update README for real-time speech-to-text application and remove spaces.GPU decorator from load_model function

README.md

@@ -1,6 +1,6 @@
 ---
-title: Asr Demo
-emoji: 🚀
+title: Real-time Speech-to-Text
+emoji: 🎙️
 colorFrom: indigo
 colorTo: gray
 sdk: gradio
@@ -9,4 +9,30 @@ app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Real-time Speech-to-Text with NeMo
+
+This is a real-time speech-to-text transcription application powered by NVIDIA NeMo and the parakeet-tdt-0.6b-v2 model.
+
+## Features
+
+- 🎙️ Web-based microphone input
+- ⚡ Real-time transcription displayed in the browser
+- 🧠 Fast inference with NeMo pre-trained model
+- 🛠️ Easy to use, no installations required
+
+## Tech Stack
+
+- Python
+- Gradio
+- NVIDIA NeMo Toolkit for ASR
+
+## How to Use
+
+1. Click the microphone button to start recording
+2. Speak clearly into your microphone
+3. The transcription will appear in real-time
+4. Click 'Clear Transcript' to start a new transcription
+
+## Note
+
+This application requires access to your microphone to function. The audio is processed in real-time and is not stored.

app.py

@@ -4,7 +4,6 @@ import torch
 import nemo.collections.asr as nemo_asr
 from omegaconf import OmegaConf
 import time
-import spaces
 
 # Check if CUDA is available
 print(f"CUDA available: {torch.cuda.is_available()}")
@@ -12,7 +11,6 @@ if torch.cuda.is_available():
     print(f"CUDA device: {torch.cuda.get_device_name(0)}")
 
 # Initialize the ASR model - removed spaces.GPU decorator due to pickling issues
-@spaces.GPU
 def load_model():
     print("Loading ASR model...")
     # Load the NVIDIA NeMo ASR model