3 ms·4 Command lines to run open source LLMs across devices with 2MB inference app1 points by 3Sophons 3y ago