MI

umisakaai

AI & ML interests

None yet

Recent Activity

new activity 2 days ago

lvladikov/LTX-2.5-Stage1-Distill-2step-LoRA:[10 Nov 2026] Checkpoint 12K released

new activity 4 days ago

NousResearch/Arete-27B:Arete-27B in CMF: a turbo file for long prompts on 16 GB cards

new activity 7 days ago

nvidia/DeepSeek-V4.1-Flash-1017-NVFP4:Add SGLang deployment instructions

View all activity

Organizations

None yet

New activity in lvladikov/LTX-2.5-Stage1-Distill-2step-LoRA 2 days ago

[10 Nov 2026] Checkpoint 12K released

#1 opened 2 days ago by

lvladikov

New activity in NousResearch/Arete-27B 4 days ago

Arete-27B in CMF: a turbo file for long prompts on 16 GB cards

#10 opened 5 days ago by

infosave

New activity in nvidia/DeepSeek-V4.1-Flash-1017-NVFP4 7 days ago

Add SGLang deployment instructions

#1 opened 10 days ago by

weiliang-nv

New activity in AuroraAI-Research/AuroraIMG2-100M 7 days ago

Demo for this model on Spaces

#1 opened 8 days ago by

multimodalart

New activity in lodestones/chroma-h3 9 days ago

Any chance we could get some example workflow(s)?

👍➕ 4

#2 opened 12 days ago by

roe2

New activity in Efficient-Large-Model/Open-H3-Regenerate-2K 9 days ago

Stage-2 refine removes subtle transient motion (e.g. eye blinks)

#7 opened 12 days ago by

ehsrzv

New activity in speridlabs/idol 11 days ago

RTX 5070 SPEED

#3 opened 13 days ago by

mithulaartigala

New activity in ElectronGirl/anima-ref-adapter 11 days ago

Thanks for the amazing model.

#1 opened 13 days ago by

CptCharname

New activity in nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B 11 days ago

NVFP4 and FP8 variants linked from the card return 404

👍 4

#8 opened 15 days ago by

shuuul

New activity in Pika-Labs/Pika-H3-Flex 11 days ago

H3 Flex now runs natively on Mac

🚀 1

#3 opened 14 days ago by

Mrbizarro

replied to HannesVonEssen's post 13 days ago

Having the warm-ups, 200 prompts and 512×512/4-step setup next to that 430 ms really helps.

replied to gianlucamazza's post 13 days ago

Thanks for spelling out that +0.076 was over the always-on arm, not cosine.

replied to kumarkjha's post 13 days ago

Counting decoded frames rather than interpolated ones makes the FPS comparison much clearer.

New activity in XHToken/Spark-X2.6-4B 14 days ago

[HER Hack-Astron # 7] [Open Source AI for Good] 易讀橋 PlainBridge

#19 opened 17 days ago by

chillg-research

New activity in Comfy-Org/MOSS-Image 17 days ago

Add MOSS-Image edit LoRA (rank 64, bf16)

#11 opened 18 days ago by

ComfyUI-Wiki

New activity in linoyts/Qwen-Image-2.1-DMD2-2steps 17 days ago

Add Qwen-Image-2.1 official space vs 2-step LoRA comparison (t2i + i2i)

#1 opened 20 days ago by

linoyts

New activity in Nanbeige/Nanbeige4.3-33B 17 days ago

KV Optimizations?

👍👀 4

#8 opened 19 days ago by

bobbytaylor

New activity in nvidia/Wan2.2-T2V-A14B-Diffusers-NVFP4 19 days ago

Fix trtllm-serve command for TensorRT-LLM ToT Main

#2 opened 28 days ago by

yiguo-nv

New activity in gradium/phonon-31m-en 21 days ago

cn version

#3 opened 23 days ago by

next111

New activity in openbmb/MiniCPM-V-4.7-35B-A3B 21 days ago

Lookahead Expert Scheduling (LES) for streaming from SSD drive (MoE-SSD)

👍 1

#15 opened 25 days ago by

ssvb

MI