README.md

---
library_name: transformers
tags:
- biology
- medical
datasets:
- Gift-Z/UshukelaDiabetesTrainingData
language:
- en
metrics:
- accuracy
- bertscore
- bleu
- rouge
base_model:
- Qwen/Qwen3-8B
pipeline_tag: question-answering
---

# Model Card for Ushukela Diabetes LLM - A domain specific LLM for diabetes care and management
UshukelaDiabetesLLM

<!-- Provide a quick summary of what the model is/does. -->



## Model Details

### Model Description

<!-- Provide a longer summary of what this model is. -->

This study was initiated to develop a domain-specific Large Language Model (LLM) aimed at improving diabetes care and management, addressing the unique challenges faced by South African patients, caregivers, and medical practitioners. To achieve this, local data was collected and supplemented with benchmark and medical Hugging Face datasets. Medical specialised BioMedLM (2.7B) and BioMistral-7B pre-trained LLMs were selected as base models, along with Qwen3-8B (a non-specialised LLM). Parameter-Efficient Fine-Tuning (PEFT) techniques, prompt tuning and Quantised Low-Rank Adaptation (QLoRA) were applied, with Retrieval-Augmented Generation (RAG) applied on the best-performing LLM. The fine-tuned LLMs were evaluated by comparing their respective performance with Diabetica-7B, a diabetes specialised model. 

The final collected dataset consisted of 18,079 processed question-answer pairs (14% of which were artificially generated to reflect the South African context), along with 1,596 documents. The data covered topics such as medication, management (diet and exercise), diagnosis and screening, and general diabetes-related content, ensuring comprehensiveness. When evaluating fill-in-the-blanks and multiple-choice question-answer pairs, Qwen QLoRA outperformed all LLMs, achieving ROUGE-1 of 0.793, ROUGE-L of 0.792, and BERT Score F1 of 0.940. However, Diabetica achieved a higher BLEU of 0.465, while Qwen3-8B. achieved a BLEU of 0.365. For multiple-choice questions only, Qwen3-7B LLMs performed consistently better, with Qwen3-7B QLoRA achieving the highest accuracy of 80.7%. For short and long answer questions, BioMistral-7B QLoRA showed slightly superior performance, with all LLMs performing above 0.800. The results of this study show promising applications of LLMs in healthcare. 

- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]

### Model Sources [optional]

<!-- Provide the basic links for the model. -->

- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]

## Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

### Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

[More Information Needed]

### Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

[More Information Needed]

### Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

[More Information Needed]

## Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

[More Information Needed]

### Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

## How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

## Training Details

### Training Data

<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

[More Information Needed]

### Training Procedure

<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->

#### Preprocessing [optional]

[More Information Needed]


#### Training Hyperparameters

- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->

#### Speeds, Sizes, Times [optional]

<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->

[More Information Needed]

## Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

### Testing Data, Factors & Metrics

#### Testing Data

<!-- This should link to a Dataset Card if possible. -->

[More Information Needed]

#### Factors

<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->

[More Information Needed]

#### Metrics

<!-- These are the evaluation metrics being used, ideally with a description of why. -->

[More Information Needed]

### Results

[More Information Needed]

#### Summary



## Model Examination [optional]

<!-- Relevant interpretability work for the model goes here -->

[More Information Needed]

## Environmental Impact
Fine-tuning LLMs, even when using open-source models, incurs financial costs and has environmental implications that should be considered. Developing the LLM for the South African context required over 1,200 compute units, a notable amount. These units were utilised for data preparation, experimentation, fine-tuning, and model evaluation. The total cost of the project is estimated at approximately US$100 (excluding value-added tax), with an associated carbon emission of 45.6 kg CO₂ equivalent

Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).

- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]

## Technical Specifications [optional]

### Model Architecture and Objective

[More Information Needed]

### Compute Infrastructure

[More Information Needed]

#### Hardware

[More Information Needed]

#### Software

[More Information Needed]

## Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

[More Information Needed]

**APA:**

[More Information Needed]

## Glossary [optional]

<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->

[More Information Needed]

## More Information [optional]

[More Information Needed]

## Model Card Authors [optional]

[More Information Needed]

## Model Card Contact

[More Information Needed]