Noctus
@noctus91
Benchmarking for fun | Ambassador @MistralAI |
Open to interesting projects & collaborations ๐ฉ
๊ฐ์
August 2025
356 ํ๋ก์ ์ค 1.3K ํฌ
So youโre telling me I can swap my local LFM2.5-2.6B from F16 to QAD Q4_0 and go from:
5.4 GB โ 1.6 GB
21 โ 64 tok/s
3.0s โ 1.2s tool-call latency
while keeping ~97% of BF16 performance?
@LiquidAI what did you just do ๐ญ
๋ ๋ณด๊ธฐ
