Expired80% confidenceFactExact time
For a 4096×4096 weight matrix, LoRA may only need to train two 4096×8 matrices, reducing the number of trainable parameters from approximately 16 million to approximately 65,000, a reduction of 99.6%.
6
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
VerifiedThe rank r of the low-rank matrices in LoRA is typically set to 8 or 16, reducing the number of trainable parameters from d×k to r×(d+k), a reduction of up to hundreds of times.78% similarUnverified对LLaMA-7B所有注意力层应用r=8的LoRA，可训练参数量从70亿降至约400万，压缩比超过1750倍71% similarUnverifiedLoRA可以将可训练参数量减少99%以上71% similarVerifiedFor a 7-billion-parameter model, LoRA typically requires training less than 1% of the total parameters, and GPU memory requirements can be reduced by more than 60%.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/47802API
curl https://kongchang.com/api/v1/knowledge/claims/47802MCP
get_claim(id=47802)