# # Copyright (C) 2019 The Android Open Source Project # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # import itertools import random def dequantize(x, scale, offset): return (x - offset) * scale def quantize(x, scale, offset): return max(-128, min(127, int(round(x / scale)) + offset)) def create_test(input0_scale, input0_offset, input1_scale, input1_offset, output_scale, output_offset): def sub_quantized(a, b): a_dequantized = dequantize(a, input0_scale, input0_offset) b_dequantized = dequantize(b, input1_scale, input1_offset) return quantize(a_dequantized - b_dequantized, output_scale, output_offset) values = [-128, -127, -126, -125, -124, -123, 122, 123, 124, 125, 126, 127] inputs = list(itertools.product(values, values)) input0_values, input1_values = zip(*inputs) output_values = [sub_quantized(a, b) for a, b in inputs] size = len(output_values) input0 = Input("input0", "TENSOR_QUANT8_ASYMM_SIGNED", "{%d}, %g, %d" % (size, input0_scale, input0_offset)) input1 = Input("input1", "TENSOR_QUANT8_ASYMM_SIGNED", "{%d}, %g, %d" % (size, input1_scale, input1_offset)) activation = 0 output0 = Output("output0", "TENSOR_QUANT8_ASYMM_SIGNED", "{%d}, %g, %d" % (size, output_scale, output_offset)) model = Model().Operation("SUB", input0, input1, activation).To(output0) Example({ input0: input0_values, input1: input1_values, output0: output_values, }) scales_and_offsets = [(1.0, -128), (1.0, -127), (0.01, -8), (10.0, -8)] for params in itertools.product(scales_and_offsets, scales_and_offsets, scales_and_offsets): input0_params, input1_params, output_params = params create_test(*input0_params, *input1_params, *output_params) ####################################################### input0 = Input("input0", "TENSOR_QUANT8_ASYMM_SIGNED", "{1, 2}, 1.0, -128") input1 = Input("input1", "TENSOR_QUANT8_ASYMM_SIGNED", "{2, 2}, 1.0, -128") activation = 0 output0 = Output("output0", "TENSOR_QUANT8_ASYMM_SIGNED", "{2, 2}, 1.0, -128") model = Model("quant8").Operation("SUB", input0, input1, activation).To(output0) input0_values = [-28, 72] input1_values = [-127, -126, -125, -124] output_values = [-29, 70, -31, 68] Example({ input0: input0_values, input1: input1_values, output0: output_values, }) ####################################################### shape = "{2, 4, 16, 2}, 0.5, -128" input0 = Input("input0", "TENSOR_QUANT8_ASYMM_SIGNED", shape) input1 = Input("input1", "TENSOR_QUANT8_ASYMM_SIGNED", shape) activation = 0 output0 = Output("output0", "TENSOR_QUANT8_ASYMM_SIGNED", shape) model = Model("quant8").Operation("SUB", input0, input1, activation).To(output0) input0_values = list(range(-128, 128)) input1_values = list(input0_values) random.seed(0) random.shuffle(input1_values) output_values = [max(-128, (a - b) - 128) for a, b in zip(input0_values, input1_values)] Example({ input0: input0_values, input1: input1_values, output0: output_values, }) ####################################################### # SUB, zero-sized input # Use BOX_WITH_NMS_LIMIT op to generate a zero-sized internal tensor for box cooridnates. p1 = Parameter("scores", "TENSOR_FLOAT32", "{1, 2}", [0.90, 0.10]) # scores p2 = Parameter("roi", "TENSOR_FLOAT32", "{1, 8}", [1, 1, 10, 10, 0, 0, 10, 10]) # roi o1 = Output("scoresOut", "TENSOR_FLOAT32", "{0}") # scores out o2 = Output("classesOut", "TENSOR_INT32", "{0}") # classes out tmp1 = Internal("roiOut", "TENSOR_FLOAT32", "{0, 4}") # roi out tmp2 = Internal("batchSplitOut", "TENSOR_INT32", "{0}") # batch split out model = Model("zero_sized").Operation("BOX_WITH_NMS_LIMIT", p1, p2, [0], 0.3, -1, 0, 0.4, 1.0, 0.3).To(o1, tmp1, o2, tmp2) # Use ROI_ALIGN op to convert into zero-sized feature map. layout = BoolScalar("layout", False) # NHWC i1 = Input("in", "TENSOR_FLOAT32", "{1, 1, 1, 2}") zero_sized = Internal("featureMap", "TENSOR_FLOAT32", "{0, 2, 2, 2}") model = model.Operation("ROI_ALIGN", i1, tmp1, tmp2, 2, 2, 2.0, 2.0, 4, 4, layout).To(zero_sized) # SUB op with numBatches = 0. i2 = Parameter("op", "TENSOR_FLOAT32", "{1, 2, 2, 1}", [1, 2, 3, 4]) # weights o3 = Output("out", "TENSOR_FLOAT32", "{0, 2, 2, 2}") # out model = model.Operation("SUB", zero_sized, i2, 0).To(o3) quant8_signed = DataTypeConverter().Identify({ p1: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0), p2: ("TENSOR_QUANT16_ASYMM", 0.125, 0), o1: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0), tmp1: ("TENSOR_QUANT16_ASYMM", 0.125, 0), i1: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0), zero_sized: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0), i2: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0), o3: ("TENSOR_QUANT8_ASYMM_SIGNED", 0.1, 0) }) Example({ i1: [1, 2], o1: [], o2: [], o3: [], }).AddVariations(quant8_signed, includeDefault=False)