metadata

{}

AudioLDM

AudioLDM is a latent text-to-audio diffusion model capable of generating realistic audio samples given any text input. It is available in the 🧨 Diffusers library from v0.15.0 onwards.

Model Details

AudioLDM was proposed in the paper AudioLDM: Text-to-Audio Generation with Latent Diffusion Models by Haohe Liu et al.

Inspired by Stable Diffusion, AudioLDM is a text-to-audio latent diffusion model (LDM) that learns continuous audio representations from CLAP latents. AudioLDM takes a text prompt as input and predicts the corresponding audio. It can generate text-conditional sound effects, human speech and music.

This is the large version of the AudioLDM model, with twice the number of UNet channels and head channels as the small checkpoint. The three AudioLDM checkpoints are summarised in the table below.

Table 1: Summary of the AudioLDM checkpoints. Frechet Distance (FD) indicates the similarity between generated samples and target samples, where lower is better.

Checkpoint	Training Data (h)	Training Steps	Params	Frechet Distance ↓
AudioLDM-S	137	0.6M	181M	29.48
AudioLDM-L	137	0.6M	739M	27.12
AudioLDM-L-Full	9174	1.5M	739M	23.31

Model Sources

Usage

First, install the required packages:

pip install --upgrade diffusers transformers

Text-to-Audio

For text-to-audio generation, the AudioLDMPipeline can be used to load pre-trained weights and generate text-conditional audio outputs:

from diffusers import AudioLDMPipeline
import torch

repo_id = "cvssp/audioldm-l"
pipe = AudioLDMPipeline.from_pretrained(repo_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")

prompt = "Techno music with a strong, upbeat tempo and high melodic riffs"
audio = pipe(prompt, num_inference_steps=10, audio_length_in_s=5.0).audios[0]

The resulting audio output can be saved as a .wav file:

import scipy

scipy.io.wavfile.write("techno.wav", rate=16000, data=audio)

Or displayed in a Jupyter Notebook / Google Colab:

from IPython.display import Audio

Audio(audio, rate=16000)

Tips

Prompts:

Descriptive prompt inputs work best: you can use adjectives to describe the sound (e.g. "high quality" or "clear") and make the prompt context specific (e.g., "water stream in a forest" instead of "stream").
It's best to use general terms like 'cat' or 'dog' instead of specific names or abstract objects that the model may not be familiar with.

Inference:

The quality of the predicted audio sample can be controlled by the num_inference_steps argument: higher steps give higher quality audio at the expense of slower inference.
The length of the predicted audio sample can be controlled by varying the audio_length_in_s argument.

Citation

BibTeX:

@article{liu2023audioldm,
  title={AudioLDM: Text-to-Audio Generation with Latent Diffusion Models},
  author={Liu, Haohe and Chen, Zehua and Yuan, Yi and Mei, Xinhao and Liu, Xubo and Mandic, Danilo and Wang, Wenwu and Plumbley, Mark D},
  journal={arXiv preprint arXiv:2301.12503},
  year={2023}
}