3 ms·
dense small models do not like quantization. i find 27b fp8 to be smarter albeit less knowledgable versus the 122B
by nicman23 3mo ago
dense small models do not like quantization. i find 27b fp8 to be smarter albeit less knowledgable versus the 122B