|
--- |
|
license: apache-2.0 |
|
language: en |
|
tags: |
|
- hate |
|
- speech |
|
|
|
widget: |
|
- text: "RT @ShenikaRoberts: The shit you hear about me might be true or it might be faker than the bitch who told it to ya ᙨ" |
|
|
|
--- |
|
|
|
# Dataset Collection: |
|
* The hatespeech dataset is collected from different open sources like Kaggle ,social media like Twitter. |
|
* The dataset has the two classes hatespeech and non hatespeech. |
|
* The class distribution is equal |
|
* Different strategies have been followed during the data gathering phase. |
|
* The dataset is collected from relevant sources. |
|
|
|
# distilbert-base-uncased model is fine-tuned for Hate Speech Detection |
|
* The model is fine-tuned on the dataset. |
|
* This model can be used to create the labels for academic purposes or for industrial purposes. |
|
* This model can be used for the inference purpose as well. |
|
|
|
# Data Fields: |
|
|
|
**label**: 0 - it is a hate speech, 1 - not a hate speech |
|
|
|
# Application: |
|
* This model is useful for the detection of hatespeech in the tweets. |
|
* There are numerous situations where we have tweet data but no labels, so this approach can be used to create labels. |
|
* You can fine-tune this model for your particular use cases. |
|
|
|
# Model Implementation |
|
|
|
# !pip install transformers[sentencepiece] |
|
|
|
from transformers import pipeline |
|
|
|
model_name="Sakil/distilbert_lazylearner_hatespeech_detection" |
|
|
|
classifier = pipeline("text-classification",model=model_name) |
|
|
|
classifier("!!! RT @mayasolovely: As a woman you shouldn't complain about cleaning up your house. & as a man you should always take the trash out...") |
|
|
|
# Github: [Sakil Ansari](https://github.com/Sakil786/hate_speech_detection_pretrained_model) |
