bytedance/1d-tokenizer

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1172 stars · 69 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A research repository from ByteDance containing code and pretrained model weights for 1D visual tokenizers (TiTok, TA-TiTok, FlowTok) and image generators (RAR, MaskGen). It compresses images into compact one-dimensional token representations for efficient reconstruction and text-to-image generation.

Use cases

• compress images into compact 1D token representations
• train a text-to-image masked generative model
• reproduce TiTok or RAR research results
• experiment with VQ and VAE image tokenizers
• build image generation models on compact latent tokens

When to choose

• you need state-of-the-art 1D image tokenization for generative model research
• you want pretrained tokenizer and generator checkpoints for experimentation
• you are researching compact latent representations for image generation

When to avoid

• you need a production-ready image generation service
• you want a plug-and-play library without ML research experience
• you lack GPU resources for training or inference

Facets

library · maturity active

machine-learning deep-learning image-processing llm-training machine-learning deep-learning image-processing computer-vision artificial-intelligence python image-tokenization text-to-image vq-vae research-code generative-models transformer gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="bytedance/1d-tokenizer")