Qwen1.5 7B
Alibaba · 2024-02-04 · 7.7B parameters
Qwen1.5-7B is Alibaba's 7-billion-parameter base language model from the Qwen1.5 family, released February 4, 2024. It is a dense Transformer decoder with multilingual coverage and up to 32K tokens of context, and was the first Qwen generation to ship with native Hugging Face Transformers integration. The base model is intended for downstream fine-tuning, while its chat counterpart Qwen1.5-7B-Chat is aligned for direct conversational use.