Expired50% confidenceFactExact time
A project called Needle (6.7k stars) compresses a model down to just 14MB, enabling it to run directly on a smartwatch.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
8/22/2026
First Seen
Valid until: 9/5/2026(expired)
Sources
Related Entities
Related Claims
Unverified测试者用双4090（48GB显存）工作站运行20.7B稠密模型，每秒生成27.6个Token，仅比矮星略快，但参数量小10倍以上65% similarUnverifiedUsing multi-stage builds, the production image of an Astro project can be compressed to the 10–30 MB level, containing neither the Node.js runtime nor the source code.63% similarPartially VerifiedRunning Muse Glimmer at full precision requires 55 GB of VRAM, which is compressed to under 20 GB via 4-bit quantization, allowing it to run on consumer-grade GPUs with 24 GB of VRAM.61% similarVerified32B参数规模的模型在量化后通常可以压缩到约16-20GB显存占用，一张NVIDIA RTX 4090（24GB显存）就能运行60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/786769API
curl https://kongchang.com/api/v1/knowledge/claims/786769MCP
get_claim(id=786769)