Zach Mustafa's picture

Zach Mustafa PRO

Zmu

·

AI & ML interests

None yet

Recent Activity

liked a dataset about 7 hours ago

OpenAssistant/oasst1

liked a Space about 7 hours ago

reddit-tools-HF/dataset-creator-reddit-bestofredditorupdates

liked a dataset about 7 hours ago

BIGJUTT/NSFW_Tags_25k

View all activity

Organizations

upvoted a collection 1 day ago

Qwen3.5

2 items • Updated 1 day ago • 170

upvoted an article 6 days ago

Article

NVIDIA brings agents to life with DGX Spark and Reachy Mini

+1

Jan 5

•

61

upvoted 3 collections 16 days ago

Heretic - Abliterated, Uncensored, Unrestricted POWER.

Models that have be abliterated using the HERETIC method. Done properly, this completely removed almost all censorship with no damage to the model. • 75 items • Updated 3 days ago • 30

✂️ Abliteration

Uncensored models using abliteration. See this article for more information: huggingface.co/blog/mlabonne/abliteration • 34 items • Updated Jul 25, 2025 • 156

Qwen3-VL-abliterated

15 items • Updated Dec 13, 2025 • 25

upvoted an article 17 days ago

Article

Introducing Daggr: Chain apps programmatically, inspect visually

+3

22 days ago

•

100

upvoted a collection 27 days ago

VibeVoice

Frontier Text-to-Speech Models https://microsoft.github.io/VibeVoice/ • 9 items • Updated 30 days ago • 209

upvoted a collection 29 days ago

Qwen3-TTS

7 items • Updated 29 days ago • 299

upvoted a collection about 1 month ago

DroPE

Extending the Context of Pretrained LLMs by Dropping Their Positional Embedding (https://www.arxiv.org/abs/2512.12167) • 1 item • Updated Jan 11 • 2

upvoted a collection 2 months ago

Qwen3-Omni

6 items • Updated Dec 31, 2025 • 183

upvoted an article 2 months ago

Article

New in llama.cpp: Model Management

Dec 11, 2025

•

120

upvoted 2 articles 3 months ago

Article

Transformers v5: Simple model definitions powering the AI ecosystem

+2

Dec 1, 2025

•

298

Article

Continuous batching from first principles

+1

Nov 25, 2025

•

326

upvoted 3 collections 3 months ago

Olmo 3

Artifacts for the Olmo 3 release. • 9 items • Updated Dec 23, 2025 • 163

V-JEPA 2

A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of https://ai.meta.com/blog/v-jepa-yann • 8 items • Updated Jun 13, 2025 • 192

ViDoRe Benchmark V3

ViDoRe V3 is our latest benchmark, engineered to set a new industry gold standard for multi-modal, enterprise document retrieval evaluation. • 8 items • Updated Jan 14 • 19

upvoted a paper 5 months ago

zELO: ELO-inspired Training Method for Rerankers and Embedding Models

Paper • 2509.12541 • Published Sep 16, 2025 • 6

upvoted 2 collections 6 months ago

OpenVision

27 items • Updated Aug 15, 2025 • 33

DINOv3

DINOv3: foundation models producing excellent dense features, outperforming SotA w/o fine-tuning - https://arxiv.org/abs/2508.10104 • 13 items • Updated Aug 21, 2025 • 501

upvoted a collection 7 months ago

Qwen3

Qwen's new Qwen3 models. In Unsloth Dynamic 2.0, GGUF, 4-bit and 16-bit Safetensor formats. Includes 128K Context Length variants. • 79 items • Updated 4 days ago • 261