Extending the context length to 1M tokens (qwenlm.github.io)
After the release of Qwen2.5, we heard the community’s demand for processing longer contexts.