KP

konstantinnikol's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in moonshotai/Kimi-K3 about 20 hours ago

Kimi K3 (2.78T) running on a 128 GB MacBook Pro at ~1 tok/s, experts streamed from four SSDs

3

#185 opened about 22 hours ago by

konstantinnikol

updated a Space about 22 hours ago

Configuration error

README

🚀

published a Space about 22 hours ago

Configuration error

README

🚀

New activity in moonshotai/Kimi-K3 2 days ago

Deploy guide & hardware requirements for Kimi-K3

1

#180 opened 24 days ago by

duola15

New activity in deepseek-ai/DeepSeek-V4.1-Flash 2 days ago

Runs on one RTX 5090 (31.8 GiB) + 125.7 GiB RAM via a llama.cpp fork: GGUF, report and numbers

🔥 5

2

#37 opened 7 days ago by

JigSawPT

DeepSeek V4.1 Flash VRAM requirements check here: https://huggingface.co/spaces/yash-711/deepseek-v41-flash-fit

2

#56 opened 2 days ago by

yash-711

New activity in antirez/deepseek-v4.1-flash-gguf 2 days ago

Prefill reads ~2x the experts it needs — fixing it is 1.7x on one SSD, 2.7x on three

👀 2

2

#6 opened 3 days ago by

konstantinnikol

New activity in deepseek-ai/DeepSeek-V4.1-Flash 3 days ago

Streaming this from SSD: check whether your engine reads all 384 experts per layer

#52 opened 3 days ago by

konstantinnikol

New activity in antirez/deepseek-v4.1-flash-gguf 3 days ago

Prefill reads ~2x the experts it needs — fixing it is 1.7x on one SSD, 2.7x on three

👀 2

2

#6 opened 3 days ago by

konstantinnikol