deepgeek/CosyVoice
2
0
Fork 0
mirror of https://github.com/FunAudioLLM/CosyVoice.git synced 2026-02-05 18:09:24 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix white space

Browse Source
This commit is contained in:
root
2025-10-09 14:49:22 +08:00
parent 807bb6ee0b
commit 8811e9f33a
7 changed files with 27 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/grpo/cosyvoice2/README.md
View File

@@ -36,7 +36,7 @@ Stage `0` converts raw JSONL files into the parquet format expected by veRL:
```bash
bash run.sh 0 0
```
Create two JSONL files—`train.jsonl` and `test.jsonl`.
Create two JSONL files—`train.jsonl` and `test.jsonl`.
The script will then generate two Parquet files:
```
@@ -111,7 +111,7 @@ bash run.sh 5 5
The script converts the Hugging Face checkpoint back into the format expected by the CosyVoice repository.
> [!TIP]
> However, we observed a slight accuracy drop when using the RL-trained model after conversion, compared with the Hugging Face format.
> However, we observed a slight accuracy drop when using the RL-trained model after conversion, compared with the Hugging Face format.
## Results