Edit model card

BERT-LABR unbalanced

Arabic version bert model fine tuned on LABR dataset


The model were fine-tuned on ~63000 book reviews in arabic using bert large arabic


class precision recall f1-score Support
0 0.8109 0.6832 0.7416 1670
1 0.9399 0.9689 0.9542 8541
Accuracy 0.9221 10211

How to use

You can use these models by installing torch or tensorflow and Huggingface library transformers. And you can use it directly by initializing it like this:

from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained(model_name,num_labels=2)
tokenizer = AutoTokenizer.from_pretrained(model_name)
Downloads last month

Dataset used to train mofawzy/bert-labr-unbalanced