update readme

README.md

@@ -18,116 +18,84 @@ task_ids:
 
 # Dataset Card for RepoBench-R
 
-## Table of Contents
-- [Dataset Card for RepoBench-R](#dataset-card-for-repobench-r)
-  - [Table of Contents](#table-of-contents)
-  - [Dataset Description](#dataset-description)
-    - [Dataset Summary](#dataset-summary)
-    - [Supported Tasks](#supported-tasks)
-  - [Dataset Structure](#dataset-structure)
-  - [Dataset Creation](#dataset-creation)
-    - [Curation Rationale](#curation-rationale)
-    - [Source Data](#source-data)
-      - [Initial Data Collection and Normalization](#initial-data-collection-and-normalization)
-      - [Who are the source language producers?](#who-are-the-source-language-producers)
-    - [Annotations](#annotations)
-      - [Annotation process](#annotation-process)
-      - [Who are the annotators?](#who-are-the-annotators)
-    - [Personal and Sensitive Information](#personal-and-sensitive-information)
-  - [Considerations for Using the Data](#considerations-for-using-the-data)
-    - [Social Impact of Dataset](#social-impact-of-dataset)
-    - [Discussion of Biases](#discussion-of-biases)
-    - [Other Known Limitations](#other-known-limitations)
-  - [Additional Information](#additional-information)
-    - [Licensing Information](#licensing-information)
-    - [Citation Information](#citation-information)
-    - [Contributions](#contributions)
-
 ## Dataset Description
 
 - **Homepage:** https://github.com/Leolty/repobench
 - **Paper:** https://arxiv.org/abs/2306.03091
 
-### Dataset Summary
+## Dataset Summary
 
-RepoBench-R is a subtask of [RepoBench](https://github.com/Leolty/repobench), targeting the retrieval component of a repository-level auto-completion
-system, focusing on extracting the most relevant code snippet from a project repository for next-line
+RepoBench-R is a subtask of [RepoBench](https://github.com/Leolty/repobench), targeting the retrieval component of a repository-level auto-completion system, focusing on retrieving the most relevant code snippet from a project repository for next-line
 code prediction. 
 
-### Supported Tasks 
-
-The dataset supports two programming languages, Python and Java, and contains two settings:
-
-- `cff`: short for `cross_file_first`, where the cross-file module in next line is first used in the current file.
-- `cfr`: short for `cross_file_random`, where the cross-file module in next line is not first used in the current file.
-
-For each setting, we provide `train` and `test` subset, and each subset with two levels of difficulty: `easy` and `hard`. 
-
-Suppose the number of code snippets in the context is \\(k\\), 
-
-- For the `easy` subset, we have \\(5 \leq k < 10\\).
-- For the `hard` subset, we have \\(k \geq 10\\).
-
-
-## Dataset Structure
-
-
-
-## Dataset Creation
-
-### Curation Rationale
-
-[More Information Needed]
-
-### Source Data
+## Settings
 
-#### Initial Data Collection and Normalization
+- `cff`: short for cross_file_first, indicating the cross-file module in next line is first used in the current file.
 
-[More Information Needed]
+- `cfr`: short for cross_file_random, indicating the cross-file module in next line is not first used in the current file.
 
-#### Who are the source language producers?
+## Supported Tasks 
 
-[More Information Needed]
+The dataset has 4 subsets:
 
-### Annotations
+- `python-cff`: python dataset with `cff` setting.
+- `python-cfr`: python dataset with `cfr` setting.
+- `java-cff`: java dataset with `cff` setting.
+- `java-cfr`: java dataset with `cfr` setting.
 
-#### Annotation process
+Each subset has 4 splits:
 
-[More Information Needed]
+- `train-easy`: training set with easy difficulty, where the number of code snippets in the context $$k$$ satisfies $$ 5 \leq k < 10 $$.
+- `train-hard`: training set with hard difficulty, where the number of code snippets in the context $$k$$ satisfies $$ k \geq 10 $$.
+- `test-easy`: testing set with easy difficulty.
+- `test-hard`: testing set with hard difficulty.
 
-#### Who are the annotators?
+## Loading Data
 
-[More Information Needed]
+For example, if you want to load the `test` `cross-file-first` `python` dataset with `easy` difficulty, you can use the following code:
 
-### Personal and Sensitive Information
+```python
+from datasets import load_dataset
 
-[More Information Needed]
+dataset = load_dataset("tianyang/repobench-r", "python-cff", "test-easy")
+```
 
-## Considerations for Using the Data
-
-### Social Impact of Dataset
-
-[More Information Needed]
-
-### Discussion of Biases
-
-[More Information Needed]
-
-### Other Known Limitations
-
-[More Information Needed]
-
-## Additional Information
-
-
-### Licensing Information
-
-[More Information Needed]
-
-### Citation Information
-
-[More Information Needed]
-
-### Contributions
+## Dataset Structure
 
-Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.
+```
+{
+  "repo_name": "repository name of the data point",
+  "file_path": "path/to/file",
+  "context": [
+      "snippet 1",
+      "snippet 2",
+      // ...
+      "snippet k"
+  ],
+  "import_statement": "all import statements in the file",
+  "gold_snippet_idex": 2, // the index of the gold snippet in the context list, 0~k-1
+  "code": "the code for next-line prediction",
+  "next_line": "the next line of the code"
+}
+```
+
+## Licensing Information
+
+CC BY-NC-ND 4.0
+
+## Citation Information
+
+```bibtex
+@misc{liu2023repobench,
+      title={RepoBench: Benchmarking Repository-Level Code Auto-Completion Systems}, 
+      author={Tianyang Liu and Canwen Xu and Julian McAuley},
+      year={2023},
+      eprint={2306.03091},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
+}
+```
+
+## Contributions
+
+Thanks to [@Leolty](https://github.com/Leolty) for adding this dataset.

repobench-r.py

@@ -117,20 +117,20 @@ class RepoBenchR(datasets.GeneratorBasedBuilder):
 
         return [
             datasets.SplitGenerator(
-                name=datasets.Split("train_easy"),
-                gen_kwargs={"data_dir": data_dir, "split": "train_easy"},
+                name=datasets.Split("train-easy"),
+                gen_kwargs={"data_dir": data_dir, "split": "train-easy"},
             ),
             datasets.SplitGenerator(
                 name=datasets.Split("train_hard"),
-                gen_kwargs={"data_dir": data_dir, "split": "train_hard"},
+                gen_kwargs={"data_dir": data_dir, "split": "train-hard"},
             ),
             datasets.SplitGenerator(
-                name=datasets.Split("test_easy"),
-                gen_kwargs={"data_dir": data_dir, "split": "test_easy"},
+                name=datasets.Split("test-easy"),
+                gen_kwargs={"data_dir": data_dir, "split": "test-easy"},
             ),
             datasets.SplitGenerator(
-                name=datasets.Split("test_hard"),
-                gen_kwargs={"data_dir": data_dir, "split": "test_hard"},
+                name=datasets.Split("test-hard"),
+                gen_kwargs={"data_dir": data_dir, "split": "test-hard"},
             )
         ]
 
@@ -139,7 +139,7 @@ class RepoBenchR(datasets.GeneratorBasedBuilder):
         with gzip.open(data_dir, "rb") as f:
             data = pickle.load(f)
         
-        subset, level = split.split("_")
+        subset, level = split.split("-")
 
         for i, example in enumerate(data[subset][level]):
             yield i, {