more read_evals.py improvement

src/leaderboard/filter_models.py

@@ -141,7 +141,7 @@ def flag_models(leaderboard_data: list[dict]):
         else:
             flag_key = model_data[AutoEvalColumn.fullname.name]
         if flag_key in FLAGGED_MODELS:
-            logging.info(f"Flagged model: {flag_key}")
+            # logging.info(f"Flagged model: {flag_key}") # Do we need to print out the list of flagged models? 
             issue_num = FLAGGED_MODELS[flag_key].split("/")[-1]
             issue_link = model_hyperlink(
                 FLAGGED_MODELS[flag_key],

src/leaderboard/read_evals.py

@@ -4,11 +4,13 @@ from datetime import datetime
 from json import JSONDecodeError
 import logging
 import math
-import os
+
 from dataclasses import dataclass, field
 from typing import Optional, Dict, List
 
-import dateutil
+from tqdm import tqdm
+from tqdm.contrib.logging import logging_redirect_tqdm
+
 import numpy as np
 
 from src.display.formatting import make_clickable_model
@@ -213,15 +215,16 @@ def get_raw_eval_results(results_path: str, requests_path: str, dynamic_path: st
         dynamic_data = json.load(f)
     
     results_path = Path(results_path)
-
     model_files = list(results_path.rglob('results_*.json'))
     model_files.sort(key=lambda file: parse_datetime(file.stem.removeprefix("results_")))
 
     eval_results = {}
-    for model_result_filepath in model_files:
+    # Wrap model_files iteration with tqdm for progress display
+    for model_result_filepath in tqdm(model_files, desc="Processing model files"):
         # Creation of result
         eval_result = EvalResult.init_from_json_file(model_result_filepath)
-        eval_result.update_with_request_file(requests_path)
+        with logging_redirect_tqdm():
+            eval_result.update_with_request_file(requests_path)
 
         if eval_result.full_model in dynamic_data:
             eval_result.update_with_dynamic_file_dict(dynamic_data[eval_result.full_model])