models available on hugging face at: https://huggingface.co/dhananjay2912
Plots and Hyperparameters available at
Section specific models: https://api.wandb.ai/links/dsteam1/onxkm0mp
Section specific models (pubmed finetuned): https://wandb.ai/dsteam1/LSG_BART_PUBMED/reports/Section-specific-model-report--Vmlldzo0MTc5NDg4?accessToken=zvp9qfgjwm1yfpe351zsld8gr0hw0fjpl4btxypqktv9p5aw6wuk7mzxk10a1ol2
Combiner model: https://api.wandb.ai/links/dsteam1/juww1i71
Procedure to recreate results for MediQA Chat Shared Task B for Clinical Conversation Sumamarization
Place the test set file from the ACI Demo Dataset in the folder
For Run 1 ./install.sh ./activate.sh decode_taskB_run1.sh taskB_testset4participants_inputConversations.csv
For Run 2 ./install.sh ./activate.sh decode_taskB_run2.sh taskB_testset4participants_inputConversations.csv
For Run 3 ./install.sh ./activate.sh decode_taskB_run3.sh taskB_testset4participants_inputConversations.csv
Training Script: TBD