Update app.py

app.py

@@ -1,8 +1,7 @@
 import streamlit as st
-from transformers import pipeline, AutoModelForSequenceClassification, AutoTokenizer, AutoModelForTokenClassification, AutoModelWithLMHead, Wav2Vec2ForCTC, Wav2Vec2Tokenizer
+from transformers import pipeline, AutoModelForSequenceClassification, AutoTokenizer, AutoModelForTokenClassification, AutoModelWithLMHead
 import pandas as pd
 import spacy
-import torchaudio
 
 st.set_page_config(layout="wide")
 
@@ -16,21 +15,19 @@ Birinci Dünya Savaşı sırasında Osmanlı ordusunda görev yapan Atatürk, Ç
 # Uygulama başlığı
 st.title("NLP Toolkit")
 
+
 # Model seçim
-model_list = {
-    'Metin Sınıflandırma': 'dbmdz/bert-base-turkish-cased',
+model_list = [
+     'Metin Sınıflandırma': 'dbmdz/bert-base-turkish-cased',
     'Metin Analizi': 'savasy/bert-base-turkish-ner-cased',
     'Duygu Analizi': 'akdeniz27/xlm-roberta-base-turkish-ner',
-    'Metin Oluşturma': 'dbmdz/bert-base-turkish-cased',
-    'Ses Tanıma': 'facebook/wav2vec2-large-960h'  # ASR model
-}
-
-st.sidebar.header("Select Model")
-model_checkpoint = st.sidebar.radio("", list(model_list.values()), format_func=lambda x: list(model_list.keys())[list(model_list.values()).index(x)])
+    'Metin Oluşturma': 'dbmdz/bert-base-turkish-cased'
+]
 
-task = list(model_list.keys())[list(model_list.values()).index(model_checkpoint)]
+st.sidebar.header("Select NER Model")
+model_checkpoint = st.sidebar.radio("", model_list)
 
-st.sidebar.write("For details of models: 'https://huggingface.co/WhiteAngelss/'")
+st.sidebar.write("For details of models: 'https://huggingface.co/WhiteAngelss/")
 st.sidebar.write("")
 
 if model_checkpoint == "akdeniz27/xlm-roberta-base-turkish-ner":
@@ -79,10 +76,6 @@ def load_pipeline(model_name, task_type):
         model = AutoModelWithLMHead.from_pretrained(model_name)
         tokenizer = AutoTokenizer.from_pretrained(model_name)
         return pipeline('text-generation', model=model, tokenizer=tokenizer)
-    elif task_type == "Ses Tanıma":
-        model = Wav2Vec2ForCTC.from_pretrained(model_name)
-        tokenizer = Wav2Vec2Tokenizer.from_pretrained(model_name)
-        return pipeline('automatic-speech-recognition', model=model, tokenizer=tokenizer)
 
 @st.cache_resource
 def setModel(model_checkpoint, aggregation):
@@ -164,15 +157,4 @@ if Run_Button and input_text != "":
         output = pipeline_model(input_text, max_length=50, num_return_sequences=1)
         st.subheader(f"{task} Sonuçları")
         for idx, item in enumerate(output):
-            st.write(f"Öneri {idx+1}: {item['generated_text']}")
-
-    elif task == "Ses Tanıma":
-        st.subheader("Ses Dosyası Yükle")
-        audio_file = st.file_uploader("Ses Dosyası Seç", type=["wav", "mp3"])
-
-        if audio_file is not None:
-            waveform, sample_rate = torchaudio.load(audio_file)
-            asr_pipeline = load_pipeline(model_checkpoint, task)
-            transcription = asr_pipeline(waveform)
-            st.subheader("Transkripsiyon Sonuçları")
-            st.write(transcription["text"])
+            st.write(f"Öneri {idx+1}: {item['generated_text']}")