Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
18.4
TFLOPS
12
Abr
Kasnol
Follow
shtefcs's profile picture
1 follower
ยท
4 following
AI & ML interests
None yet
Recent Activity
reacted
to
Anran-MLLM
's
post
with ๐ฅ
6 days ago
๐ Introducing PerceptionDLM โ the first multimodal diffusion LLM for parallel region perception! Most MLLMs are autoregressive, so captioning N regions costs N sequential passes. PerceptionDLM instead describes ALL masked regions in a single denoising process. ๐งฉ โจ Highlights โข โก Up to 3.4ร faster on dense multi-region captioning, with stable per-image latency โข ๐ PerceptionDLM-Base beats LLaDA-V on 15/16 multimodal benchmarks (new SOTA among open diffusion VLMs) โข ๐ New benchmark: ParaDLC-Bench โ jointly evaluates caption quality AND inference efficiency โข ๐ Code, models & benchmark all open-sourced ๐ค Models https://huggingface.co/MSALab/PerceptionDLM-Base https://huggingface.co/MSALab/PerceptionDLM ๐ Benchmark https://huggingface.co/datasets/MSALab/ParaDLC-Bench ๐ Paper: https://huggingface.co/papers/2606.19534 ๐ป Code: https://github.com/MSALab-PKU/PerceptionDLM Diffusion LLMs aren't just for text โ they unlock efficient, parallel visual perception. ๐๏ธโจ #multimodal #diffusion #VLM #perception
liked
a model
13 days ago
OBLITERATUS/gemma-4-E4B-it-OBLITERATED
liked
a model
9 months ago
nineninesix/kani-tts-370m
View all activity
Organizations
None yet
Kasnol
's models
1
Sort:ย Recently updated
Kasnol/FLUX.1-dev-LoRA-AsianMaleAthleticModel
Updated
Sep 30, 2024