Add other files from fine-tunning

3abeb35 over 5 years ago

834 Bytes

**[`microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext`](https://huggingface.co/microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext)** fine-tuned on **[`SQuAD V2`](https://rajpurkar.github.io/SQuAD-explorer/)** using **[`run_qa.py`](https://github.com/huggingface/transformers/blob/master/examples/pytorch/question-answering/run_qa.py)**

Tunning script:
```bash
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
```

	[`microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext`](https://huggingface.co/microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext) fine-tuned on [`SQuAD V2`](https://rajpurkar.github.io/SQuAD-explorer/) using [`run_qa.py`](https://github.com/huggingface/transformers/blob/master/examples/pytorch/question-answering/run_qa.py)

	Tunning script:
	```bash
	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
	```