Update app.py

app.py

@@ -9,7 +9,6 @@ from transformers import (
 )
 
 torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-
 model_id = "sudoping01/whosper-large"
 
 # Load model and create pipeline
@@ -18,7 +17,7 @@ model = WhisperForConditionalGeneration.from_pretrained(
     device_map="auto",
     use_cache=True,
     attention_dropout=0.1,
-    dropout=0.1, 
+    dropout=0.1,
     token=os.environ.get("HF_TOKEN")
 )
 
@@ -44,13 +43,13 @@ pipe = pipeline(
     batch_size=1
 )
 
-def transcribe(audio_path):
-    if audio_path is None:
+def transcribe(audio):
+    if audio is None:
         return "Please provide an audio input."
     
     try:
         result = pipe(
-            audio_path,
+            audio,
             generate_kwargs={
                 "temperature": 0.0,
                 "do_sample": False,
@@ -67,10 +66,10 @@ def transcribe(audio_path):
 demo = gr.Interface(
     fn=transcribe,
     inputs=[
-        gr.Audio(source="microphone", type="filepath", label="Audio Input")
+        gr.Audio(sources=["microphone", "upload"], type="filepath")
     ],
     outputs=gr.Textbox(label="Transcription"),
-    title="Multilingual Speech Recognition: Wolof, French, English, .. or Mix ",
+    title="Multilingual Speech Recognition: Wolof, French, English, .. or Mix",
     description="Upload an audio file or record audio to transcribe Wolof, French, or English speech...",
     theme="default",
     enable_queue=True