AI projects from Hugging Face

HEART

Phips · Kit

Upscales low-resolution images with a super-resolution network that trains on one consumer GPU and exports to ONNX. For people who train or run image upscalers.

ReFoCo

yuji-nozawa · Kit

Pretrained vision transformer checkpoints in three sizes from a research paper on training attention heads to focus on image regions. For researchers who extra…

Spark-X2.5-4B

XHToken · Kit

A compact 4.1B general-purpose language model with a native 1M-token context, over 200 languages and a hybrid attention design, tuned for coding and agent use.…

KLM 1.5

MinimaLabs · Kit

A 99M-parameter base language model trained from scratch on 100B tokens, with a hybrid attention design. Made for researchers and tinkerers who want a tiny ope…

svd-linear-attention-roberta-base-adasplash

Hooshaai · Kit

Classifies English text, such as SST-2 sentiment, with a RoBERTa Base model whose attention is made sparse by AdaSplash, trading some accuracy for lower latenc…

roberta-base-adasplash

Hooshaai · Kit

A RoBERTa Base sentiment classifier whose attention is swapped for a sparse AdaSplash version, which zeroes out 40% of attention weights to cut time and memory…

distilbert-sst2-rpc

Hooshaai · Kit

A DistilBERT text classifier fine-tuned on GLUE SST-2 that swaps in a compressed random-projection attention to halve the scoring work. An experiment for resea…

qwen-2.5-0.5b-block-transformer-hiba-wikitext2

tahamajs · Kit

Swaps 18 of the 24 attention layers in Qwen2.5-0.5B for a hierarchical block attention operator, then distills it on WikiText-2. A research checkpoint for stud…

qwen-2.5-0.5b-bam-umoe-wikitext2

tahamajs · Kit

Swaps 18 of the 24 attention layers in Qwen2.5-0.5B for a mixture-of-experts attention block, then distills it on WikiText-2. A research checkpoint for studyin…

Qwen3.8-27B-QSA-GSQ-RCO-GGUF

pentacoxian-dev · Kit

Runs Qwen3.8-27B with sparse attention on a single 24 GB GPU at a 262K-token context with image input, as GGUF files for people with a custom llama.cpp engine…

LongCat-Flash-Lite-Sparse

meituan-longcat · Kit

Mixture-of-experts text model with 69B parameters, about 3B active per token, sparse attention and a 1M-token context. For teams hosting long-context models fo…

DeepSeek-V3.2

deepseek-ai · Kit

A 685B open reasoning model from DeepSeek with sparse attention for long context and a chat template that lets it think while it calls tools. For teams that ru…

GLM-5.3-Flash-BF16

zai-org · Kit

A 320B-parameter multimodal model with 18B active that reads text and images, with an adjustable thinking budget and a hybrid attention design for long context…

ARLoop-0.6B

osunlp · Kit

Completes text as a small pretrained looped transformer with attention residuals. A base model for researchers who want to continue its training with their own…

Augmenting CNNs with attention-based aggregation

keras-io · Kit

Shows the attention map behind an image classifier's decision, using a CNN with attention-based aggregation trained on ImageNette, in a Gradio page. For machin…

ARLoop-1.2B

osunlp · Kit

An open 1.2B base language model from osunlp, a looped transformer with attention residuals, released with checkpoints from a 500B-token training run. For text…

Qwen3.8-27B QSA BF16

higa2515 · Kit

BF16 source weights of Qwen3.8-27B with learned sparse attention indexers added, plus a reference Transformers loader, for people studying or deploying Qwen Sp…

Qwen3.8-27B-QSA-GGUF

higa2515 · Kit

Runs Qwen3.8-27B with sparse attention from GGUF files in Q8_0 and Unsloth Q4, Q5 and Q6 mixes, with text, image and video input. For Linux users with NVIDIA G…

Unlimited-OCR-2

baidu · Kit

Baidu's 3.3B document parsing model for long documents, using SR-SWA attention to keep decoding cost nearly flat, with Transformers and SGLang routes. Suits te…

GLM-5.3-Flash

zai-org · Kit

zai-org's natively multimodal GLM-5 series model with 320B total and 18B active parameters, with a hybrid sparse and linear attention design and a thinking bud…