localmod
Self-hosted content moderation API that outperforms Amazon Comprehend. 100% offline, your data never leaves your server. Text + Image moderation.
- 116
- Python
- MIT
- 3 wk ago
KOKOSde
Escaping the Simulation, or Creating one.
Built from KOKOSde's public GitHub profile, Arizona . Not on Take My Repo yet, so there are no posts or updates here.
Self-hosted content moderation API that outperforms Amazon Comprehend. 100% offline, your data never leaves your server. Text + Image moderation.
Open-source behavioral Sybil attack detection for blockchain networks
A fully functional conversational AI system for restaurant phone ordering using Twilio, Whisper, LLM, and RAG
Offline Nsight Systems SQLite explainer for LLM inference hotspots
ONNX/MLIR compiler workbench for transformer inference graphs: analysis, safe rewrites, ORT validation, and optional lowering.
Production-ready Agentforce reliable RAG agent for Salesforce scratch orgs
Cross-Layer Transcoder training for Vision-Language Models - Adapted from EleutherAI for VLMs
Cross-Layer Transcoder (CLT) library for extracting sparse interpretable features from transformer models. Batched processing, memory-efficient, works with any LLM/VLM.
4.76x Faster Attribution Graph Generation for LLMs and VLMs - Achieves 79% speedup by eliminating Python loops and vectorizing GPU operations. Works with GPT, LLaMA, Qwen, LLaVA, CLIP
Physics-based road & track annotation engine for AV/robotaxi teams and researchers (C++20 core, Python+Go frontends).
VLM-ready attribution graph tooling with PLT/CLT support, spatial heatmaps, labels, and steering.
SageMaker deployment and benchmarking for Hugging Face models
PyPI package for optimized sparse feature extraction from transformer models. Batched processing, memory-efficient chunking, works with any LLM/VLM.
A tool to fetch PancakeSwap round results and odds for BNB prediction game