Kumru-2B-TUS-Medical

- Prompt
- -
Model description
Kumru-2B Turkish Medical Specialty Exam (TUS) Model
A fine-tuned version of Kumru-2B on the Turkish Medical Specialty Exam (TUS) dataset. Specialized for Turkish medical question answering.
Features
- Turkish medical question answering (Q&A)
- Trained on TUS exam questions
- Efficient fine-tuning with QLoRA + Unsloth
- Low memory usage with 4-bit quantization (load_in_4bit=True)
- TextStreamer integration for smooth output
- Optimized output with max_seq_length=512, max_new_tokens=300
Training Details
- Base model: vngrs-ai/Kumru-2B
- Dataset: turkerberkdonmez/TUSGPT-TR-Medical-Dataset-v1
- Method: QLoRA + Unsloth
- LoRA r=16, alpha=16
- Epochs: 1
Trigger words
You should use medical to trigger the image generation.
You should use TUS to trigger the image generation.
You should use Kumru-2B to trigger the image generation.
Download model
Download them in the Files & versions tab.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for anilsrml/Kumru-2B-TUS-Medical
Base model
vngrs-ai/Kumru-2B