Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!

"Card Beam Animation" by BL/S® Studio codepen.io/blacklead-studio/…

sqrtVINS Robust and Ultrafast Square-Root Filter-based 3D Motion Tracking github.com/rpng/sqrtVINS

just shipped this beam effect built entirely in @framer with the new vector effects - no code components turns out you can create proper laser beam animations now without touching any overrides here's how it works:

here is maya1, our open source voice model: We’re building the future of voice intelligence @mayaresearch_ai team is incredible; amazing work by the team. remarkable moment.

you can design any voice and add emotions

UI/UX Designers, I recently found a website where you can easily get access to hundreds of SVG logos in seconds. 🤩 No more hunting on Google for fake PNG logos. Bookmark it for later 💜

These explainer micro-animations are chef's kiss. → airform.design

Want to understand B-trees better? Try btree.app and bplustree.app. These are standalone sandboxes of the visuals I built for my "B-trees and database indexes" article. Helpful for learning B-tree insertion, search, and node splits.

i am starting a new mini design movement with ascii art - hero section for agentmail

Even invoices should be thoughtful

Two 23 year old Indians just dropped the #2 open-weight AI voice model in the world, trained purely on free credits! Maya1 is #20 globally, better than even Google's best. 3B params, runs on one GPU and does 20+ emotions with < 100ms latency You can just do things.

[ICCV 2025] ACE-G is an architecture and pre-training scheme to improve generalization for scene coordinate regression-based visual relocalization. github.com/nianticspatial/ac…

3D engine for web with an online editor

Open source! We’ve fine-tuned Whisper models to handle arbitrary audio chunks and compressed them with ANNA. This enables optimized inference on @NVIDIA and @Apple devices with 2x faster time-to-first-token and full streaming support. Clone the repo and start building

NVSim: Novel View Synthesis Simulator for Large Scale Indoor Navigation Contributions: • Novel View Synthesis Simulator (NVSim): We propose a new framework that scalably and automatically constructs large-scale indoor environments from only common traversal image sequences. • Floor-Aware Gaussian Splatting: To solve the artifact problem that occurs in floor regions during 3D scene representation, we introduce a robust floor segmentation technique and a Floor-aware Loss. • Mesh-Free Traversability Checking: We propose a method to infer traversability using only rendered views and a zero-shot vision model, without an explicit 3D mesh, and automatically construct a topological navigation graph.

Trace walls over 3D LiDAR scans in @pascal_app editor

Thrilled to share our work, IGGT: Instance-Grounded Geometry Transformer! ✨ 🔧 End-to-End Unified Model 📊 Large-Scale Dataset InsScene-15K 🔌 Instance-Grounded Scene Understanding 🎯 Support Multi-Applications (tracking, segmentation, grounding) lifuguan.github.io/IGGT_offi…

Create animated mockups for product teasers

I finally cracked it. Bank SEC filings are chaos: • messy data • many edge cases • all are unique Standardizing them has always been a pain. Not anymore. Now, when a bank drops its 10-K or 10-Q, it’s parsed and standardized on @findatasets in under 2 seconds.