---
license: apache-2.0
base_model: distilbert-base-uncased
tags:
- generated_from_trainer
model-index:
- name: hausa-distilbert
  results: []
language:
- ha
metrics:
- glue
- accuracy: 0.7323943661971831
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# hausa-distilbert

This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on a mixed dataset of different sources.

## Model description

This AI model analyzes Hausa language sentences and classifies their sentiment as positive or negative. 

Here are some key details:

* **Task:** Sentiment analysis (positive/negative)
* **Language:** Hausa
* **Output:** Sentiment classification for each sentence

**Note:**  Hausa NLP resources are still under development. 

## Intended uses & limitations

This AI model offers sentiment analysis (positive/negative) for Hausa sentences, making it useful for tasks like social media monitoring or customer feedback analysis. However, due to limitations in Hausa language resources and the inherent challenges of sentiment analysis,  
production use is not currently recommended.  For important tasks,  consider this model as a starting point and combine its results with human expertise for the most reliable sentiment understanding.

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0

### Training results

Using the SST-2 metrics, the following result was achieved:
- accuracy: 0.7323943661971831

### Framework versions

- Transformers 4.40.1
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1