--- license: apache-2.0 base_model: FireRedTeam/FireRedTTS3 pipeline_tag: text-to-speech tags: - sglang-omni - fireredtts3 - text-to-speech - voice-cloning --- # FireRedTTS3-Instruct for SGLang Omni This repository contains the officially released HF-format FireRedTTS3-Instruct checkpoint for inference with [SGLang Omni](https://github.com/sgl-project/sglang-omni). It supports reference-audio voice cloning, instruction-guided voice design, semantic editing, and acoustic editing. Audio output is 24 kHz mono, with WAV responses and streaming PCM available through the `/v1/audio/speech` API. Our model code and original checkpoint release: - [FireRedTTS3 source code](https://github.com/FireRedTeam/FireRedTTS3) - [Original FireRedTTS3 weights and model card](https://huggingface.co/FireRedTeam/FireRedTTS3) This card documents the SGLang Omni loading and serving workflow. ## Install SGLang Omni Use Linux with an NVIDIA CUDA GPU and Python 3.12. Follow the [SGLang Omni installation guide](https://github.com/sgl-project/sglang-omni/blob/main/docs/get_started/installation.md) for the CUDA, UCX, and FlashAttention prerequisites, then install from source: ```bash git clone https://github.com/sgl-project/sglang-omni.git cd sglang-omni pip install uv uv venv .venv -p 3.12 source .venv/bin/activate uv pip install --prerelease=allow -e '.[firered-tts3]' ``` Run the following commands from the `sglang-omni` repository root. ## Download this checkpoint Download the official HF-format checkpoint from [FireRedTeam/FireRedTTS3-Instruct-HF](https://huggingface.co/FireRedTeam/FireRedTTS3-Instruct-HF). ```bash hf download FireRedTeam/FireRedTTS3-Instruct-HF \ --local-dir ./models/FireRedTTS3-Instruct ``` Download the complete repository, including configuration, tokenizer, and `redae/` assets. Loading only the root `model.safetensors` file is insufficient. If you already have the complete checkpoint locally, use that directory directly. ## Start the server ```bash CUDA_VISIBLE_DEVICES=0 sgl-omni serve \ --model-path ./models/FireRedTTS3-Instruct \ --config examples/configs/firered_tts3_instruct.yaml \ --allowed-local-media-path "$PWD/docs/_static/audio" \ --port 8000 ``` Wait for `/health` to return HTTP 200 before sending requests. Initial model loading, compilation, and warmup can take several minutes. The reference clips below are included in the SGLang Omni repository. Local paths are resolved on the **server**, relative to its working directory. For your own clips, allow their directory with `--allowed-local-media-path` and supply the matching transcript for cloning. ## Voice cloning Provide exactly one reference recording and its actual transcript. The example uses the repository's `ref_voice.wav`: ```bash curl --fail-with-body -X POST http://localhost:8000/v1/audio/speech \ -H 'Content-Type: application/json' \ -d '{ "model": "FireRedTTS3-Instruct", "input": "Hello, welcome to SGLang Omni.", "references": [{ "audio_path": "docs/_static/audio/ref_voice.wav", "text": "It was the night before my birthday. Hooray! It’s almost here! It may not be a holiday, but it’s the best day of the year." }], "seed": 42 }' --output clone.wav ``` Omit `task_type` for cloning, or set it to `Base`. The value `tts` is not an accepted serving task name. ## Voice design Generate a voice from a description, without a reference recording: ```bash curl --fail-with-body -X POST http://localhost:8000/v1/audio/speech \ -H 'Content-Type: application/json' \ -d '{ "model": "FireRedTTS3-Instruct", "input": "今天天气很好,我们一起去公园散步吧。", "task_type": "VoiceDesign", "instructions": "一个年轻女性的温柔嗓音,语速稍慢,带一点俏皮。", "seed": 42 }' --output voice_design.wav ``` ## Semantic editing Use the reference recording as the source to edit. Its content supplies the input, so `input` may be empty; no reference transcript is required: ```bash curl --fail-with-body -X POST http://localhost:8000/v1/audio/speech \ -H 'Content-Type: application/json' \ --data-binary @- --output semantic_edit.wav <<'JSON' { "model": "FireRedTTS3-Instruct", "input": "", "task_type": "SemanticEdit", "instructions": "Replace 'birthday' with 'anniversary'.", "ref_audio": "docs/_static/audio/ref_voice.wav", "seed": 42 } JSON ``` For substitutions, use the quoted form `Replace 'source' with 'target'.` The quoted example above was verified by transcribing the generated audio; omitting the quotes did not perform the intended replacement in this test. The `x-gen-text` response header contains text with edit masks, so the target word may appear only in the audio. Validate the rendered audio rather than requiring the target word to appear in that header. ## Acoustic editing Adjust the source recording's speed, pitch, or volume. This example requests 0.8x speed through the model's native acoustic-edit task: ```bash curl --fail-with-body -X POST http://localhost:8000/v1/audio/speech \ -H 'Content-Type: application/json' \ -d '{ "model": "FireRedTTS3-Instruct", "input": "", "task_type": "AcousticEdit", "ref_audio": "docs/_static/audio/ref_voice.wav", "speed": 0.8, "seed": 42 }' --output acoustic_edit.wav ``` Raw acoustic instructions must follow the trained templates rather than free-form descriptions. For pitch or volume, use, for example, `"instructions": "shift the pitch by 2 steps"` or `"instructions": "adjust the volume to 0.8"`. An explicit `instructions` string takes precedence over composed acoustic instructions. Without explicit `instructions`, the composed `acoustic_attribute` must be supported (`Speed`, `Pitch`, or `Volume`) and its value must be a finite number. Set `acoustic_attribute` and `acoustic_value` under `stage_params.preprocessing`. For example: ```json { "stage_params": { "preprocessing": {"acoustic_attribute": "Pitch", "acoustic_value": 2} } } ``` The top-level `speed` field supplies the Speed value for `AcousticEdit`; it does not change cloning or voice design. Edit sources are capped at 20 seconds. VoiceDesign and SemanticEdit generate text before audio, so their first-audio latency differs from cloning. ## Streaming To stream audio, add `"stream": true` and `"response_format": "pcm"` to the request and use `curl -N`. The response contains raw **24 kHz, signed 16-bit little-endian, mono PCM**. Save it as `output.pcm` and convert it with: ```bash ffmpeg -f s16le -ar 24000 -ac 1 -i output.pcm output.wav ``` The server also exposes `/v1/audio/speech/stream` over WebSocket. See the [WebSocket protocol](https://github.com/sgl-project/sglang-omni/blob/main/docs/basic_usage/tts.md#websocket-speech-streaming) for `session.config`, `input.text`, `input.commit`, and `input.done` events. ## Offline Python SDK Save this as `offline_tts.py` in the SGLang Omni repository root and run `CUDA_VISIBLE_DEVICES=0 python offline_tts.py`. This starts the pipeline locally without an HTTP server. ```python import asyncio import multiprocessing as mp from pathlib import Path from sglang_omni.client import Client, GenerateRequest from sglang_omni.models.firered_tts3.config import FireRedTTS3PipelineConfig from sglang_omni.pipeline.mp_runner import MultiProcessPipelineRunner async def main() -> None: config = FireRedTTS3PipelineConfig( model_path="./models/FireRedTTS3-Instruct", ) config.stage_named("talker").factory.flow_mode = "fp32" runner = MultiProcessPipelineRunner(config) await runner.start(timeout=600) try: client = Client(runner.coordinator) result = await client.speech( GenerateRequest( prompt={ "input": "今天天气很好,我们一起去公园散步吧。", "task_type": "VoiceDesign", "instructions": "一个年轻女性的温柔嗓音,语速稍慢,带一点俏皮。", "seed": 42, }, stream=False, ), request_id="demo-1", ) Path("offline.wav").write_bytes(result.audio_bytes) finally: await runner.stop() if __name__ == "__main__": mp.set_start_method("spawn", force=True) asyncio.run(main()) ``` ## Serving notes - The current deployment supports a single GPU (`tp_size=1`). - The `input` text is limited to 400 characters. Editing requests may use an empty `input` as shown above. - Instruct does not use the Base model's language tag. - Submit spoken-form text when normalization is needed. The FastText and LLM text-normalization setup in our standalone pipeline is not used in this quick start. - Put flow-solver settings under `stage_params.talker`, for example `{"n_timesteps": 10, "inference_cfg": 2.0, "max_patches": 200}`. Unknown top-level fields are not forwarded as model settings. - HTTP, batch, and WebSocket requests do not forward `extra_params`. The SDK `GenerateRequest.extra_params` field remains available for SDK requests. - `seed` controls the per-request flow sampler. Different precisions, hardware, or batching schedules can produce different results with the same seed. - For our published model-quality results, see the [FireRedTTS3 model card](https://huggingface.co/FireRedTeam/FireRedTTS3). SGLang Omni runtime accuracy and performance are evaluated separately. - Keep the example Instruct configuration's `flow_mode: fp32` for the supported default. The offline example sets it explicitly because it does not load the example YAML. - `temperature`, `top_p`, and `top_k` affect the VoiceDesign/SemanticEdit text phase. They do not control continuous-latent audio generation. See the [FireRedTTS3 cookbook](https://github.com/sgl-project/sglang-omni/blob/main/docs/cookbook/firered_tts3.md) for the full request parameters and deployment configuration. ## Usage Disclaimer - The project incorporates zero-shot voice cloning functionality; Please note that this capability is intended **solely for academic research purposes**. - **DO NOT** use this model for **ANY illegal activities**❗️❗️ - The developers assume no liability for any misuse of this model. - If you identify any instances of **abuse**, **misuse**, or **fraudulent** activities related to this project, **please report them to our team immediately.** ## Citation ```bib @article{fireredtts3, title = {FireRedTTS3: Unified Speech Generation and Editing with Semantically Enriched Speech Representations}, author = {FireRed Team}, journal = {arXiv preprint}, year = {2026}, } ``` ## Acknowledgements - [Qwen3](https://github.com/QwenLM/Qwen3) and [Qwen2-Audio](https://github.com/QwenLM/Qwen2-Audio) for the language model and audio understanding foundations - DiTAR for the patch-level diffusion autoregressive formulation - [X-Codec](https://github.com/zhenye234/xcodec) for the discriminator design used in RedAE training - [CAM++](https://modelscope.cn/models/iic/speech_campplus_sv_en_voxceleb_16k) for speaker embedding extraction - [fastText](https://fasttext.cc/docs/en/language-identification.html) for automatic language identification - [WeTextProcessing](https://github.com/wenet-e2e/WeTextProcessing) (wetext) for the Chinese / English text normalization front-end ## License FireRedTTS3 is released under [Apache-2.0](https://github.com/FireRedTeam/FireRedTTS3/blob/main/LICENSE). See our [original model card](https://huggingface.co/FireRedTeam/FireRedTTS3) for usage guidance and model evaluation details.