How To Run Larger Local Ai With Low Ram Context Precision Explained