Skip to content

Commit 66cf6d5

Browse files
committed
docs: update for samples
1 parent ec9b3d7 commit 66cf6d5

File tree

1 file changed

+16
-1
lines changed

1 file changed

+16
-1
lines changed

README.md

+16-1
Original file line numberDiff line numberDiff line change
@@ -851,7 +851,7 @@ if __name__ == "__main__":
851851
852852
![Finetune Model Choice](images/finetune-model-choice.jpg)
853853
854-
#### 数据集信息
854+
#### 数据集示例 1
855855
856856
由 Unit Eval + OSS Instruct 数据集构建而来:
857857
@@ -863,6 +863,21 @@ if __name__ == "__main__":
863863
864864
测试视频:[开源 AI 辅助编程方案:Unit Mesh 端到端打通 v0.0.1 版本](https://www.bilibili.com/video/BV1si4y1h7Vw/)
865865
866+
在 Unit Eval [0.3.2](https://github.com/unit-mesh/unit-eval/releases/tag/v0.3.2) 版本里
867+
868+
组成如下:
869+
870+
```python
871+
# Merge and shuffle records from different files
872+
merge_jsonl(
873+
output_file=merged_file,
874+
input_files=[oss_instruction, 'code_bugfix_cleaned_5K.json', 'codeGPT_CN_cleaned_20K.json',
875+
'code_summarization_CN_cleaned_10K.json', 'java-code-completion.jsonl', 'java-test-gen.jsonl',
876+
'kotlin-completion-11929.jsonl', 'kotlin-java-comments-3715.jsonl'],
877+
lines_per_file=[4000, 4000, 15000, 8000, 5000, 3000, 3000, 2000 ]
878+
)
879+
```
880+
866881
#### 参数示例:
867882
868883
```bash

0 commit comments

Comments
 (0)