Update app.py

app.py

@@ -10,17 +10,13 @@ import whisper
 
 import matplotlib as plt
 
-whisper_model = whisper.load_model('large-v2')      # Whisper 모델을 불러오기
+# whisper_model = whisper.load_model('large-v2')      # Whisper 모델을 불러오기
 
 
 path = "Hyeonsieun/NTtoGT_1epoch"
 tokenizer = T5Tokenizer.from_pretrained(path)
 model = T5ForConditionalGeneration.from_pretrained(path)
 
-BATCH_SIZE = 8
-FILE_LIMIT_MB = 1000
-YT_LENGTH_LIMIT_S = 3600  # limit to 1 hour YouTube files
-
 def do_correction(text, model, tokenizer):
     input_text = f"translate the text pronouncing the formula to a LaTeX equation: {text}"
     inputs = tokenizer.encode(
@@ -46,7 +42,13 @@ def do_correction(text, model, tokenizer):
     )
     return corrected_sentence
 
+# corrected_sentence = do_correction(sentence, model, tokenizer)
+
+
+gr.Interface(fn=yt_do_correction, inputs="text", outputs="text")
+
 
+'''
 pipe = pipeline(
     task="automatic-speech-recognition",
     model=MODEL_NAME,
@@ -179,4 +181,5 @@ yt_transcribe = gr.Interface(
 with demo:
     gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
 
-demo.launch(enable_queue=True)
+demo.launch(enable_queue=True)
+'''