v1.0.0

This release marks the end of llm-benchmark as a standalone launcher. In the future it will be expected to use https://github.com/swiss-ai/pretrain launcher, and only use this repo to analyse results or something like that (to be defined).