why run "python run_squad.py" doesn't work? - nlp

Question

I want fine tune on squad with huggingface run_squad.py, but meet the following question:
1, when I use "--do_train" without "True" as following code, after 20 minutes runing,there is no models in output_dir:
!python run_squad.py \
--model_type bert \
--model_name_or_path bert-base-uncased \
--output_dir models/bert/ \
--data_dir data/squad \
--overwrite_output_dir \
--overwrite_cache \
--do_train \
--train_file train-v2.0.json \
--version_2_with_negative \
--do_lower_case \
--do_eval \
--predict_file dev-v2.0.json \
--per_gpu_train_batch_size 2 \
--learning_rate 3e-5 \
--num_train_epochs 2.0 \
--max_seq_length 384 \
--doc_stride 128 \
--threads 10 \
--save_steps 5000
2, when I use "--do_train=True" as following code, the error message is "run_squad.py: error: argument --do_train: ignored explicit argument 'True'":
!python run_squad.py \
--model_type bert \
--model_name_or_path bert-base-uncased \
--output_dir models/bert/ \
--data_dir data/squad \
--overwrite_output_dir \
--overwrite_cache \
--do_train=True \
--train_file train-v2.0.json \
--version_2_with_negative \
--do_lower_case \
--do_eval \
--predict_file dev-v2.0.json \
--per_gpu_train_batch_size 2 \
--learning_rate 3e-5 \
--num_train_epochs 2.0 \
--max_seq_length 384 \
--doc_stride 128 \
--threads 10 \
--save_steps 5000
3, when I use "--do_train True" as following code, the error message is "run_squad.py: error: unrecognized arguments: True":
!python run_squad.py \
--model_type bert \
--model_name_or_path bert-base-uncased \
--output_dir models/bert/ \
--data_dir data/squad \
--overwrite_output_dir \
--overwrite_cache \
--do_train True \
--train_file train-v2.0.json \
--version_2_with_negative \
--do_lower_case \
--do_eval \
--predict_file dev-v2.0.json \
--per_gpu_train_batch_size 2 \
--learning_rate 3e-5 \
--num_train_epochs 2.0 \
--max_seq_length 384 \
--doc_stride 128 \
--threads 10 \
--save_steps 5000
I run code in colab with GPU: Tesla P100-PCIE-16GB
Judging by the running time, I think the code didn't through training process, but I don't know how to set parameters in order to let training go.what should I do?