Spaces:
Runtime error
Runtime error
# Pretrained Models Dependency | |
The models dependency of Amphion are as follows (sort alphabetically): | |
- [Pretrained Models Dependency](#pretrained-models-dependency) | |
- [Amphion Singing BigVGAN](#amphion-singing-bigvgan) | |
- [Amphion Speech HiFi-GAN](#amphion-speech-hifi-gan) | |
- [ContentVec](#contentvec) | |
- [WeNet](#wenet) | |
- [Whisper](#whisper) | |
- [RawNet3](#rawnet3) | |
The instructions about how to download them is displayed as follows. | |
## Amphion Singing BigVGAN | |
We fine-tune the official BigVGAN pretrained model with over 120 hours singing voice data. The fine-tuned checkpoint can be downloaded [here](https://huggingface.co/amphion/BigVGAN_singing_bigdata). You need to download the `400000.pt` and `args.json` files into `Amphion/pretrained/bigvgan`: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ bivgan | |
β β β£ 400000.pt | |
β β β£ args.json | |
``` | |
## Amphion Speech HiFi-GAN | |
We trained our HiFi-GAN pretrained model with 685 hours speech data. Which can be downloaded [here](https://huggingface.co/amphion/hifigan_speech_bigdata). You need to download the whole folder of `hifigan_speech` into `Amphion/pretrained/hifigan`. | |
``` | |
Amphion | |
β£ pretrained | |
β β£ hifigan | |
β β β£ hifigan_speech | |
β β β β£ log | |
β β β β£ result | |
β β β β£ checkpoint | |
β β β β£ args.json | |
``` | |
## Amphion DiffWave | |
We trained our DiffWave pretrained model with 125 hours speech data and around 80 hours of singing voice data. Which can be downloaded [here](https://huggingface.co/amphion/diffwave). You need to download the whole folder of `diffwave` into `Amphion/pretrained/diffwave`. | |
``` | |
Amphion | |
β£ pretrained | |
β β£ diffwave | |
β β β£ diffwave_speech | |
β β β β£ samples | |
β β β β£ checkpoint | |
β β β β£ args.json | |
``` | |
## ContentVec | |
You can download the pretrained ContentVec model [here](https://github.com/auspicious3000/contentvec). Note that we use the `ContentVec_legacy-500 classes` checkpoint. Assume that you download the `checkpoint_best_legacy_500.pt` into the `Amphion/pretrained/contentvec`. | |
``` | |
Amphion | |
β£ pretrained | |
β β£ contentvec | |
β β β£ checkpoint_best_legacy_500.pt | |
``` | |
## WeNet | |
You can download the pretrained WeNet model [here](https://github.com/wenet-e2e/wenet/blob/main/docs/pretrained_models.md). Take the `wenetspeech` pretrained checkpoint as an example, assume you download the `wenetspeech_u2pp_conformer_exp.tar` into the `Amphion/pretrained/wenet`. Unzip it and modify its configuration file as follows: | |
```sh | |
cd Amphion/pretrained/wenet | |
### Unzip the expt dir | |
tar -xvf wenetspeech_u2pp_conformer_exp.tar.gz | |
### Specify the updated path in train.yaml | |
cd 20220506_u2pp_conformer_exp | |
vim train.yaml | |
# TODO: Change the value of "cmvn_file" (Line 2) to the absolute path of the `global_cmvn` file. (Eg: [YourPath]/Amphion/pretrained/wenet/20220506_u2pp_conformer_exp/global_cmvn) | |
``` | |
The final file struture tree is like: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ wenet | |
β β β£ 20220506_u2pp_conformer_exp | |
β β β β£ final.pt | |
β β β β£ global_cmvn | |
β β β β£ train.yaml | |
β β β β£ units.txt | |
``` | |
## Whisper | |
The official pretrained whisper checkpoints can be available [here](https://github.com/openai/whisper/blob/e58f28804528831904c3b6f2c0e473f346223433/whisper/__init__.py#L17). In Amphion, we use the `medium` whisper model by default. You can download it as follows: | |
```bash | |
cd Amphion/pretrained | |
mkdir whisper | |
cd whisper | |
wget https://openaipublic.azureedge.net/main/whisper/models/345ae4da62f9b3d59415adc60127b97c714f32e89e936602e85993674d08dcb1/medium.pt | |
``` | |
The final file structure tree is like: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ whisper | |
β β β£ medium.pt | |
``` | |
## RawNet3 | |
The official pretrained RawNet3 checkpoints can be available [here](https://huggingface.co/jungjee/RawNet3). You need to download the `model.pt` file and put it in the folder. | |
The final file structure tree is like: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ rawnet3 | |
β β β£ model.pt | |
``` | |
# (Optional) Model Dependencies for Evaluation | |
When utilizing Amphion's Evaluation Pipelines, terminals without access to `huggingface.co` may encounter error messages such as "OSError: Can't load tokenizer for ...". To work around this, the dependant models for evaluation can be pre-prepared and stored here, at `Amphion/pretrained`, and follow [this README](../egs/metrics/README.md#troubleshooting) to configure your environment to load local models. | |
The dependant models of Amphion's evaluation pipeline are as follows (sort alphabetically): | |
- [Evaluation Pipeline Models Dependency](#optional-model-dependencies-for-evaluation) | |
- [bert-base-uncased](#bert-base-uncased) | |
- [facebook/bart-base](#facebookbart-base) | |
- [roberta-base](#roberta-base) | |
- [wavlm](#wavlm) | |
The instructions about how to download them is displayed as follows. | |
## bert-base-uncased | |
To load `bert-base-uncased` locally, follow [this link](https://huggingface.co/bert-base-uncased) to download all files for `bert-base-uncased` model, and store them under `Amphion/pretrained/bert-base-uncased`, conforming to the following file structure tree: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ bert-base-uncased | |
β β β£ config.json | |
β β β£ coreml | |
β β β β£ fill-mask | |
β β β β£ float32_model.mlpackage | |
β β β β£ Data | |
β β β β£ com.apple.CoreML | |
β β β β£ model.mlmodel | |
β β β£ flax_model.msgpack | |
β β β£ LICENSE | |
β β β£ model.onnx | |
β β β£ model.safetensors | |
β β β£ pytorch_model.bin | |
β β β£ README.md | |
β β β£ rust_model.ot | |
β β β£ tf_model.h5 | |
β β β£ tokenizer_config.json | |
β β β£ tokenizer.json | |
β β β£ vocab.txt | |
``` | |
## facebook/bart-base | |
To load `facebook/bart-base` locally, follow [this link](https://huggingface.co/facebook/bart-base) to download all files for `facebook/bart-base` model, and store them under `Amphion/pretrained/facebook/bart-base`, conforming to the following file structure tree: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ facebook | |
β β β£ bart-base | |
β β β β£ config.json | |
β β β β£ flax_model.msgpack | |
β β β β£ gitattributes.txt | |
β β β β£ merges.txt | |
β β β β£ model.safetensors | |
β β β β£ pytorch_model.bin | |
β β β β£ README.txt | |
β β β β£ rust_model.ot | |
β β β β£ tf_model.h5 | |
β β β β£ tokenizer.json | |
β β β β£ vocab.json | |
``` | |
## roberta-base | |
To load `roberta-base` locally, follow [this link](https://huggingface.co/roberta-base) to download all files for `roberta-base` model, and store them under `Amphion/pretrained/roberta-base`, conforming to the following file structure tree: | |
``` | |
Amphion | |
β£ pretrained | |
β β£ roberta-base | |
β β β£ config.json | |
β β β£ dict.txt | |
β β β£ flax_model.msgpack | |
β β β£ gitattributes.txt | |
β β β£ merges.txt | |
β β β£ model.safetensors | |
β β β£ pytorch_model.bin | |
β β β£ README.txt | |
β β β£ rust_model.ot | |
β β β£ tf_model.h5 | |
β β β£ tokenizer.json | |
β β β£ vocab.json | |
``` | |
## wavlm | |
The official pretrained wavlm checkpoints can be available [here](https://huggingface.co/microsoft/wavlm-base-plus-sv). The file structure tree is as follows: | |
``` | |
Amphion | |
β£ wavlm | |
β β£ config.json | |
β β£ preprocessor_config.json | |
β β£ pytorch_model.bin | |
``` |