A Unified Tokenizer for Visual Generation and Understanding
active 2025-02-27 → 2026-04-22 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 420 days from 2025-02-27 to 2026-04-22. Pushes: 21 total, peak 6 in a day. Pull requests: 2 total, peak 2 in a day. Issues: 42 total, peak 5 in a day. Comments: 59 total, peak 8 in a day. Stars: 377 total, peak 43 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| machuofan | 57 | 18 | 1 | 31 |
| VJJJJJJ1 | 11 | 0 | 0 | 5 |
| Jason3900 | 8 | 0 | 0 | 6 |
| Masaaki-75 | 5 | 0 | 0 | 4 |
| zgcmx | 4 | 0 | 0 | 2 |
| DavisMeee | 4 | 0 | 0 | 1 |
| AndrewYuweiChen | 4 | 0 | 0 | 2 |
| codeview3d | 3 | 0 | 0 | 2 |
| yvrjsharma | 3 | 0 | 0 | 2 |
| sunset-clouds | 2 | 0 | 0 | 1 |
| NielsRogge | 2 | 0 | 0 | 1 |
| ThisisBillhe | 2 | 0 | 0 | 0 |
| ls-kelvin | 2 | 0 | 0 | 1 |
| iFighting | 2 | 2 | 0 | 0 |
| JingfengYao | 2 | 0 | 0 | 0 |
| Redtides0 | 2 | 0 | 0 | 0 |
| Amber530 | 1 | 0 | 0 | 0 |
| HalvesChen | 1 | 0 | 0 | 0 |
| thesby | 1 | 0 | 0 | 0 |
| RyenZhu | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#34machuofan2026-04-22 00:13Clarification on LLM backbone used
- Issue#32songweii2026-02-17 04:15Causal attention for factorization blocks?
- Issue#31Amber5302025-12-08 04:03ask for model weight based on clip
- Issue#29hyomin142025-11-27 02:26Question on Training Config for UniTok MLLM Pre-training
- Issue#28zgcmx2025-11-26 04:33Difference between MCQ and product quantization?
- Issue comment#28zgcmx2025-11-26 04:33Difference between MCQ and product quantization?
- Issue#28zgcmx2025-11-25 05:54Difference between MCQ and product quantization?
- Issue comment#5zgcmx2025-11-17 11:53Multi-Codebook Quantization
- Issue comment#25machuofan2025-10-15 21:21ValueError: The following `model_kwargs` are not used by the model: ['images', 'image_sizes']
- Issue comment#20machuofan2025-10-15 21:16Training Config for UniTok MLLM
- Issue comment#5sunset-clouds2025-08-30 01:44Multi-Codebook Quantization
- Issue#24sunset-clouds2025-08-30 01:37Extremely large codebook size 2^56
- Issue#23DavisMeee2025-08-26 10:01Incorrect Var lock defination
- Issue comment#22DavisMeee2025-08-25 07:51CUDA Version
- Issue#22DavisMeee2025-08-25 07:51CUDA Version
- Issue comment#21machuofan2025-08-23 08:48Could you please provide 'timm' version?
- Issue#21DavisMeee2025-08-22 18:48Could you please provide 'timm' version?
- Issue comment#20machuofan2025-08-13 09:33Training Config for UniTok MLLM
- Issue#19RyenZhu2025-07-11 02:47Encountering 'staticmethod' object is not callable error when using unitok library on Windows
- Issue comment#15ls-kelvin2025-05-07 13:14Training Code for UniTok mllm
- Issue comment#17Masaaki-752025-05-06 12:41LlamaGen text-to-image training and inference code
- Issue comment#17Masaaki-752025-05-06 10:39LlamaGen training code
- Issue comment#17machuofan2025-05-06 10:07LlamaGen training code
- Issue comment#11machuofan2025-05-06 09:53The training code for generation task seems exist bug.
- Issue comment#17Masaaki-752025-05-06 04:59LlamaGen training code
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 377 stars here means stars gained during the window, not the repo's star count.