sm1rk commited on
Commit
73538dd
·
verified ·
1 Parent(s): 32c587d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -17
README.md CHANGED
@@ -4,6 +4,7 @@ size_categories:
4
  - n<1K
5
  task_categories:
6
  - summarization
 
7
  pretty_name: MCE
8
  dataset_info:
9
  features:
@@ -38,34 +39,23 @@ tags:
38
 
39
  # Modified ClassEval (MCE) Dataset
40
 
41
- This dataset is a modification of the ClassEval benchmark, designed for evaluating code summarization models beyond the function level. It explores the impact of class and repository contexts on summary quality. The dataset includes modifications for evaluating at both class and repository levels.
42
 
43
- [Paper: Code Summarization Beyond Function Level](https://hf.co/papers/2502.16704)
44
 
45
  **Dataset Structure:**
46
 
47
- The dataset contains samples with the following features:
48
 
49
  * `class_id`: Identifier for the class.
50
  * `class_code`: The code of the class.
51
- * `skeleton`: (Likely refers to class structure, needs clarification in the original dataset)
52
  * `method_code`: The code of a method within the class.
53
  * `method_summary`: The summary of the method.
54
 
55
 
56
  **Dataset Splits:**
57
 
58
- * `test`: 410 examples.
59
 
60
- **Tags:**
61
-
62
- * `code summarization`
63
- * `python`
64
- * `function-level`
65
- * `class-level`
66
- * `ClassEval`
67
-
68
-
69
- This dataset is part of a broader study investigating code summarization at various levels (function, class, and repository). The complete study details, code, and evaluation results are available on GitHub: [https://github.com/kilimanj4r0/code-summarization-beyond-function-level](https://github.com/kilimanj4r0/code-summarization-beyond-function-level).
70
-
71
- **Note:** The repository also includes a modified CodeSearchNet dataset for repository-level summarization experiments.
 
4
  - n<1K
5
  task_categories:
6
  - summarization
7
+ - text-generation
8
  pretty_name: MCE
9
  dataset_info:
10
  features:
 
39
 
40
  # Modified ClassEval (MCE) Dataset
41
 
42
+ This dataset is a modification of the ClassEval benchmark, designed for evaluating code summarization models beyond the function level. It explores the impact of function and class contexts on summary quality. The dataset includes modifications for evaluating at both function and class levels.
43
 
44
+ Paper: [Code Summarization Beyond Function Level](https://hf.co/papers/2502.16704)
45
 
46
  **Dataset Structure:**
47
 
48
+ The dataset contains samples with the following fields:
49
 
50
  * `class_id`: Identifier for the class.
51
  * `class_code`: The code of the class.
52
+ * `skeleton`: The structured blueprint for the class, including class-level and function-level information.
53
  * `method_code`: The code of a method within the class.
54
  * `method_summary`: The summary of the method.
55
 
56
 
57
  **Dataset Splits:**
58
 
59
+ * `test`: 410 samples (400 + 10 samples for few-shot learing)
60
 
61
+ This dataset is part of a broader study investigating code summarization at various levels (function, class, and repository). The complete study details, code, and evaluation results are available on GitHub: [https://github.com/kilimanj4r0/code-summarization-beyond-function-level](https://github.com/kilimanj4r0/code-summarization-beyond-function-level).