A real test suite which fully drives both Julia and Emacs and runs in a CI environment (like GitHub Actions) wouldn’t hurt, either.
Москвичам пообещали тепло17:31
。WPS极速下载页对此有专业解读
RYS-XLargeAfter testing several smaller models (Llama’s and smaller Qwen2’s), I set up the config for Qwen2-72B and let it sweep. Each $(i, j)$ configuration took a few minutes: load the re-layered model, run the math probe, run the EQ probe, record the scores, move on. Days of continuous GPU time on the 4090s. But far less compute than a fine tune! In fact, I didn’t even have the hardware needed for a LORA fine-tune on just 48GB of VRAM.，详情可参考手游
<img src="/media/core/1x1clear.gif"，详情可参考博客