multimodalart/higgs-audio-v3-tts-4b-transformers

codec load: trust_remote_code=True

Use transformers 5.x key_mapping for weight remap (fixes MISSING weights)

Fix tie_weights signature for transformers 5.x

Upload README.md with huggingface_hub

Upload modeling_higgs_multimodal_qwen3.py with huggingface_hub

Upload configuration_higgs_multimodal_qwen3.py with huggingface_hub

Upload config.json with huggingface_hub

Copy tokenizer config, chat template, index, license

Server-side copy weights + tokenizer.json from bosonai/higgs-audio-v3-tts-4b

initial commit