You can deploy this model and start earning money today!

vyshnaviii123/
qwen-medical-qa

Name: qwen-medical-qa
Availability: InStock

safetensors

Views

No views yet

Qwen2-0.5B Medical Q&A (Fine-tuned)

A Qwen2-0.5B-Instruct model fine-tuned on medical health advice data using QLoRA.

Training Details

Base model: Qwen/Qwen2-0.5B-Instruct
Dataset: medalpaca/medical_meadow_health_advice
Method: QLoRA (4-bit quantization + LoRA r=8)
Hardware: Kaggle free GPU (Tesla T4)
Epochs: 1 | Steps: 100
Final loss: 3.02 → being improved with more data

How to use

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("vyshnaviii123/qwen-medical-qa")
tokenizer = AutoTokenizer.from_pretrained("vyshnaviii123/qwen-medical-qa")

input = tokenizer("Q: What helps with a sore throat?\nA:", return_tensors="pt")
output = model.generate(**input, max_new_tokens=100)
print(tokenizer.decode(output[0]))

Limitations

Trained on limited data — not for real medical advice.