Keep on to blur preview images; turn off to show them clearly
It’s a misconception to think that popular implementations of AI inference pipelines using transformer-based LLMs are somehow leaving huge amounts of performance on the table as a result of engineering mistakes or sloppiness.
Former Quant Investor, now building @lumera (formerly called Pastel Network) | My Open Source Projects: https://t.co/9qbOCDlaqM
Finally, we've drastically improved the performance of using LSPs. Python and TypeScript LSPs are now faster by default. Memory is dynamically configured based on available RAM. We've also fixed a number of memory leaks and improved memory usage.
The best way to code with AI.
We've updated our web interface.
Finally, we've drastically improved the performance of using LSPs. Python and TypeScript LSPs are now faster by default. Memory is dynamically configured based on available RAM. We've also fixed a number of memory leaks and improved memory usage.
We've shipped several quality-of-life improvements to Cursor! Details below...
You can now plan with one model and implement with another. Build the plan locally or in the cloud.
RT @gradypb: 1/ We @sequoia have always said that we’re only as good as our next investment. It’s also true that we’re only as good as our…
Co-founder & CEO @HuggingFace 🤗, the open and collaborative platform for AI builders
implicit in this is the assumption that we can’t and won’t build fast enough on earth. that we’ll have to go to the most expensive, farthest place imaginable to do it instead. sci-fi usually imagines humanity leaving earth because we built too much — too many factories, people, pollution, etc. the irony is that the real driver of space colonization might be the opposite: we can't build enough.
investing @a16z // curating https://t.co/ssslqn6eo7