Why Qwen3.8 27B Looked Brilliant in Testing but Failed to Ship My AI Newspaper
中文摘要
Qwen 3.8 27B 在基准测试中表现完美，但在执行 AI 报纸生成任务时，因运行二十多分钟后超时而失败。
English Summary
Qwen 3.8 27B passed benchmarks perfectly but failed a real-world AI newspaper task by timing out after twenty minutes of processing.
Original Excerpt
Qwen 3.8 27b on a DGX Spark: perfect on my benchmark, and it ran for twenty minutes on the real job, timed out and published nothing… Continue reading on Towards Deep Learning »