[Model] Add Qwen3 and Qwen3MoE#15289

Merged

vllm-bot merged 11 commits intovllm-project:mainfrom

YamPengLi:qwen3

Apr 7, 2025

Commits on Apr 3, 2025

Initial commit for Qwen3
YamPengLi
committed
Update Qwen3 model registration in the model registry to correct identifiers for Qwen3 and Qwen3MoE models.
YamPengLi
committed
Refactor Qwen3 model to inherit from Qwen2Model, adding support for custom decoder layer types and simplifying the architecture. Removed unused imports and parameters for clarity.
YamPengLi
committed
Refactor Qwen3MoE model to streamline parameter handling by adding prefix support for various components, enhancing clarity and maintainability. Removed unused imports and parameters to simplify th…
YamPengLi
committed
Add Qwen3 and Qwen3MoE models to the model registry with online availability set to false, enhancing the model catalog.
YamPengLi
committed
Add packed_modules_mapping to Qwen3ForCausalLM for LoRA supporting
YamPengLi
committed
Updata docs
YamPengLi
committed
Update docs/source/models/supported_models.md

YamPengLi
and
DarkLight1337
authored
Update Qwen2Model to set default decoder_layer_type to Qwen2DecoderLayer for improved clarity and usability.
YamPengLi
committed
Merge remote-tracking branch 'origin/main' into qwen3
YamPengLi
committed

Commits on Apr 7, 2025

Add min_transformers_version
YamPengLi
committed