View source on GitHub |
Please use tf.quantization.quantize
instead.
tf.compat.v1.quantize_v2(
input,
min_range,
max_range,
T,
mode='MIN_COMBINED',
name=None,
round_mode='HALF_AWAY_FROM_ZERO',
narrow_range=False,
axis=None,
ensure_minimum_range=0.01
)