I'm computing topic models through scikit-learn with this script (I'm starting with a dataset "df" which has one document per row in the column "Text")
from sklearn.decomposition import LatentDirichletAllocation
#Applying LDA
# the vectorizer object will be used to transform text to vector form
vectorizer = CountVectorizer(max_df=int(0.9*len(df)), min_df=int(0.01*len(df)), token_pattern='\w+|\$[\d\.]+|\S+')
# apply transformation
tf = vectorizer.fit_transform(df.Text).toarray()
# tf_feature_names tells us what word each column in the matric represents
tf_feature_names = vectorizer.get_feature_names()
number_of_topics = 6
model = LatentDirichletAllocation(n_components=number_of_topics, random_state=0)
model.fit(tf)
I'm interested in comparing models with different number of topics (kind of from 2 to 20 topics) through a coherence measure. How can I do it?
My idea is to extract the CLS token for all the text in the DB and save it in CSV or somewhere else. So when a new text comes in, instead of using the Cosine Similarity/JAccard/MAnhattan/Euclidean or other distances, I have to use some approximation like LSH, ANN (ANNOY, sklearn.neighbor) or the one given here faiss . How can that be done? I have my code as:
PyTorch:
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
model = BertModel.from_pretrained('bert-base-uncased')
input_ids = torch.tensor(tokenizer.encode("Hello, I am a text")).unsqueeze(0) # Batch size 1
outputs = model(input_ids)
last_hidden_states = outputs[0] # The last hidden-state is the first element of the output tuple
Using Tensorflow:
import tensorflow as tf
from transformers import BertTokenizer, TFBertModel
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
model = TFBertModel.from_pretrained('bert-base-uncased')
input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :] # Batch size 1
outputs = model(input_ids)
last_hidden_states = outputs[0] # The last hidden-state is the first element of the output tuple
and I think can get the CLS token as: (Please correct if wrong)
last_hidden_states = outputs[0]
cls_embedding = last_hidden_states[0][0]
Please tell me if it's the right way to use and how can I use any of the LSH, ANNOT, faiss or something like that?
So for every text, there'll a 768 length vector and we can create a N(No of texts 10M)x768 matrix, how can I find the Index of top-5 data points (texts) which are most similar to the given image/embedding/data point?
The code below as an example for analyzing massive corpus. I want to restrict the term-document matrix to 1000 most frequent unigrams, but changing the max-features parameter to n only return the first n unigrams. Any suggestion?
from sklearn.feature_extraction.text import TfidfVectorizer
import pandas as pd
corpus = ['Hi my name is Joe.', 'Hi my name is Donald.']
vectorizer = TfidfVectorizer(max_features=3)
X = vectorizer.fit_transform(corpus).todense()
df = pd.DataFrame(X, columns=vectorizer.get_feature_names())
df.to_csv('test.csv')
I am assuming that this is a problem in your example, but according to the sklearn documentation for TfidfVectorizer ssays the following for max_features:
If not None, build a vocabulary that only consider the top max_features ordered by term frequency across the corpus.
It might be that the first n terms are considered when there are words of equal frequency, but otherwise it should return the correct result. If it still does not work, I strongly suggest to open a bug report in the sklearn repository. However, you can also manually construct a vocabulary yourself (with your own interpretation of "frequency", by setting the vocabulary option:
vocabulary: Mapping or iterable, default=None
Either a Mapping (e.g., a dict) where keys are terms and values are indices in the feature matrix, or an iterable over terms. If not given, a vocabulary is determined from the input documents.
I have a csv with both categorical and float dtypes. I want to do the following:
For each categorical column i will use pandas to compute the unique values (pd.unique()) that are present in the column. say u_l for a column
I will use the len(u_l) to decide upon the dimension of embeddings that i use for a particular categorical column that i want i embed (this step is the reason i cannot use tensorflow_transform)
I want to create some stateful node that can map category (token) value to embeddings index thus subsequently i can lookup the embedding from embeddings matrix that i created in step 2
I dont know how to go about doing it currently. A very inelegant solution i can see is using tensorflow_datasets:
encoder = tfds.features.text.TokenTextEncoder(u_l,decode_token_separator=' ')
concatenate the entire column using space delimiter (c_l) (c_l is one string now) and then using encoder.encode(c_l)
This is a very basic thing that i think tensorflow would be able to do relatively easily. Please guide me to the right solution
If you want to use your word corpus as embedding like if you have corpus as this :
corpus :
"This pasta is good"
"This pasta is very good"
and you want to use embedding you can use Tokenizer of TF see this. It will create a dict containing words as keys and index as value like in above corpus dict looks like :
word_index = {"this" : 1, "pasta" : 2, "good" : 3, "very" : 4}
you can avoid stopwords.
Now you can make word embedding vector using these word_index dict so that it looks like
For corpus 1 : [1, 2, 3]
For corpus 2 : [1, 2, 4, 3]
Enough talk let see some code : Also define oov_token for out of vocabulary words.
You can do like this :
vocab_size = 10000
embedding_dim = 16
max_length = 120
trunc_type='post'
oov_tok = "<OOV>"
from tensorflow.keras.preprocessing.text import Tokenizer
tokenizer = Tokenizer(num_words = vocab_size, oov_token=oov_tok)
tokenizer.fit_on_texts(training_sentences)
word_index = tokenizer.word_index
sequences = tokenizer.texts_to_sequences(training_sentences) # This will create word embedding vector
padded = pad_sequences(sequences,maxlen=max_length, truncating=trunc_type) # This will padd zeros according to `trunc_type`, here add zeros in last
testing_sequences = tokenizer.texts_to_sequences(testing_sentences)
testing_padded = pad_sequences(testing_sequences,maxlen=max_length)
Also see this GitHub code of me hope it will help
I have read the CNN Tutorial on the TensorFlow and I am trying to use the same model for my project.
The problem is now in data reading. I have around 25000 images for training and around 5000 for testing and validation each. The files are in png format and I can read them and convert them into the numpy.ndarray.
The CNN example in the tutorials use a queue to fetch the records from the file list provided. I tried to create my own such binary file by reshaping my images into 1-D array and attaching a label value in the front of it. So my data looks like this
[[1,12,34,24,53,...,105,234,102],
[12,112,43,24,52,...,115,244,98],
....
]
The single row of the above array is of length 22501 size where the first element is the label.
I dumped the file to using pickle and the tried to read from the file using the
tf.FixedLengthRecordReader to read from the file as demonstrated in example
I am doing the same things as given in the cifar10_input.py to read the binary file and putting them into the record object.
Now when I read from the files the labels and the image values are different. I can understand the reason for this to be that pickle dumps the extra information of braces and brackets also in the binary file and they change the fixed length record size.
The above example uses the filenames and pass it to a queue to fetch the files and then the queue to read a single record from the file.
I want to know if I can pass the numpy array as defined above instead of the filenames to some reader and it can fetch records one by one from that array instead of the files.
Probably the easiest way to make your data work with the CNN example code is to make a modified version of read_cifar10() and use it instead:
Write out a binary file containing the contents of your numpy array.
import numpy as np
images_and_labels_array = np.array([[...], ...], # [[1,12,34,24,53,...,102],
# [12,112,43,24,52,...,98],
# ...]
dtype=np.uint8)
images_and_labels_array.tofile("/tmp/images.bin")
This file is similar to the format used in CIFAR10 datafiles. You might want to generate multiple files in order to get read parallelism. Note that ndarray.tofile() writes binary data in row-major order with no other metadata; pickling the array will add Python-specific metadata that TensorFlow's parsing routines do not understand.
Write a modified version of read_cifar10() that handles your record format.
def read_my_data(filename_queue):
class ImageRecord(object):
pass
result = ImageRecord()
# Dimensions of the images in the dataset.
label_bytes = 1
# Set the following constants as appropriate.
result.height = IMAGE_HEIGHT
result.width = IMAGE_WIDTH
result.depth = IMAGE_DEPTH
image_bytes = result.height * result.width * result.depth
# Every record consists of a label followed by the image, with a
# fixed number of bytes for each.
record_bytes = label_bytes + image_bytes
assert record_bytes == 22501 # Based on your question.
# Read a record, getting filenames from the filename_queue. No
# header or footer in the binary, so we leave header_bytes
# and footer_bytes at their default of 0.
reader = tf.FixedLengthRecordReader(record_bytes=record_bytes)
result.key, value = reader.read(filename_queue)
# Convert from a string to a vector of uint8 that is record_bytes long.
record_bytes = tf.decode_raw(value, tf.uint8)
# The first bytes represent the label, which we convert from uint8->int32.
result.label = tf.cast(
tf.slice(record_bytes, [0], [label_bytes]), tf.int32)
# The remaining bytes after the label represent the image, which we reshape
# from [depth * height * width] to [depth, height, width].
depth_major = tf.reshape(tf.slice(record_bytes, [label_bytes], [image_bytes]),
[result.depth, result.height, result.width])
# Convert from [depth, height, width] to [height, width, depth].
result.uint8image = tf.transpose(depth_major, [1, 2, 0])
return result
Modify distorted_inputs() to use your new dataset:
def distorted_inputs(data_dir, batch_size):
"""[...]"""
filenames = ["/tmp/images.bin"] # Or a list of filenames if you
# generated multiple files in step 1.
for f in filenames:
if not gfile.Exists(f):
raise ValueError('Failed to find file: ' + f)
# Create a queue that produces the filenames to read.
filename_queue = tf.train.string_input_producer(filenames)
# Read examples from files in the filename queue.
read_input = read_my_data(filename_queue)
reshaped_image = tf.cast(read_input.uint8image, tf.float32)
# [...] (Maybe modify other parameters in here depending on your problem.)
This is intended to be a minimal set of steps, given your starting point. It may be more efficient to do the PNG decoding using TensorFlow ops, but that would be a larger change.
In your question, you specifically asked:
I want to know if I can pass the numpy array as defined above instead of the filenames to some reader and it can fetch records one by one from that array instead of the files.
You can feed the numpy array to a queue directly, but it will be a more invasive change to the cifar10_input.py code than my other answer suggests.
As before, let's assume you have the following array from your question:
import numpy as np
images_and_labels_array = np.array([[...], ...], # [[1,12,34,24,53,...,102],
# [12,112,43,24,52,...,98],
# ...]
dtype=np.uint8)
You can then define a queue that contains the entire data as follows:
q = tf.FIFOQueue([tf.uint8, tf.uint8], shapes=[[], [22500]])
enqueue_op = q.enqueue_many([image_and_labels_array[:, 0], image_and_labels_array[:, 1:]])
...then call sess.run(enqueue_op) to populate the queue.
Another—more efficient—approach would be to feed records to the queue, which you could do from a parallel thread (see this answer for more details on how this would work):
# [With q as defined above.]
label_input = tf.placeholder(tf.uint8, shape=[])
image_input = tf.placeholder(tf.uint8, shape=[22500])
enqueue_single_from_feed_op = q.enqueue([label_input, image_input])
# Then, to enqueue a single example `i` from the array.
sess.run(enqueue_single_from_feed_op,
feed_dict={label_input: image_and_labels_array[i, 0],
image_input: image_and_labels_array[i, 1:]})
Alternatively, to enqueue a batch at a time, which will be more efficient:
label_batch_input = tf.placeholder(tf.uint8, shape=[None])
image_batch_input = tf.placeholder(tf.uint8, shape=[None, 22500])
enqueue_batch_from_feed_op = q.enqueue([label_batch_input, image_batch_input])
# Then, to enqueue a batch examples `i` through `j-1` from the array.
sess.run(enqueue_single_from_feed_op,
feed_dict={label_input: image_and_labels_array[i:j, 0],
image_input: image_and_labels_array[i:j, 1:]})
I want to know if I can pass the numpy array as defined above instead
of the filenames to some reader and it can fetch records one by one
from that array instead of the files.
tf.py_func, that wraps a python function and uses it as a TensorFlow operator, might help. Here's an example.
However, since you've mentioned that your images are stored in png files, I think the simplest solution would be to replace this:
reader = tf.FixedLengthRecordReader(record_bytes=record_bytes)
result.key, value = reader.read(filename_queue)
with this:
result.key, value = tf.WholeFileReader().read(filename_queue))
value = tf.image.decode_jpeg(value)