Open-Orca/OpenOrca-Platypus2-13B

OpenOrca-Platypus2-13B is a 13 billion parameter language model based on the Llama 2 transformer architecture, created by Open-Orca through a merge of Platypus2-13B and OpenOrcaxOpenChat-Preview2-13B. This model demonstrates enhanced performance in logical reasoning and complex problem-solving, achieving an average of 64.56 on the HuggingFace Leaderboard benchmarks. It is particularly strong in areas like LSAT Logical Reasoning and BigBench-Hard, making it suitable for tasks requiring advanced reasoning capabilities.