llamatheollama

not really “better” in a universal way - pandoc is still the hammer when you need serious format coverage and reliability. markitdown is solid when you mostly want raw text out for agents.

markitme is just aimed at a different thing: markdown you’d actually want to read or drop in obsidian / a wiki / a repo readme. pretty mode, frontmatter, wikilinks, batch folder stuff with a toc and light cross-linking, optional local ollama if you want. if you’re optimizing for ingestion pipelines, markitdown or pandoc might be the move. if you’re optimizing for “this looks like a real note,” that’s what i built it for

Talos is a custom FPGA-based hardware accelerator built from the ground up to execute Convolutional Neural Networks with extreme efficiency. It isn't just a reimplementation of existing software logic in hardware; it is a rethinking of how deep learning inference should work at the circuit level.

Most deep learning frameworks are built for flexibility. They handle dynamic graphs, varying batch sizes, and a multitude of layer types. Talos takes the opposite approach. It strips away the runtime, the scheduler, and the operating system overhead to expose the raw compute capability of the FPGA. By implementing the entire inference pipeline in SystemVerilog, we achieve deterministic, cycle-accurate control over every calculation.

This is a really thoughtful direction. The “overlay instead of another workspace” idea resonates a lot, especially the screen-as-context inversion.

Curious about one thing: where local LLMs feel “good enough” today vs where you still fall back to remote models

The perf work + installers make this feel way more real than most agent demos. Nice job shipping something people can actually try.