mrq/vall-e
5
9
Fork 0
Code Issues 8 Pull Requests Packages Projects Releases Wiki Activity

cleanup

Browse Source
This commit is contained in:
mrq 2023-10-06 10:13:54 -05:00
parent 3db7e7dea1
commit 1fd91b6437
3 changed files with 11 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
README.md
View File

@ -44,12 +44,7 @@ Each model file has a barebones trainer and inference routine.
My pre-trained weights can be acquired from [here](https://huggingface.co/ecker/vall-e).
For example:
```
git lfs clone --exclude "*.h5" https://huggingface.co/ecker/vall-e ./data/ # remove the '--exclude "*.h5"' if you wish to also download the libre dataset.
python -m vall_e "The birch canoe slid on the smooth planks." "./path/to/an/utterance.wav" --out-path="./output.wav" yaml="./data/config.yaml"
```
A script to setup a proper environment and download the weights can be invoked with `./scripts/setup.sh`
## Train
@ -60,7 +55,9 @@ Training is very dependent on:
### Pre-Processed Dataset
A "libre" dataset can be found [here](https://huggingface.co/ecker/vall-e/blob/main/data.h5). Simply place it in the same folder as your `config.yaml`, and ensure its `dataset.use_hdf5` is set to `True`.
A "libre" dataset can be found [here](https://huggingface.co/ecker/vall-e) under `data.tar.gz`.
A script to setup a proper environment and train can be invoked with `./scripts/setup-training.sh`
### Leverage Your Own Dataset

7
scripts/setup-training.sh Executable file
View File

@ -0,0 +1,7 @@
#!/bin/bash
`dirname $0`/setup.sh
wget -P ./training/valle/ "https://huggingface.co/ecker/vall-e/resolve/main/data.tar.gz"
wget -P ./training/valle/ "https://huggingface.co/ecker/vall-e/resolve/main/.cache.tar.gz"
tar -xzf ./training/valle/data.tar.gz -C "./training/valle/" data.h5
tar -xzf ./training/valle/.cache.tar.gz -C "./training/valle/"

5
scripts/setup.sh
View File

@ -5,9 +5,4 @@ pip3 install -e .
mkdir -p ./training/valle/ckpt/ar+nar-retnet-8/
wget -P ./training/valle/ckpt/ar+nar-retnet-8/ "https://huggingface.co/ecker/vall-e/resolve/main/ckpt/ar%2Bnar-retnet-8/fp32.pth"
wget -P ./training/valle/ "https://huggingface.co/ecker/vall-e/resolve/main/data.tar.gz"
wget -P ./training/valle/ "https://huggingface.co/ecker/vall-e/resolve/main/.cache.tar.gz"
wget -P ./training/valle/ "https://huggingface.co/ecker/vall-e/raw/main/config.yaml"
tar -xzf ./training/valle/data.tar.gz -C "./training/valle/" data.h5
tar -xzf ./training/valle/.cache.tar.gz -C "./training/valle/"