Create whisper2_asr.py

whisper2_asr.py

@@ -0,0 +1,35 @@
+import gradio as gr
+import requests
+
+# Set the model ID for Whisper small English model
+model_id = "openai/whisper-small.en"
+
+# Function to send audio to Hugging Face Inference API and get transcription
+def transcribe(audio):
+    if audio is None:
+        return "No audio provided."
+
+    API_URL = f"https://api-inference.huggingface.co/models/{model_id}"
+    headers = {"Authorization": "Bearer YOUR_HUGGINGFACE_API_TOKEN"}
+
+    # Read and send the audio file
+    with open(audio, "rb") as f:
+        data = f.read()
+    response = requests.post(API_URL, headers=headers, data=data)
+
+    # Return the transcription or error
+    if response.status_code == 200:
+        return response.json().get("text", "No text returned.")
+    else:
+        return f"Error: {response.status_code} - {response.text}"
+
+# Gradio Interface
+interface = gr.Interface(
+    fn=transcribe,
+    inputs=gr.Audio(source="upload", type="filepath", label="Upload Audio"),
+    outputs=gr.Textbox(label="Transcribed Text"),
+    title="Speech Recognition with Whisper",
+    description="Upload an audio file and get the transcribed text using OpenAI Whisper (small.en)."
+)
+
+interface.launch()