github.com/ShinMegamiBoson/lookatmyquant Inkling-2-Mini in 5.4 GB Memory
LookAtMyQuant compresses thinkingmachines/Inkling-2-Mini from 18.89 GB to 5.38 GB and 8.87 GB using its compiler and proximal fine-tuning. The 5.38 GB build outscores Unsloth's 9.14 GB UD-Q2_K_XL on GSM8K and HumanEval at a 41% smaller size, and both builds run through a new Apple Silicon runtime.