ver 1.2

app.py

@@ -7,6 +7,7 @@ import torchaudio
 import numpy as np
 import gradio as gr
 from uroman import uroman
+import concurrent.futures
 from pydub import AudioSegment
 from datasets import load_dataset
 from IPython.display import Audio
@@ -176,11 +177,21 @@ def generate_audio_with_pause(srt_file_path):
     
     return final_audio
 
-def srt_to_audio(srt_file):
-    audio_data = generate_audio_with_pause(srt_file.name)
-    output_path = os.path.join(cache_dir, 'output.wav')
-    torchaudio.save(output_path, torch.tensor(audio_data).unsqueeze(0), 16000)
-    return output_path
+def srt_to_audio_multi(srt_files):
+    output_paths = []
+    
+    def process_file(srt_file):
+        audio_data = generate_audio_with_pause(srt_file.name)
+        output_path = os.path.join(cache_dir, f'output_{os.path.basename(srt_file.name)}.wav')
+        torchaudio.save(output_path, torch.tensor(audio_data).unsqueeze(0), 16000)
+        return output_path
+    
+    with concurrent.futures.ThreadPoolExecutor() as executor:
+        futures = [executor.submit(process_file, srt_file) for srt_file in srt_files]
+        for future in concurrent.futures.as_completed(futures):
+            output_paths.append(future.result())
+    
+    return output_paths
 
 # UI display
 css = '''
@@ -191,10 +202,10 @@ with gr.Blocks(css=css) as demo:
         """<h1>SRT to Audio Tool</h1>""",
         elem_id="title",
     )
-    inp = gr.File(label="Upload SRT file", file_count="single", type="file")
-    out = gr.Audio(label="Generated Audio", type="filepath")
-    
-    inp.change(fn=srt_to_audio, inputs=inp, outputs=out)
+    inp = gr.File(label="Upload SRT files", file_count="multiple", type="filepath")
+    out = gr.File(label="Generated Audio Files", file_count="multiple", type="filepath")
+
+    inp.change(fn=srt_to_audio_multi, inputs=inp, outputs=out)
 
 if __name__ == "__main__":
     demo.launch()