Both models are claimed to be a significant leap forward compared to Llama 2. Meta says that Llama 3 8B outperforms other open models like Mistral 7B and Google's Gemma 7B on several benchmarks ...
Some results have been hidden because they may be inaccessible to you