Custom model configs that cannot be expressed via HuggingFace Transformers alone, following the same pattern as vllm.transformers_utils.configs.
Modules:
| Name | Description |
|---|---|
audio8_tts | Audio8 TTS config registration with transformers AutoConfig. |
auk | Config for the AuK instruction-driven audio generation model. |
cosyvoice3 | |
fish_speech | Fish Speech S2 Pro config registration with transformers AutoConfig. |
glm_tts | GLM-TTS config registration with transformers AutoConfig. |
higgs_audio_v3 | Configuration class for higgs-audio v3 (HiggsMultimodalQwen3) in vllm-omni. |
mammoth_moda2 | |
ming_flash_omni | Configuration for Ming-flash-omni-2.0 model |
minimax_music3 | HuggingFace config for the MiniMax Music 3 composite repository. |
omnivoice | OmniVoice configuration for vLLM-Omni two-stage pipeline. |
sensenova_u1 | HuggingFace-style configuration classes for SenseNova-U1. |
voxcpm2 | |
voxtral_tts | |