even different quantizers? Use cases:\r\n- less compression in MLP vs attention\r\n- wild mix of methods to get that extra 1/1000s of perplexity\r\n\r\nIf supported, what should be the way? \r\n1) from_pretrained() calling the right quntizer as it loads each module?\r\n1) custom HFQuantizer subclass that applies methods as they fit the modules", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28703). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Thanks @poedator for your comments and ideas - I think the way forward would be to extend the `xxxQuantizer` by adding new arguments in the corresponding quantization config object - e.g. `quantize_mlp_only` . I feel mixing different quantization approaches (your second point) might be a bit too much of an edge case but contributors can always create a new quantizer for it `MixedQuantizer` with a `MixedQuantizationConfig`. ", "I will merge the commits of this PR directly in https://github.com/huggingface/transformers/pull/26610 to properly credit @poedator from his great work ! closing this - thanks @ArthurZucker for the review and offline discussions! " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28703/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28703", "html_url": "https://github.com/huggingface/transformers/pull/28703", "diff_url": "https://github.com/huggingface/transformers/pull/28703.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28703.patch", "merged_at": null }

{ "login": "Iron-Bound", "id": 7122848, "node_id": "MDQ6VXNlcjcxMjI4NDg=", "avatar_url": "https://avatars.githubusercontent.com/u/7122848?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Iron-Bound", "html_url": "https://github.com/Iron-Bound", "followers_url": "https://api.github.com/users/Iron-Bound/followers", "following_url": "https://api.github.com/users/Iron-Bound/following{/other_user}", "gists_url": "https://api.github.com/users/Iron-Bound/gists{/gist_id}", "starred_url": "https://api.github.com/users/Iron-Bound/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Iron-Bound/subscriptions", "organizations_url": "https://api.github.com/users/Iron-Bound/orgs", "repos_url": "https://api.github.com/users/Iron-Bound/repos", "events_url": "https://api.github.com/users/Iron-Bound/events{/privacy}", "received_events_url": "https://api.github.com/users/Iron-Bound/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @Iron-Bound, thanks for raising this issue! \r\n\r\nWhat do you get as output if you import numpy and check the version in a python session? \r\n\r\n```py\r\nimport numpy as np\r\nprint(np.__version__)\r\n```", "let me try to find the container taht I had the issue on, side note I think this may be due to conda changing package paths" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28702/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ahmedkooli", "id": 56259512, "node_id": "MDQ6VXNlcjU2MjU5NTEy", "avatar_url": "https://avatars.githubusercontent.com/u/56259512?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ahmedkooli", "html_url": "https://github.com/ahmedkooli", "followers_url": "https://api.github.com/users/ahmedkooli/followers", "following_url": "https://api.github.com/users/ahmedkooli/following{/other_user}", "gists_url": "https://api.github.com/users/ahmedkooli/gists{/gist_id}", "starred_url": "https://api.github.com/users/ahmedkooli/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ahmedkooli/subscriptions", "organizations_url": "https://api.github.com/users/ahmedkooli/orgs", "repos_url": "https://api.github.com/users/ahmedkooli/repos", "events_url": "https://api.github.com/users/ahmedkooli/events{/privacy}", "received_events_url": "https://api.github.com/users/ahmedkooli/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @ahmedkooli, thanks for raising this issue! \r\n\r\nDigging into this, it seems this is coming from argparser itself, and is due to [this line](https://github.com/python/cpython/blob/d2da4e417ed9e6217e925e1df2820a2bf090efb3/Lib/argparse.py#L2325). As the option `text_column_names` starts with `text_column_na` and it's selected. \r\n\r\nFor example - just using argparser: \r\n\r\n```py\r\n>>> import argparse\r\n>>> parser = argparse.ArgumentParser()\r\n>>> parser.add_argument('--foo', default=None)\r\n>>> parser.parse_args(['--fo', 'a'])\r\nNamespace(foo='a')\r\n```", "Thanks for the answer :) Would you suggest trying to enforce it in the transformers library or should I take the discussion up in the python repo directly?", "@ahmedkooli As it's not something that we've encountered causing issues for our users, enforcing this isn't a feature we'd add at the moment. If it becomes something that is requested by many people (I'll measure as 👍 on this comment) or is a common pain point then we can revisit. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28701/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "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.github.com/users/nileshkokane01/followers", "following_url": "https://api.github.com/users/nileshkokane01/following{/other_user}", "gists_url": "https://api.github.com/users/nileshkokane01/gists{/gist_id}", "starred_url": "https://api.github.com/users/nileshkokane01/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/nileshkokane01/subscriptions", "organizations_url": "https://api.github.com/users/nileshkokane01/orgs", "repos_url": "https://api.github.com/users/nileshkokane01/repos", "events_url": "https://api.github.com/users/nileshkokane01/events{/privacy}", "received_events_url": "https://api.github.com/users/nileshkokane01/received_events", "type": "User", "site_admin": false }

[]

[]

[ "I do not know why the repo-consistency and setup_and_quality is failing. I think I broke compatibility for ViT I guess. @NielsRogge any clue? " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28700/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28700", "html_url": "https://github.com/huggingface/transformers/pull/28700", "diff_url": "https://github.com/huggingface/transformers/pull/28700.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28700.patch", "merged_at": null }

{ "login": "mturetskii", "id": 96064903, "node_id": "U_kgDOBbnVhw", "avatar_url": "https://avatars.githubusercontent.com/u/96064903?v=4", "gravatar_id": "", "url": "https://api.github.com/users/mturetskii", "html_url": "https://github.com/mturetskii", "followers_url": "https://api.github.com/users/mturetskii/followers", "following_url": "https://api.github.com/users/mturetskii/following{/other_user}", "gists_url": "https://api.github.com/users/mturetskii/gists{/gist_id}", "starred_url": "https://api.github.com/users/mturetskii/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/mturetskii/subscriptions", "organizations_url": "https://api.github.com/users/mturetskii/orgs", "repos_url": "https://api.github.com/users/mturetskii/repos", "events_url": "https://api.github.com/users/mturetskii/events{/privacy}", "received_events_url": "https://api.github.com/users/mturetskii/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28699). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28699/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28699", "html_url": "https://github.com/huggingface/transformers/pull/28699", "diff_url": "https://github.com/huggingface/transformers/pull/28699.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28699.patch", "merged_at": 1706269974000 }

{ "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.github.com/users/pradeepdev-1995/followers", "following_url": "https://api.github.com/users/pradeepdev-1995/following{/other_user}", "gists_url": "https://api.github.com/users/pradeepdev-1995/gists{/gist_id}", "starred_url": "https://api.github.com/users/pradeepdev-1995/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pradeepdev-1995/subscriptions", "organizations_url": "https://api.github.com/users/pradeepdev-1995/orgs", "repos_url": "https://api.github.com/users/pradeepdev-1995/repos", "events_url": "https://api.github.com/users/pradeepdev-1995/events{/privacy}", "received_events_url": "https://api.github.com/users/pradeepdev-1995/received_events", "type": "User", "site_admin": false }

[]

[]

[ "cc @ArthurZucker - possibly an issue for the tokenizers library?", "I am not sure I have ever seen the `pretokenizer` argument 😅 \r\nYou should do `tokenizer._tokenizer.pre_tokenizer = Whitespace()`", "@ArthurZucker \r\ncan you show me the full code, please?\r\n\r\nI am getting this error\r\n```\r\nModuleNotFoundError: No module named 'tokenizer._tokenizer'\r\n```", "```diff\r\n from transformers import AutoTokenizer\r\n tokenizer = AutoTokenizer.from_pretrained(\"mistralai/Mistral-7B-Instruct-v0.2\", trust_remote_code=True,\r\n use_fast=False)\r\n tokenizer.pad_token = tokenizer.eos_token\r\n tokenizer.padding_side = \"right\"\r\n sentence = \"Transformers tokenization testing\"\r\n tokenized_sentence = tokenizer.tokenize(sentence)\r\n print(\"without WhitespaceSplit\")\r\n print(tokenized_sentence)\r\n \r\n from tokenizers.pre_tokenizers import WhitespaceSplit\r\n- tokenizer = AutoTokenizer.from_pretrained(\"mistralai/Mistral-7B-Instruct-v0.2\", trust_remote_code=True,pretokenizer=WhitespaceSplit(), use_fast=False)\r\n+ tokenizer = AutoTokenizer.from_pretrained(\"mistralai/Mistral-7B-Instruct-v0.2\", trust_remote_code=True,use_fast=True)\r\n+ tokenizer._tokenizer.pre_tokenizer = WhitespaceSplit()\r\n tokenizer.pad_token = tokenizer.eos_token\r\n tokenizer.padding_side = \"right\"\r\n tokenized_sentence = tokenizer.tokenize(sentence)\r\n print(\"with WhitespaceSplit \")\r\n print(tokenized_sentence)\r\n``` \r\n🤗 ", "@ArthurZucker \r\n\r\n```\r\nAttributeError: 'LlamaTokenizer' object has no attribute '_tokenizer'\r\n```\r\nSince I am using mistral model which is Llama architecture based i think.\r\nany way to solve this?", "If you are using a slow tokenizer this cannot work. Updated the script to use fast", "@ArthurZucker \r\nNot working\r\n\r\n", "That is because you still have a normalizer:\r\n```python \r\nfrom tokenizers import normalizers\r\ntokenizer._tokenizer.normalizer = normalizers.Sequence([])\r\n```\r\n", "Not worked as expected\r\n@ArthurZucker \r\n\r\n", "I am not sure I understand, if you follow the logic of the tokenization [see the doc here](https://huggingface.co/docs/tokenizers/pipeline#pretokenization) the pretokenization will split, but the final tokens are not given by the pre tokenized text:\r\n\r\n> The role of the model is to split your “words” into tokens, using the rules it has learned. It’s also responsible for mapping those tokens to their corresponding IDs in the vocabulary of the model.\r\n\r\nthus `Transformers` is split into it's known tokens", "@ArthurZucker \r\n\r\nSo splitting the sentence **\"Transformers tokenization testing\"** into **[\"Transformers\", \"tokenization\", \"testing\"]** is not possible using **mistralai/Mistral-7B-Instruct-v0.2** tokenizer?", "You can just use `sentence.split(\" \")` but no, the token `Transformers` is not part of the vocab so it's not a token " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28698/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "wwx007121", "id": 13541369, "node_id": "MDQ6VXNlcjEzNTQxMzY5", "avatar_url": "https://avatars.githubusercontent.com/u/13541369?v=4", "gravatar_id": "", "url": "https://api.github.com/users/wwx007121", "html_url": "https://github.com/wwx007121", "followers_url": "https://api.github.com/users/wwx007121/followers", "following_url": "https://api.github.com/users/wwx007121/following{/other_user}", "gists_url": "https://api.github.com/users/wwx007121/gists{/gist_id}", "starred_url": "https://api.github.com/users/wwx007121/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/wwx007121/subscriptions", "organizations_url": "https://api.github.com/users/wwx007121/orgs", "repos_url": "https://api.github.com/users/wwx007121/repos", "events_url": "https://api.github.com/users/wwx007121/events{/privacy}", "received_events_url": "https://api.github.com/users/wwx007121/received_events", "type": "User", "site_admin": false }

[]

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[ { "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false } ]

[ "Hi @wwx007121, thanks for raising an issue! \r\n\r\nCould you give some more details about exactly the bug that is occurring i.e. the error being encountered (including full traceback) and a minimal code snippet to reproduce the issue? \r\n\r\ncc @ydshieh ", "> Hi @wwx007121, thanks for raising an issue!\r\n> \r\n> Could you give some more details about exactly the bug that is occurring i.e. the error being encountered (including full traceback) and a minimal code snippet to reproduce the issue?\r\n> \r\n> cc @ydshieh\r\n\r\n```\r\n model_id = \"openai/whisper-large-v3\"\r\n pretrain_model = AutoModelForSpeechSeq2Seq.from_pretrained(\r\n model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True, cache_dir=model_cache\r\n )\r\n pretrain_model.to(device)\r\n print(\"load model done\")\r\n\r\n processor = AutoProcessor.from_pretrained(model_id, cache_dir=model_cache)\r\n```\r\nmodel in cache was downloaded in others process which shared same docker environments.\r\n\r\n```\r\nTraceback (most recent call last):\r\n File \"script/get_whisper_result.py\", line 28, in <module>\r\n processor = AutoProcessor.from_pretrained(model_id, cache_dir=model_cache)\r\n File \"/opt/miniconda/lib/python3.8/site-packages/transformers/models/auto/processing_auto.py\", line 313, in from_pretrained\r\n return processor_class.from_pretrained(\r\n File \"/opt/miniconda/lib/python3.8/site-packages/transformers/processing_utils.py\", line 464, in from_pretrained\r\n processor_dict, kwargs = cls.get_processor_dict(pretrained_model_name_or_path, **kwargs)\r\n File \"/opt/miniconda/lib/python3.8/site-packages/transformers/processing_utils.py\", line 308, in get_processor_dict\r\n resolved_processor_file = cached_file(\r\n File \"/opt/miniconda/lib/python3.8/site-packages/transformers/utils/hub.py\", line 425, in cached_file\r\n raise EnvironmentError(\r\nOSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like distil-whisper/distil-large-v2 is not the path to a directory containing a file named processor_config.json.\r\nCheckout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'.\r\n```", "Hi @wwx007121 , it looks like I should modify the condition indeed, thank you for reporting this.\r\n\r\nHowever, to make sure, I would really like to be able to reproduce the issue. So far, I am doing the following, which should be the situation you described, but this code snippet works without any error.\r\n\r\nCould you describe in more detail how to reproduce it, please?\r\n\r\nYou mentioned that `cache was downloaded in others process`. When running the provided code example, **is the connection cut/disabled?**\r\n\r\n```python\r\nfrom transformers import AutoProcessor, AutoModelForSpeechSeq2Seq\r\n\r\nmodel_id = \"openai/whisper-large-v3\"\r\n\r\nmodel_cache = \"my_cache\"\r\n\r\npretrain_model = AutoModelForSpeechSeq2Seq.from_pretrained(\r\n model_id, use_safetensors=True, cache_dir=model_cache\r\n)\r\n\r\nprocessor = AutoProcessor.from_pretrained(model_id, cache_dir=model_cache)\r\n\r\n```", "Well, I tried to disable the internet connection and I can reproduce the issue. I will open a PR to fix it, thanks again for reporting", "@wwx007121\r\n\r\nThe fix is merged into `main`. Thanks again!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28697/reactions", "total_count": 1, "+1": 1, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ThibaultLengagne", "id": 11950126, "node_id": "MDQ6VXNlcjExOTUwMTI2", "avatar_url": "https://avatars.githubusercontent.com/u/11950126?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ThibaultLengagne", "html_url": "https://github.com/ThibaultLengagne", "followers_url": "https://api.github.com/users/ThibaultLengagne/followers", "following_url": "https://api.github.com/users/ThibaultLengagne/following{/other_user}", "gists_url": "https://api.github.com/users/ThibaultLengagne/gists{/gist_id}", "starred_url": "https://api.github.com/users/ThibaultLengagne/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ThibaultLengagne/subscriptions", "organizations_url": "https://api.github.com/users/ThibaultLengagne/orgs", "repos_url": "https://api.github.com/users/ThibaultLengagne/repos", "events_url": "https://api.github.com/users/ThibaultLengagne/events{/privacy}", "received_events_url": "https://api.github.com/users/ThibaultLengagne/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hello,\r\n\r\nThanks @stevhliu and @Sarapuce for the review. I have fixed every remark. \r\n\r\nSeems good to go :+1: ", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28696). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28696/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28696", "html_url": "https://github.com/huggingface/transformers/pull/28696", "diff_url": "https://github.com/huggingface/transformers/pull/28696.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28696.patch", "merged_at": 1706551669000 }

{ "login": "tomaarsen", "id": 37621491, "node_id": "MDQ6VXNlcjM3NjIxNDkx", "avatar_url": "https://avatars.githubusercontent.com/u/37621491?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tomaarsen", "html_url": "https://github.com/tomaarsen", "followers_url": "https://api.github.com/users/tomaarsen/followers", "following_url": "https://api.github.com/users/tomaarsen/following{/other_user}", "gists_url": "https://api.github.com/users/tomaarsen/gists{/gist_id}", "starred_url": "https://api.github.com/users/tomaarsen/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tomaarsen/subscriptions", "organizations_url": "https://api.github.com/users/tomaarsen/orgs", "repos_url": "https://api.github.com/users/tomaarsen/repos", "events_url": "https://api.github.com/users/tomaarsen/events{/privacy}", "received_events_url": "https://api.github.com/users/tomaarsen/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28695/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28695", "html_url": "https://github.com/huggingface/transformers/pull/28695", "diff_url": "https://github.com/huggingface/transformers/pull/28695.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28695.patch", "merged_at": 1706175293000 }

{ "login": "yusyel", "id": 25446622, "node_id": "MDQ6VXNlcjI1NDQ2NjIy", "avatar_url": "https://avatars.githubusercontent.com/u/25446622?v=4", "gravatar_id": "", "url": "https://api.github.com/users/yusyel", "html_url": "https://github.com/yusyel", "followers_url": "https://api.github.com/users/yusyel/followers", "following_url": "https://api.github.com/users/yusyel/following{/other_user}", "gists_url": "https://api.github.com/users/yusyel/gists{/gist_id}", "starred_url": "https://api.github.com/users/yusyel/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/yusyel/subscriptions", "organizations_url": "https://api.github.com/users/yusyel/orgs", "repos_url": "https://api.github.com/users/yusyel/repos", "events_url": "https://api.github.com/users/yusyel/events{/privacy}", "received_events_url": "https://api.github.com/users/yusyel/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28694). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28694/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28694", "html_url": "https://github.com/huggingface/transformers/pull/28694", "diff_url": "https://github.com/huggingface/transformers/pull/28694.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28694.patch", "merged_at": 1706191598000 }

{ "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.github.com/users/nileshkokane01/followers", "following_url": "https://api.github.com/users/nileshkokane01/following{/other_user}", "gists_url": "https://api.github.com/users/nileshkokane01/gists{/gist_id}", "starred_url": "https://api.github.com/users/nileshkokane01/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/nileshkokane01/subscriptions", "organizations_url": "https://api.github.com/users/nileshkokane01/orgs", "repos_url": "https://api.github.com/users/nileshkokane01/repos", "events_url": "https://api.github.com/users/nileshkokane01/events{/privacy}", "received_events_url": "https://api.github.com/users/nileshkokane01/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28693/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28693", "html_url": "https://github.com/huggingface/transformers/pull/28693", "diff_url": "https://github.com/huggingface/transformers/pull/28693.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28693.patch", "merged_at": null }

{ "login": "pedrogengo", "id": 27240528, "node_id": "MDQ6VXNlcjI3MjQwNTI4", "avatar_url": "https://avatars.githubusercontent.com/u/27240528?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pedrogengo", "html_url": "https://github.com/pedrogengo", "followers_url": "https://api.github.com/users/pedrogengo/followers", "following_url": "https://api.github.com/users/pedrogengo/following{/other_user}", "gists_url": "https://api.github.com/users/pedrogengo/gists{/gist_id}", "starred_url": "https://api.github.com/users/pedrogengo/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pedrogengo/subscriptions", "organizations_url": "https://api.github.com/users/pedrogengo/orgs", "repos_url": "https://api.github.com/users/pedrogengo/repos", "events_url": "https://api.github.com/users/pedrogengo/events{/privacy}", "received_events_url": "https://api.github.com/users/pedrogengo/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Maybe we should have a different error message for these models?" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28692/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28692", "html_url": "https://github.com/huggingface/transformers/pull/28692", "diff_url": "https://github.com/huggingface/transformers/pull/28692.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28692.patch", "merged_at": null }

{ "login": "minmie", "id": 40080081, "node_id": "MDQ6VXNlcjQwMDgwMDgx", "avatar_url": "https://avatars.githubusercontent.com/u/40080081?v=4", "gravatar_id": "", "url": "https://api.github.com/users/minmie", "html_url": "https://github.com/minmie", "followers_url": "https://api.github.com/users/minmie/followers", "following_url": "https://api.github.com/users/minmie/following{/other_user}", "gists_url": "https://api.github.com/users/minmie/gists{/gist_id}", "starred_url": "https://api.github.com/users/minmie/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/minmie/subscriptions", "organizations_url": "https://api.github.com/users/minmie/orgs", "repos_url": "https://api.github.com/users/minmie/repos", "events_url": "https://api.github.com/users/minmie/events{/privacy}", "received_events_url": "https://api.github.com/users/minmie/received_events", "type": "User", "site_admin": false }

[]

[]

[ "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!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28691/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ningziwen", "id": 8747309, "node_id": "MDQ6VXNlcjg3NDczMDk=", "avatar_url": "https://avatars.githubusercontent.com/u/8747309?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ningziwen", "html_url": "https://github.com/ningziwen", "followers_url": "https://api.github.com/users/ningziwen/followers", "following_url": "https://api.github.com/users/ningziwen/following{/other_user}", "gists_url": "https://api.github.com/users/ningziwen/gists{/gist_id}", "starred_url": "https://api.github.com/users/ningziwen/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ningziwen/subscriptions", "organizations_url": "https://api.github.com/users/ningziwen/orgs", "repos_url": "https://api.github.com/users/ningziwen/repos", "events_url": "https://api.github.com/users/ningziwen/events{/privacy}", "received_events_url": "https://api.github.com/users/ningziwen/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Seems related to this recent change. https://github.com/huggingface/transformers/pull/28447", "Found the PT commit adding this interface. https://github.com/pytorch/pytorch/commit/df14650f0b14b80db132b0c1797dc595fbee1054\r\n\r\nThis is only added in PT 2.0. PT 1.13.1 does not have it. https://github.com/pytorch/pytorch/blob/49444c3e546bf240bed24a101e747422d1f8a0ee/torch/nn/functional.py#L4808\r\n\r\nEven if PT 1.10 is not supported from 4.37.0, PT 1.13.1 is still supported right?", "Hi @ningziwen thank you, indeed I did not add a guard on torch>=2.0. Will fix, thank you.", "Hi @ningziwen, this is fixed on main with https://github.com/huggingface/transformers/pull/28774. Thank you!" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28690/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "tamanna-mostafa", "id": 156403336, "node_id": "U_kgDOCVKGiA", "avatar_url": "https://avatars.githubusercontent.com/u/156403336?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tamanna-mostafa", "html_url": "https://github.com/tamanna-mostafa", "followers_url": "https://api.github.com/users/tamanna-mostafa/followers", "following_url": "https://api.github.com/users/tamanna-mostafa/following{/other_user}", "gists_url": "https://api.github.com/users/tamanna-mostafa/gists{/gist_id}", "starred_url": "https://api.github.com/users/tamanna-mostafa/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tamanna-mostafa/subscriptions", "organizations_url": "https://api.github.com/users/tamanna-mostafa/orgs", "repos_url": "https://api.github.com/users/tamanna-mostafa/repos", "events_url": "https://api.github.com/users/tamanna-mostafa/events{/privacy}", "received_events_url": "https://api.github.com/users/tamanna-mostafa/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @tamanna-mostafa, thanks for raising an issue! \r\n\r\nBased on the error message, it looks as though the weights for the peft file are corrupted or possibly empty. Outside of the script, are you able to run the following: \r\n\r\n```py\r\nfrom peft import PeftModel \r\n\r\nmodel = PeftModel.from_pretrained(\"/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k\")\r\n```\r\n\r\nIf you look at the size of the files and shards for `/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k`, what do you see? \r\n\r\nNote: your script won't run because `get_args` doesn't return anything. Also, you don't need to pass sharing or safe serialization paramters when saving the tokenizer. ", "@amyeroberts \r\nThanks for your comments. When I run the code you suggested, I get this error:\r\n```\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/files_t/debig_amy.py\", line 4, in <module>\r\n model = PeftModel.from_pretrained(model_id)\r\nTypeError: PeftModel.from_pretrained() missing 1 required positional argument: 'model_id'\r\n```\r\n>If you look at the size of the files and shards for /mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k, what do you see?\r\n\r\n```\r\n(ml_v4) ubuntu@ip-172-31-8-218:/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k$ ls -lh *\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 18:13 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 18:13 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 29 19:19 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 18:13 added_tokens.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 18:14 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 18:13 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 18:13 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 18:13 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 18:13 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 18:13 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 18:14 zero_to_fp32.py\r\n\r\ncheckpoint-100:\r\ntotal 2.4M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 23 21:32 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 23 21:32 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 23 21:32 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 23 21:32 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 23 21:33 global_step100\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 23 21:34 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 23 21:34 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 23 21:34 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 23 21:32 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 23 21:32 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 23 21:32 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 23 21:32 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 4.4K Jan 23 21:34 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 23 21:32 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 23 21:34 zero_to_fp32.py\r\n\r\ncheckpoint-200:\r\ntotal 2.4M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 00:47 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 00:47 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 00:47 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 00:47 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 00:48 global_step200\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 00:49 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 00:49 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 00:49 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 00:47 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 00:47 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 00:47 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 00:47 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 7.9K Jan 24 00:49 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 00:47 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 00:49 zero_to_fp32.py\r\n\r\ncheckpoint-300:\r\ntotal 2.4M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 04:02 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 04:02 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 04:02 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 04:02 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 04:03 global_step300\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 04:04 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 04:04 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 04:04 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 04:02 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 04:02 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 04:02 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 04:02 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 12K Jan 24 04:04 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 04:02 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 04:04 zero_to_fp32.py\r\n\r\ncheckpoint-400:\r\ntotal 2.4M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 07:17 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 07:17 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 07:17 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 07:17 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 07:18 global_step400\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 07:19 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 07:19 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 07:19 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 07:17 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 07:17 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 07:17 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 07:17 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 15K Jan 24 07:19 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 07:17 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 07:19 zero_to_fp32.py\r\n\r\ncheckpoint-500:\r\ntotal 2.5M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 10:32 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 10:32 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 10:32 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 10:32 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 10:33 global_step500\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 10:34 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 10:34 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 10:34 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 10:32 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 10:32 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 10:32 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 10:32 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 19K Jan 24 10:34 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 10:32 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 10:34 zero_to_fp32.py\r\n\r\ncheckpoint-600:\r\ntotal 2.5M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 13:47 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 13:47 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 13:47 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 13:47 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 13:48 global_step600\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 13:50 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 13:50 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 13:50 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 13:47 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 13:47 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 13:47 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 13:47 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 22K Jan 24 13:50 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 13:47 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 13:50 zero_to_fp32.py\r\n\r\ncheckpoint-700:\r\ntotal 2.5M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 17:03 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 17:03 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 48 Jan 24 17:03 adapter_model.safetensors\r\n-rw-rw-r-- 1 ubuntu ubuntu 133 Jan 24 17:03 added_tokens.json\r\ndrwxrwxr-x 2 ubuntu ubuntu 6.0K Jan 24 17:04 global_step700\r\n-rw-rw-r-- 1 ubuntu ubuntu 14 Jan 24 17:05 latest\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_0.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_1.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_2.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_3.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_4.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_5.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_6.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 16K Jan 24 17:05 rng_state_7.pth\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.1K Jan 24 17:05 scheduler.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 829 Jan 24 17:03 special_tokens_map.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.8M Jan 24 17:03 tokenizer.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 482K Jan 24 17:03 tokenizer.model\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.9K Jan 24 17:03 tokenizer_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 26K Jan 24 17:05 trainer_state.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.7K Jan 24 17:03 training_args.bin\r\n-rwxrw-r-- 1 ubuntu ubuntu 24K Jan 24 17:05 zero_to_fp32.py\r\n\r\nfinal_checkpoint:\r\ntotal 20M\r\n-rw-rw-r-- 1 ubuntu ubuntu 5.1K Jan 24 18:14 README.md\r\n-rw-rw-r-- 1 ubuntu ubuntu 676 Jan 24 18:14 adapter_config.json\r\n-rw-rw-r-- 1 ubuntu ubuntu 20M Jan 24 18:14 adapter_model.safetensors\r\n\r\nglobal_step736:\r\ntotal 14G\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 31M Jan 24 18:14 bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_0_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_1_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_2_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_3_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_4_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_5_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_6_mp_rank_00_model_states.pt\r\n-rw-rw-r-- 1 ubuntu ubuntu 1.7G Jan 24 18:14 zero_pp_rank_7_mp_rank_00_model_states.pt\r\n```\r\nAlso, if I take `LLAMA2 7b ` as the base model, then the `merge_peft_adaptors_gpu.py` script works fine if I put `final_checkpoint` as the PEFT model path. Hence, in this case too (mistral 7b), I tried running the `merge_peft_adaptors_gpu.py` script with `final_checkpoint` as the PEFT model path. Then I get this error:\r\n\r\n```\r\nLoading checkpoint shards: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:04<00:00, 1.53s/it]\r\nLoading PEFT: /mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k/final_checkpoint\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/scripts_hb/merge_peft_adaptors_gpu.py\", line 51, in <module>\r\n main()\r\n File \"/mnt/efs/data/tammosta/scripts_hb/merge_peft_adaptors_gpu.py\", line 38, in main\r\n model = PeftModel.from_pretrained(base_model, args.peft_model_path)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 354, in from_pretrained\r\n model.load_adapter(model_id, adapter_name, is_trainable=is_trainable, **kwargs)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 698, in load_adapter\r\n load_result = set_peft_model_state_dict(self, adapters_weights, adapter_name=adapter_name)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/utils/save_and_load.py\", line 241, in set_peft_model_state_dict\r\n load_result = model.load_state_dict(peft_model_state_dict, strict=False)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/torch/nn/modules/module.py\", line 2152, in load_state_dict\r\n raise RuntimeError('Error(s) in loading state_dict for {}:\\n\\t{}'.format(\r\nRuntimeError: Error(s) in loading state_dict for PeftModelForCausalLM:\r\n size mismatch for base_model.model.model.layers.0.mlp.gate_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 8]).\r\n size mismatch for base_model.model.model.layers.0.mlp.up_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 8]).\r\n. . .\r\n```\r\n\r\nThis issue is the same as reported in this ticket: https://github.com/huggingface/transformers/issues/28688 . Hence, you can close this issue if you don't want to keep duplicate tickets open.\r\nI'm trying to understand what's wrong with using fine tuned Mistral 7b as the base model. " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28689/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "tamanna-mostafa", "id": 156403336, "node_id": "U_kgDOCVKGiA", "avatar_url": "https://avatars.githubusercontent.com/u/156403336?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tamanna-mostafa", "html_url": "https://github.com/tamanna-mostafa", "followers_url": "https://api.github.com/users/tamanna-mostafa/followers", "following_url": "https://api.github.com/users/tamanna-mostafa/following{/other_user}", "gists_url": "https://api.github.com/users/tamanna-mostafa/gists{/gist_id}", "starred_url": "https://api.github.com/users/tamanna-mostafa/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tamanna-mostafa/subscriptions", "organizations_url": "https://api.github.com/users/tamanna-mostafa/orgs", "repos_url": "https://api.github.com/users/tamanna-mostafa/repos", "events_url": "https://api.github.com/users/tamanna-mostafa/events{/privacy}", "received_events_url": "https://api.github.com/users/tamanna-mostafa/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Hi @tamanna-mostafa, thanks for raising this issue! \r\n\r\nCould you list the files saved under `/data/DPO_output_mistral_32k`? ", "@amyeroberts \r\nHi, thanks for your comment. Below is what I see when I run `ls `in the folder, `DPO_output_mistral_32k` :\r\n```\r\nubuntu@ip-172-31-8-218:/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k$ ls \r\nREADME.md adapter_model.safetensors checkpoint-100 checkpoint-300 checkpoint-500 checkpoint-700 global_step736 special_tokens_map.json tokenizer.model training_args.bin\r\nadapter_config.json added_tokens.json checkpoint-200 checkpoint-400 checkpoint-600 final_checkpoint latest tokenizer.json tokenizer_config.json zero_to_fp32.py\r\n```", "Could you share how you're loading the model? If you're using adapters, then I'd expect this pattern: \r\n\r\n```py\r\nfrom transformers import AutoModelForCausalLM\r\n\r\nmodel_id = \"{MISTRAL_CHECKPOINT}\"\r\ndpo_model_id = \"/data/DPO_output_mistral_32k\"\r\nmodel = AutoModelForCausalLM.from_pretrained(model_id)\r\nmodel.load_adapter(dpo_model_id)\r\n```", "You mean how I'm loading the DPO model on docker? Here are the steps:\r\n```\r\nmodel=/data/DPO_output_mistral_32k\r\nvolume=/mnt/efs/data/tammosta/files_t:/data\r\nnum_shard=8\r\ndocker run --gpus all --shm-size 1g -p 172.31.8.218:80:80 -v $volume ghcr.io/huggingface/text-generation-inference:1.1.0 --model-id $model --num-shard $num_shard --max-input-length 4095 --max-total-tokens 12000\r\n\r\n```\r\n\r\n\r\nJust in case, this is the command I ran for the DPO training:\r\n```\r\naccelerate launch --config_file ./accelerate_configs/ds_zero3.yaml rlhf_dpo.py \\\r\n--model_name_or_path=\"/mnt/efs/data/tammosta/files_t/output_sft_32k\" \\\r\n--output_dir=\"/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k\" \\\r\n--data_path=\"/mnt/efs/data/tammosta/files_t/DPO_data_rbs_clean_AIF.json\" \\\r\n--use_lamma2_peft_config False \\\r\n--beta 0.1 \\\r\n--optimizer_type adamw_hf \\\r\n--learning_rate 1e-6 \\\r\n--warmup_steps 50 \\\r\n--per_device_train_batch_size 1 \\\r\n--per_device_eval_batch_size 1 \\\r\n--gradient_accumulation_steps 8 \\\r\n--lora_alpha 16 \\\r\n--lora_dropout 0.05 \\\r\n--lora_r 8 \\\r\n--max_prompt_length 2048 \\\r\n--max_length 4096 \\\r\n--num_train_epochs 4 \\\r\n--logging_steps 20 \\\r\n--save_steps 100 \\\r\n--save_total_limit 8 \\\r\n--eval_steps 50 \\\r\n--gradient_checkpointing True \\\r\n--report_to \"wandb\"\r\n```", "It uses adapters in DPO training.", "Is there anything wrong in the way I'm loading the model on docker?", "@tamanna-mostafa No, I don't think so. From the current error and the files in the model repo I currently think there's two possible causes: \r\n* How the model is being loaded in the `text_generation_server` package\r\n* How the model is being saved out in the `rlhf_dpo.py` script.\r\n\r\nFor a model with adapter weights, I'd expect the adapter weights repo to look something like this: https://huggingface.co/ybelkada/opt-350m-lora/tree/main\r\n\r\nCould you share the contents of the `adapter_config.json`? ", "@amyeroberts \r\n>How the model is being loaded in the text_generation_server package\r\n\r\nIn my understanding, I used the below steps to load the model (prior to running docker):\r\n\r\n```\r\nmodel=/data/DPO_output_mistral_32k\r\nvolume=/mnt/efs/data/tammosta/files_t:/data\r\n```\r\n\r\nHere is the contents of the `adapter_config.json`:\r\n\r\n```\r\n{\r\n \"alpha_pattern\": {},\r\n \"auto_mapping\": null,\r\n \"base_model_name_or_path\": \"/mnt/efs/data/tammosta/files_t/output_sft_32k\",\r\n \"bias\": \"none\",\r\n \"fan_in_fan_out\": false,\r\n \"inference_mode\": true,\r\n \"init_lora_weights\": true,\r\n \"layers_pattern\": null,\r\n \"layers_to_transform\": null,\r\n \"loftq_config\": {},\r\n \"lora_alpha\": 16.0,\r\n \"lora_dropout\": 0.05,\r\n \"megatron_config\": null,\r\n \"megatron_core\": \"megatron.core\",\r\n \"modules_to_save\": null,\r\n \"peft_type\": \"LORA\",\r\n \"r\": 8,\r\n \"rank_pattern\": {},\r\n \"revision\": null,\r\n \"target_modules\": [\r\n \"v_proj\",\r\n \"q_proj\",\r\n \"up_proj\",\r\n \"down_proj\",\r\n \"gate_proj\",\r\n \"o_proj\",\r\n \"k_proj\"\r\n ],\r\n \"task_type\": \"CAUSAL_LM\"\r\n```", "I'm also pasting the last 3 sections from the `rlhf_dpo.py` script. \r\n\r\n```\r\n # 5. initialize the DPO trainer\r\n dpo_trainer = DPOTrainer(\r\n model,\r\n model_ref,\r\n args=training_args,\r\n beta=script_args.beta,\r\n train_dataset=train_dataset,\r\n eval_dataset=eval_dataset,\r\n tokenizer=tokenizer,\r\n peft_config=peft_config,\r\n max_prompt_length=script_args.max_prompt_length,\r\n max_length=script_args.max_length,\r\n )\r\n\r\n # 6. train\r\n dpo_trainer.train()\r\n dpo_trainer.save_model(script_args.output_dir)\r\n\r\n # 7. save\r\n output_dir = os.path.join(script_args.output_dir, \"final_checkpoint\")\r\n dpo_trainer.model.save_pretrained(output_dir)\r\n```", "@amyeroberts \r\nHi, did you have a chance to take a look? thanks", "I'm going cc in @younesbelkada here, who knows more about the DPO trainer and expected values in the configs :) ", "Hi @tamanna-mostafa \r\nThanks for the issue! \r\nin order to run the trained adapter with TGI using Docker, you need to first merge the adapter weights into the base model, and push / save the merged weights somewhere either on the Hub or locally.\r\n\r\nBy merging the adapter weights you make sure to convert the trained model into a standalone transformers model so that it becomes compatible with TGI. Please see: https://huggingface.co/docs/peft/main/en/conceptual_guides/lora#merge-lora-weights-into-the-base-model to understand what merging means.\r\n\r\nTo merge the model, run:\r\n\r\n```python\r\nfrom peft import AutoPeftModelForCausalLM\r\n\r\nmodel = AutoPeftModelForCausalLM.from_pretrained(model_id)\r\nmodel = model.merge_and_unload()\r\n# at this point the model is a standalone transformers model\r\nmodel.push_to_hub(xxx)\r\n```", "Hi @younesbelkada \r\n\r\nThanks a lot for your comment. As the base model, I used mistral 7b that I fine-tuned with my own preference data. I ran the following code to merge:\r\n```\r\nfrom transformers import AutoModelForCausalLM\r\nfrom peft import PeftModel\r\nimport torch\r\n\r\n#base_model = \"/mnt/efs/data/tammosta/files_t/output_sft_32k\"\r\nbase_model = AutoModelForCausalLM.from_pretrained(\r\n \"/mnt/efs/data/tammosta/files_t/output_sft_32k\",\r\n return_dict=True,\r\n torch_dtype=torch.float16,\r\n trust_remote_code=True,\r\n #**device_arg\r\n )\r\npeft_model_id = \"/mnt/efs/data/tammosta/files_t/DPO_output_32k_Test/final_checkpoint\"\r\nmodel = PeftModel.from_pretrained(base_model, peft_model_id)\r\nmerged_model = model.merge_and_unload()\r\nmerged_model.save_pretrained(\"/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k_merged\")\r\n```\r\nHowever, I'm getting the following error:\r\n```\r\nLoading checkpoint shards: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:01<00:00, 2.77it/s]\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/files_t/merge_peft_tammosta.py\", line 14, in <module>\r\n model = PeftModel.from_pretrained(base_model, peft_model_id)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 354, in from_pretrained\r\n model.load_adapter(model_id, adapter_name, is_trainable=is_trainable, **kwargs)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/peft_model.py\", line 698, in load_adapter\r\n load_result = set_peft_model_state_dict(self, adapters_weights, adapter_name=adapter_name)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/utils/save_and_load.py\", line 241, in set_peft_model_state_dict\r\n load_result = model.load_state_dict(peft_model_state_dict, strict=False)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/torch/nn/modules/module.py\", line 2152, in load_state_dict\r\n raise RuntimeError('Error(s) in loading state_dict for {}:\\n\\t{}'.format(\r\nRuntimeError: Error(s) in loading state_dict for PeftModelForCausalLM:\r\n size mismatch for base_model.model.model.layers.0.mlp.gate_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 16]).\r\n size mismatch for base_model.model.model.layers.0.mlp.up_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 16]).\r\n size mismatch for base_model.model.model.layers.0.mlp.down_proj.lora_A.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([16, 14336]).\r\n size mismatch for base_model.model.model.layers.1.mlp.gate_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 16]).\r\n . . .\r\n```\r\n\r\nIt looks the weights in the base model doesn't match with that in the PEFT model. Could you please suggest a possible way of debugging this?", "@tamanna-mostafa have you used DeepSpeed to train your adapters by any chance?", "Can you also try:\r\n```python\r\nfrom transformers import AutoModelForCausalLM\r\nfrom peft import AutoPeftModelForCausalLM\r\nimport torch\r\n\r\npeft_model_id = \"/mnt/efs/data/tammosta/files_t/DPO_output_32k_Test/final_checkpoint\"\r\nmodel = AutoPeftModelForCausalLM.from_pretrained(peft_model_id, torch_dtype=torch.float16,)\r\nmodel = model.merge_and_unload()\r\nmodel.save_pretrained(\"/mnt/efs/data/tammosta/files_t/DPO_output_mistral_32k_merged\")\r\n```", "Hi @younesbelkada ,\r\nI used `DeepSpeed `to fine tune the mistral 7b (the base model).\r\nI used `accelerate launch` to train the DPO model (PEFT model).\r\nWhen I run the suggested code, I get:\r\n\r\n```\r\nLoading checkpoint shards: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:01<00:00, 2.64it/s]\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/files_t/merge_peft_tammosta_2.py\", line 6, in <module>\r\n model = AutoPeftModelForCausalLM.from_pretrained(peft_model_id, torch_dtype=torch.float16,)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/peft/auto.py\", line 115, in from_pretrained\r\n tokenizer_exists = file_exists(\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py\", line 110, in _inner_fn\r\n validate_repo_id(arg_value)\r\n File \"/opt/conda/envs/ml_v4/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py\", line 158, in validate_repo_id\r\n raise HFValidationError(\r\nhuggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': '/mnt/efs/data/tammosta/files_t/DPO_output_32k_Test/final_checkpoint'. Use `repo_type` argument if needed.\r\n```", "@tamanna-mostafa it seems that behavior is a duplicate of https://github.com/huggingface/peft/issues/1430 - can you try to pass a relative path instead and run the script from the final checkpoint folder ? I'll submit a fix on PEFT", "Using the relative path, I've the issue of `size mismatch`:\r\n```\r\n(ml_v2) ubuntu@ip-172-31-32-104:/mnt/efs/data/tammosta/files_t$ python hf_test_2.py\r\nLoading checkpoint shards: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:01<00:00, 2.18it/s]\r\nTraceback (most recent call last):\r\n File \"/mnt/efs/data/tammosta/files_t/hf_test_2.py\", line 6, in <module>\r\n model = AutoPeftModelForCausalLM.from_pretrained(peft_model_id, torch_dtype=torch.float16,)\r\n File \"/opt/conda/envs/ml_v2/lib/python3.10/site-packages/peft/auto.py\", line 127, in from_pretrained\r\n return cls._target_peft_class.from_pretrained(\r\n File \"/opt/conda/envs/ml_v2/lib/python3.10/site-packages/peft/peft_model.py\", line 354, in from_pretrained\r\n model.load_adapter(model_id, adapter_name, is_trainable=is_trainable, **kwargs)\r\n File \"/opt/conda/envs/ml_v2/lib/python3.10/site-packages/peft/peft_model.py\", line 698, in load_adapter\r\n load_result = set_peft_model_state_dict(self, adapters_weights, adapter_name=adapter_name)\r\n File \"/opt/conda/envs/ml_v2/lib/python3.10/site-packages/peft/utils/save_and_load.py\", line 241, in set_peft_model_state_dict\r\n load_result = model.load_state_dict(peft_model_state_dict, strict=False)\r\n File \"/opt/conda/envs/ml_v2/lib/python3.10/site-packages/torch/nn/modules/module.py\", line 2153, in load_state_dict\r\n raise RuntimeError('Error(s) in loading state_dict for {}:\\n\\t{}'.format(\r\nRuntimeError: Error(s) in loading state_dict for PeftModelForCausalLM:\r\n size mismatch for base_model.model.model.layers.0.mlp.gate_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 16]).\r\n size mismatch for base_model.model.model.layers.0.mlp.up_proj.lora_B.default.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([14336, 16]).\r\n```\r\nI suspect it might be the PEFT config I'm using during DPO training. In the DPO training command of the mistral 7b SFT model, if I use the same PEFT config as that used for DPO-training LLAMA2 7b SFT model, then I don't have this `size mismatch` issue in the adaptor merge. \r\n\r\nHere's the PEFT config used for DOP-training LLAMA 2 7b SFT model:\r\n\r\n \r\n```\r\npeft_config = LoraConfig(\r\n r=script_args.lora_r,\r\n lora_alpha=script_args.lora_alpha,\r\n lora_dropout=script_args.lora_dropout,\r\n target_modules=[\r\n \"q_proj\",\r\n \"v_proj\",\r\n \"k_proj\",\r\n \"out_proj\",\r\n \"fc_in\",\r\n \"fc_out\",\r\n \"wte\",\r\n ],\r\n bias=\"none\",\r\n task_type=\"CAUSAL_LM\",\r\n )\r\n print(f\"peft_config: {peft_config}\")\r\n\r\n\r\n```\r\n\r\nWhat peft_config should I use to DPO-train a Mistral 7b SFT model? \r\nCan I use the same PEFT config as is used for DPO-training a LLAMA2 7b model (as pasted above)?\r\n", "@younesbelkada \r\nIt would be very helpful if you kindly share your thoughts on this.", "Hi @tamanna-mostafa \r\nI am going to cc @pacman100 as he is more familiar than I am with respect to interactions between DeepSpeed and PEFT" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28688/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "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://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false }

[]

[]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28687). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.", "Still need to make sure that \"Whisper default language detection behavior\" stays the same. See: https://huggingface.co/openai/whisper-large-v3/discussions/71#65b78d5a34297095bcecfe78", "**Update**:\r\n\r\nThis PR now does two additonal things:\r\n- `generation_config` is always preferred over `config.json`\r\n- By default multilingual Whisper does language detection followed by transcription\r\n\r\ncc @sanchit-gandhi ", "- [x] TODO(Patrick) Make sure to run all slow tests" ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28687/reactions", "total_count": 1, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 1 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28687", "html_url": "https://github.com/huggingface/transformers/pull/28687", "diff_url": "https://github.com/huggingface/transformers/pull/28687.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28687.patch", "merged_at": 1706702527000 }

{ "login": "FoamoftheSea", "id": 50897218, "node_id": "MDQ6VXNlcjUwODk3MjE4", "avatar_url": "https://avatars.githubusercontent.com/u/50897218?v=4", "gravatar_id": "", "url": "https://api.github.com/users/FoamoftheSea", "html_url": "https://github.com/FoamoftheSea", "followers_url": "https://api.github.com/users/FoamoftheSea/followers", "following_url": "https://api.github.com/users/FoamoftheSea/following{/other_user}", "gists_url": "https://api.github.com/users/FoamoftheSea/gists{/gist_id}", "starred_url": "https://api.github.com/users/FoamoftheSea/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/FoamoftheSea/subscriptions", "organizations_url": "https://api.github.com/users/FoamoftheSea/orgs", "repos_url": "https://api.github.com/users/FoamoftheSea/repos", "events_url": "https://api.github.com/users/FoamoftheSea/events{/privacy}", "received_events_url": "https://api.github.com/users/FoamoftheSea/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Note that fix-copies made changes to DETA model because it has the \"# copied from\" comments over methods taken from Deformable DETR. If we don't want to change that model, we could instead remove those comments and revert the changes to that file.", "Hi @FoamoftheSea, thanks for working on this!\r\n\r\nThere's another open PR which addresses enabling gradient checkpointing for DETA -- #28615. What I'd suggest is removing the relevant `# Copied from` headers in DETA and the changes in its modeling file. \r\n\r\nThis way we can include both of your respective contributions without clashes.", "@amyeroberts All done!", "No problem! 😀", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/transformers/pr_28686). All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28686/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28686", "html_url": "https://github.com/huggingface/transformers/pull/28686", "diff_url": "https://github.com/huggingface/transformers/pull/28686.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28686.patch", "merged_at": 1706523041000 }

{ "login": "rwightman", "id": 5702664, "node_id": "MDQ6VXNlcjU3MDI2NjQ=", "avatar_url": "https://avatars.githubusercontent.com/u/5702664?v=4", "gravatar_id": "", "url": "https://api.github.com/users/rwightman", "html_url": "https://github.com/rwightman", "followers_url": "https://api.github.com/users/rwightman/followers", "following_url": "https://api.github.com/users/rwightman/following{/other_user}", "gists_url": "https://api.github.com/users/rwightman/gists{/gist_id}", "starred_url": "https://api.github.com/users/rwightman/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/rwightman/subscriptions", "organizations_url": "https://api.github.com/users/rwightman/orgs", "repos_url": "https://api.github.com/users/rwightman/repos", "events_url": "https://api.github.com/users/rwightman/events{/privacy}", "received_events_url": "https://api.github.com/users/rwightman/received_events", "type": "User", "site_admin": false }

[]

[]

[ "Code instances where this either definitely a concern, or likely (depending on ranges involved).\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/llama/modeling_llama.py#L130-L131\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/llama/modeling_llama.py#L140\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/llama/modeling_llama.py#L168\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/llama/modeling_llama.py#L195\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/examples/research_projects/bertabs/modeling_bertabs.py#L265-L266\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/codegen/modeling_codegen.py#L55-L59\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/conditional_detr/modeling_conditional_detr.py#L437-L455\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/conditional_detr/modeling_conditional_detr.py#L496-L509\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/ctrl/modeling_ctrl.py#L47-L60\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deformable_detr/modeling_deformable_detr.py#L494-L495\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deformable_detr/modeling_deformable_detr.py#L620-L621\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deformable_detr/modeling_deformable_detr.py#L1542-L1543\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deprecated/transfo_xl/modeling_transfo_xl.py#L945-L946\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deta/modeling_deta.py#L404-L405\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deta/modeling_deta.py#L529-L530\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/deta/modeling_deta.py#L1453-L1454\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/detr/modeling_detr.py#L438-L439\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/falcon/modeling_falcon.py#L151-L152\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/falcon/modeling_falcon.py#L180\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/falcon/modeling_falcon.py#L208\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/fastspeech2_conformer/modeling_fastspeech2_conformer.py#L823-L826\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/fsmt/modeling_fsmt.py#L1349-L1351\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/funnel/modeling_funnel.py#L235-L267\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/fuyu/image_processing_fuyu.py#L687-L690\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/gpt_neox/modeling_gpt_neox.py#L547\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/gpt_neox/modeling_gpt_neox.py#L576\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/gpt_neox/modeling_gpt_neox.py#L604\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/gpt_neox_japanese/modeling_gpt_neox_japanese.py#L255\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/gptj/modeling_gptj.py#L60\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/idefics/modeling_idefics.py#L480\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/kosmos2/modeling_kosmos2.py#L776-L780\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/m2m_100/modeling_m2m_100.py#L114-\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mask2former/modeling_mask2former.py#L863-L864\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mask2former/modeling_mask2former.py#L2132-L2133\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/maskformer/modeling_maskformer.py#L1354-L1355\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mega/modeling_mega.py#L172-L174\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mistral/modeling_mistral.py#L109-L110\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mistral/modeling_mistral.py#L109-L110\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mixtral/modeling_mixtral.py#L202-L203\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/mpt/modeling_mpt.py#L69-L70\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/musicgen/modeling_musicgen.py#L129-L131\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/nezha/modeling_nezha.py#L153-L155\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/nllb_moe/modeling_nllb_moe.py#L167-L169\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/oneformer/modeling_oneformer.py#L2803-L2804\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/pegasus_x/modeling_pegasus_x.py#L112-L113\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/persimmon/modeling_persimmon.py#L61-L62\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/persimmon/modeling_persimmon.py#L90-L91\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/persimmon/modeling_persimmon.py#L118-L119\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/phi/modeling_phi.py#L99\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/phi/modeling_phi.py#L128\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/phi/modeling_phi.py#L156\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/qwen2/modeling_qwen2.py#L116\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/seamless_m4t/modeling_seamless_m4t.py#L417\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/seamless_m4t/modeling_seamless_m4t.py#L1024-L1026\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/seamless_m4t_v2/modeling_seamless_m4t_v2.py#L980-L982\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/speech_to_text/modeling_speech_to_text.py#L133-L136\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/speecht5/modeling_speecht5.py#L316-L318\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/speecht5/modeling_speecht5.py#L406-L407\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/swin2sr/modeling_swin2sr.py#L293-L296\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/swinv2/modeling_swinv2.py#L449-L451\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/table_transformer/modeling_table_transformer.py#L374-L375\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/trocr/modeling_trocr.py#L88-L90\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/wav2vec2_bert/modeling_wav2vec2_bert.py#L315\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/wav2vec2_conformer/modeling_wav2vec2_conformer.py#L445-L448\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/xglm/modeling_xglm.py#L160-L162\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/src/transformers/models/xlnet/modeling_xlnet.py#L1023-L1038\r\nhttps://github.com/huggingface/transformers/blob/8278b1538ecc89dad8ebca510a31a86bc8645edb/examples/research_projects/visual_bert/modeling_frcnn.py#L171-L192", "If we look at the original Llama code, this issue is avoided.\r\n\r\nhttps://github.com/facebookresearch/llama/blob/ef351e9cd9496c579bf9f2bb036ef11bdc5ca3d2/llama/model.py#L100-L104\r\n```python\r\n freqs = 1.0 / (theta ** (torch.arange(0, dim, 2)[: (dim // 2)].float() / dim))\r\n t = torch.arange(end, device=freqs.device) # type: ignore\r\n freqs = torch.outer(t, freqs).float() # type: ignore\r\n freqs_cis = torch.polar(torch.ones_like(freqs), freqs) # complex64\r\n```\r\n\r\nFor GPT-NeoX, it's also avoided, but problematic in transformers (see above).\r\n\r\nhttps://github.com/EleutherAI/gpt-neox/blob/63991555ec082c8f80c475f851d008193b10008c/megatron/model/positional_embeddings.py#L27-L34\r\n```\r\n t = torch.arange(x.shape[seq_dim], device=x.device).type_as(self.inv_freq)\r\n sinusoid_inp = torch.einsum(\"i,j->ij\", t, self.inv_freq)\r\n if self.precision == torch.bfloat16:\r\n sinusoid_inp = sinusoid_inp.float()\r\n sin, cos = sinusoid_inp.sin(), sinusoid_inp.cos()\r\n if self.precision == torch.bfloat16:\r\n sin, cos = sin.bfloat16(), cos.bfloat16()\r\n emb = torch.cat((sin, cos), dim=-1)\r\n```", "I believe this is the problem being seen in this issue https://github.com/microsoft/DeepSpeed/issues/4932 and also seeing now this may be a dupe of https://github.com/huggingface/transformers/issues/28596\r\n", "Related to this possible concern with the zero.Init() overriding dtype for arange (and I did confirm this is a problem with a test bench), there's also an overlapping issue that's been brought up before in e.g. #25681 but I don't think fully addressed as that improvement focused on rescaling at runtime for larger seq len, this one is due to zero.Init() overriding the device arg for tensor creation fns and having the init done on a non-CPU device.\r\n\r\nWhen a library like DeepSpeed forces the calculation of the cached RoPe sin/cos/freq values onto the GPU it is wrong compared to the CPU calcs due to a rather nasty combo of floating point ops that differ enough to have a significant impact (div, pow, outer product, convert to low precision), ~5e-4 in float16 and 2e-3 eps for Llama. This results in model logit values differing by close to 1.0. This is with the calcs forced to float32 (so explicitly avoiding doing them in low precision), even doing the calculations in double precision is not enough to avoid problematic differences between GPU and CPU. \r\n\r\nThe only approach that seems viable is ensuring the init of those constants are always done on CPU (requires extra workarounds to prevent DeepSpeed from forcing onto GPU) and then at the very last step before they're used, do the cast to computation dtype. I trialed an approach that's related to an Eleuther workaround in their lib, but it likely has some breaking concerns with other use cases like tracing, etc.\r\nhttps://github.com/microsoft/DeepSpeed/issues/4932#issuecomment-1911277956\r\n\r\nEDIT: also think we should be forcing RoPE embeddings to be applied in float32 instead of default computation dtype. I think the original Llama is doing this but transformers is not.", "First things off: should we run RoPE in FP32? Are buffers a problem?\r\n\r\nI've done a quick perplexity benchmark of `meta-llama/Llama-2-7b-hf` with BF16 (from `main`) vs BF16 with RoPE being computed in FP32 without buffers (from [this commit](https://github.com/gante/transformers/commit/ade030977df3e2939556b4f9db317151bed7bbe4)), adapting the ppl benchmark scripts from [this comment](https://github.com/huggingface/transformers/pull/26933#issuecomment-1789646553). \r\n\r\nHere are the results:\r\n\r\n\r\nWe can see a very tiny PPL upgrade, almost negligible. This comes at the expense of an equally small increase in GPU memory requirements.\r\n\r\n\r\n\r\nOn the latency side, we see that going beyond the original context length is much more expensive -- the new `sin`/`cos` must to be computed in FP32, which is more expensive.\r\n\r\n👉 To me, this indicates that changing RoPE computations to FP32 is not worth it. Happy to do more experiments, if you suspect this logic may be flawed in some settings/other models 🤗 (cc @ArthurZucker, we've chatted about this a few days ago)\r\n\r\n👉 @rwightman could this difference be more pronounced in DeepSpeed? I have no DeepSpeed experience. \r\n\r\n\r\n", "Regarding the Deepspeed init issues, going to open a PR to fix it 💪 ", "@gante are you sure the perplexity test is representative of a wide enough range of use? In the particular users case, I was testing with their input vectors and the logits are significantly different with embeddings calc in bfloat16. Computing pos embeds in float32 on CPU and applying just the calculated embedding in lower precision wasn't too bad... \r\n\r\nSimilarly comparing the embedding values themselves, fully calculated on a different device, or calculated in low precision, the differences in the embedding floats is well beyond a range I'd be comfortable with... might be something we'd want to consider allowing the user to make their own tradeoffs via config... ", "@rwightman Absolutely not, the data/model space to test on is too large to measure! \r\n\r\nHowever, since the latency penalty of applying the change in all cases is non-negligible and exposing the ability to recompute the buffer in fp32 adds yet another flag, I'd like to have a reproducible example of a failure case in `transformers` -- at least to fully understand what's going on. That's why I added the note that I'm fully open to run more experiments, as long as I have some pointers.\r\n\r\nThere are many competing requests. Sadly, it's hard to find the time to do a proper deep dive to find a failure mode 🤗 ", "@gante understood, with the cached values the runtime latency for the 'okay' case I described should be non-existent though... namely,\r\n1. calculate the values on the cpu, in float32 as a rule\r\n2. cast to the usage dtype, eg store sin/cos embeds in bfloat16 on the target device)\r\n\r\nThere would be no memory or runtime overhead (other than when a new seq length is switched to), but the pos embed values would be significantly closer to their intended values.", "@gante from my personal test, changing the inv_freq to float32 can increase performances on MMLU of about 20points.\r\nThere are a few things to test:\r\n- `inv_freq` is a buffer, it's saved, then casted to the `dtype` you ask. But if you load with `torch_dytpe` then it's going to be a float32 casted to the `torch_dytpe`. \r\n- cos and sin can be computed in float32 or query type. Computing in float prevents overflow (known) of rope, and mostly does not slow down\r\n- q and k can be upcasted to float32, sin and cos computed in float32 then cast back everything.\r\n- vs mixed precision where q and k are not upcasted, but sin and cos are in float32.\r\n\r\n\r\nI don't think perplexity is something we should ever use for these kind of tests, but rather proper generations / bench: \r\n- perplexity will be similar as this only affect the logits a tad bit \r\n- both generations will make sense. But the distribution are little by little shifted, and even small answers can be affected. ", "Also, perplexity is an average score, I'm not overly familiar with the typical test data, but I assume it's probably not pushing corner cases? well formed? \r\n\r\nWhat I was looking at comparing some forward pass outputs with different cpu vs gpu, bfloat16 vs float16 vs float32 precisions for computing those sin/cos embeds, the differences were significant. The logit (output of the model) differences could also be quite significant, but I was looking at worst case, not average logit diffs, the mean is pretty unintersting, most logits were close, but the worst ones were well outside the range I'd consider reasonable... and it's the worst case that cause networks to blow up...", "Perhaps my message above came across incorrectly 😅 \r\n\r\nI trust what you wrote, that it should be computed in FP32. I meant that we should have a few concrete failure examples to a) test against and prevent regressions and b) document why we made modeling changes (especially ones that increase HW requirements). Since I had yet to come across this particular issue and wasn't aware of the type of numerical issue (systemic drift vs infrequent failure), a few extra pointers were needed to speed things up. A reproducible script would be even better. We do request this on our contributors' issues 🤗 \r\n\r\nNow I have a more precise target, which facilitates search: infrequent large differences. I'm going to dig up a clear example and open a PR to add the corresponding RoPE fix and tests.", "An example of a failure case is below. \r\n\r\n```py\r\nfrom transformers import AutoModelForCausalLM\r\nimport torch\r\n\r\nmodel_1 = AutoModelForCausalLM.from_pretrained(\r\n \"HuggingFaceM4/tiny-random-LlamaForCausalLM\",\r\n device_map=\"auto\",\r\n torch_dtype=torch.bfloat16,\r\n)\r\nmodel_2 = AutoModelForCausalLM.from_pretrained(\r\n \"HuggingFaceM4/tiny-random-LlamaForCausalLM\",\r\n device_map=\"auto\",\r\n).to(torch.bfloat16)\r\n\r\n# `torch_dtype=...` doesn't cast explicitly set types, `.to(...)` does\r\nassert model_1.model.layers[0].self_attn.rotary_emb.inv_freq.dtype == torch.float32\r\nassert model_2.model.layers[0].self_attn.rotary_emb.inv_freq.dtype == torch.bfloat16\r\n\r\n# sequence length smaller than the initialized length (2048) -> no problem\r\ninput_ids = torch.randint(0, 32000, (1, 1024)).to(\"cuda\")\r\nmodel_1_out = model_1(input_ids)\r\nmodel_2_out = model_2(input_ids)\r\nassert torch.allclose(model_1_out.logits, model_2_out.logits)\r\n\r\n# sequence length larger than the initialized length (2048) -> problem\r\n# why? larger than initialized length -> sin/cos have to be recomputed -> the different type of non-permanent buffers\r\n# will have an impact\r\ninput_ids = torch.randint(0, 32000, (1, 2049)).to(\"cuda\")\r\nmodel_1_out = model_1(input_ids)\r\nmodel_2_out = model_2(input_ids)\r\nassert torch.allclose(model_1_out.logits, model_2_out.logits)\r\n```\r\n\r\nIt is extremely easy to find the bug when `.to()` is used instead of `torch_dtype` -- but only after the original sequence length on `main`, due to the existing order of operations. Anything that fiddles with types at initialization time (like DeepSpeed) will run into the problem immediately, even before breaking the sequence length.\r\n\r\nThe same perplexity script can also find the problem, using the `.to()` method to cast the model:\r\n\r\n\r\n👉 moving forward with the PR to ensure this OP stays in FP32 and these artefacts are no longer present ", "super nice 🤗 " ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28685/reactions", "total_count": 4, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 4, "rocket": 0, "eyes": 0 }

{ "login": "stevhliu", "id": 59462357, "node_id": "MDQ6VXNlcjU5NDYyMzU3", "avatar_url": "https://avatars.githubusercontent.com/u/59462357?v=4", "gravatar_id": "", "url": "https://api.github.com/users/stevhliu", "html_url": "https://github.com/stevhliu", "followers_url": "https://api.github.com/users/stevhliu/followers", "following_url": "https://api.github.com/users/stevhliu/following{/other_user}", "gists_url": "https://api.github.com/users/stevhliu/gists{/gist_id}", "starred_url": "https://api.github.com/users/stevhliu/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/stevhliu/subscriptions", "organizations_url": "https://api.github.com/users/stevhliu/orgs", "repos_url": "https://api.github.com/users/stevhliu/repos", "events_url": "https://api.github.com/users/stevhliu/events{/privacy}", "received_events_url": "https://api.github.com/users/stevhliu/received_events", "type": "User", "site_admin": false }

[]

[]

[]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28684/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "url": "https://api.github.com/repos/huggingface/transformers/pulls/28684", "html_url": "https://github.com/huggingface/transformers/pull/28684", "diff_url": "https://github.com/huggingface/transformers/pull/28684.diff", "patch_url": "https://github.com/huggingface/transformers/pull/28684.patch", "merged_at": 1706123940000 }

{ "login": "cohml", "id": 62400541, "node_id": "MDQ6VXNlcjYyNDAwNTQx", "avatar_url": "https://avatars.githubusercontent.com/u/62400541?v=4", "gravatar_id": "", "url": "https://api.github.com/users/cohml", "html_url": "https://github.com/cohml", "followers_url": "https://api.github.com/users/cohml/followers", "following_url": "https://api.github.com/users/cohml/following{/other_user}", "gists_url": "https://api.github.com/users/cohml/gists{/gist_id}", "starred_url": "https://api.github.com/users/cohml/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/cohml/subscriptions", "organizations_url": "https://api.github.com/users/cohml/orgs", "repos_url": "https://api.github.com/users/cohml/repos", "events_url": "https://api.github.com/users/cohml/events{/privacy}", "received_events_url": "https://api.github.com/users/cohml/received_events", "type": "User", "site_admin": false }

[ { "id": 2155169140, "node_id": "MDU6TGFiZWwyMTU1MTY5MTQw", "url": "https://api.github.com/repos/huggingface/transformers/labels/trainer", "name": "trainer", "color": "2ef289", "default": false, "description": "" }, { "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" } ]

[]

[ "@cohml \r\nhave you tried adding the `disable_tqdm` parameter to your `TrainingArguments` ? \r\n\r\nyou can read more about it here : https://huggingface.co/docs/transformers/v4.37.1/en/main_classes/trainer#transformers.TrainingArguments.disable_tqdm", "Oh man, that's exactly what I was after! Thanks so much @not-lain for pointing it out! I will close this issue now." ]

{ "url": "https://api.github.com/repos/huggingface/transformers/issues/28683/reactions", "total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0 }

{ "login": "ydshieh", "id": 2521628, "node_id": "MDQ6VXNlcjI1MjE2Mjg=", "avatar_url": "https://avatars.githubusercontent.com/u/2521628?v=4", "gravatar_id": "", "url": "https://api.github.com/users/ydshieh", "html_url": "https://github.com/ydshieh", "followers_url": "https://api.github.com/users/ydshieh/followers", "following_url": "https://api.github.com/users/ydshieh/following{/other_user}", "gists_url": "https://api.github.com/users/ydshieh/gists{/gist_id}", "starred_url": "https://api.github.com/users/ydshieh/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/ydshieh/subscriptions", "organizations_url": "https://api.github.com/users/ydshieh/orgs", "repos_url": "https://api.github.com/users/ydshieh/repos", "events_url": "https://api.github.com/users/ydshieh/events{/privacy}", "received_events_url": "https://api.github.com/users/ydshieh/received_events", "type": "User", "site_admin": false }

[]