add dataset

hf_cmmu_dataset.py

@@ -1,46 +0,0 @@
-from datasets import DatasetBuilder, Features, Value, Sequence, DatasetInfo
-import json
-import os.path as osp
-
-class HfCmmuDataset(DatasetBuilder):
-    def _info(self):
-        return DatasetInfo(
-            features=Features({
-                "type": Value("string"),
-                "class": Value("string"),
-                "grade_band": Value("string"),
-                "difficulty": Value("string"),
-                "answer": Value("string"),
-                "question_info": Value("string"),
-                "solution_info": Value("string"),
-                "options": Sequence(Value("string")),
-                "sub_questions": Sequence(Value("string")),
-                "id": Value("string"),
-                "images": Sequence(Value("string")),
-                "split": Value("string")
-            }),
-        )
-
-    def _split_generators(self, dl_manager):
-        # Define the splits here if needed
-        return []
-
-    def _generate_examples(self):
-        # Here, you'll need to write code to read and yield each example
-        # from your jsonl files.
-        jsonl_files = [
-            "biology.jsonl",
-            "chemistry.jsonl",
-            "geography.jsonl",
-            "history.jsonl",
-            "math.jsonl",
-            "physics.jsonl",
-            "politics.jsonl"
-        ]
-        for file_path in jsonl_files:  # Add all your jsonl files here
-            with open(osp.join('val', file_path), "r", encoding="utf-8") as f:
-                for idx, line in enumerate(f):
-                    record = json.loads(line)
-                    yield idx, record
-
-dataset = HfCmmuDataset()