Save tokenizer ,model and load

 Hi all

-----------------------------------------

from google.colab import drive

drive.mount('/content/drive')

cd /content/drive/MyDrive/HUPHealth

import joblib

file_path = "model.joblib"

joblib.dump(model, file_path)

-------------------------------------------

from tensorflow.keras.models import load_model

emotion_model = load_model("HUPYogadeep.h5") 

--------------------------------------

 from keras.models import load_model

 hupmodel= load_model("hup_sentimental_lstm.h5")

------------------------------

from keras.models import load_model

model.save("hupmodel18.h5")

-------------------------------

from sklearn.preprocessing import LabelEncoder

import joblib

labelencoder = LabelEncoder()

y = labelencoder.fit_transform(df['cyberbullying_type'])

joblib.dump(labelencoder, 'huplabelencoder.pkl')

from tensorflow.keras.utils import to_categorical

y1 = to_categorical(y)

--------------------------------------------

import nltk

from nltk.corpus import stopwords

from nltk.stem.porter import PorterStemmer

import pandas as pd

import re

import joblib

nltk.download('stopwords')

port_stem = PorterStemmer()

def stemming(content):

    stemmed_content = re.sub('[^a-zA-Z]',' ',content)

    stemmed_content = stemmed_content.lower()

    stemmed_content = stemmed_content.split()

    stemmed_content = [port_stem.stem(word) for word in stemmed_content if not word in stopwords.words('english')]

    stemmed_content = ' '.join(stemmed_content)

    return stemmed_content

hupstr=input('Enter the tweet')

df1 = pd.DataFrame({'text': [hupstr]})

df1['text']  = df1['text'].apply(stemming)

from keras.models import load_model

##hupmodel= load_model("hupmodel17.h5")

hupmodel= load_model("my_model.keras")

loaded_vectorizer = joblib.load('huptokenizer.joblib')

X2 = loaded_vectorizer.texts_to_sequences(df1['text'])

X2 = pad_sequences(X2,maxlen=337)

hupprediction1 = hupmodel.predict_on_batch(np.stack(X2))

labelencoder = joblib.load('huplabelencoder.pkl')

huplabel1 = labelencoder.inverse_transform(np.argmax(hupprediction1, axis=1))

hupresult1 = ''.join(huplabel1)

print(hupresult1)

------------

hupprediction1 = model.predict([data3]).argmax(axis=1)

OR

hupprediction1 = model.predict_on_batch(np.stack(data3))

Text to Vector , Preprocessing and Loading to mode

 Hi

import nltk

from nltk.corpus import stopwords

from nltk.stem.porter import PorterStemmer

import pandas as pd

import re

nltk.download('stopwords')

port_stem = PorterStemmer()

def stemming(content):

    stemmed_content = re.sub('[^a-zA-Z]',' ',content)

    stemmed_content = stemmed_content.lower()

    stemmed_content = stemmed_content.split()

    stemmed_content = [port_stem.stem(word) for word in stemmed_content if not word in stopwords.words('english')]

    stemmed_content = ' '.join(stemmed_content)

    return stemmed_content

hupstr=input('Enter the tweet')

df1 = pd.DataFrame({'text': [hupstr]})

df1['text']  = df1['text'].apply(stemming)

from keras.models import load_model

hupmodel= load_model("hupmodel17.h5")

loaded_vectorizer = joblib.load('huptokenizer.joblib')

X2 = loaded_vectorizer.texts_to_sequences(df1['text'])

X2 = pad_sequences(X2,maxlen=337)

#X_test_transformed = loaded_vectorizer.transform(hupinput)

#X_test_dense = X_test_transformed.toarray()

hupprediction1 = hupmodel.predict_on_batch(np.stack(X2))

labelencoder = joblib.load('huplabelencoder.pkl')

huplabel1 = labelencoder.inverse_transform(np.argmax(hupprediction1, axis=1))

hupresult1 = ''.join(huplabel1)

print(hupresult1)

Print Tensorflow version

 Hi 

Try this

import tensorflow as tf print(tf.__version__) import keras print(keras.__version__)

How Flask file shows templates and images

 Hi

Normal Flask code look like

--------------

from flask import Flask, redirect, url_for, request

app = Flask(__name__)

 

 

@app.route('/success/<name>')

def success(name):

    return 'Hi %s' % name

 

 

@app.route('/login', methods=['POST', 'GET'])

def login():

    if request.method == 'POST':

        user = request.form['name']

        return redirect(url_for('success', name=user))

    else:

        user = request.args.get('name')

        return redirect(url_for('success', name=user))

 

 

if __name__ == '__main__':

    app.run(debug=True)

--------------

If you need template page, your flask file should return the following code

return render_template('template.html', result=hupresult)

Your template.html should be saved inside templates folder

and the code template.html should look like

<!DOCTYPE html>

<html lang="en">

<head>

    <meta charset="UTF-8">

    <meta name="viewport" content="width=device-width, initial-scale=1.0">

    <title>Flask Image Example</title>

</head>

<body>

    <h1>{{ result }}</h1>

    <!-- Display the image from the static folder -->

    <img src="{{ url_for('static', filename='bg.jpg') }}" alt="Your Image">

</body>

</html>

The image bg.jpg should be saved inside 'static' folder

Face Emotion Detection - Colab

 Hi all

Step1 : Open the colab

Step2 : File -> Save a copy in Drive

Step3 : Work with the copy

Link

https://colab.research.google.com/drive/1K-XhE_qW0VLWrPmjycz8kriRYmA77M33?usp=sharing

Face emotion dataset

 Hi all

https://drive.google.com/drive/folders/1PaAA1P_5O8bNV42AHxNEHyzmQuMsR3kW?usp=sharing

DIST Google form outputs

 Hi all

Please upload these screenshots

1. Python program without ML for sentiment analysis

2. using SET option 

3. Using ML for sentiment analysis

4. Graph showing algorithm performance - 7 agorithms

5. Use of Pickle

6. Use of Savemodel

7. Covid data analysis - Using ML for sentiment analysis

8. Deep learning - Face emotion recognition 

9. Flask Basics

Show graph in Python for Test accuracy and Train accuracy o various ML algorithms

 Hi all,

Please see the code.

import matplotlib.pyplot as plt

import numpy as np

# Example data (replace this with your actual data)

algorithms = ['KNN', 'ANN', 'Decision Tree', 'Ada Boost', 'Random forest', 'Naive Bayes', 'Log Regression']

test_accuracies = [0.85, 0.92, 0.78, 0.88, 0.95, 0.89, 0.91]

train_accuracies = [0.95, 0.97, 0.88, 0.93, 0.98, 0.92, 0.96]

# Set up bar positions

bar_width = 0.35

index = np.arange(len(algorithms))

# Create bar chart

fig, ax = plt.subplots()

bar1 = ax.bar(index, test_accuracies, bar_width, label='Test Accuracy')

bar2 = ax.bar(index + bar_width, train_accuracies, bar_width, label='Train Accuracy')

# Add labels, title, and legend

ax.set_xlabel('Machine Learning Algorithms')

ax.set_ylabel('Accuracy')

ax.set_title('Test and Train Accuracies of Machine Learning Algorithms')

ax.set_xticks(index + bar_width / 2)

ax.set_xticklabels(algorithms)

ax.legend()

# Show the plot

plt.show()

Deploy ML Model using input from user

 Hi,

Please find the code

data=[]

marks=int(input("enter your marks"))

data.append(marks)

maths=int(input("enter marks for maths"))

data.append(maths)

qa=int(input("enter marks for qa"))

data.append(qa)

pg=int(input("enter marks for programing"))

data.append(pg)

ndata=[]

ndata.append(data)

result=knn.predict(ndata)

if(result==0):

  print("not placed")

else:

  print("placed")

Deploy model

 Hi,

Please find the code

input1=[[60,80,90,6]]

result=knn.predict(input1)

print(result)

Shared Datasets

 Hi,

Please find the shared dataset.

https://drive.google.com/drive/folders/1Ht32d3wP-lIJ4QrZsjQHFwU_hnGal0YI?usp=sharing