ByT5-base fine-tuned for Hate Speech Detection (on Tweets)

ByT5 base fine-tuned on tweets hate speech detection dataset for Sequence Classification downstream task.

Details of ByT5 - Base ?

ByT5 is a tokenizer-free version of Google's T5 and generally follows the architecture of MT5 . ByT5 was only pre-trained on mC4 excluding any supervised training with an average span-mask of 20 UTF-8 characters. Therefore, this model has to be fine-tuned before it is useable on a downstream task. ByT5 works especially well on noisy text data, e.g. , google/byt5-base significantly outperforms mt5-base on TweetQA . Paper: ByT5: Towards a token-free future with pre-trained byte-to-byte models Authors: Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel

Details of the downstream task (Sequence Classification as Text generation) - Dataset ?

tweets_hate_speech_detection

The objective of this task is to detect hate speech in tweets. For the sake of simplicity, we say a tweet contains hate speech if it has a racist or sexist sentiment associated with it. So, the task is to classify racist or sexist tweets from other tweets.

Formally, given a training sample of tweets and labels, where label ‘1’ denotes the tweet is racist/sexist and label ‘0’ denotes the tweet is not racist/sexist, your objective is to predict the labels on the given test dataset.

• Data Instances:

The dataset contains a label denoting is the tweet a hate speech or not

{'label': 0,  # not a hate speech
 'tweet': ' @user when a father is dysfunctional and is so selfish he drags his kids into his dysfunction.   #run'}

• Data Fields:

label : 1 - it is a hate speech, 0 - not a hate speech

tweet : content of the tweet as a string

• Data Splits:

The data contains training data with 31962 entries

Test set metrics ?

We created a representative test set with the 5% of the entries.

The dataset is so imbalanced and we got a F1 score of 79.8

Model in Action ?

git clone https://github.com/huggingface/transformers.git
pip install -q ./transformers

from transformers import AutoTokenizer, T5ForConditionalGeneration

ckpt = 'Narrativa/byt5-base-tweet-hate-detection'

tokenizer = AutoTokenizer.from_pretrained(ckpt)
model = T5ForConditionalGeneration.from_pretrained(ckpt).to("cuda")

def classify_tweet(tweet):

    inputs = tokenizer([tweet], padding='max_length', truncation=True, max_length=512, return_tensors='pt')
    input_ids = inputs.input_ids.to('cuda')
    attention_mask = inputs.attention_mask.to('cuda')
    output = model.generate(input_ids, attention_mask=attention_mask)
    return tokenizer.decode(output[0], skip_special_tokens=True)
    
    
classify_tweet('here goes your tweet...')

Created by: Narrativa

About Narrativa: Natural Language Generation (NLG) | Gabriele, our machine learning-based platform, builds and deploys natural language solutions. #NLG #AI