czlll commited on
Commit
94a3106
·
verified ·
1 Parent(s): ea98bb2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -7
README.md CHANGED
@@ -49,11 +49,12 @@ dataset_info:
49
  LOC-BENCH is a dataset specifically designed for evaluating code localization methods in software repositories.
50
  LOC-BENCH provides a diverse set of issues, including bug reports, feature requests, security vulnerabilities, and performance optimizations.
51
 
52
- To reproduce the results reported in the [paper](https://arxiv.org/abs/2503.09089), please refer to [`czlll/Loc-Bench_V0`](https://huggingface.co/datasets/czlll/Loc-Bench_V0).
53
 
54
- Loc-Bench is a refined benchmark for evaluating code localization methods.
55
- Compared to the V0, it improves data quality by filtering out examples that do not involve any function-level code modifications.
56
- We recommend using this dataset for a more accurate and reliable evaluation of code localization performance.
 
57
 
58
  The table below shows the distribution of categories in the dataset.
59
 
@@ -65,14 +66,12 @@ The table below shows the distribution of categories in the dataset.
65
  | Security Vulnerability | 29 |
66
 
67
 
68
- Code: https://github.com/gersteinlab/LocAgent
69
-
70
  ## 🔧 How to Use
71
  You can easily load LOC-BENCH using Hugging Face's datasets library:
72
  ```
73
  from datasets import load_dataset
74
 
75
- dataset = load_dataset("czlll/Loc-Bench_V2", split="test")
76
 
77
  ```
78
  ## 📄 Citation
 
49
  LOC-BENCH is a dataset specifically designed for evaluating code localization methods in software repositories.
50
  LOC-BENCH provides a diverse set of issues, including bug reports, feature requests, security vulnerabilities, and performance optimizations.
51
 
52
+ Please refer to [**`Loc-Bench_V1`**](https://huggingface.co/datasets/czlll/Loc-Bench_V1) for evaluating code localization methods and for easy comparison with our approach.
53
 
54
+ Code: https://github.com/gersteinlab/LocAgent
55
+
56
+ ## 📊 Details
57
+ Compared to the [V0](https://huggingface.co/datasets/czlll/Loc-Bench_V0.1), it filters out examples that do not involve function-level code modifications and then supplements the dataset to restore it to the original size of 660 examples.
58
 
59
  The table below shows the distribution of categories in the dataset.
60
 
 
66
  | Security Vulnerability | 29 |
67
 
68
 
 
 
69
  ## 🔧 How to Use
70
  You can easily load LOC-BENCH using Hugging Face's datasets library:
71
  ```
72
  from datasets import load_dataset
73
 
74
+ dataset = load_dataset("czlll/Loc-Bench_V0.2", split="test")
75
 
76
  ```
77
  ## 📄 Citation