Initial Draft

app.py

@@ -0,0 +1,48 @@
+import streamlit as st
+import speech_recognition as sr
+from streaming_stt_nemo import Model
+
+if "sv_load_flash-attn" not in st.session_state:
+    subprocess.run(
+                    "pip install flash-attn --no-build-isolation",
+                    env={"FLASH_ATTENTION_SKIP_CUDA_BUILD": "TRUE"},
+                    shell=True,
+                  )
+    st.session_state.sv_load_flash-attn = True
+
+def main():
+    st.title("Speech to Text Converter")
+
+    # Create a recognizer object
+    r = sr.Recognizer()
+    
+    # Create a microphone object
+    mic = sr.Microphone()
+
+    # Continuously listen for audio input
+    with mic as source:
+        st.write("Listening...")
+        while True:
+            try:
+                # Adjust the ambient noise threshold for better results
+                r.adjust_for_ambient_noise(source)
+
+                # Listen for audio input
+                audio = r.listen(source)
+
+                # Convert audio to text
+                default_lang = "en"
+                engines = {
+                                default_lang: Model(default_lang)
+                          }
+                model = engines[default_lang]
+                text = model.stt_file(audio)[0]
+
+                # Display the converted text
+                st.write("You said:", text)
+
+            except sr.UnknownValueError:
+                pass
+
+if __name__ == "__main__":
+    main()