Update app.py

app.py

@@ -6,12 +6,7 @@ from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM, BitsAndB
 WHISPER_MODEL = "openai/whisper-large-v3"
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
 
-whisper_quant_config = BitsAndBytesConfig(
-    load_in_4bit=True,
-    bnb_4bit_use_double_quant=True,
-    bnb_4bit_compute_dtype=torch.bfloat16,
-    bnb_4bit_quant_type="nf4"
-)
+processor = AutoProcessor.from_pretrained(MODEL_NAME)
 
 whisper_model = AutoModelForSpeechSeq2Seq.from_pretrained(
     WHISPER_MODEL,
@@ -21,11 +16,11 @@ whisper_model = AutoModelForSpeechSeq2Seq.from_pretrained(
 
 whisper_tokenizer = AutoTokenizer.from_pretrained(WHISPER_MODEL)
 transcriber = pipeline(
-    task="automatic-speech-recognition",
-    model=whisper_model,
-    tokenizer=whisper_tokenizer,
-    chunk_length_s=30,
-    device=DEVICE
+    "automatic-speech-recognition",
+    model=model,
+    tokenizer=processor.tokenizer,
+    feature_extractor=processor.feature_extractor,
+    device=0 if torch.cuda.is_available() else "cpu",
 )
 
 # LLaMA Model Optimization