Katehuuh

> Updated the readme . Which ver of Pytorch do you have , we are running with 2.7.0 Same `torch==2.7.0,` like my Install: > pip install torch==2.7.0+cu128 torchvision torchaudio --index-url...

Mergekit doesn't have a direct prompts per-layer _analysis_; But this might help: - `positive_prompts:` from `mergekit-moe` is closest; it allows creating MoE models where different experts handle specific types of...

> I couldn't get flash-attn=2.8.0.post2 to work with cuda 12.8. Did anyone get 12.8 to work? 😔 - `flash_attn-2.8.0 - linux`: https://github.com/mjun0812/flash-attention-prebuild-wheels/releases - Else, I've built a prebuilt whl 2.8.**1...

.codex/config.toml of Codex exec ignores --skip-git-repo-check(openai/codex/issues/3511) For example `%USERPROFILE%\.codex\config.toml` (full access): ```toml # not affected by exec... skip_git_repo_check = true # — Global defaults — model = "gpt-5-codex" model_reasoning_effort =...