Tango2 Full

🚀 Tango 2: Aligning Diffusion-based Text-to-Audio Generative Models through Direct Preference Optimization

We developed Tango 2 based on Tango for text-to-audio generation. Tango 2 was initialized with the Tango-full-ft checkpoint and trained using Direct Preference Optimization (DPO) on audio-alpaca, a pairwise text-to-audio preference dataset. Tango-2-full was trained on an extended version of Audio-alpaca.

Read the paper

🚀 Quick Start

Download and Generate Audio

Download the Tango 2 model and generate audio from a text prompt:

import IPython
import soundfile as sf
from tango import Tango

tango = Tango("declare-lab/tango2-full")

prompt = "An audience cheering and clapping"
audio = tango.generate(prompt)
sf.write(f"{prompt}.wav", audio, samplerate=16000)
IPython.display.Audio(data=audio, rate=16000)

The model will be automatically downloaded and saved in cache. Subsequent runs will load the model directly from cache.

Adjusting Generation Steps

The generate function uses 100 steps by default to sample from the latent diffusion model. We recommend using 200 steps for generating better quality audios. This comes at the cost of increased run-time.

prompt = "Rolling thunder with lightning strikes"
audio = tango.generate(prompt, steps=200)
IPython.display.Audio(data=audio, rate=16000)

Batch Generation

Use the generate_for_batch function to generate multiple audio samples for a batch of text prompts:

prompts = [
    "A car engine revving",
    "A dog barks and rustles with some clicking",
    "Water flowing and trickling"
]
audios = tango.generate_for_batch(prompts, samples=2)

This will generate two samples for each of the three text prompts.

💻 Usage Examples

Basic Usage

import IPython
import soundfile as sf
from tango import Tango

tango = Tango("declare-lab/tango2-full")

prompt = "An audience cheering and clapping"
audio = tango.generate(prompt)
sf.write(f"{prompt}.wav", audio, samplerate=16000)
IPython.display.Audio(data=audio, rate=16000)

Advanced Usage

# Generate audio with 200 steps for better quality
prompt = "Rolling thunder with lightning strikes"
audio = tango.generate(prompt, steps=200)
IPython.display.Audio(data=audio, rate=16000)

# Generate multiple audio samples for a batch of text prompts
prompts = [
    "A car engine revving",
    "A dog barks and rustles with some clicking",
    "Water flowing and trickling"
]
audios = tango.generate_for_batch(prompts, samples=2)

📦 Installation

Our code is released here: https://github.com/declare-lab/tango

Please follow the instructions in the repository for installation, usage and experiments.

📄 License

This project is licensed under the CC BY-NC-SA 4.0 license.

📚 Documentation

Model Information