gemma-2b-pubmed-classifier

This lora was made for educational purposes as part of an upcoming series. The model returns a series of classifications in CSV. More information will be released 7/1

I make no claims about medical accuracy.

Trained on owaiskha9654/PubMed_MultiLabel_Text_Classification_Dataset_MeSH

Prompt Template:

<|im_start|>system
{system}<|im_end|>
<|im_start|>user
{user}<|im_end|>
<|im_start|>assistant

Example Usage

System: Given the title and abstract of a medical research paper, classify it into the most relevant MeSH (Medical Subject Headings) major terms.

Input: Title: Expression of p53 and coexistence of HPV in premalignant lesions and in cervical cancer. \n Abstract: Fifty-four paraffin embedded tissue sections from patients with dysplasia (21 cases) and with cervical cancer (33 cases) were analysed. HPV was detected and identified in two stages. Firstly, using mixed starters, chosen genomic DNA sequences were amplified; secondly the material thus obtained was analyzed by hybridization method using oligonucleotyde 31-P labelled probe. HPVs of type 6, 11, 16, 18, 33 were identified.

Output: Cervical Intraepithelial Neoplasia, DNA, Viral, Female, Humans, Precancerous Conditions, Tumor Suppressor Protein p53, Viral Load, Virology