Update app.py

app.py

@@ -1,10 +1,8 @@
 import gradio as gr
 from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
 import torch
-import requests
-import os
 
-model_id = "distil-whisper/distil-large-v2"
+model_id = "distil-whisper/distil-large-v3"
 
 device = "cuda:0" if torch.cuda.is_available() else "cpu"
 torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
@@ -17,26 +15,27 @@ model.to(device)
 torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
 processor = AutoProcessor.from_pretrained(model_id)
 
-pipe = pipeline(
-    "automatic-speech-recognition",
-    model=model,
-    tokenizer=processor.tokenizer,
-    feature_extractor=processor.feature_extractor,
-    max_new_tokens=128,
-    torch_dtype=torch_dtype,
-    device=device,
-)
+
 
 def transcribe_audio(audio_file):
-    recorded_filename = audio_file.name
-    if os.path.exists(recorded_filename):
-        results = pipe(recorded_filename)
-        return results["text"]
-    else:
-        return "Error: No audio file uploaded."
-
-inputs = gr.Audio(sources="upload", type="filepath")
+    pipe = pipeline(
+        "automatic-speech-recognition",
+        model=model,
+        tokenizer=processor.tokenizer,
+        feature_extractor=processor.feature_extractor,
+        max_new_tokens=128,
+        torch_dtype=torch_dtype,
+        device=device,
+    )
+    results = pipe(audio_file)
+    return results["text"]
+
+inputs = [
+    gr.Audio(sources="upload", type="filepath"),
+]
 outputs = gr.Textbox()
 
-interface = gr.Interface(fn=transcribe_audio, inputs=inputs, outputs=outputs, title="Audio Transcription App")
+interface = gr.Interface(
+    fn=transcribe_audio, inputs=inputs, outputs=outputs, title="Audio Transcription App"
+)
 interface.launch()