Update app.py

app.py

@@ -3,25 +3,29 @@ import os
 import json
 import pandas as pd
 import random
+import shutil
 from datasets import load_dataset
 from huggingface_hub import HfApi
 
 dataset = load_dataset("intersteller2887/Turing-test-dataset", split="train")
 
-print(os.getcwd())
-print(dataset[0])
+target_audio_dir = "/home/user/app/audio"
+os.makedirs(target_audio_dir, exist_ok=True)
 
-all_data_audio_paths = [
-    item["audio"]["path"] 
-    for item in dataset 
-    if item["audio"] and "path" in item["audio"]
-]
+local_audio_paths = []
 
-"""valid_audio_paths = [path for path in all_data_audio_paths if os.path.exists(path)]
-print(f"Total valid audio files: {len(valid_audio_paths)}")"""
+for item in dataset:
+    src_path = item["audio"]["path"]
+    if src_path and os.path.exists(src_path):
+        filename = os.path.basename(src_path)
+        dst_path = os.path.join(target_audio_dir, filename)
+        if not os.path.exists(dst_path):
+            shutil.copy(src_path, dst_path)
+        local_audio_paths.append(dst_path)
 
+all_data_audio_paths = local_audio_paths
 
-sample1_audio_path = next((p for p in all_data_audio_paths if p.endswith("/home/user/.cache/huggingface/hub/datasets--intersteller2887--Turing-test-dataset/snapshots/5cd5dd6c5badbbce22a66b3b49c4224751b10375/data/bf_train_L2_T01_1_glm.wav")), None)
+sample1_audio_path = next((p for p in all_data_audio_paths if p.endswith("sample1.wav")), None)
 print(sample1_audio_path)
 
 # ==============================================================================
@@ -491,8 +495,8 @@ with gr.Blocks(theme=gr.themes.Soft(), css=".gradio-container {max-width: 960px
 # 程序入口 (Entry Point)
 # ==============================================================================
 if __name__ == "__main__":
-    if not os.path.exists("audio"):
-        os.makedirs("audio")
+    """if not os.path.exists("audio"):
+        os.makedirs("audio")"""
     # A quick check to see if we're in a deployed Space, to avoid local errors.
     if "SPACE_ID" in os.environ:
         print("Running in a Hugging Face Space, checking for audio files...")