How to use get_skips method in Slash

Best Python code snippet using slash

keras_model.py

Source:keras_model.py

...18        self.embedding_size = embedding_size19        self.window_size = window_size20        self.model = None21        self.embeddings = None22    def get_skips(self, docs):23        """24        Formats the data and generates negative samples.25        :param docs: list; a list of documents; each document is a list of sentences;26        a sentence is a list of tokens (strings)27        :return: tuple; contains the center and context words, and the corresponding labels28        """29        sampling_table = make_sampling_table(self.vocab_size)30        center_words, context_words, labels = [], [], []31        for doc in docs:32            tokens = [token for sent in doc for token in sent]33            pairs, labels_ = skipgrams(tokens,34                                        self.vocab_size,35                                        window_size=self.window_size,36                                        sampling_table=sampling_table)37            try:38                center, context = zip(*pairs)39            except ValueError:40                continue41            center_words += center42            context_words += context43            labels += labels_44        return center_words, context_words, labels45    def w2v_model(self, learning_rate):46        """47        Generates the neural architecture for the word2vec skip-gram model48        :return: keras.models.Model(); the word2vec model49        """50        # Add the input and embedding layers51        input_center = Input((1,))52        input_context = Input((1,))53        self.embeddings = Embedding(self.vocab_size, self.embedding_size, input_length=1, name="Embeddings")54        # Get the center and context embeddings55        center = self.embeddings(input_center)56        center = Reshape((self.embedding_size, 1))(center)57        context = self.embeddings(input_context)58        context = Reshape((self.embedding_size, 1))(context)59        # Calculate the linear activations60        # dot_product = Concatenate([center, context], mode="dot", dot_axes=1)61        dot_product = dot([center, context], axes=1, normalize=False)62        dot_product = Reshape((1,))(dot_product)63        # Sigmoid activations64        output = Dense(1, activation="sigmoid")(dot_product)65        # Define the model66        model = Model(input=[input_center, input_context], output=output)67        optimizer = RMSprop(lr=learning_rate, rho=0.9, epsilon=None, decay=0.0)68        model.compile(loss="binary_crossentropy", optimizer=optimizer)69        return model70    def train(self, docs, num_batches=2000, learning_rate=0.001, verbose=True):71        """72        Optimizes the model on the training data73        :param docs: list; a sequence of documents; each document is a list of sentences;74        a sentence is a list of tokens (strings)75        :param num_batches: int; the number of (center, context) pairs to use in training76        :param verbose: Boolean; if true, prints the loss druing training77        """78        # Get the data and the model79        center_words, context_words, labels = self.get_skips(docs)80        self.model = self.w2v_model(learning_rate)81        # Randomly sample pair/label82        loss = []83        for batch in range(num_batches):84            idx = np.random.randint(0, len(center_words)-1)85            center_word = np.array([center_words[idx]])86            context_word = np.array([context_words[idx]])87            label = np.array([labels[idx]])88            loss += [self.model.train_on_batch([center_word, context_word], label)]89            # Print the loss every 1000 batches90            if len(loss) >= 1000 and verbose:91                print(batch, sum(loss)/1000)92                loss = []93    def get_embedding_array(self):...

analogy.py

Source:analogy.py

...43    embeds = Format(args.format).load(args.embeddings, args.lossy, args.mmap)44    with open(args.input) as queries:45        for query in queries:46            query_a, query_b, query_c = query.strip().split()47            skips = get_skips(query_a, query_b, query_c, args.include)48            res = embeds.analogy(query_a,49                                 query_b,50                                 query_c,51                                 k=args.k,52                                 skip=skips)53            if res is None:54                print(55                    f"Could not compute for: {query_a} : {query_b}, {query_c} : ? ",56                    file=sys.stderr)57            else:58                print("\n".join(f"{ws.word} {ws.similarity}" for ws in res))59def get_skips(  # pylint: disable=missing-function-docstring60        query_a: str, query_b: str, query_c: str,61        includes: List[str]) -> Set[str]:62    if includes == []:63        return {query_c, query_b, query_a}64    skips = set()65    if 'a' not in includes:66        skips.add(query_a)67    if 'b' not in includes:68        skips.add(query_b)69    if 'c' not in includes:70        skips.add(query_b)71    return skips72if __name__ == '__main__':73    main()

aoc10.py

Source:aoc10.py

...36            difs[3] += 137            break38    return difs[1] * difs[3]3940def get_skips(arr):41    arr.append(0)42    arr.append(max(arr) + 3)43    arr.sort()44    result = []45    group_len = 046    for i in range(1, len(arr) - 1):47        if arr[i] - arr[i - 1] == 3 or arr[i + 1] - arr[i] == 3:48            if group_len > 0:49                if group_len == 1:50                    result.append(1)51                elif group_len == 2:52                    result.append(3)53                elif group_len == 3:54                    result.append(6)55                group_len = 056        else:57            group_len += 158    return result5960def part2(adapters):61    skips = get_skips(adapters)62    result = 163    for choose in range(1, len(skips) + 1):64        choices = list(itertools.combinations(skips, choose))65        possibilities = list(map(math.prod, choices))66        result += sum(possibilities)67    return result6869def main():70    data = get_data('aoc10.txt')7172    # Part 173    print(f'PART 1--Answer: {part1(data)}')7475    # Part 2
...

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.