Muiruriscode / Spam-detector-using-python

Tensorflow Spam Detector model

Install tflite-model-maker

Install the package to create the spam detection model

!pip install -q tflite-model-maker

Import dependencies

import numpy as np
import os

from tflite_model_maker import configs
from tflite_model_maker import ExportFormat
from tflite_model_maker import model_spec
from tflite_model_maker import text_classifier
from tflite_model_maker.text_classifier import DataLoader

Get Data From csv file

Fetch data and store it in a file named comment-spam=extras.csv. Give the url to fetch data from and assign extract to false since dta is not zipped.

data_file = tf.keras.utils.get_file(fname='comment-spam-extras.csv',
                                    origin='https://storage.googleapis.com/jmstore/TensorFlowJS/EdX/code/6.5/jm_blog_comments_extras.csv',
                                    extract=False
                                    )

Model Specifications

Assign specifications to be used to train the modekl

# use average_word_evec model
spec = model_spec.get('average_word_vec')
# use 2000 words for training
spec.num_words = 2000
# Give the length to be used per token
spec.seq_len = 20
# Give the dimensions used for training
spec.wordvec_dim = 7

data = DataLoader.from_csv(
    filename=data_file,
    text_column='commenttext',
    label_column='spam',
    model_spec=spec,
    delimiter=',',
    shuffle=True,
    is_training=True
)

Train the Data

# Use 90% for training and 10% for testing
train_data, test_data = data.split(0.9)

model = text_classifier.create(train_data, model_spec=spec, epochs=50)

Export the Model

Export the data to /tmp/js_export in a tfjs format vith LABEL and VOCAB included

model.export(export_dir='/tmp/js_export', export_format=[ExportFormat.TFJS, ExportFormat.LABEL, ExportFormat.VOCAB])

Zip the Model for download

!zip -r /tmp/js_export/ModelFiles.zip /tmp/js_export/