tf-Overview - 覚書

e-quantize the 'inputs' tensor of type float via global float scalars minfake_quant_with_min_max_vars_gradient(...): Compute gradien

更新日：2018-07-26 16:28:26