# Fixing Memory Issue - Windows Virtual Memory ## Problem ``` OSError: The paging file is too small for this operation to complete. (os error 1455) ``` Your system needs more virtual memory to load the large AI models (1.21GB+). ## Solution: Increase Windows Virtual Memory ### Step-by-Step Instructions: 1. **Open System Properties** - Press `Windows Key + Pause/Break` OR - Right-click "This PC" → Properties → Advanced system settings 2. **Access Virtual Memory Settings** - Click "Advanced" tab - Under "Performance", click "Settings..." - Click "Advanced" tab again - Under "Virtual memory", click "Change..." 3. **Configure Virtual Memory** - **Uncheck** "Automatically manage paging file size for all drives" - Select your C: drive (or main drive) - Select "Custom size" - Set values: - **Initial size (MB):** 8192 (8 GB) - **Maximum size (MB):** 16384 (16 GB) - Click "Set" - Click "OK" on all dialogs 4. **Restart Your Computer** - This is required for changes to take effect 5. **Try Running the App Again** ```powershell python app.py ``` ## Alternative: Quick Fix (Temporary) If you can't change virtual memory settings, try these: ### Option A: Close Other Programs - Close all browsers, apps, and programs - This frees up RAM - Then try running the app again ### Option B: Use Smaller Model (Code Change) Edit `classifier/config.py` to use a smaller model if available. ### Option C: Run with Priority ```powershell # Run Python with higher priority Start-Process python -ArgumentList "app.py" -WindowStyle Normal -Wait ``` ## Checking Current Virtual Memory To see your current settings: 1. Follow steps 1-2 above 2. Note the current "Total paging file size" at the bottom Typical recommendations: - **Minimum:** 1.5x your RAM - **Recommended:** 2-3x your RAM - **For ML models:** At least 8-16 GB ## After Fixing Once virtual memory is increased and computer is restarted: ```powershell cd "C:\Users\Tarak Jha\OneDrive - Coast to Coast Logistics\Desktop\HEALTHBOT\health-query-classifier" python app.py ``` The models should load successfully!