Unverified50% confidenceFactExact time
The confidence column shipped by Largha-150M-v1 has an ECE of 0.2519, and the built-in recalibration column (temperature scaling) reduces it to 0.0168.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedIn this configuration, Qwen3 27B achieves approximately 70 T/s natively in llama.cpp, with a measured peak of 76 T/s.56% similarUnverifiedThe full weights of LLaMA 2 (70B parameter version) amount to approximately 130 GB.55% similarUnverifiedThe TDP of the R7 7730U is typically around 15W.52% similarUnverifiedOn GPU fp16 (RTX 5090), Largha-150M-v1 reports a p50 latency of 21.0 ms per decision.52% similarUnverifiedThe H100's peak arithmetic intensity is approximately 295 FLOPs/Byte; when the batch size exceeds this value it is compute-bound, and when it is below this value it is memory-bound.52% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1016987API
curl https://kongchang.com/api/v1/knowledge/claims/1016987MCP
get_claim(id=1016987)