bit-width reduction

model compression (3302) large language model (13587) straight-through estimator (25) parameter quantization (20) weight quantization (137) mixed-precision quantization (40) low-bit quantization (39) parameter initialization (21) parameter compression (26) quantization-aware training (50)