Update app.py

app.py

@@ -3,23 +3,35 @@ os.system("pip install git+https://github.com/openai/whisper.git")
 
 import gradio as gr
 import whisper
+model = whisper.load_model("large")
 
+import time
 
-def speech_to_text(tmp_filename,model_size):
-   
-    model = whisper.load_model(model_size)
-    result = model.transcribe(tmp_filename, fp16=False)
- 
-    return result["text"]
+def transcribe(audio):
+    # load audio for 30 seconds
+    audio = whisper.load_audio(audio)
+    audio = whisper.pad_or_trim(audio)
+
+    # make log-Mel spectrogram and move to device as the model
+    mel = whisper.log_mel_spectrogram(audio).to(model.device)
+
+    # detect the spoken language
+    _, probs = model.detect_language(mel)
+    print(f"Detected language: {max(probs, key=probs.get)}")
+
+    # decoding the audio
+    options = whisper.DecodingOptions(fp16 = False)
+    result = whisper.decode(model, mel, options)
+    print(result.text)
+    return result.text
     
 gr.Interface(
-    title = 'Whisper: Speech to text Model by OpenAI', 
-    fn=speech_to_text, 
+    title = 'Speech to Text with OpenAI (large)', 
+    fn=transcribe, 
     inputs=[
-        gr.Audio(source="microphone", type="filepath"),
-        gr.Dropdown(choices=["tiny","base","small","medium","large"]),
+        gr.inputs.Audio(source="microphone", type="filepath")
     ],
     outputs=[
-        "text"
+        "textbox"
     ],
     live=True).launch()