Category
Explore by categories
Runs Stable Diffusion XL behind a Midjourney-style interface, hiding samplers, model swaps, and LoRA weights. A built-in GPT2 expander rewrites prompts into richer styling, and it works fully offline on as little as 4GB of Nvidia VRAM.
Swaps a face from a single photo onto a live webcam feed or video in real time, using the inswapper_128 model with GFPGAN enhancement. Runs on NVIDIA, Apple Silicon, and Intel GPUs, with a built-in filter that blocks explicit or sensitive media.
Browser-based editor for inspecting, editing, optimizing and publishing 3D Gaussian splats. Runs entirely in the browser with live preview, localization support, and export/publishing workflows — no install required, aimed at quick iteration and lightweight delivery.
Provides a diffusion-model studio for image, video, audio-video, editing, LoRA, and full training workflows so many model families share one inference and training framework.
Reworks AUTOMATIC1111's Stable Diffusion WebUI onto a custom backend that auto-manages GPU memory to speed inference and cut VRAM use. Adds native FLUX support with NF4/GGUF quantization and a UNetPatcher framework for model-agnostic extensions.
Produces real-time 3D reconstructions from multi-view images using Gaussian splatting, with on-device training and interactive viewing across native desktops, Android, and the browser. Uses WebGPU and the Burn ML framework to ship dependency-free binaries, a CLI, live training visualization, and streaming .ply support.
Provides local inference, fine-tuning, and a server/CLI for vision–language and omni (image/audio/video) models via MLX. Supports multi-image chat, audio/video inputs, activation quantization (CUDA), TurboQuant KV cache, and LoRA/QLoRA fine-tuning for on-device workflows.
Segments each PDF page into 11 labeled regions — titles, tables, formulas, figures, footnotes and more — and recovers reading order. Offers two engines: an accurate VGT visual model (~0.96 F1) or a faster CPU-only LightGBM ensemble.
Turns a UI screenshot into structured elements so a vision LLM can act without HTML or accessibility trees. A fine-tuned detector finds interactable icons; a caption model describes their function, lifting GPT-4V grounding on ScreenSpot and Mind2Web.
Trains a 65M-parameter vision-language model from scratch in ~2 hours on one RTX 3090, about 3 RMB (~$0.40) of GPU rental. Connects a frozen SigLIP2 encoder to a small MiniMind LLM via a two-layer MLP projector; full PyTorch code for pretraining and SFT.
Turns PDFs and images into clean Markdown with a 7B vision-language model, keeping tables, equations, handwriting, and multi-column reading order while removing headers and footers. Runs on one 12GB+ GPU at about 1/32 the cost of GPT-4o APIs.
High-resolution image and video generation codebase and models that run with far lower compute and memory than typical diffusion systems. Uses linear-attention DiT variants, aggressive latent compression, and inference-scaling to support text-to-image (up to 4K), fast one/few-step generation, and efficient video pipelines.