| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455145614571458145914601461146214631464146514661467146814691470147114721473147414751476147714781479148014811482148314841485148614871488148914901491149214931494149514961497149814991500150115021503150415051506150715081509151015111512151315141515151615171518151915201521152215231524152515261527152815291530153115321533153415351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593159415951596159715981599160016011602160316041605160616071608160916101611161216131614161516161617161816191620162116221623162416251626162716281629163016311632163316341635163616371638163916401641164216431644164516461647164816491650165116521653165416551656165716581659166016611662166316641665166616671668166916701671167216731674167516761677167816791680168116821683168416851686168716881689169016911692169316941695169616971698169917001701170217031704170517061707170817091710171117121713171417151716171717181719172017211722172317241725172617271728172917301731173217331734173517361737173817391740174117421743174417451746174717481749175017511752175317541755175617571758175917601761176217631764176517661767176817691770177117721773177417751776177717781779178017811782178317841785178617871788178917901791179217931794179517961797179817991800180118021803180418051806180718081809181018111812181318141815181618171818181918201821182218231824182518261827182818291830183118321833183418351836183718381839184018411842184318441845184618471848184918501851185218531854185518561857185818591860186118621863186418651866186718681869187018711872187318741875187618771878187918801881188218831884188518861887188818891890189118921893189418951896189718981899190019011902190319041905190619071908190919101911191219131914191519161917191819191920192119221923192419251926192719281929193019311932193319341935 |
- #!/usr/bin/env python3
- #
- # SPDX-FileCopyrightText: Copyright 2010-2023 Arm Limited and/or its affiliates <open-source-office@arm.com>
- #
- # SPDX-License-Identifier: Apache-2.0
- #
- # 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
- #
- # 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 os
- import sys
- import argparse
- from packaging import version
- from test_settings import TestSettings
- from conv_settings import ConvSettings
- from pooling_settings import PoolingSettings
- from softmax_settings import SoftmaxSettings
- from svdf_settings import SVDFSettings
- from add_mul_settings import AddMulSettings
- from lstm_settings import LSTMSettings
- from fully_connected_settings import FullyConnectedSettings
- import tensorflow as tf
- def parse_args() -> argparse.Namespace:
- parser = argparse.ArgumentParser(description="Generate input and refererence output data for unittests."
- " It can regenerate all data, load all stored data or a combination of it.")
- parser.add_argument('--dataset', type=str, default=None, help="Name of generated test set.")
- parser.add_argument('--regenerate-weights', action='store_true', help="Regenerate and store new weights.")
- parser.add_argument('--regenerate-input', action='store_true', help="Regenerate and store new input.")
- parser.add_argument('--regenerate-biases', action='store_true', help="Regenerate and store new biases.")
- parser.add_argument('-a', '--regenerate-all', action='store_true', help="Regenerate and store all data.")
- parser.add_argument('-t',
- '--testtype',
- type=str,
- default=None,
- choices=[
- 'conv', 'depthwise_conv', 'avgpool', 'maxpool', 'fully_connected', 'softmax', 'svdf', 'add',
- 'mul', 'lstm'
- ],
- help='Type of test. There are the operators that have unit tests.')
- parser.add_argument('--run-all-testsets',
- action='store_true',
- help="Run the script for all existing test "
- "sets. Regenerate all, partially all or no input data (output may still change, depending on"
- " changes in script) depending on regenerate flags. If used together with the -t flag, only"
- " tests of that type will be run.")
- parser.add_argument('--schema-file', type=str, help="Path to schema file. This may be needed for some tests.")
- parser.add_argument('--interpreter', type=str, default='tensorflow', choices=['tensorflow', 'tflite_runtime',
- 'tflite_micro'],
- help="Use interpreter from tensorflow or tflite_runtime. See README for more info.")
- return parser.parse_args()
- def load_testdata_sets(regenerate_input, regenerate_weights, regenerate_biases, schema_file, interpreter) -> dict:
- """
- Add all new testdata sets here
- """
- testdata_sets = {}
- type_of_test = 'conv'
- dataset = 'basic'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=1,
- out_ch=1,
- x_in=5,
- y_in=8,
- w_x=2,
- w_y=4,
- stride_x=1,
- stride_y=1,
- pad=False,
- interpreter=interpreter)
- dataset = 'stride2pad1'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=1,
- out_ch=1,
- x_in=7,
- y_in=7,
- w_x=3,
- w_y=3,
- stride_x=2,
- stride_y=2,
- pad=True,
- interpreter=interpreter)
- dataset = 'kernel1x1'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=19,
- out_ch=7,
- x_in=7,
- y_in=5,
- w_x=1,
- w_y=1,
- stride_x=1,
- stride_y=1,
- pad=False,
- bias_min=TestSettings.INT8_MIN,
- bias_max=TestSettings.INT8_MAX,
- out_activation_min=-126,
- out_activation_max=127,
- batches=2,
- interpreter=interpreter)
- dataset = 'kernel1x1_stride_x'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=9,
- out_ch=5,
- x_in=7,
- y_in=4,
- w_x=1,
- w_y=1,
- stride_x=3,
- stride_y=1,
- pad=False,
- out_activation_min=-126,
- out_activation_max=127,
- batches=2,
- interpreter=interpreter)
- dataset = 'kernel1x1_stride_x_y'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=23,
- out_ch=15,
- randmin=0,
- x_in=7,
- y_in=6,
- w_x=1,
- w_y=1,
- stride_x=2,
- stride_y=2,
- pad=False,
- out_activation_min=-6,
- out_activation_max=127,
- batches=3,
- interpreter=interpreter)
- dataset = 'kernel1x1_stride_x_y_1'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=4,
- w_x=1,
- w_y=1,
- stride_x=2,
- stride_y=2,
- pad=False,
- out_activation_min=-126,
- out_activation_max=127,
- batches=2,
- interpreter=interpreter)
- dataset = 'kernel1x1_stride_x_y_2'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=4,
- w_x=1,
- w_y=1,
- stride_x=3,
- stride_y=3,
- pad=False,
- out_activation_min=-126,
- out_activation_max=127,
- batches=2,
- interpreter=interpreter)
- dataset = 'conv_3'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=1,
- x_in=10,
- y_in=49,
- w_x=4,
- w_y=10,
- stride_x=1,
- stride_y=2,
- pad=True,
- out_activation_min=-127,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_1_x_n_1' # left and right pad, no non-padded elements
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=3,
- x_in=2,
- y_in=1,
- w_x=3,
- w_y=1,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-127,
- out_activation_max=127,
- batches=2,
- interpreter=interpreter)
- dataset = 'conv_1_x_n_2' # no pad
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=3,
- x_in=296,
- y_in=1,
- w_x=48,
- w_y=1,
- stride_x=2,
- stride_y=1,
- pad=False,
- out_activation_min=-111,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_1_x_n_3'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=1,
- x_in=296,
- y_in=1,
- w_x=48,
- w_y=1,
- stride_x=2,
- stride_y=1,
- pad=True,
- out_activation_min=-111,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_1_x_n_4' # 0 left pad, 1 right pad
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=4,
- x_in=16,
- y_in=1,
- w_x=3,
- w_y=1,
- stride_x=2,
- stride_y=1,
- pad=True,
- out_activation_min=-88,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_1_x_n_5'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=1,
- x_in=17,
- y_in=1,
- w_x=3,
- w_y=1,
- stride_x=3,
- stride_y=1,
- pad=True,
- out_activation_min=-88,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_2'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=4,
- x_in=6,
- y_in=3,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-101,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_4' # batches > 2
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=3,
- x_in=5,
- y_in=5,
- w_x=2,
- w_y=3,
- stride_x=2,
- stride_y=2,
- pad=False,
- out_activation_min=-109,
- out_activation_max=127,
- batches=3,
- interpreter=interpreter)
- dataset = 'conv_5'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=128,
- out_ch=1,
- x_in=128,
- y_in=1,
- w_x=3,
- w_y=3,
- stride_x=4,
- stride_y=4,
- pad=True,
- out_activation_min=-88,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'conv_out_activation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=2,
- x_in=3,
- y_in=3,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-61,
- out_activation_max=107,
- interpreter=interpreter)
- dataset = 'conv_dilation_golden'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=1,
- batches=2,
- out_ch=3,
- x_in=6,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-128,
- out_activation_max=127,
- dilation_x=3,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'conv_2x2_dilation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=10,
- y_in=10,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=False,
- out_activation_min=-61,
- out_activation_max=107,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'conv_2x3_dilation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=3,
- y_in=3,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-61,
- out_activation_max=107,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'conv_3x2_dilation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=3,
- y_in=3,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-61,
- out_activation_max=107,
- dilation_x=3,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'conv_2x2_dilation_5x5_input'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=5,
- y_in=5,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-61,
- out_activation_max=107,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'conv_3x3_dilation_5x5_input'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=9,
- y_in=11,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=True,
- out_activation_min=-61,
- out_activation_max=107,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'int16xint8'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=4,
- x_in=7,
- y_in=8,
- w_x=2,
- w_y=4,
- stride_x=2,
- stride_y=3,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-13335,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'requantize_s64'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=3,
- y_in=2,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=TestSettings.INT16_MAX,
- int16xint8=True,
- bias_min=-0x300,
- bias_max=0x9fff,
- interpreter=interpreter)
- dataset = 'int16xint8_dilation_1'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=32,
- y_in=32,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=TestSettings.INT16_MAX,
- int16xint8=True,
- bias_min=-0x300,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'int16xint8_dilation_2'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=4,
- x_in=7,
- y_in=8,
- w_x=2,
- w_y=4,
- stride_x=1,
- stride_y=1,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-13335,
- out_activation_max=32767,
- int16xint8=True,
- dilation_x=2,
- dilation_y=2,
- interpreter=interpreter)
- dataset = 'int16xint8_dilation_3'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=4,
- x_in=7,
- y_in=8,
- w_x=2,
- w_y=4,
- stride_x=1,
- stride_y=1,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-13335,
- out_activation_max=32767,
- int16xint8=True,
- dilation_x=2,
- interpreter=interpreter)
- type_of_test = 'depthwise_conv'
- dataset = 'depthwise_2'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=9,
- x_in=6,
- y_in=5,
- w_x=3,
- w_y=4,
- stride_x=2,
- stride_y=2,
- pad=True,
- out_activation_min=-73,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'depthwise_kernel_3x3'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=5,
- w_x=3,
- w_y=3,
- stride_x=2,
- stride_y=2,
- pad=True,
- bias_min=TestSettings.INT8_MIN,
- bias_max=TestSettings.INT8_MAX,
- out_activation_min=-104,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'depthwise_kernel_3x3_null_bias'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=5,
- w_x=3,
- w_y=3,
- stride_x=2,
- stride_y=2,
- pad=True,
- generate_bias=False,
- out_activation_min=-104,
- out_activation_max=127,
- interpreter=interpreter)
- dataset = 'depthwise_eq_in_out_ch'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=250,
- out_ch=250,
- x_in=7,
- y_in=5,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=True,
- interpreter=interpreter)
- dataset = 'depthwise_sub_block'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=9,
- out_ch=9,
- x_in=7,
- y_in=5,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- interpreter=interpreter)
- dataset = 'depthwise_x_stride'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=9,
- out_ch=9,
- x_in=7,
- y_in=5,
- w_x=2,
- w_y=2,
- stride_x=2,
- stride_y=1,
- pad=False,
- interpreter=interpreter)
- dataset = 'depthwise_out_activation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=3,
- x_in=6,
- y_in=5,
- w_x=3,
- w_y=4,
- pad=False,
- out_activation_min=-45,
- out_activation_max=103,
- interpreter=interpreter)
- dataset = 'depthwise_mult_batches'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=3,
- x_in=3,
- y_in=5,
- w_x=2,
- w_y=4,
- stride_x=2,
- stride_y=2,
- pad=True,
- batches=2,
- interpreter=interpreter)
- dataset = 'depthwise_null_bias_0'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=2,
- x_in=4,
- y_in=5,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=True,
- generate_bias=False,
- batches=1,
- interpreter=interpreter)
- dataset = 'depthwise_null_bias_1'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=16,
- x_in=4,
- y_in=5,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=True,
- generate_bias=False,
- batches=1,
- interpreter=interpreter)
- dataset = 'depthwise_dilation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=3,
- out_ch=9,
- x_in=6,
- y_in=5,
- w_x=3,
- w_y=4,
- stride_x=2,
- stride_y=2,
- pad=True,
- out_activation_min=-70,
- out_activation_max=127,
- dilation_x=2,
- dilation_y=3,
- interpreter=interpreter)
- dataset = 'dw_int16xint8'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=8,
- x_in=9,
- y_in=5,
- w_x=3,
- w_y=4,
- stride_x=3,
- stride_y=2,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-21111,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_dilation'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=4,
- out_ch=8,
- x_in=9,
- y_in=5,
- w_x=4,
- w_y=4,
- stride_x=1,
- stride_y=1,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-32700,
- dilation_x=3,
- dilation_y=2,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_mult4'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=2,
- out_ch=8,
- x_in=4,
- y_in=5,
- w_x=3,
- w_y=4,
- stride_x=3,
- stride_y=2,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-32767,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=4,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-17000,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_multiple_batches_uneven_buffers'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=5,
- y_in=5,
- w_x=3,
- w_y=3,
- stride_x=1,
- stride_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-17000,
- out_activation_max=32767,
- int16xint8=True,
- batches=3,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_multiple_batches_uneven_buffers_null_bias'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=4,
- y_in=4,
- w_x=3,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-17000,
- out_activation_max=32767,
- int16xint8=True,
- batches=3,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_test_bias'
- nbr_of_out_channels = 8
- bias = [i for i in range(nbr_of_out_channels)]
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=nbr_of_out_channels,
- x_in=4,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-17000,
- out_activation_max=32767,
- int16xint8=True,
- generate_bias=bias,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_null_bias'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=4,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=1,
- stride_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-17000,
- out_activation_max=32767,
- int16xint8=True,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_stride'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=4,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=2,
- stride_y=2,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- batches=2,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=16000,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_stride_null_bias'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=8,
- out_ch=8,
- x_in=4,
- y_in=4,
- w_x=2,
- w_y=2,
- stride_x=2,
- stride_y=2,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- batches=2,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=16000,
- int16xint8=True,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_spill'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=4,
- w_x=3,
- w_y=3,
- stride_x=2,
- stride_y=1,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- batches=3,
- out_activation_min=-30000,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'dw_int16xint8_fast_spill_null_bias'
- testdata_sets[dataset] = ConvSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=5,
- out_ch=5,
- x_in=4,
- y_in=4,
- w_x=3,
- w_y=3,
- stride_x=2,
- stride_y=1,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- batches=3,
- out_activation_min=-30000,
- out_activation_max=32767,
- int16xint8=True,
- generate_bias=False,
- interpreter=interpreter)
- type_of_test = 'fully_connected'
- dataset = 'fully_connected'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=10,
- out_ch=6,
- x_in=2,
- y_in=1,
- batches=3,
- interpreter=interpreter)
- dataset = 'fully_connected_mve_0'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=16,
- out_ch=9,
- x_in=1,
- y_in=1,
- batches=1,
- interpreter=interpreter)
- dataset = 'fully_connected_mve_1'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=20,
- out_ch=4,
- x_in=1,
- y_in=1,
- batches=1,
- interpreter=interpreter)
- dataset = 'fully_connected_null_bias_0'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=33,
- out_ch=5,
- batches=2,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'fully_connected_out_activation'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=10,
- out_ch=4,
- out_activation_min=-70,
- out_activation_max=100,
- interpreter=interpreter)
- dataset = 'fully_connected_int16'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=7,
- out_ch=11,
- x_in=3,
- y_in=3,
- batches=2,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-9999,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'fully_connected_int16_big'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=7,
- out_ch=11,
- x_in=10,
- y_in=10,
- batches=3,
- out_activation_min=-1444,
- out_activation_max=32767,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'fc_int16_slow'
- testdata_sets[dataset] = FullyConnectedSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- in_ch=7,
- out_ch=11,
- x_in=10,
- y_in=8,
- batches=3,
- randmin=(TestSettings.INT16_MAX - 100),
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- type_of_test = 'avgpool'
- dataset = 'avgpooling'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=20,
- x_in=22,
- y_in=12,
- stride_x=9,
- stride_y=5,
- w_x=6,
- w_y=5,
- pad=True,
- interpreter=interpreter)
- dataset = 'avgpooling_1'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=3,
- x_in=9,
- y_in=5,
- stride_x=1,
- stride_y=2,
- w_x=9,
- w_y=5,
- pad=False,
- interpreter=interpreter)
- dataset = 'avgpooling_2'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=5,
- x_in=12,
- y_in=1,
- stride_x=1,
- stride_y=2,
- w_x=3,
- w_y=1,
- pad=True,
- interpreter=interpreter)
- dataset = 'avgpooling_3'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=9,
- y_in=1,
- stride_x=2,
- stride_y=1,
- w_x=1,
- w_y=1,
- pad=False,
- interpreter=interpreter)
- dataset = 'avgpooling_4'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=1,
- y_in=20,
- stride_x=1,
- stride_y=3,
- w_x=1,
- w_y=3,
- pad=True,
- interpreter=interpreter)
- dataset = 'avgpooling_5'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=1,
- x_in=3,
- y_in=3,
- stride_x=1,
- stride_y=1,
- w_x=1,
- w_y=3,
- pad=True,
- relu6=True,
- interpreter=interpreter)
- dataset = 'avgpooling_int16'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=17,
- x_in=6,
- y_in=4,
- stride_x=2,
- stride_y=1,
- w_x=2,
- w_y=3,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'avgpooling_int16_1'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=9,
- y_in=1,
- stride_x=2,
- stride_y=1,
- w_x=1,
- w_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'avgpooling_int16_2'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=20,
- x_in=9,
- y_in=1,
- stride_x=2,
- stride_y=1,
- w_x=1,
- w_y=1,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'avgpooling_int16_3'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=21,
- x_in=1,
- y_in=20,
- stride_x=1,
- stride_y=3,
- w_x=1,
- w_y=3,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- type_of_test = 'maxpool'
- dataset = 'maxpooling'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=8,
- x_in=22,
- y_in=12,
- stride_x=9,
- stride_y=5,
- w_x=6,
- w_y=5,
- pad=True,
- interpreter=interpreter)
- dataset = 'maxpooling_1'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=3,
- x_in=9,
- y_in=5,
- stride_x=1,
- stride_y=2,
- w_x=9,
- w_y=5,
- pad=False,
- interpreter=interpreter)
- dataset = 'maxpooling_2'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=5,
- x_in=12,
- y_in=1,
- stride_x=1,
- stride_y=2,
- w_x=3,
- w_y=1,
- pad=True,
- interpreter=interpreter)
- dataset = 'maxpooling_3'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=9,
- y_in=1,
- stride_x=2,
- stride_y=1,
- w_x=1,
- w_y=1,
- pad=False,
- interpreter=interpreter)
- dataset = 'maxpooling_4'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=1,
- y_in=20,
- stride_x=1,
- stride_y=3,
- w_x=1,
- w_y=3,
- pad=True,
- interpreter=interpreter)
- dataset = 'maxpooling_5'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=20,
- x_in=1,
- y_in=1,
- stride_x=1,
- stride_y=1,
- w_x=1,
- w_y=1,
- pad=True,
- interpreter=interpreter)
- dataset = 'maxpooling_6'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=17,
- x_in=1,
- y_in=5,
- stride_x=1,
- stride_y=3,
- w_x=3,
- w_y=4,
- pad=True,
- interpreter=interpreter)
- dataset = 'maxpooling_7'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=1,
- x_in=4,
- y_in=2,
- stride_x=2,
- stride_y=2,
- w_x=2,
- w_y=2,
- pad=False,
- relu6=True,
- interpreter=interpreter)
- dataset = 'maxpool_int16'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=4,
- y_in=3,
- stride_x=2,
- stride_y=2,
- w_x=2,
- w_y=2,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'maxpool_int16_1'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=2,
- x_in=4,
- y_in=5,
- stride_x=2,
- stride_y=1,
- w_x=3,
- w_y=3,
- pad=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-30000,
- out_activation_max=30000,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'maxpool_int16_2'
- testdata_sets[dataset] = PoolingSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=3,
- x_in=7,
- y_in=7,
- stride_x=1,
- stride_y=1,
- w_x=3,
- w_y=3,
- pad=False,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-30000,
- out_activation_max=30000,
- int16xint8=True,
- interpreter=interpreter)
- type_of_test = 'softmax'
- dataset = 'softmax'
- testdata_sets[dataset] = SoftmaxSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- x_in=5,
- y_in=2,
- interpreter=interpreter)
- dataset = 'softmax_s16'
- testdata_sets[dataset] = SoftmaxSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- x_in=10,
- y_in=3,
- int16xint8=True,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- interpreter=interpreter)
- dataset = 'softmax_s8_s16'
- testdata_sets[dataset] = SoftmaxSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- x_in=12,
- y_in=2,
- inInt8outInt16=True,
- interpreter=interpreter)
- type_of_test = 'svdf'
- dataset = 'svdf'
- testdata_sets[dataset] = SVDFSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=2,
- number_inputs=2,
- rank=8,
- memory_size=8,
- input_size=3,
- number_units=3,
- interpreter=interpreter)
- dataset = 'svdf_1'
- testdata_sets[dataset] = SVDFSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=3,
- number_inputs=2,
- rank=1,
- memory_size=2,
- input_size=7,
- number_units=5,
- interpreter=interpreter)
- dataset = 'svdf_2'
- testdata_sets[dataset] = SVDFSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=3,
- number_inputs=2,
- rank=2,
- memory_size=2,
- input_size=7,
- number_units=5,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'svdf_3'
- testdata_sets[dataset] = SVDFSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=1,
- number_inputs=2,
- rank=1,
- memory_size=2,
- input_size=20,
- number_units=12,
- generate_bias=False,
- interpreter=interpreter)
- dataset = 'svdf_int8'
- testdata_sets[dataset] = SVDFSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=1,
- number_inputs=2,
- rank=1,
- memory_size=2,
- input_size=20,
- number_units=12,
- generate_bias=False,
- int8_time_weights=True,
- interpreter=interpreter)
- type_of_test = 'add'
- dataset = 'add'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=8,
- x_in=4,
- y_in=4,
- randmin=TestSettings.INT8_MIN,
- randmax=TestSettings.INT8_MAX,
- interpreter=interpreter)
- dataset = 'add_s16'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=8,
- x_in=4,
- y_in=4,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'add_s16_spill'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=7,
- x_in=5,
- y_in=3,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=-2000,
- out_activation_max=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- type_of_test = 'mul'
- dataset = 'mul'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=8,
- x_in=4,
- y_in=5,
- randmin=TestSettings.INT8_MIN,
- randmax=TestSettings.INT8_MAX,
- interpreter=interpreter)
- dataset = 'mul_s16'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=8,
- x_in=5,
- y_in=4,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=TestSettings.INT16_MAX,
- int16xint8=True,
- interpreter=interpreter)
- dataset = 'mul_s16_spill'
- testdata_sets[dataset] = AddMulSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- channels=7,
- x_in=5,
- y_in=7,
- randmin=TestSettings.INT16_MIN,
- randmax=TestSettings.INT16_MAX,
- out_activation_min=TestSettings.INT16_MIN,
- out_activation_max=1000,
- int16xint8=True,
- interpreter=interpreter)
- type_of_test = 'lstm'
- dataset = 'lstm_1'
- testdata_sets[dataset] = LSTMSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=1,
- time_steps=10,
- number_inputs=22,
- number_units=11,
- time_major=True,
- interpreter=interpreter)
- dataset = 'lstm_2'
- testdata_sets[dataset] = LSTMSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=2,
- time_steps=9,
- number_inputs=6,
- number_units=7,
- time_major=False,
- interpreter=interpreter)
- dataset = 'lstm_one_time_step'
- testdata_sets[dataset] = LSTMSettings(dataset,
- type_of_test,
- regenerate_weights,
- regenerate_input,
- regenerate_biases,
- schema_file,
- batches=3,
- time_steps=1,
- number_inputs=22,
- number_units=3,
- time_major=False,
- interpreter=interpreter)
- return testdata_sets
- def main():
- if version.parse(tf.__version__) < TestSettings.REQUIRED_MINIMUM_TENSORFLOW_VERSION:
- print("Unsupported tensorflow version, ", version.parse(tf.__version__))
- return 1
- args = parse_args()
- testdataset = args.dataset
- test_type = args.testtype
- schema_file = args.schema_file
- regenerate_input = args.regenerate_input
- regenerate_weights = args.regenerate_weights
- regenerate_biases = args.regenerate_biases
- if args.regenerate_all:
- regenerate_biases = True
- regenerate_weights = True
- regenerate_input = True
- testdata_sets = load_testdata_sets(regenerate_input,
- regenerate_weights,
- regenerate_biases,
- schema_file,
- args.interpreter)
- if args.run_all_testsets:
- for testset_name, testset_generator in testdata_sets.items():
- if test_type and testset_generator.test_type != test_type:
- continue
- print("Generating testset {}..".format(testset_name))
- testset_generator.generate_data()
- print()
- # Check that all testsets have been loaded.
- found_test_data_sets = []
- directory = 'TestCases/TestData'
- for dir in next(os.walk(directory))[1]:
- found_test_data_sets.append(dir)
- for testset_name in found_test_data_sets:
- if testset_name not in testdata_sets:
- print("WARNING: Testset {} in {} was not loaded".format(testset_name, directory))
- elif testdataset:
- try:
- generator = testdata_sets[testdataset]
- except KeyError:
- print("WARNING: testset {} not in testset list".format(testdataset))
- if test_type == 'conv' or test_type == 'depthwise_conv':
- generator = ConvSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'fully_connected':
- generator = FullyConnectedSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'avgpool' or test_type == 'maxpool':
- generator = PoolingSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'softmax':
- generator = SoftmaxSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'svdf':
- generator = SVDFSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'add' or test_type == 'mul':
- generator = AddMulSettings(testdataset, test_type, True, True, True, schema_file)
- elif test_type == 'lstm':
- generator = LSTMSettings(testdataset, test_type, True, True, True, schema_file)
- else:
- raise RuntimeError("Please specify type of test with -t")
- generator.generate_data()
- else:
- raise RuntimeError("Please select testdataset or use --run-all-testsets")
- return 0
- if __name__ == '__main__':
- sys.exit(main())
|
