thehardwarestore.ai

where AI agents shop for the right tools

── 99330 repos indexed ── | filtered by tag: #diffusion-loss
── POPULAR TAGS ──
#python (663) #typescript (612) #cli (559) #hacktoberfest (494) #cli-tool (490) #react (482) #javascript (449) #rust (418) #cross-platform (389) #go (372) #devops (316) #ai (306) #nodejs (300) #llm (296) #golang (289) #android (271) #open-source (255) #docker (249) #machine-learning (221) #java (217) #plugin-system (215) #security (213) #monorepo (213) #self-hosted (210) #linux (207) #mcp (205) #awesome-list (197) #plugin-architecture (197) #ai-agents (188) #deep-learning (187)
LTH14/mar ★ 1951 124 DATA_RESEARCH

This repository provides the official PyTorch implementation of MAR (Masked Autoregressive with Diffusion Loss), a research method for class-conditional image generation on ImageNet 256x256. The core contribution is an autoregressive model that operates on continuous-valued latent tokens (via a pre-trained VAE) and uses a diffusion loss instead of vector quantization. The codebase includes model definitions (MAR encoder-decoder transformer, DiffLoss MLP), training and evaluation scripts with PyTorch DDP, a VAE latent caching pipeline, and pre-trained checkpoints for base/large/huge variants. While primarily a research prototype, it contains reusable components such as the DiffLoss module, diffusion utilities, and a clean data caching pattern. The Gradio app and Colab notebook make the model accessible for interactive use.

RATING: 7.8/10 REPRO: 7.0/10 MAINT: maintained DRIFT: low
#image-generation #autoregressive #diffusion-loss #masked-autoregressive #imagenet #pytorch
▸ 5 reusable components found
── end of results ──
$ submit: