SALMONN: Speech Audio Language Music Open Neural Network

🚀🚀 Welcome to the repo of SALMONN!

SALMONN is a large language model (LLM) enabling speech, audio events, and music inputs, which is developed by the Department of Electronic Engineering at Tsinghua University and ByteDance. Instead of speech-only input or audio-event-only input, SALMONN can perceive and understand all kinds of audio inputs and therefore obtain emerging capabilities such as multilingual speech recognition and translation and audio-speech co-reasoning. This can be regarded as giving the LLM "ears" and cognitive hearing abilities, which makes SALMONN a step towards hearing-enabled artificial general intelligence.

🌟 Structure

The model architecture of SALMONN is shown below. A window-level Q-Former is used as the connection module to fuse the outputs from a Whisper speech encoder and a BEATs audio encoder as augmented audio tokens, which are aligned with the LLM input space. The LoRA adaptor aligns the augmented LLM input space with its output space. The text prompt is used to instruct SALMONN to answer open-ended questions about the general audio inputs and the answers are in the LLM text responses.

⚡️ Demos

Compared with traditional speech and audio processing tasks such as speech recognition and audio caption, SALMONN leverages the general knowledge and cognitive abilities of the LLM to achieve a cognitively oriented audio perception, which dramatically improves the versatility of the model and the richness of the task. In addition, SALMONN is able to follow textual commands and even spoken commands with a relatively high degree of accuracy. Since SALMONN only uses training data based on textual commands, listening to spoken commands is also a cross-modal emergent ability.

Here are some examples of SALMONN.

Audio	Response
gunshots.wav
duck.wav
music.wav

Datasets

• Download raw audio files from here
  • Put downloaded directory path into data_prefix of config
  • contains 1.4TB of audio

    168G WavCaps
    165G audiocaps
    110G GigaSpeech
    58G LibriSpeech
    3.7G MusicNet
    2.0G Clotho
    

• Download annotation files from here
  • place the jsons under data directory.
  • NOTE: Only train split will be released to public.

🌈 How to train a model

For SALMONN-13B v1, you need to use the following dependencies:

1. Our environment: The python version is 3.9.17, and other required packages can be installed with the following command: pip install -r requirements.txt.
2. Download whisper large v2 to whisper_path.
3. Download Fine-tuned BEATs_iter3+ (AS2M) (cpt2) to beats_path.
4. Download vicuna 13B v1.1 to llama_path.
5. Running with python3 train.py --cfg-path configs/config.yaml
6. You may try --dryrun for loading dataset and dummy small model.

🌈 How to inference in CLI

1. Same as How to train a model: 1-4.
2. Download salmonn v1 to ckpt.
3. Running with python3 cli_inference.py --cfg-path configs/decode_config.yaml Now you can input wav_path and prompt. Enjoy yourself !

🌈 How to launch a web demo

1. Same as How to train a model: 1-4.
2. Download salmonn v1 to ckpt.
3. Running with python3 web_demo.py --cfg-path configs/decode_config.yaml

👀 Team

Team Tsinghua: Wenyi Yu, Changli Tang, Guangzhi Sun, Chao Zhang

Team ByteDance: Xianzhao Chen, Wei Li, Tian Tan, Lu Lu, Zejun Ma

✨ Citation

If you find SALMONN / video-SALMONN useful, please cite the paper:

@inproceedings{
tang2024salmonn,
title={{SALMONN}: Towards Generic Hearing Abilities for Large Language Models},
author={Changli Tang and Wenyi Yu and Guangzhi Sun and Xianzhao Chen and Tian Tan and Wei Li and Lu Lu and Zejun MA and Chao Zhang},
booktitle={The Twelfth International Conference on Learning Representations},
year={2024},
url={https://openreview.net/forum?id=14rn7HpKVk}
}