Type: huggingface.text_to_audio.AudioLDM2
Namespace: huggingface.text_to_audio
Description
Generates audio using the AudioLDM2 model based on text prompts. audio, generation, AI, text-to-audio
Use cases:
- Create custom sound effects based on textual descriptions
- Generate background audio for videos or games
- Produce audio content for multimedia projects
- Explore AI-generated audio for creative sound design
Properties
| Property | Type | Description | Default |
|---|---|---|---|
| prompt | str |
A text prompt describing the desired audio. | The sound of a hammer hitting a wooden surface. |
| negative_prompt | str |
A text prompt describing what you don’t want in the audio. | Low quality. |
| num_inference_steps | int |
Number of denoising steps. More steps generally improve quality but increase generation time. | 200 |
| audio_length_in_s | float |
The desired duration of the generated audio in seconds. | 10.0 |
| num_waveforms_per_prompt | int |
Number of audio samples to generate per prompt. | 3 |
| seed | int |
Seed for the random number generator. Use -1 for a random seed. | 0 |
Outputs
| Output | Type | Description |
|---|---|---|
| output | audio |
Metadata
Related Nodes
Browse other nodes in the huggingface.text_to_audio namespace.