joeyism

import tensorflow_hub as hub
import tensorflow as tf

words_to_embed = ["dog", "cat", "sloth"] 

elmo = hub.Module("https://tfhub.dev/google/elmo/2")
embedding_tensor = elmo(words_to_embed) # <-- removed other params

with tf.Session() as sess:
  sess.run(tf.global_variables_initializer())

joeyism

import tensorflow_hub as hub
import tensorflow as tf

words_to_embed = ["dog", "cat", "sloth"] # <-- it's a list now, and the name changed

elmo = hub.Module("https://tfhub.dev/google/elmo/2")
embedding_tensor = elmo(words_to_embed, as_dict=True)["word_emb"] # <-- passing in a list instead of [word]

with tf.Session() as sess:
  sess.run(tf.global_variables_initializer())

joeyism

import tensorflow_hub as hub
import tensorflow as tf

word_to_embed = "dog"

elmo = hub.Module("https://tfhub.dev/google/elmo/2")
embedding_tensor = elmo([word_to_embed], as_dict=True)["word_emb"] # Use as_dict because I want the whole dict, then I select "word_emb"

with tf.Session() as sess:
  sess.run(tf.global_variables_initializer())

joeyism

from cifar import Cifar
from tqdm import tqdm
import tensornets as nets
import tensorflow as tf
import numpy as np
import helper

learning_rate = 0.00001
batch_size = 16
no_of_epochs = 100

joeyism

import tensornets as nets
import tensorflow as tf

learning_rate = 0.00001
n_classes = 10
image_size = 224

inputs = tf.placeholder(tf.float32, [None, image_size, image_size, 3])
outputs = tf.placeholder(tf.float32, [None, n_classes])

joeyism

from cifar import Cifar
from tqdm import tqdm
import pretrained
import numpy as np
import tensorflow as tf
import helper


n_classes = 10
learning_rate = 0.00001

joeyism

from cifar import Cifar
from tqdm import tqdm
import pretrained
import tensorflow as tf
import helper


n_classes = 10
learning_rate = 0.00001
batch_size = 16

joeyism

import pretrained
import tensorflow as tf


n_classes = 10
learning_rate = 0.00001
batch_size = 16

conv5 = tf.layers.flatten(pretrained.maxpool5)

joeyism

import pretrained
import tensorflow as tf


n_classes = 10
learning_rate = 0.00001

conv5 = tf.layers.flatten(pretrained.maxpool5)

weights = tf.Variable(tf.zeros([9216, n_classes]), name="output_weight")

joeyism

import numpy as np
import pickle
import os
import math


def __extract_file__(fname):
    with open(fname, 'rb') as fo:
        d = pickle.load(fo, encoding='bytes')
    return d