File size: 1,080 Bytes
e15fe5a 45a1c50 1abd670 45a1c50 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
---
license: mit
---
# Amphion Text-to-Audio Pretrained Models
We provide the following pretrained checkpoints for you to use, specifically:
Two [AudioLDM](https://github.com/open-mmlab/Amphion/tree/main/egs/tta/audioldm) pretrained checkpoints with corresponding [AutoencoderKL](https://github.com/open-mmlab/Amphion/tree/main/egs/tta/autoencoderkl) checkpoints trained on AudioCaps.
## Quick Start
To utilize the pretrained models, just run the following commands:
### Step1: Download the checkpoint
```bash
git lfs install
git clone https://huggingface.co/amphion/text_to_audio
```
### Step2: Clone the Amphion's Source Code of GitHub
```bash
git clone https://github.com/open-mmlab/Amphion.git
```
### Step3: Specify the checkpoint's path
Use the soft link to specify the downloaded checkpoint in the first step:
```bash
cd Amphion
mkdir -p ckpts
ln -s ../../../text_to_speech/tta ckpts/
```
### Step4: Inference
You can follow the inference part of [this recipe](https://github.com/open-mmlab/Amphion/tree/main/egs/tta/RECIPE.md) to generate audio from text.
|