Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+jigsaw-toxic-comment-classification-challenge/test.csv/test.csv filter=lfs diff=lfs merge=lfs -text
+jigsaw-toxic-comment-classification-challenge/train.csv/train.csv filter=lfs diff=lfs merge=lfs -text
+toxicity.keras filter=lfs diff=lfs merge=lfs -text

.github/workflows/update_space.yml

@@ -0,0 +1,27 @@
+name: Update Space
+
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  update:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/setup-python@v2
+      with:
+        python-version: "3.8"
+
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install gradio
+
+    - name: Update Space
+      run: |
+        gradio deploy main
+      env:
+        GRADSPACE_APP_FILE: Toxicity.py
+        GRADSPACE_TITLE: Hate_Speech_Detection

README.md

@@ -1,12 +1,6 @@
 ---
-title: Hate Speech Detection
-emoji: 🐢
-colorFrom: purple
-colorTo: green
+title: Hate_Speech_Detection
+app_file: Toxicity.py
 sdk: gradio
 sdk_version: 4.15.0
-app_file: app.py
-pinned: false
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Toxicity.py

@@ -0,0 +1,304 @@
+#!/usr/bin/env python
+# coding: utf-8
+
+# # 0. Install Dependencies and Bring in Data
+%pip install tensorflow pandas matplotlib scikit-learn
+# In[2]:
+
+
+import os
+import pandas as pd
+import tensorflow as tf
+import numpy as np
+
+
+# In[3]:
+
+
+df = pd.read_csv(os.path.join('jigsaw-toxic-comment-classification-challenge','train.csv', 'train.csv'))
+
+
+# In[4]:
+
+
+df.head()
+
+
+# # 1. Preprocess
+
+# In[5]:
+
+
+get_ipython().system('pip list')
+
+
+# In[6]:
+
+
+from tensorflow.keras.layers import TextVectorization
+
+
+# In[7]:
+
+
+X = df['comment_text']
+y = df[df.columns[2:]].values
+
+
+# In[8]:
+
+
+MAX_FEATURES = 200000 # number of words in the vocab
+
+
+# In[9]:
+
+
+vectorizer = TextVectorization(max_tokens=MAX_FEATURES,
+                               output_sequence_length=1800,
+                               output_mode='int')
+
+
+# In[10]:
+
+
+vectorizer.adapt(X.values)
+
+
+# In[11]:
+
+
+vectorized_text = vectorizer(X.values)
+
+
+# In[12]:
+
+
+#MCSHBAP - map, chache, shuffle, batch, prefetch  from_tensor_slices, list_file
+dataset = tf.data.Dataset.from_tensor_slices((vectorized_text, y))
+dataset = dataset.cache()
+dataset = dataset.shuffle(160000)
+dataset = dataset.batch(16)
+dataset = dataset.prefetch(8) # helps bottlenecks
+
+
+# In[13]:
+
+
+train = dataset.take(int(len(dataset)*.7))
+val = dataset.skip(int(len(dataset)*.7)).take(int(len(dataset)*.2))
+test = dataset.skip(int(len(dataset)*.9)).take(int(len(dataset)*.1))
+
+
+# # 2. Create Sequential Model
+
+# In[14]:
+
+
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import LSTM, Dropout, Bidirectional, Dense, Embedding
+
+
+# In[15]:
+
+
+model = Sequential()
+# Create the embedding layer 
+model.add(Embedding(MAX_FEATURES+1, 32))
+# Bidirectional LSTM Layer
+model.add(Bidirectional(LSTM(32, activation='tanh')))
+# Feature extractor Fully connected layers
+model.add(Dense(128, activation='relu'))
+model.add(Dense(256, activation='relu'))
+model.add(Dense(128, activation='relu'))
+# Final layer 
+model.add(Dense(6, activation='sigmoid'))
+
+
+# In[16]:
+
+
+model.compile(loss='BinaryCrossentropy', optimizer='Adam')
+
+
+# In[17]:
+
+
+model.summary()
+
+
+# In[18]:
+
+
+history = model.fit(train, epochs=1, validation_data=val)
+
+
+# In[19]:
+
+
+from matplotlib import pyplot as plt
+
+
+# In[20]:
+
+
+plt.figure(figsize=(8,5))
+pd.DataFrame(history.history).plot()
+plt.show()
+
+
+# # 3. Make Predictions
+
+# In[21]:
+
+
+# Make Predictions
+input_text = vectorizer(['You freaking suck! I am going to hit you.'])  # Make sure to pass a list
+res = model.predict(input_text)
+
+
+# In[63]:
+
+
+# Convert the predictions to binary values based on the threshold (0.5)
+binary_predictions = (res > 0.5).astype(int)
+print(binary_predictions)
+
+
+# In[64]:
+
+
+batch_X, batch_y = test.as_numpy_iterator().next()
+
+
+# In[65]:
+
+
+(model.predict(batch_X) > 0.5).astype(int)
+
+
+# In[66]:
+
+
+res.shape
+
+
+# # 4. Evaluate Model
+
+# In[67]:
+
+
+from tensorflow.keras.metrics import Precision, Recall, CategoricalAccuracy
+
+
+# In[68]:
+
+
+pre = Precision()
+re = Recall()
+acc = CategoricalAccuracy()
+
+
+# In[69]:
+
+
+for batch in test.as_numpy_iterator(): 
+    # Unpack the batch 
+    X_true, y_true = batch
+    # Make a prediction 
+    yhat = model.predict(X_true)
+    
+    # Flatten the predictions
+    y_true = y_true.flatten()
+    yhat = yhat.flatten()
+    
+    pre.update_state(y_true, yhat)
+    re.update_state(y_true, yhat)
+    acc.update_state(y_true, yhat)
+
+
+# In[81]:
+
+
+print(f'Precision: {pre.result().numpy()}, Recall:{re.result().numpy()}, Accuracy:{acc.result().numpy()}')
+
+
+# # 5. Test and Gradio
+
+# In[72]:
+
+
+get_ipython().system('pip install gradio jinja2')
+
+
+# In[73]:
+
+
+import tensorflow as tf
+import gradio as gr
+
+
+# In[85]:
+
+
+model.save('toxicity.keras')
+
+
+# In[86]:
+
+
+model = tf.keras.models.load_model('toxicity.keras')
+
+
+# In[87]:
+
+
+input_str = vectorizer('hey i freaken hate you!')
+
+
+# In[88]:
+
+
+res = model.predict(np.expand_dims(input_str,0))
+
+
+# In[89]:
+
+
+res
+
+
+# In[90]:
+
+
+def score_comment(comment):
+    vectorized_comment = vectorizer([comment])
+    results = model.predict(vectorized_comment)
+    
+    text = ''
+    for idx, col in enumerate(df.columns[2:]):
+        text += '{}: {}\n'.format(col, results[0][idx]>0.5)
+    
+    return text
+
+
+# In[93]:
+
+
+interface = gr.Interface(fn=score_comment, 
+                         inputs="textbox",  # Use "textbox" instead of gr.inputs.Textbox
+                         outputs="text",
+                        )
+
+
+# In[94]:
+
+
+interface.launch(share=True)
+
+
+# In[ ]:
+
+
+
+

jigsaw-toxic-comment-classification-challenge/test.csv/test.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2513ce4abb98c4d1d216e3ca0d4377d57589a0989aa8c06a840509a16c786e8
+size 60354593

jigsaw-toxic-comment-classification-challenge/train.csv/train.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd4084611bd27c939ba98e5e63bc3e5a2c1a4e99477dcba46c829e4c986c429d
+size 68802655

requirements.txt

@@ -0,0 +1,5 @@
+tensorflow
+pandas
+numpy
+scikit-learn
+matplotlib

toxicity.keras

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb801227b73177f73c564d8fc07a4d1a5931ec82245ae2cf6c8394f069d62500
+size 26006432