Your 4-Bit Model Can Now Beat Its 16-Bit Original
Quantization-Aware Healing distills a 4-bit model from the pre-compression teacher, winning 7 of 9 benchmarks over its bf16 source.
#quantization
1 post
1 post
Quantization-Aware Healing distills a 4-bit model from the pre-compression teacher, winning 7 of 9 benchmarks over its bf16 source.