dependabot[bot] opened a new pull request, #33943: URL: https://github.com/apache/beam/pull/33943
Bumps [transformers](https://github.com/huggingface/transformers) from 4.30.0 to 4.48.0. <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/huggingface/transformers/releases";>transformers's releases</a>.</em></p> <blockquote> <h2>v4.48.0: ModernBERT, Aria, TimmWrapper, ColPali, Falcon3, Bamba, VitPose, DinoV2 w/ Registers, Emu3, Cohere v2, TextNet, DiffLlama, PixtralLarge, Moonshine</h2> <h2>New models</h2> <h3>ModernBERT</h3> <p>The ModernBert model was proposed in <a href="https://arxiv.org/abs/2412.13663";>Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference</a> by Benjamin Warner, Antoine Chaffin, Benjamin Clavié, Orion Weller, Oskar Hallström, Said Taghadouini, Alexis Galalgher, Raja Bisas, Faisal Ladhak, Tom Aarsen, Nathan Cooper, Grifin Adams, Jeremy Howard and Iacopo Poli.</p> <p>It is a refresh of the traditional encoder architecture, as used in previous models such as <a href="https://huggingface.co/docs/transformers/en/model_doc/bert";>BERT</a> and <a href="https://huggingface.co/docs/transformers/en/model_doc/roberta";>RoBERTa</a>.</p> <p>It builds on BERT and implements many modern architectural improvements which have been developed since its original release, such as:</p> <ul> <li><a href="https://huggingface.co/blog/designing-positional-encoding";>Rotary Positional Embeddings</a> to support sequences of up to 8192 tokens.</li> <li><a href="https://arxiv.org/abs/2208.08124";>Unpadding</a> to ensure no compute is wasted on padding tokens, speeding up processing time for batches with mixed-length sequences.</li> <li><a href="https://arxiv.org/abs/2002.05202";>GeGLU</a> Replacing the original MLP layers with GeGLU layers, shown to improve performance.</li> <li><a href="https://arxiv.org/abs/2004.05150v2";>Alternating Attention</a> where most attention layers employ a sliding window of 128 tokens, with Global Attention only used every 3 layers.</li> <li><a href="https://github.com/Dao-AILab/flash-attention";>Flash Attention</a> to speed up processing.</li> <li>A model designed following recent <a href="https://arxiv.org/abs/2401.14489";>The Case for Co-Designing Model Architectures with Hardware</a>, ensuring maximum efficiency across inference GPUs.</li> <li>Modern training data scales (2 trillion tokens) and mixtures (including code ande math data)</li> </ul> <p><img src="https://github.com/user-attachments/assets/4256c0b1-9b40-4d71-ac42-fc94827d5e9d"; alt="image" /></p> <ul> <li>Add ModernBERT to Transformers by <a href="https://github.com/warner-benjamin";><code>@warner-benjamin</code></a> in <a href="https://redirect.github.com/huggingface/transformers/issues/35158";>#35158</a></li> </ul> <h3>Aria</h3> <p>The Aria model was proposed in <a href="https://huggingface.co/papers/2410.05993";>Aria: An Open Multimodal Native Mixture-of-Experts Model</a> by Li et al. from the Rhymes.AI team.</p> <p>Aria is an open multimodal-native model with best-in-class performance across a wide range of multimodal, language, and coding tasks. It has a Mixture-of-Experts architecture, with respectively 3.9B and 3.5B activated parameters per visual token and text token.</p> <ul> <li>Add Aria by <a href="https://github.com/aymeric-roucher";><code>@aymeric-roucher</code></a> in <a href="https://redirect.github.com/huggingface/transformers/issues/34157";>#34157</a> <img src="https://github.com/user-attachments/assets/ef41fcc9-2c5f-4a75-ab1a-438f73d3d7e2"; alt="image" /></li> </ul> <h3>TimmWrapper</h3> <p>We add a <code>TimmWrapper</code> set of classes such that timm models can be loaded in as transformer models into the library.</p> <p>Here's a general usage example:</p> <pre lang="py"><code>import torch from urllib.request import urlopen from PIL import Image from transformers import AutoConfig, AutoModelForImageClassification, AutoImageProcessor <p>checkpoint = "timm/resnet50.a1_in1k" img = Image.open(urlopen( '<a href="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png";>https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png</a>' ))</p> <p>image_processor = AutoImageProcessor.from_pretrained(checkpoint) </tr></table> </code></pre></p> </blockquote> <p>... (truncated)</p> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/huggingface/transformers/commit/6bc0fbcfa7acb6ac4937e7456a76c2f7975fefec";><code>6bc0fbc</code></a> [WIP] Emu3: add model (<a href="https://redirect.github.com/huggingface/transformers/issues/33770";>#33770</a>)</li> <li><a href="https://github.com/huggingface/transformers/commit/59e28c30fa3a91213f569bccef73f082afa8c656";><code>59e28c3</code></a> Fix flex_attention in training mode (<a href="https://redirect.github.com/huggingface/transformers/issues/35605";>#35605</a>)</li> <li><a href="https://github.com/huggingface/transformers/commit/7cf6230e25078742b21907ae49d1542747606457";><code>7cf6230</code></a> push a fix for now</li> <li><a href="https://github.com/huggingface/transformers/commit/d6f446ffa79811d35484d445bc5c7932e8a536d6";><code>d6f446f</code></a> when filtering we can't use the convert script as we removed them</li> <li><a href="https://github.com/huggingface/transformers/commit/8ce1e9578af6151e4192d59c345e2ad86ee789d4";><code>8ce1e95</code></a> [test-all]</li> <li><a href="https://github.com/huggingface/transformers/commit/af2d7caff393cf8881396b73d92d0595b6a3b2ae";><code>af2d7ca</code></a> Add Moonshine (<a href="https://redirect.github.com/huggingface/transformers/issues/34784";>#34784</a>)</li> <li><a href="https://github.com/huggingface/transformers/commit/42b8e7916b6b6dff5cb77252286db1aa07b7b41e";><code>42b8e79</code></a> ModernBert: reuse GemmaRotaryEmbedding via modular + Integration tests (<a href="https://redirect.github.com/huggingface/transformers/issues/35459";>#35459</a>)</li> <li><a href="https://github.com/huggingface/transformers/commit/e39c9f7a78fa2960a7045e8fc5a2d96b5d7eebf1";><code>e39c9f7</code></a> v4.48-release</li> <li><a href="https://github.com/huggingface/transformers/commit/8de7b1ba8d126a6fc9f9bcc3173a71b46f0c3601";><code>8de7b1b</code></a> Add flex_attn to diffllama (<a href="https://redirect.github.com/huggingface/transformers/issues/35601";>#35601</a>)</li> <li><a href="https://github.com/huggingface/transformers/commit/1e3ddcb2d0380d0d909a44edc217dff68956ec5e";><code>1e3ddcb</code></a> ModernBERT bug fixes (<a href="https://redirect.github.com/huggingface/transformers/issues/35404";>#35404</a>)</li> <li>Additional commits viewable in <a href="https://github.com/huggingface/transformers/compare/v4.30.0...v4.48.0";>compare view</a></li> </ul> </details> <br /> [](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 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 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) You can disable automated security fix PRs for this repo from the [Security Alerts page](https://github.com/apache/beam/network/alerts). </details> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@beam.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
![https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png"](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png"){kind=link}
