init

app.py

@@ -14,6 +14,8 @@ from transformers.pipelines.audio_utils import ffmpeg_read
 MODEL_NAME = "kotoba-tech/kotoba-whisper-v2.0"
 BATCH_SIZE = 16
 CHUNK_LENGTH_S = 15
+EXAMPLE = "sample_diarization_japanese.mp3"
+
 # device setting
 if torch.cuda.is_available():
     torch_dtype = torch.bfloat16
@@ -89,6 +91,7 @@ mf_transcribe = gr.Interface(
     title=title,
     description=description,
     allow_flagging="never",
+    examples=[EXAMPLE]
 )
 file_transcribe = gr.Interface(
     fn=transcribe,
@@ -100,6 +103,7 @@ file_transcribe = gr.Interface(
     title=title,
     description=description,
     allow_flagging="never",
+    examples=[EXAMPLE]
 )
 with demo:
     gr.TabbedInterface([mf_transcribe, file_transcribe], ["Microphone", "Audio file"])