update mmmlu_lite dataload from oss

liushz · Nov 1, 2024 · 3930fc7 · 3930fc7
1 parent 47186b9
commit 3930fc7
Show file tree

Hide file tree

Showing 4 changed files with 19 additions and 12 deletions.
diff --git a/opencompass/configs/datasets/mmmlu_lite/README.md b/opencompass/configs/datasets/mmmlu_lite/README.md
@@ -31,11 +31,8 @@ MMMLU contains the MMLU test set translated into the following locales:
 
 
 ## How to Use
-Download file from [link](https://hf-mirror.com/datasets/openai/MMMLU)
 
 ```python
 from datasets import load_dataset
-ds = load_dataset("openai/MMMLU", "default")
-from datasets import load_dataset
-ds = load_dataset("openai/MMMLU", "by_language")
+ds = load_dataset("opencompass/mmmlu_lite", "AR_XY")
 ```
diff --git a/opencompass/configs/datasets/mmmlu_lite/mmmlu_lite_gen_c51a84.py b/opencompass/configs/datasets/mmmlu_lite/mmmlu_lite_gen_c51a84.py
@@ -95,8 +95,7 @@
         dict(
             abbr=f'openai_m{_name}',
             type=MMMLULiteDataset,
-            # path='opencompass/mmmlu_lite',
-            path = './data/mmmlu_lite',
+            path='opencompass/mmmlu_lite',
             name=f'openai_m{_name}',
             reader_cfg=mmmlu_lite_reader_cfg,
             infer_cfg=mmmlu_lite_infer_cfg,

diff --git a/opencompass/datasets/mmmlu.py b/opencompass/datasets/mmmlu.py
@@ -2,7 +2,7 @@
 # yapf: disable
 
 import json
-import os
+import os.path as osp
 
 from datasets import Dataset, DatasetDict, load_dataset
 
@@ -43,10 +43,12 @@ class MMMLULiteDataset(BaseDataset):
 
     @staticmethod
     def load(path: str, name: str):
+        path = get_data_path(path, local_mode=False)
         dataset = DatasetDict()
-        path = os.path.join(path, name + '.jsonl')
-        dataset_list = []
-        with open(path, 'r') as f:
-            dataset_list = [json.loads(line) for line in f.readlines()]
-        dataset['test'] = Dataset.from_list(dataset_list)
+        name = name.split('_')[-1]
+        raw_data = []
+        filename = osp.join(path, name, 'test.jsonl')
+        with open(filename, encoding='utf-8') as f:
+            raw_data = [json.loads(line) for line in f.readlines()]
+        dataset['test'] = Dataset.from_list(raw_data)
         return dataset
diff --git a/opencompass/utils/datasets_info.py b/opencompass/utils/datasets_info.py
@@ -301,6 +301,11 @@
         "ms_id": "",
         "hf_id": "",
         "local": "./data/cmo.jsonl",
+    },
+    "opencompass/mmmlu_lite": {
+        "ms_id": "",
+        "hf_id": "",
+        "local": "./data/mmmlu_lite",
     }
 }
 
@@ -309,6 +314,10 @@
         "url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/mmlu.zip",
         "md5": "761310671509a239e41c4b717f7fab9c",
     },
+    "/mmmlu_lite": {
+        "url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/mmmlu_lite.zip",
+        "md5": "a776af1220e1826fd0608eda1bc4425e",
+    },
     "/gpqa/": {
         "url": "http://opencompass.oss-cn-shanghai.aliyuncs.com/datasets/data/gpqa.zip",
         "md5": "2e9657959030a765916f1f2aca29140d",