Upload app.py with huggingface_hub

app.py

@@ -0,0 +1,56 @@
+import gradio as gr
+import torch
+import librosa
+import numpy as np
+from transformers import AutoProcessor, AutoModelForCTC
+
+# Load model and processor
+print("Loading model...")
+processor = AutoProcessor.from_pretrained("HAMMALE/mms-darija-finetuned")
+model = AutoModelForCTC.from_pretrained("HAMMALE/mms-darija-finetuned")
+
+def transcribe_audio(audio_file):
+    try:
+        # Load audio
+        if audio_file is None:
+            return "Please upload an audio file."
+        
+        # Load and preprocess audio
+        audio, sr = librosa.load(audio_file, sr=16000)
+        
+        # Handle very short audio
+        if len(audio) < 1600:  # Less than 0.1 seconds
+            return "Audio too short. Please upload a longer audio file."
+        
+        # Process with model
+        inputs = processor(audio, sampling_rate=16000, return_tensors="pt")
+        
+        # Inference
+        with torch.no_grad():
+            logits = model(**inputs).logits
+        
+        predicted_ids = torch.argmax(logits, dim=-1)
+        transcription = processor.batch_decode(predicted_ids)[0]
+        
+        return transcription if transcription.strip() else "No transcription generated."
+        
+    except Exception as e:
+        return f"Error processing audio: {str(e)}"
+
+# Create Gradio interface
+demo = gr.Interface(
+    fn=transcribe_audio,
+    inputs=gr.Audio(type="filepath", label="Upload Darija Audio"),
+    outputs=gr.Textbox(label="Transcription", placeholder="Transcription will appear here..."),
+    title="🎤 Darija Speech Recognition",
+    description="Upload an audio file in Moroccan Arabic (Darija) and get the transcription. This model was fine-tuned on the Darija Bible dataset.",
+    article="Model: [HAMMALE/mms-darija-finetuned](https://huggingface.co/HAMMALE/mms-darija-finetuned)",
+    examples=[
+        # You can add example audio files here if you have them
+    ],
+    cache_examples=False,
+    theme=gr.themes.Soft()
+)
+
+if __name__ == "__main__":
+    demo.launch()