Datasets:

Modalities:
Text
Formats:
csv
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
mshihabr commited on
Commit
0f56a5d
·
verified ·
1 Parent(s): d799d49

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +69 -3
README.md CHANGED
@@ -1,3 +1,69 @@
1
- ---
2
- license: cc-by-nc-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-4.0
3
+ tags:
4
+ - coding
5
+ - agents
6
+ ---
7
+ # SWE-PolyBench
8
+
9
+ [SWE-PolyBench](https://github.com/amazon-science/SWE-PolyBench) is a multi language repo level software engineering benchmark. Currently it includes 4 languages: Python, Java, Javascript, and Typescript. The number of instances in each language is:
10
+
11
+ Javascript: 1017
12
+
13
+ Typescript: 729
14
+
15
+ Python: 199
16
+
17
+ Java: 165
18
+
19
+ ## Datasets
20
+
21
+ There are total *two* datasets available under SWE-PolyBench. `AmazonScience/SWE-PolyBench` is the full dataset and `AmazonScience/SWE-PolyBench_500` is the stratified sampled dataset with 500 instances.
22
+
23
+ ## Leaderboard
24
+
25
+ We evaluated several open source coding agents/models on this dataset and report them in our [leaderboard]().
26
+
27
+ ## Submit
28
+
29
+ To submit your predictions on this dataset, please follow this [README](https://github.com/amazon-science/SWE-PolyBench/blob/submission/README.md)
30
+
31
+ ## Languages
32
+
33
+ The text of the dataset is primarily English.
34
+
35
+ ## Dataset Structure
36
+
37
+ An example row from the dataset includes the following columns:
38
+ ```
39
+ instance_id: (str) - A formatted instance identifier, usually as repo_owner__repo_name-PR-number.
40
+ patch: (str) - The gold patch, the patch generated by the PR (minus test-related code), that resolved the issue.
41
+ repo: (str) - The repository owner/name identifier from GitHub.
42
+ base_commit: (str) - The commit hash of the repository representing the HEAD of the repository before the solution PR is applied.
43
+ hints_text: (str) - Comments made on the issue prior to the creation of the solution PR’s first commit creation date.
44
+ created_at: (str) - The creation date of the pull request.
45
+ test_patch: (str) - A test-file patch that was contributed by the solution PR.
46
+ problem_statement: (str) - The issue title and body.
47
+ F2P: (str) - A json list of strings that represent the set of tests resolved by the PR and tied to the issue resolution.
48
+ P2P: (str) - A json list of strings that represent tests that should pass before and after the PR application.
49
+ language: (str) - The programming language
50
+ Dockerfile: (str) - The instance level dockerfile
51
+ test_command: (str) - The test command used to get F2P and P2P
52
+ task_category: (str) - The problem classification (Bug Fix, Refactoring, Feature)
53
+ is_no_nodes: (bool) - Helpful info for evaluating retrieval metrics
54
+ is_func_only: (bool) - Helpful info for evaluating retrieval metrics
55
+ is_class_only: (bool) - Helpful info for evaluating retrieval metrics
56
+ is_mixed: (bool) - Helpful info for evaluating retrieval metrics
57
+ num_func_changes: (int) - Helpful info for evaluating retrieval metrics
58
+ num_class_changes: (int) - Helpful info for evaluating retrieval metrics
59
+ num_nodes: (int) - Helpful info for evaluating retrieval metrics
60
+ is_single_func: (bool) - Helpful info for evaluating retrieval metrics
61
+ is_single_class: (bool) - Helpful info for evaluating retrieval metrics
62
+ modified_nodes: (bool) - Helpful info for evaluating retrieval metrics
63
+ ```
64
+
65
+ ## Citation
66
+
67
+ If you find our work useful please cite:
68
+ ```
69
+ ```