This should make the use case a bit clearer. It's basically a starting point / wrapper of a few tools when you know you'll probably build something custom later so want to invest 0 time in the beginning but need something that's workable: https://www.differentiated.io/blog/the-modern-document-proce...
That doesn't really answer my question. Like, I have a website, and I have many references; I also use LLM embeddings for nearest-neighbors recommendations of references to each other.
What might this... do... for me? Don't dump a bunch of JS which is how I would 'do' whatever it does. What does it do? Like, can I dump the URL 'https://pmc.ncbi.nlm.nih.gov/articles/PMC4543385/' into it and get out nice usable clean text of the abstract, say? What about a complicated PDF like https://gwern.net/doc/psychiatry/anxiety/2025-he.pdf (these are the last two references I added)? What do I get? Do I have to install the whole darn thing just to see what it does?
With Llama 3 & Phi-3 just being released and achieving incredible benchmark results it makes sense.
However, Deepmind is doing some really cool experiments with different architectures. Recently they have applied the Griffin architecture to Gemma:
- Griffin combines gated linear recurrences with local attention to optimize performance on long sequences.
- This achieves comparable results to larger models with far fewer training tokens.
- During inference, Griffin provides MUCH higher throughput compared to Transformer based models
