A post-training method reduces 99.5% of FLUX.1-dev's transformer weights to ternary values and reports roughly comparable text-to-image quality with large storage and memory savings.
Zeroq: A novel zero shot quantization framework
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
1.58-bit FLUX
A post-training method reduces 99.5% of FLUX.1-dev's transformer weights to ternary values and reports roughly comparable text-to-image quality with large storage and memory savings.