---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
widget:
- output:
    url: images/Screenshot from 2025-08-20 00-50-33.png
  text: None
  parameters:
    negative_prompt: None
base_model: ProsusAI/finbert
instance_prompt: null
license: other
license_name: useless
license_link: LICENSE
---
# Emotion

<Gallery />

## Model description 

Emotion Recognition Model (BERT-based)
📌 Overview

This is a BERT-based emotion recognition model that I created purely for educational and learning purposes.
The model was trained as part of my journey to understand transformers, distillation, GPU management, fine-tuning, and Hugging Face workflows.

⚙️ How I built it

I started with a pretrained BERT model.

I experimented with layer distillation (copying a few layers into a smaller student model).

I trained it on an emotion classification dataset to predict different emotional states from text.

I focused on hands-on practice: learning about tokenization, GPU memory issues, checkpointing, and model saving&#x2F;loading.

⚠️ Disclaimer

This model is not production-ready.

It is not optimized for real-world use.

It should not be used for commercial, fine-tuning, or deployment purposes.

It was built only as a learning exercise to explore Hugging Face and model training.

💡 Purpose

To help me (and maybe others) understand how Hugging Face works.

To practice model distillation and fine-tuning techniques.

To learn the workflow of pushing models to Hugging Face Hub.

🚫 Limitations

Accuracy and reliability are not guaranteed.

Not suitable for critical applications (mental health, customer service, etc.).

Limited number of layers and trained on a small dataset.


## Download model


[Download](/Abdullah6395/Text_Emotion_Recognition/tree/main) them in the Files & versions tab.