Projects
A few things I’ve built or am still tinkering with.
lgtmaybe #
A PR reviewer that works with six model providers behind one --provider flag. Bedrock and Vertex use GitHub OIDC, so the cloud providers don’t need static keys in your secrets. The same Python core ships as a PyPI CLI and a GitHub Action, with AI doing most of the typing. I wrote up how I built it, the full docs live at lgtmaybe.coles.codes, and the source is MattJColes/lgtmaybe.
The homelab #
A few Raspberry Pis, a NAS, a Framework Desktop, a DGX Spark and a stack of Dell OptiPlexes. Tailscale ties it together, with Docker containers running the self-hosted services and local AI experiments.
Fine-tuning local vision LLMs #
I fine-tune open-source vision models such as Qwen with Unsloth on my DGX Spark and Framework Desktop. Mostly for fun: I want to know how well a model I can run at home reads images and dense documents.