and got an f1 validation score of 5.8 . Can you please provide instructions on how to migrate the model to huggingface?
**Additional context**
I am able to load and evaluate the model from https://huggingface.co/deepset/roberta-base-squad2 and reach the published f1.
Thanks in advanced
Question I've train roberta base on squad v2 using
python -u transformers/examples/pytorch/question-answering/run_qa.py --model_name_or_path FARM/saved_models/bert-english-qa-tutorial/ --train_file squad2/train-v2.0.json --validation_file squad2/dev-v2.0.json --do_eval --per_device_train_batch_size 4 --max_seq_length 384 --doc_stride 128 --output_dir /tmp/debug_squad/ --tokenizer_name FARM/saved_models/bert-english-qa-tutorial --version_2_with_negative --overwrite_output_dir