Eval bug: MiMo-V2.6-Distill-Qwen-9B chat template misdetected as Qwen3-Coder (tool calls never complete)

这个报错通常出现在用 llama-server 跑 MiMo-V2.6-Distill-Qwen-9B(GGUF)并调用 OpenAI 兼容 tools API 时,模型自带的 chat template 被误判为 Qwen3-Coder 模板,导致工具调用永远走不到正常结束。优先排查 llama.
![[Bug][ROCm/gfx942]: DeepSeek-V4-Flash silent retrieval corruption for prompts ≥ ~4-5k tokens (AITER sparse indexer)](https://www.chat-gpts.plus/wp-content/uploads/2026/09/52109-0c477c70-768x403.jpg)





![[Bug]: rag/res/term.freq is loaded by term_weight but not shipped — every lowercase Latin token gets the same IDF](https://www.chat-gpts.plus/wp-content/uploads/2026/09/18414-d3efad56-768x403.jpg)
![[Bug]: Memory extraction stores an empty valid_at for semantic items, and an unparseable timestamp is written through unchanged](https://www.chat-gpts.plus/wp-content/uploads/2026/09/18415-7895e527-768x403.jpg)
