README.md

---
language: en
license: mit
datasets:
  - glue/rte
tags:
  - text-classification
  - glue
  - bert
---

# BertLinearClassifier for RTE

This model is a fine-tuned version of `bert-base-uncased` on the RTE task from GLUE.

## Model Architecture

I've implemented a custom architecture with multiple linear layers on top of BERT:

- Uses BERT base model for feature extraction
- Multiple linear layers with ReLU activations instead of attention
- Simple and efficient classification approach

## Usage

This model can be used for textual entailment classification (whether one text logically follows from another).