[PR #22026] [CLOSED] chore(deps): bump accelerate from 1.8.1 to 1.12.0 #97457

New Issue

GiteaMirror · 2026-05-15T23:54:31-05:00

GiteaMirror commented

2026-05-15 23:54:31 -05:00

📋 Pull Request Information

Original PR: https://github.com/open-webui/open-webui/pull/22026
Author: @dependabot[bot]
Created: 3/1/2026
Status: ❌ Closed

Base: dev ← Head: dependabot/uv/dev/accelerate-1.12.0

📝 Commits (1)

5140fa0 chore(deps): bump accelerate from 1.8.1 to 1.12.0

📊 Changes

1 file changed (+1365 additions, -1010 deletions)

View changed files

📝 uv.lock (+1365 -1010)

📄 Description

Bumps accelerate from 1.8.1 to 1.12.0.

Release notes

Sourced from accelerate's releases.

v1.12.0: Deepspeed Ulysses/ALST

Deepspeed Ulysses/ALST integration

Deepspeed Ulysses/ALST is an efficient way of training on long sequences by employing sequence parallelism and attention head parallelism. You can learn more about this technology in this paper https://arxiv.org/abs/2506.13996 or this deepspeed tutorial https://www.deepspeed.ai/tutorials/ulysses-alst-sequence-parallelism/.

To enable Deepspeed Ulysses, you first need to create ParallelismConfig and setting sp related args:
parallelism_config = ParallelismConfig(
    sp_backend="deepspeed",
    sp_size=2,
    sp_handler=DeepSpeedSequenceParallelConfig(...),
)
Then, you need to make sure to compute the correct loss as described on our docs
        ...
        losses_per_rank = torch.distributed.nn.functional.all_gather(loss, group=sp_group)
        good_tokens = (shift_labels != -100).view(-1).sum()
        good_tokens_per_rank = torch.distributed.nn.functional.all_gather(good_tokens, group=sp_group)
        total_loss = sum(
            losses_per_rank[rank] * good_tokens_per_rank[rank]
            for rank in range(sp_world_size)
            if good_tokens_per_rank[rank] > 0
        )
        total_good_tokens = sum(good_tokens_per_rank)
        loss = total_loss / max(total_good_tokens, 1)
Thanks @S1ro1 for starting this work and for @stas00 for finishing this work. Also thanks @kashif for adding docs and reviewing/testing this PR !

This feature will also be available in HF Trainer thanks for this PR from @stas00: huggingface/transformers#41832

Minor changes

Remove warning for cpu_ram_efficient_loading by @SunMarc in huggingface/accelerate#3816

update typo in bnb quantisation 4bit flag docstring by @hbraith in huggingface/accelerate#3828

ArXiv -> HF Papers by @qgallouedec in huggingface/accelerate#3834

Fix typo in broadcast_object_list docstring by @wsntxxn in huggingface/accelerate#3823

[Bug] Update torch.optim.Optimizer parameter states after tensor parallelism by @naomili0924 in huggingface/accelerate#3835

use self hosted runner by @SunMarc in huggingface/accelerate#3841

device type helper by @kashif in huggingface/accelerate#3843

New Contributors

@hbraith made their first contribution in huggingface/accelerate#3828

... (truncated)

Commits

61ff833 Release: v1.12.0
e3655b9 device type helper (#3843)
c7e59dd Deepspeed Ulysses/ALST integration (#3817)
e8f241f use self hosted runner (#3841)
00b1b18 [Bug] Update torch.optim.Optimizer parameter states after tensor parallelism ...
a73fd3a Fix typo in broadcast_object_list docstring (#3823)
a08fb32 Change arxiv to hf papers (#3834)
61bcdaa update bnb quantisation 4bit flag (#3828)
a12beee Remove warning for cpu_ram_efficient_loading (#3816)
0a0c989 v1.12.0dev
Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

@dependabot rebase will rebase this PR
@dependabot recreate will recreate this PR, overwriting any edits that have been made to it
@dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
@dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
@dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
@dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/open-webui/open-webui/pull/22026 **Author:** [@dependabot[bot]](https://github.com/apps/dependabot) **Created:** 3/1/2026 **Status:** ❌ Closed **Base:** `dev` ← **Head:** `dependabot/uv/dev/accelerate-1.12.0` --- ### 📝 Commits (1) - [`5140fa0`](https://github.com/open-webui/open-webui/commit/5140fa04843adcefded95e84facf77980234c50b) chore(deps): bump accelerate from 1.8.1 to 1.12.0 ### 📊 Changes **1 file changed** (+1365 additions, -1010 deletions) <details> <summary>View changed files</summary> 📝 `uv.lock` (+1365 -1010) </details> ### 📄 Description Bumps [accelerate](https://github.com/huggingface/accelerate) from 1.8.1 to 1.12.0. <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/huggingface/accelerate/releases">accelerate's releases</a>.</em></p> <blockquote> <h2>v1.12.0: Deepspeed Ulysses/ALST</h2> <h2>Deepspeed Ulysses/ALST integration</h2> <p>Deepspeed Ulysses/ALST is an efficient way of training on long sequences by employing sequence parallelism and attention head parallelism. You can learn more about this technology in this paper <a href="https://arxiv.org/abs/2506.13996">https://arxiv.org/abs/2506.13996</a> or this deepspeed tutorial <a href="https://www.deepspeed.ai/tutorials/ulysses-alst-sequence-parallelism/">https://www.deepspeed.ai/tutorials/ulysses-alst-sequence-parallelism/</a>.</p>  <p>To enable Deepspeed Ulysses, you first need to create <code>ParallelismConfig</code> and setting <code>sp</code> related args:</p> <pre lang="python"><code>parallelism_config = ParallelismConfig( sp_backend="deepspeed", sp_size=2, sp_handler=DeepSpeedSequenceParallelConfig(...), ) </code></pre> <p>Then, you need to make sure to compute the correct loss as described on our <a href="https://huggingface.co/docs/accelerate/main/en/concept_guides/sequence_parallelism">docs</a></p> <pre lang="python"><code> ... losses_per_rank = torch.distributed.nn.functional.all_gather(loss, group=sp_group) good_tokens = (shift_labels != -100).view(-1).sum() good_tokens_per_rank = torch.distributed.nn.functional.all_gather(good_tokens, group=sp_group) total_loss = sum( losses_per_rank[rank] * good_tokens_per_rank[rank] for rank in range(sp_world_size) if good_tokens_per_rank[rank] > 0 ) total_good_tokens = sum(good_tokens_per_rank) loss = total_loss / max(total_good_tokens, 1) </code></pre> <p>Thanks <a href="https://github.com/S1ro1"><code>@S1ro1</code></a> for starting this work and for <a href="https://github.com/stas00"><code>@stas00</code></a> for finishing this work. Also thanks <a href="https://github.com/kashif"><code>@kashif</code></a> for adding docs and reviewing/testing this PR !</p> <p>This feature will also be available in HF Trainer thanks for this PR from <a href="https://github.com/stas00"><code>@stas00</code></a>: <a href="https://redirect.github.com/huggingface/transformers/pull/41832">huggingface/transformers#41832</a></p> <h2>Minor changes</h2> <ul> <li>Remove warning for <code>cpu_ram_efficient_loading</code> by <a href="https://github.com/SunMarc"><code>@SunMarc</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3816">huggingface/accelerate#3816</a></li> <li>update typo in bnb quantisation 4bit flag docstring by <a href="https://github.com/hbraith"><code>@hbraith</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3828">huggingface/accelerate#3828</a></li> <li>ArXiv -> HF Papers by <a href="https://github.com/qgallouedec"><code>@qgallouedec</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3834">huggingface/accelerate#3834</a></li> <li>Fix typo in broadcast_object_list docstring by <a href="https://github.com/wsntxxn"><code>@wsntxxn</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3823">huggingface/accelerate#3823</a></li> <li>[Bug] Update torch.optim.Optimizer parameter states after tensor parallelism by <a href="https://github.com/naomili0924"><code>@naomili0924</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3835">huggingface/accelerate#3835</a></li> <li>use self hosted runner by <a href="https://github.com/SunMarc"><code>@SunMarc</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3841">huggingface/accelerate#3841</a></li> <li>device type helper by <a href="https://github.com/kashif"><code>@kashif</code></a> in <a href="https://redirect.github.com/huggingface/accelerate/pull/3843">huggingface/accelerate#3843</a></li> </ul> <h2>New Contributors</h2> <ul> <li><a href="https://github.com/hbraith"><code>@hbraith</code></a> made their first contribution in <a href="https://redirect.github.com/huggingface/accelerate/pull/3828">huggingface/accelerate#3828</a></li> </ul>  </blockquote> <p>... (truncated)</p> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/huggingface/accelerate/commit/61ff83394592db1052be9e64ab46beb21d794d69"><code>61ff833</code></a> Release: v1.12.0</li> <li><a href="https://github.com/huggingface/accelerate/commit/e3655b932cfb7992d7c453af51e9eda752807c39"><code>e3655b9</code></a> device type helper (<a href="https://redirect.github.com/huggingface/accelerate/issues/3843">#3843</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/c7e59dd7b3ae20c876cf277808620cf52ad716f9"><code>c7e59dd</code></a> Deepspeed Ulysses/ALST integration (<a href="https://redirect.github.com/huggingface/accelerate/issues/3817">#3817</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/e8f241f8784580784b2e88099d6687942879ed43"><code>e8f241f</code></a> use self hosted runner (<a href="https://redirect.github.com/huggingface/accelerate/issues/3841">#3841</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/00b1b1810944a5422330522a46e3044f1e6d1529"><code>00b1b18</code></a> [Bug] Update torch.optim.Optimizer parameter states after tensor parallelism ...</li> <li><a href="https://github.com/huggingface/accelerate/commit/a73fd3a19652dc61c3a8da6d33032f85f309370c"><code>a73fd3a</code></a> Fix typo in broadcast_object_list docstring (<a href="https://redirect.github.com/huggingface/accelerate/issues/3823">#3823</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/a08fb323904e1c1947258cf75d498b81ff7af801"><code>a08fb32</code></a> Change arxiv to hf papers (<a href="https://redirect.github.com/huggingface/accelerate/issues/3834">#3834</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/61bcdaa45d35a77478decc78ce204b54dc108812"><code>61bcdaa</code></a> update bnb quantisation 4bit flag (<a href="https://redirect.github.com/huggingface/accelerate/issues/3828">#3828</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/a12beee389f6bd37cfae0aba233db03f375f7f80"><code>a12beee</code></a> Remove warning for <code>cpu_ram_efficient_loading</code> (<a href="https://redirect.github.com/huggingface/accelerate/issues/3816">#3816</a>)</li> <li><a href="https://github.com/huggingface/accelerate/commit/0a0c9895bd71a8df48f59b383e02c926b7409ce6"><code>0a0c989</code></a> v1.12.0dev</li> <li>Additional commits viewable in <a href="https://github.com/huggingface/accelerate/compare/v1.8.1...v1.12.0">compare view</a></li> </ul> </details> <br /> [![Dependabot compatibility score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=accelerate&package-manager=uv&previous-version=1.8.1&new-version=1.12.0)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores) Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting `@dependabot rebase`. [//]: # (dependabot-automerge-start) [//]: # (dependabot-automerge-end) --- <details> <summary>Dependabot commands and options</summary> <br /> You can trigger Dependabot actions by commenting on this PR: - `@dependabot rebase` will rebase this PR - `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it - `@dependabot show <dependency name> ignore conditions` will show all of the ignore conditions of the specified dependency - `@dependabot ignore this major version` will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this minor version` will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this dependency` will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself) </details> --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

GiteaMirror added the pull-request label 2026-05-15 23:54:31 -05:00

GiteaMirror closed this issue

2026-05-15 23:54:34 -05:00

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/open-webui#97457