LoMoGan

7 ms·

1.
DiffRatio: A SOTA one-step Diffusion model with 50% less GPU memory(arxiv.org)
1 points by LoMoGan 8mo ago|0 comments
2.
DiffRatio – A One-Step Diffusion Model with SOTA quality and 50% less memory(arxiv.org)
4 points by LoMoGan 8mo ago|1 comments
3.
by LoMoGan 8mo ago
This paper was posted on arXiv today. It shows surprisingly strong results on ImageNet at 512 resolution (FID 1.41) with one-step generation, while requiring 50% less training-time memory. Do you think this could become the next standard tr
4.
Show HN: PageIndex Chat – AI reader for long documents
2 points by LoMoGan 9mo ago|0 comments
5.
by LoMoGan 10mo ago
Yeah good point!
6.
by LoMoGan 10mo ago
Yeah lolll, this is the memory you need
7.
Show HN: ChatIndex – A Lossless Memory System for AI Agents
17 points by LoMoGan 10mo ago|5 comments
8.
An API for Chating with Nvidia 10-Q Report(github.com)
2 points by LoMoGan 10mo ago|0 comments
9.
Show HN: ChatIndex – an open-sourced long-context managment system(github.com)
4 points by LoMoGan 10mo ago|0 comments
10.
by LoMoGan 11mo ago
With the rise of vision-language models (VLMs) (such as Qwen-VL and GPT-4.1), new end-to-end OCR models like DeepSeek-OCR have emerged. These models jointly understand visual and textual information, enabling direct interpretation of PDFs w
11.
Do we still need OCR when we can build a pure vision-based AI agent(pageindex.ai)
4 points by LoMoGan 11mo ago|1 comments
12.
by LoMoGan 1y ago
Sounds interesting, will try it out.
13.
by LoMoGan 1y ago
Interesting, is this based on an external Vector DB to store and process the PDF?