negedng

model = get_model()
model.compile("adam", "binary_crossentropy", metrics=["accuracy"])
model.fit(ds1_train, epochs=3, validation_data=ds1_val)

negedng

shapes=([200],())
ds1_train = (
    ds1_train
    .map(encode_map_fn)
    .map(truncate)
    .cache()
    .shuffle(1024)
    .padded_batch(32, shapes)
    .prefetch(tf.data.experimental.AUTOTUNE)
)

negedng

from collections import Counter
# Building vocabulary set for tokenizer
tokenizer = tfds.features.text.Tokenizer()

vocabulary_counter = Counter()
for text_tensor, _ in ds1_train:
  some_tokens = tokenizer.tokenize(text_tensor.numpy())
  vocabulary_counter.update(some_tokens)

vocabulary = vocabulary_counter.most_common(max_features-2)

negedng

(ds1_train, ds1_val), ds1_info = tfds.load('imdb_reviews',
                                           split=['train', 'test'],
                                           as_supervised=True,
                                           with_info=True)

negedng

model.compile("adam", "binary_crossentropy", metrics=["accuracy"])
model.fit(x_train, y_train, batch_size=32, epochs=3, validation_data=(x_val, y_val))

negedng

(x_train, y_train), (x_val, y_val) = keras.datasets.imdb.load_data(
    num_words=max_features
)

x_train = keras.preprocessing.sequence.pad_sequences(x_train, maxlen=maxlen)
x_val = keras.preprocessing.sequence.pad_sequences(x_val, maxlen=maxlen)

negedng

import numpy as np
from tensorflow import keras
from tensorflow.keras import layers

max_features = 20000  # Only consider the top 20k words
maxlen = 200  # Only consider the first 200 words of each movie review

def get_model():
    # Input for variable-length sequences of integers
    inputs = keras.Input(shape=(None,), dtype="int32")

negedng

import tensorflow as tf
from tensorflow.keras.applications.imagenet_utils import decode_predictions
from skimage import io
from skimage.transform import resize
import numpy as np

image_url = 'https://upload.wikimedia.org/wikipedia/commons/b/bb/Kittyply_edit1.jpg'
img = io.imread(image_url)
img_resized = resize(img, (224,224,3))

negedng

from keras.layers import Layer
import tensorflow as tf

class LabelLimitLayer(Layer):

    def __init__(self, labels, topn, **kwargs):
        self.labels = labels
        self.topn = topn
        super(LabelLimitLayer, self).__init__(**kwargs)

negedng

def lin_map_matrix(FV, W):
    """For ß basis in V and γ basis in W,
    returns with the [T]^γ_ß matrix of the transformation T,
    which transforms v in V to fv in FV"""
    return np.transpose([np.linalg.solve(np.transpose(W), fv) for fv in FV])


def linmap_predict(v, M, V, W):
    return np.matmul(np.transpose(W),np.matmul(M,np.linalg.solve(np.transpose(V),v)))