build(deps): bump the dependencies group with 5 updates

[dependabot]

Bumps the dependencies group with 5 updates:

Package	From	To
ctranslate2	4.2.1	4.3.0
transformers	4.40.2	4.41.0
pytest	8.2.0	8.2.1
pytest-asyncio	0.23.6	0.23.7
pylint	3.2.0	3.2.2

Updates ctranslate2 from 4.2.1 to 4.3.0

Release notes

Sourced from ctranslate2's releases.

CTranslate2 4.3.0

New features

• Support phi-3 (8k and 128k) (#1700 and #1680)

Fixes and improvements

• Fix regression Flash Attention (#1695)

Changelog

Sourced from ctranslate2's changelog.

v4.3.0 (2024-05-17)

New features

• Support phi-3 (8k and 128k) (#1700 and #1680)

Fixes and improvements

• Fix regression Flash Attention (#1695)

Commits

• 173a0d1 bump version 4.3.0 (#1701)
• a1eaa71 support phi-3 128k (#1700)
• 580c685 fix regression flash attention (#1695)
• 9d54f5d Add phi3 converter (#1680)
• See full diff in compare view

Updates transformers from 4.40.2 to 4.41.0

Release notes

Sourced from transformers's releases.

v4.41.0: Phi3, JetMoE, PaliGemma, VideoLlava, Falcon2, FalconVLM & GGUF support

New models

Phi3

The Phi-3 model was proposed in Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone by Microsoft.

TLDR; Phi-3 introduces new ROPE scaling methods, which seems to scale fairly well! A 3b and a Phi-3-mini is available in two context-length variants—4K and 128K tokens. It is the first model in its class to support a context window of up to 128K tokens, with little impact on quality.

• Phi-3 by @​gugarosa in huggingface/transformers#30423

JetMoE

JetMoe-8B is an 8B Mixture-of-Experts (MoE) language model developed by Yikang Shen and MyShell. JetMoe project aims to provide a LLaMA2-level performance and efficient language model with a limited budget. To achieve this goal, JetMoe uses a sparsely activated architecture inspired by the ModuleFormer. Each JetMoe block consists of two MoE layers: Mixture of Attention Heads and Mixture of MLP Experts. Given the input tokens, it activates a subset of its experts to process them. This sparse activation schema enables JetMoe to achieve much better training throughput than similar size dense models. The training throughput of JetMoe-8B is around 100B tokens per day on a cluster of 96 H100 GPUs with a straightforward 3-way pipeline parallelism strategy.

• Add JetMoE model by @​yikangshen in huggingface/transformers#30005

PaliGemma

PaliGemma is a lightweight open vision-language model (VLM) inspired by PaLI-3, and based on open components like the SigLIP vision model and the Gemma language model. PaliGemma takes both images and text as inputs and can answer questions about images with detail and context, meaning that PaliGemma can perform deeper analysis of images and provide useful insights, such as captioning for images and short videos, object detection, and reading text embedded within images.

More than 120 checkpoints are released see the collection here !

• Add PaliGemma by @​molbap in huggingface/transformers#30814

VideoLlava

Video-LLaVA exhibits remarkable interactive capabilities between images and videos, despite the absence of image-video pairs in the dataset.

💡 Simple baseline, learning united visual representation by alignment before projection With the binding of unified visual representations to the language feature space, we enable an LLM to perform visual reasoning capabilities on both images and videos simultaneously. 🔥 High performance, complementary learning with video and image Extensive experiments demonstrate the complementarity of modalities, showcasing significant superiority when compared to models specifically designed for either images or videos.

• Add Video Llava by @​zucchini-nlp in huggingface/transformers#29733

Falcon 2 and FalconVLM:

... (truncated)

Commits

• 4c6c45b Release: v4.41.0
• e9a8041 update release script (#30880)
• 0a9300f Support arbitrary processor (#30875)
• 57edd84 [whisper] fix multilingual fine-tuning (#30865)
• 977ce58 Fix dependencies for image classification example (#30842)
• 3802e78 Enable device map (#30870)
• 57c965a Remove deprecated logic and warnings (#30743)
• 3d7d3a8 TEST: Add llama logits tests (#30835)
• 15c74a2 Fix VideoLlava imports (#30867)
• 4e17e7d TST / Quantization: Reverting to torch==2.2.1 (#30866)
• Additional commits viewable in compare view

Updates pytest from 8.2.0 to 8.2.1

Release notes

Sourced from pytest's releases.

8.2.1

pytest 8.2.1 (2024-05-19)

Improvements

• #12334: Support for Python 3.13 (beta1 at the time of writing).

Bug Fixes

• #12120: Fix [PermissionError]{.title-ref} crashes arising from directories which are not selected on the command-line.
• #12191: Keyboard interrupts and system exits are now properly handled during the test collection.
• #12300: Fixed handling of 'Function not implemented' error under squashfuse_ll, which is a different way to say that the mountpoint is read-only.
• #12308: Fix a regression in pytest 8.2.0 where the permissions of automatically-created .pytest_cache directories became rwx------ instead of the expected rwxr-xr-x.

Trivial/Internal Changes

• #12333: pytest releases are now attested using the recent Artifact Attestation support from GitHub, allowing users to verify the provenance of pytest's sdist and wheel artifacts.

Commits

• 66ff8df Prepare release version 8.2.1
• 3ffcfd1 Merge pull request #12340 from pytest-dev/backport-12334-to-8.2.x
• 0b28313 [8.2.x] Add Python 3.13 (beta) support
• f3dd93a [8.2.x] Attest package provenance (#12335)
• bb5a125 [8.2.x] Spelling (#12331)
• f179bf2 Merge pull request #12327 from pytest-dev/backport-12325-to-8.2.x
• 2b671b5 [8.2.x] cacheprovider: fix .pytest_cache not being world-readable
• 65ab7cb Merge pull request #12324 from pytest-dev/backport-12320-to-8.2.x
• 4d5fb7d Merge pull request #12319 from pytest-dev/backport-12311-to-8.2.x
• cbe5996 [8.2.x] changelog: document unittest 8.2 change as breaking
• Additional commits viewable in compare view

Updates pytest-asyncio from 0.23.6 to 0.23.7

Release notes

Sourced from pytest-asyncio's releases.

pytest-asyncio 0.23.7

0.23.7 (2024-05-19)

• Silence deprecation warnings about unclosed event loops that occurred with certain CPython patch releases #817

Known issues

As of v0.23, pytest-asyncio attaches an asyncio event loop to each item of the test suite (i.e. session, packages, modules, classes, functions) and allows tests to be run in those loops when marked accordingly. Pytest-asyncio currently assumes that async fixture scope is correlated with the new event loop scope. This prevents fixtures from being evaluated independently from the event loop scope and breaks some existing test suites (see #706). For example, a test suite may require all fixtures and tests to run in the same event loop, but have async fixtures that are set up and torn down for each module. If you're affected by this issue, please continue using the v0.21 release, until it is resolved.

Commits

• eb63d5a docs: Prepared for release of v0.23.7.
• da04a7a Build(deps): Bump exceptiongroup in /dependencies/default
• 00c667a Build(deps): Bump pytest from 8.1.1 to 8.2.0 in /dependencies/default
• 3bd9cd8 [docs] Add changelog entry.
• 15544f0 Revert GitHub Actions and Tox changes.
• 6316b28 Deduplicate simplefilter snippet.
• 3ffdfc5 asyncio.run(port_afinalizer())
• 0107fd7 Remove extra space.
• 3bf700a Fix GH Action mapping.
• d15dc31 Fix 3109/3108 typo.
• Additional commits viewable in compare view

Updates pylint from 3.2.0 to 3.2.2

Commits

• 769ffd2 Bump pylint to 3.2.2, update changelog (#9658)
• 98c5af9 Fix false-positive with contextmanager missing cleanup (#9654) (#9657)
• 9a9db8f Update astroid to 3.2.2 (#9655) (#9656)
• 9223172 Bump pylint to 3.2.1, update changelog
• 926547b [trailing-comma-tuple] Fix enabling with message control locally when disable...
• 1498675 Fix linterstats.get_module_message_count() (#9146) (#9648)
• aed496a Fix FP for possibly-used-before-assignment with assert_never() (#9645) (#...
• 9dae975 [Backport maintenance/3.2.x] Add --prefer-stubs=y option (#9646)
• a03ceae Add --prefer-stubs=y option (#9632)
• b2ea316 [Backport maintenance/3.2.x] Don't emit incorrect-variance for type parameter...
• 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 merge will merge this PR after your CI passes on it
• @dependabot squash and merge will squash and merge this PR after your CI passes on it
• @dependabot cancel merge will cancel a previously requested merge and block automerging
• @dependabot reopen will reopen this PR if it is closed
• @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore <dependency name> major version will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself)
• @dependabot ignore <dependency name> minor version will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself)
• @dependabot ignore <dependency name> will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself)
• @dependabot unignore <dependency name> will remove all of the ignore conditions of the specified dependency
• @dependabot unignore <dependency name> <ignore condition> will remove the ignore condition of the specified dependency and ignore conditions