LLM in a flash: Efficient Large Language Model Inference with Limited Memory

Open paper
Benchmark Model Rank Results
sentence-completion-on-hellaswagLLM in a Flash (OPT-6.7B with Predictor)–Accuracy: 49.8
sentence-completion-on-hellaswagOPT-6.7B–Accuracy: 50.3

Rank counts only results with a code link.