enhance dataset loading script with multithreading and tqdm

olympiads-ref.py

@@ -18,8 +18,10 @@
 import re
 import json
 from pathlib import Path
+from concurrent.futures import ThreadPoolExecutor
 
 import datasets
+from tqdm import tqdm
 from huggingface_hub import HfApi
 
 
@@ -39,22 +41,22 @@ _LICENSE = ""
 
 
 class OlympiadReferenceDataset(datasets.GeneratorBasedBuilder):
-    VERSION = datasets.Version("0.0.1")
+    VERSION = datasets.Version("0.0.2")
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self._hfapi = HfApi()
-        self.pattern = re.compile(r'.*/segmented/[^/]+\.jsonl$')
+        self.pattern = re.compile(r'.*/segmented/.+\.jsonl$')
 
     def _info(self):
         features = datasets.Features(
-            {   
-                "problem_type": datasets.Value("string"),
-                "problem_label": datasets.Value("string"),
+            {
                 "problem": datasets.Value("string"),
                 "solution": datasets.Value("string"),
                 "year": datasets.Value("int32"),
                 "tier": datasets.Value("int32"),
+                "problem_type": datasets.Value("string"),
+                "problem_label": datasets.Value("string"),
                 "resource_path": datasets.Value("string")
             }
         )
@@ -73,7 +75,18 @@ class OlympiadReferenceDataset(datasets.GeneratorBasedBuilder):
         repo_files = self._hfapi.list_repo_files(repo_id="AI-MO/olympiads-ref", repo_type="dataset")
         seg_jsonl_files = [s for s in repo_files if self.pattern.match(s)]
 
-        data_files = [(sjf, dl_manager.extract(dl_manager.download(data_root_path / sjf))) for sjf in seg_jsonl_files]
+        data_files = []
+        with ThreadPoolExecutor(max_workers=12) as executor:
+            futures = [
+                executor.submit(
+                    lambda sjf: (sjf, dl_manager.download_and_extract(data_root_path / sjf)),
+                    sjf
+                )
+                for sjf in seg_jsonl_files
+            ]
+
+            for future in tqdm(futures, desc="Downloading data files"):
+                data_files.append(future.result())
 
         return [
             datasets.SplitGenerator(