upload

README.md

@@ -58,6 +58,9 @@ pip install -e .
 ## USLM Models
 This version of USLM is trained on the LibriTTS dataset, so the performance is not optimal due to data limitations.
 
+| Model| Dataset |Discription|
+|:----|:----:|:----|
+|[USLM_libri](https://huggingface.co/fnlp/USLM/resolve/main/USLM_libritts/)|LibriTTS|USLM trained on LibriTTS dataset |
 
 
 ## Zero-shot TTS Using USLM
@@ -76,8 +79,8 @@ Download pre-trained USLM models:
 uslm_dir="ckpt/uslm/"
 mkdir -p ${uslm_dir}
 cd ${uslm_dir}
-wget "https://huggingface.co/fnlp/USLM/resolve/main/USLM_ls960/USLM.pt"
-wget "https://huggingface.co/fnlp/USLM/resolve/main/USLM_ls960/unique_text_tokens.k2symbols" 
+wget "https://huggingface.co/fnlp/USLM/resolve/main/USLM_libritts/USLM.pt"
+wget "https://huggingface.co/fnlp/USLM/resolve/main/USLM_libritts/unique_text_tokens.k2symbols" 
 cd -
 ```
 
@@ -101,4 +104,17 @@ python3 bin/infer.py --output-dir ${out_dir}/ \
 or you can directly run inference.sh
 ``` bash
 bash inference.sh
+```
+
+## Citation
+If you use this code or result in your paper, please cite our work as:
+```Tex
+@misc{zhang2023speechtokenizer,
+      title={SpeechTokenizer: Unified Speech Tokenizer for Speech Language Models}, 
+      author={Xin Zhang and Dong Zhang and Shimin Li and Yaqian Zhou and Xipeng Qiu},
+      year={2023},
+      eprint={2308.16692},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
+}
 ```

images/README.md

@@ -0,0 +1 @@
+