ksdkamesh99

# Create a tflite model object from TFLite Converter

tfmodel = converter.convert()

# Save TFLite model into a .tflite file 

open("degree.tflite","wb").write(tfmodel)

ksdkamesh99

# We need to create a TFLite Converter Object from model we created

converter = tensorflow.lite.TFLiteConverter.from_keras_model(model=model)

ksdkamesh99

# Import tensorflow library

import tensorflow

#Save the model to saved_model.pbtxt

tensorflow.keras.models.save_model(model,'saved_model.pbtxt')

ksdkamesh99

# Import the tensorflow libraries which are used to create a sequential model and add a dense layer

import tensorflow as tf
import numpy as np
from tensorflow.python.keras.models import Sequential
from tensorflow.python.keras.layers import Dense

# Create random arrays x1,y1 as training data

x1=np.array([50,100,150,200,250,300,350])

ksdkamesh99

'''If we want to divide our data frame into 2 parts one belongs to setosa
and other belongs to other than setosa you need to add a boolean series with dataframe
equal to setosa and mention the names in compare_intra() function
'''

compare_intra_report=sv.compare_intra(train,train['species']=='Iris-setosa',['setosa','other'])

# If you want to display the report then use show_html function

compare_intra_report.show_html()

ksdkamesh99

#We use compare function to do EDA Comparision between 2 dataframes i.e train,test

compare_report=sv.compare(train,test)

# To display the analysis of this report you need to use show_html() from sweetviz library

compare_report.show_html()

ksdkamesh99

# analyze the train dataframe using analyze() function

report=sv.analyze(train)

# To display the analysis of this dataframe you need to use show_html() from sweetviz library

report.show_html()

ksdkamesh99

# Import sweetviz and pandas libraries

import sweetviz as sv
import pandas as pd

# Now read the dataset using pd.read_csv() function

data = pd.read_csv('IRIS.csv')

# Split first 125 entries/rows into train data frame

ksdkamesh99

%%time
y_pred_torch=model_torch.predict(np.array(x_test))

ksdkamesh99

%%time
model_torch.to('cuda')