KaneAI - World’s First E2E Software Testing Agent.Book a Demo Join TestMu Conference

How to use _build_combination method in Slash

Best Python code snippet using slash

ModelHelper.py

Source:ModelHelper.py

...29            self._build_model_vq()30        elif self._settings.identifier == 'vanilla':31            self._build_model_vanilla()32        elif self._settings.identifier == 'combination':33            self._build_combination()34            self._input_tensor = self._misc.set_probabilities('model_checkpoint/vq_graph/p_s_m.mat')35        # elif self.settings.identifier == 'restore':36        #     self._build_combination()37    def _build_model_vq(self):38        """39        Build your model for training. All of the architecture40        is defined here.41        :param scale_softmax: Scaling for the WTA-layer (Winner-Takes-All)42        :param codebook_size: Size of the codebook43        """44        # Start here to define your network45        # ------------------------------------------------------------------46        num_neurons = 51247        activation = tf.nn.relu48        regulizer = None49        if self._settings.l2_loss_reg:50            regulizer = tf.contrib.layers.l2_regularizer(scale=self._settings.scale_l2)51        with tf.variable_scope('base_network'):52            fc = tf.layers.dense(self.features, num_neurons, activation=activation, kernel_regularizer=regulizer)53            # fc1_bn = tf.layers.batch_normalization(fc1, training=self.train, center=False, scale=False)54            fc = tf.layers.batch_normalization(fc, training=self.train)55            # fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)56            fc = tf.layers.dense(fc, num_neurons, activation=activation, kernel_regularizer=regulizer)57            # fc2_bn = tf.layers.batch_normalization(fc2, training=self.train, center=False, scale=False)58            fc = tf.layers.batch_normalization(fc, training=self.train)59            # fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)60            fc = tf.layers.dense(fc, num_neurons, activation=activation, kernel_regularizer=regulizer)61            # fc3_bn = tf.layers.batch_normalization(fc3, training=self.train, center=False, scale=False)62            fc = tf.layers.batch_normalization(fc, training=self.train)63            # fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)64            #65            fc = tf.layers.dense(fc, num_neurons, activation=activation, kernel_regularizer=regulizer)66            # fc4_bn = tf.layers.batch_normalization(fc4, training=self.train, center=False, scale=False)67            fc = tf.layers.batch_normalization(fc, training=self.train)68            fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)69            #70            # fc = tf.layers.dense(fc, num_neurons, activation=activation)71            # # # fc3_bn = tf.layers.batch_normalization(fc3, training=self.train, center=False, scale=False)72            # fc = tf.layers.batch_normalization(fc, training=self.train)73            # fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)74            # # #75            # fc = tf.layers.dense(fc, num_neurons, activation=activation)76            # # # fc4_bn = tf.layers.batch_normalization(fc4, training=self.train, center=False, scale=False)77            # fc = tf.layers.batch_normalization(fc, training=self.train)78            # fc = tf.layers.dropout(fc, rate=self._settings.do_rate, training=self.train)79            # # WTA-layer starts here80            out = tf.layers.dense(fc, self._settings.codebook_size, activation=tf.nn.sigmoid,81                                  kernel_regularizer=regulizer)82            out_scaled = tf.scalar_mul(self._settings.scale_soft, out)83            # output without softmax84            self.logits = out_scaled85            # output with soft, be aware use a name 'nn_output' for the output node!86            self.inference = tf.nn.softmax(self.logits, name='nn_output')87            # Low rank matrix factorization88            # (c.f. LOW-RANK MATRIX FACTORIZATION FOR DEEP NEURAL NETWORK TRAINING WITH89            # HIGH-DIMENSIONAL OUTPUT TARGETS)90            # out_1 = tf.layers.dense(fc4_bn, 512, activation=None)91            # out = tf.layers.dense(out_1, self._settings.codebook_size, activation=None)92            # out_scaled = tf.scalar_mul(self._settings.scale_soft, out)93            # # output without softmax94            # self.logits = out_scaled95            # # output with soft, be aware use a name 'nn_output' for the output node!96            # self.inference = tf.nn.softmax(self.logits, name='nn_output')97        # ------------------------------------------------------------------98        # end of definition of network99    def _build_model_vanilla(self):100        num_neurons = 512101        with tf.variable_scope('vanilla_network'):102            fc1 = tf.layers.dense(self.features, num_neurons, activation=tf.nn.relu)103            fc1_bn = tf.layers.batch_normalization(fc1, training=self.train, center=False, scale=False)104            fc1_dropout = tf.layers.dropout(fc1_bn, rate=0.25, training=self.train)105            fc2 = tf.layers.dense(fc1_dropout, num_neurons, activation=tf.nn.relu)106            fc2_bn = tf.layers.batch_normalization(fc2, training=self.train, center=False, scale=False)107            fc2_dropout = tf.layers.dropout(fc2_bn, rate=0.25, training=self.train)108            #109            fc3 = tf.layers.dense(fc2_dropout, num_neurons, activation=tf.nn.relu)110            fc3_bn = tf.layers.batch_normalization(fc3, training=self.train, center=False, scale=False)111            fc3_dropout = tf.layers.dropout(fc3_bn, rate=0.25, training=self.train)112            fc4 = tf.layers.dense(fc3_dropout, num_neurons, activation=tf.nn.relu)113            fc4_bn = tf.layers.batch_normalization(fc4, training=self.train, center=False, scale=False)114            fc4_dropout = tf.layers.dropout(fc4_bn, rate=0.25, training=self.train)115            fc5 = tf.layers.dense(fc4_dropout, num_neurons, activation=tf.nn.relu)116            fc5_bn = tf.layers.batch_normalization(fc5, training=self.train, center=False, scale=False)117            fc5_dropout = tf.layers.dropout(fc5_bn, rate=0.25, training=self.train)118            self.logits = tf.layers.dense(fc5_dropout, self._settings.num_labels, activation=None)119            # output with soft, be aware use a name 'nn_output' for the output node!120            self.inference = tf.nn.softmax(self.logits, name='nn_output')121    def _build_combination(self):122        # we combine the nnvq and vanilla network123        num_neurons = 512124        # with tf.variable_scope('scaling_network'):125        #     fc1_scale = tf.layers.dense(self.features, 1, activation=tf.nn.sigmoid)126        #     self.scale = 35 * fc1_scale127        #     # self.scale = tf.Print(self.scale, [tf.reduce_min(self.scale)])128        # first we build the nnvq network129        with tf.variable_scope('base_network'):130            fc1 = tf.layers.dense(self.features, num_neurons, activation=tf.nn.relu)131            fc1_bn = tf.layers.batch_normalization(fc1, training=self.train, center=False, scale=False)132            fc1_dropout = tf.layers.dropout(fc1_bn, rate=0.25, training=self.train)133            fc2 = tf.layers.dense(fc1_dropout, num_neurons, activation=tf.nn.relu)134            fc2_bn = tf.layers.batch_normalization(fc2, training=self.train, center=False, scale=False)135            fc2_dropout = tf.layers.dropout(fc2_bn, rate=0.25, training=self.train)...

cartesian.py

Source:cartesian.py

...22        return returned23    def check(self, iterator):24        names = list(self.sets)25        sets = [self.sets[name] for name in names]26        expected = sorted((self._build_combination(names, combination) for combination in itertools.product(*sets)), key=lambda d: sorted(d.items()))27        got = sorted(iterator, key=lambda d: sorted(d.items()))28        assert got == expected29    def _build_combination(self, names, combination):30        returned = {}31        for name, value in zip(names, combination):32            returned[name] = value33        for assign_source, assign_target in self._assigns:34            returned[assign_target] = returned[assign_source]35        return returned36class SetMaker(object):37    def __init__(self, cartesian, name):38        super(SetMaker, self).__init__()39        self.cartesian = cartesian40        self.name = name41    def make_set(self, size=3):42        assert self.name not in self.cartesian.sets43        returned = self.cartesian.sets[self.name] = ["{}{}".format(self.name, i) for i in range(size)]...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.