Mistral 7B is a 7.3B parameter model that: - Outperforms Llama 2 13B on all benchmarks - Outperforms Llama 1 34B on many benchmarks - Approaches CodeLlama 7B performance on code, while remaining good
--- license: apache-2.0 language: - en - fr - it - de - es --- This the synthetic dataset used for training ```https://huggingface.co/urchade/gliner_multi_pii-v1```. You can get it by browsing the fi