Qwen/Qwen2-7B-Instruct

Qwen/Qwen2-7B-Instruct is a 7.6 billion parameter instruction-tuned causal language model developed by Qwen, based on the Transformer architecture. It supports an extensive context length of up to 131,072 tokens, utilizing YARN for long text processing. This model demonstrates strong performance across language understanding, generation, multilingual tasks, coding, mathematics, and reasoning benchmarks, often surpassing other open-source models in its class.