Downloads · 30 days
17
1% of all-time downloads
pubmed_bert_squadv2
franklu/pubmed_bert_squadv2
pubmed_bert_squadv2 is a question answering model from franklu. Use it when the input is a question plus a passage. It is set up for transformers.
microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext fine-tuned on SQuAD V2 using runqa.py
All-time downloads
3.2K
Public
Repo size
1.3 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin436 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- Model type
- bert
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Jul 9, 2021
- SHA
- 3abeb356
Try a prompt
Spaces
Model card
From the Hugging Face model README
microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext fine-tuned on SQuAD V2 using run_qa.py
Tunning script:
BASE_MODEL=microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext
OUTPUT_DIR=~/Documents/projects/tunned_models/ms_pubmed_bert_squadv2/
python run_qa.py \
--model_name_or_path $BASE_MODEL\
--dataset_name squad_v2 \
--do_train \
--do_eval \
--version_2_with_negative \
--per_device_train_batch_size 12 \
--learning_rate 3e-5 \
--num_train_epochs 2 \
--max_seq_length 384 \
--doc_stride 128 \
--output_dir $OUTPUT_DIR
- Task
- Question Answering
- Library
- transformers
- Type
- bert
- Created
- Mar 2, 2022
- Updated
- Jul 9, 2021