[New Model] Add TADA (Hume AI) TTS model support

Model

TADA by Hume AI — a speech-language model that unifies speech synthesis with text generation using 1:1 text-speech token alignment built on Llama 3.2.

Architecture

• Encoder (HumeAI/tada-codec): encodes reference audio into aligned token sequences
• LLM (TadaForCausalLM): autoregressive generation, available in 1B and 3B-ML (multilingual) variants
• Generates the complete speech segment per text token in a single AR step (dynamic duration/prosody)

This would likely be a 2-stage pipeline in vllm-omni (AR generation → codec decode), similar to Qwen3 TTS / Fish Speech.

References

• GitHub: https://github.com/HumeAI/tada
• Requested in https://github.com/vllm-project/vllm-omni/issues/1795#issuecomment-4048885352

cc @ashok-arora — would you be interested in taking a look at this?