HumanEval is saturated: new coding LLM benchmark released b bigcode-bench.github.io 1 eitanturok 2 years ago 0 comments
Are there certain types of failures that pre-training alone cannot help with? g gradientscience.org 2 eitanturok 2 years ago 0 comments
LLMs: Speeding up ALiBi by 3-5x with a hardware-efficient implementation p pli.princeton.edu 2 eitanturok 2 years ago 1 comments
