textlog
๐Ÿค”profilefollowHai I am kureplied to๐Ÿคธprofilefollowcreative technologist creator of textlog:
Ah thank you. I need to learn this stuff more. I have an desktop at home with AMD 5800X, AMD 5700XT, and 64 GB ram running fedora server that I ssh into using tailscale. I thought I would use it to remotely run all the models but turns out my build is anemic at best for LLM. I use it to offload generating videos from my spoken audio offloading work from my laptop which is interesting to me. Claude wrote the bash scripts and I asked it to use podman containers as an experiment. It is fun to try.
๐Ÿคธprofilefollowcreative technologist creator of textlogreplied to๐Ÿค”profilefollowHai I am ku:
Itโ€™s not running it locally, using OpenRouteropenrouter.aiOpenRouterThe unified interface for every model. Find the best models & prices for your prompts it offers 1K free calls per day for certain models, this is one of them, I use it for thefrontpage.devThe Front PageThe Front PageA newspaper-style front page for Hacker News. for the summaries and this now, most of the time it stays below that limit, if it exceeds it turns to the paid version of the model but still itโ€™s cheap.
join the communityorbrowse more notes