open-llm-leaderboard-old/details_facebook__opt-13b

Dataset Card for Evaluation run of facebook/opt-13b Dataset Summary Dataset automatically created during the evaluation run of model facebook/opt-13b on the Open LLM Leaderboard. The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task. The dataset has been created from 4 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_facebook__opt-13b.