Comment by kouteiheika
18 hours ago
Note that these quants are not quantized uniformly, so 4-bit isn't actually a "true" 4-bit here, so these observations won't necessarily hold up to other quants which might be done differently.
18 hours ago
Note that these quants are not quantized uniformly, so 4-bit isn't actually a "true" 4-bit here, so these observations won't necessarily hold up to other quants which might be done differently.
It looks like they tested Q4_K_M which should be just the standard K-quant without any imatrix calibration. The smaller ones are indeed dynamic though.