Palmyra-Med-70B: Specialized LLM for Healthcare
Palmyra-Med-70B is a 70 billion parameter large language model developed by Writer, meticulously designed to address the unique linguistic and knowledge demands of the medical and life sciences sectors. Fine-tuned from Palmyra-X-003 with an 8192 token context window, this model has been trained on an extensive collection of high-quality biomedical data, ensuring precise domain-specific accuracy and fluency.
Key Capabilities
- Leading Biomedical Performance: Achieves an average score of 85.87% on biomedical benchmarks, outperforming models like GPT-4, Claude Opus, Gemini, and Med-PaLM-2 base model, as well as medically trained human test-takers.
- Specialized Training: Utilizes Direct Preference Optimization (DPO) and a custom, diverse medical instruction dataset for enhanced performance in the healthcare field.
- Clinical Data Analysis: Excels at analyzing and summarizing complex clinical notes, EHR data, and discharge summaries, extracting key information for concise, structured summaries.
- Advanced Clinical Entity Recognition: Identifies key medical concepts such as diseases, symptoms, medications, procedures, and anatomical structures from unstructured text.
Intended Use Cases
- Research and Development: Ideal for non-commercial and research use in English within the biomedical domain.
- Information Retrieval: Enhances data analysis and knowledge discovery from EHRs, research articles, and other biomedical sources.
- Clinical Decision Support (Research): Supports applications like pharmacovigilance and medical research by providing deep understanding of medical terminology.
Note: Palmyra-Med-70B is intended for non-commercial and research use only and should not be used for direct patient care, clinical decision support, or professional medical purposes without extensive additional validation and regulatory compliance.