Model Card for Model ID
Model Details
Model Description
A Transformer-based model fine-tuned for abstractive summarization of medical reports such as discharge summaries, clinical notes, or diagnostic findings. Useful for common people,non medical background person or EHR platforms seeking concise insights from lengthy medical documentation.
- Developed by: Abhinav S Kangale
- Model type: Transformer (encoder-decoder architecture)
- Language(s) (NLP): Enlish(medical domain)
Model Sources [optional]
Uses
Direct Use Automatic summarization of medical documents, such as:
Doctor's notes
Patient discharge summaries
Radiology/imaging reports
Out-of-Scope Use
Not suitable for non-English or non-medical texts.
Should not be used without human review in clinical decision-making or life-critical situations
Bias, Risks, and Limitations
The model may:
Miss critical medical terms if phrased unusually.
Introduce hallucinated or misleading summaries.
Perform poorly on non-standard or shorthand-heavy notes.
Training Details
Training Hyperparameters
-Training Hyperparameters
Precision: fp16 mixed precision
Max input length: 1024
Epochs: 3
Batch size: 8
Optimizer: AdamW-->
Results
[More Information Needed]
Summary
Model Examination [optional]
[More Information Needed]
Environmental Impact
- Hardware Type: Hardware Type: 1x NVIDIA A100 40GB
- Hours used: Hours Used: ~6 hours
Citation [optional]
@misc{cikista2025summarizer, title={Cikista Medical Report Summarizer}, author={AbinavSK}, year={2025}, howpublished={\url{https://huggingface.co/AbhinavSK/cikista-medical-report-summarizer}}, }