Скачать или смотреть KERAS preprocessing Layers (2/3)

Create your TensorFlow Data Pipeline for Text preprocessing and work w/ latest KERAS preprocessing Layers.

Learn the latest TensorFlow functions and KERAS preprocessing layers to parallelize preprocessing your input data and training your Word2vec model (Part 2 of 3).

To perform efficient batching for the potentially large number of training examples, use the tf.data.Dataset API. The official KERAS /TF2 Jupyter NB.

Use the Keras Subclassing API and accelerate your TF model on CPU / GPU significantly. Code in real time with me your first optimized Word2Vec model in pure TensorFlow (no Cython at all, d.h. no gensim). Learn TensorFlow on a familiar DEMO. With learn-in-parallel Colab /Jupyter NB provided.

TensorFlow tutorial on Word2Vec w/ TF 2.7:
https://www.tensorflow.org/tutorials/...

COLAB / Jupyter NB (official code from KERAS, TensorFlow):
https://colab.research.google.com/git...

KERAS preprocessing Layer:
https://keras.io/guides/preprocessing...

#bert
#sbert
#tensorflow2
#keras
#3dvisualization
#word2vec
#cosinesimilarity
#vectorspace
#similarity
#nlproc
#deeplearning
#machinelearningwithpython
#jupyterlab
#pythonprogramming

00:00 Word2Vec TF code / tutorial
00:44 Create TF Data Pipeline for text preprocessing with KERAS
06:45 Real time code execution of a Word2Vec DEMO project
10:11 TextVectorization
15:45 Our optimized Data for Word2Vec model
18:27 KERAS subclassing API
23:10 Our Word2Vec model in pure TF 2.7 code
25:37 Train our Word2Vec model with 25 epochs (watch real time)
25:49 Result