Xuyang Ge
Xuyang Ge
Hello! Lots in the repository have changed since the publish of Llama Scope, etc. We'll recently fix the documentation.
I'm sorry, but the current README and examples are actually outdated. We'll update them as soon as we have enough capacity. Currently we recommend to use [uv](https://docs.astral.sh/uv/guides/projects/#managing-dependencies) as the package...
Sorry for the late reply. > 1. Can you tell me what command I should use to build environment using `uv`? Once you have `uv` installed (following the instructions on...
Have you tried the script above in this issue? The example script has not been up-to-date yet.
Hi, It seems to be some bugs in the current train runner. It doesn't fit non-pre-generated datasets. I'll push a fix asap.
Hello, this should be fixed with #85 . Also, you can try separately generating activations and training the SAE, which can drastically improve the training speed as long as you...
Helps a lot in understanding how TypeScript type guards work.
感谢反馈！之前几个月整个仓库结构变动较大，我们近期会把 examples 和 README.md 修正好。
Sorry for bothering but IMO this feature will have a wide range of uses, e.g. stop auto-regressive generation of LLM when encountering EOS. Would really appreciate it if this could...