|
--- |
|
license: mit |
|
datasets: |
|
- QuyenAnhDE/Diseases_Symptoms |
|
language: |
|
- en |
|
pipeline_tag: text-generation |
|
library_name: transformers |
|
--- |
|
|
|
# MedLLM |
|
|
|
## Model Description |
|
MedLLM is a language model designed to assist healthcare professionals and patients by providing detailed information about the symptoms of various diseases. This model can help in preliminary assessments and educational purposes by offering accurate and concise symptom descriptions. |
|
|
|
## Model Details |
|
- **Architecture**: Transformer-based architecture optimized for medical text. |
|
- **Training Data**: Trained on a diverse set of medical texts, including clinical notes, research articles, and symptom databases, to ensure a comprehensive understanding of disease symptoms. |
|
- **Training Procedure**: Fine-tuned using supervised learning techniques with a focus on accuracy and relevancy in medical contexts. Hyperparameters were adjusted for optimal performance in natural language understanding and generation. |
|
|
|
## Usage |
|
```python |
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
|
|
model_name = "shivvamm/MedLLM" # Change to your repo name |
|
model = AutoModelForCausalLM.from_pretrained(model_name) |
|
tokenizer = AutoTokenizer.from_pretrained(model_name) |
|
|
|
inputs = tokenizer("What are the symptoms of diabetes?", return_tensors="pt") |
|
outputs = model.generate(**inputs) |
|
response = tokenizer.decode(outputs[0], skip_special_tokens=True) |
|
print(response) |
|
``` |
|
|
