bump(deps-dev): bump transformers from 5.5.4 to 5.9.0

[dependabot]

Bumps transformers from 5.5.4 to 5.9.0.

Release notes

Sourced from transformers's releases.

Release v5.9.0

New Model additions

Cohere2Moe

Command A+ is a Mixture-of-Experts (MoE) language model from Cohere that features a hybrid attention pattern combining sliding window and full attention layers. The model incorporates both shared and routed experts and supports a very large context window for processing extensive text sequences.

Links: Documentation

• Add new cohere2_moe model (#46115) by @​Cyrilvallez in #46115

Parakeet tdt (#44171)

• Parakeet tdt (#44171) by @​lmaksym

HRM-Text

HRM-Text is an improved autoregressive language-modeling variant of the Hierarchical Reasoning Model (HRM) that uses a hierarchical recurrent forward pass with two transformer stacks - one for slow, abstract planning (H) and one for fast, detailed computation (L) - reused inside a nested recurrence. It features PrefixLM attention where instruction tokens attend bidirectionally while response tokens attend causally, per-head sigmoid output gates, and parameterless RMSNorm. The model is designed as a base language model without instruction tuning or chat templates.

Links: Documentation | Paper

• Add hrm text (#46025) by @​abcd1927 in #46025

Breaking changes

The text_embeds input for SAM3, EdgeTAM, and SAM3-Lite-Text models now expects full text embeddings instead of just pooler outputs, aligning with other models in the library — users must update their inputs accordingly.

• 🚨Fix memory leaks caused by lru decorators in vision models (#45922) by @​yonigozlan

Audio

Audio support was expanded with the addition of AudioFlamingoNext model checkpoints and improved compilability of audio/vision encoders via standalone pure functions. Additional improvements include better error messaging when loading audio from video files and new documentation for audio/video processors.

• user friendly error when loading audio from video (#45221) by @​eustlb in #45221
• [docs] adding audio/video processors (#45795) by @​stevhliu in #45795
• Support Audio Flamingo Next checkpoints (#44830) by @​lashahub in #44830
• Extract dynamic vision/audio tensors into standalone pure functions (#45396) by @​IlyasMoutawwakil in #45396

Generation

Fixed generation issues including inputs_embeds and per_layer_inputs handling for Gemma4, an AttributeError in RAG's generate() caused by missing config fields, and flaky VLM generation tests by blocking special image tokens during sampling.

• Fix Gemma4 generation from inputs_embeds and per_layer_inputs (#46049) by @​Cyrilvallez in #46049
• Fix AttributeError in RAG generate() for missing config fields (#46035) by @​Sriniketh24 in #46035

... (truncated)

Commits

• 0a2757d release v5.9.0
• e370a7f fix cohere2 tp_plan for release
• f59ffd1 Add new cohere2_moe model (#46115)
• 4f41f26 [loading] Free up tensors faster inside ConversionOps (#46110)
• d5dd7ee Fix post processing RF-DETR (#46041)
• 0b25f8c [serve] Support for reasoning (#45690)
• 0df9b7f Fix Gemma4 generation from inputs_embeds and per_layer_inputs (#46049)
• 38a8b55 Parakeet tdt (#44171)
• 3428030 Remove mask visualization tool from masking_utils.py (#46066)
• dda0650 user friendly error when loading audio from video (#45221)
• Additional commits viewable in compare view

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[Nicola Franco (franconicola)]

LGTM