how to use this in my own NER training

[tongbc]

as i said,3ks

[shizhediao]

You might want to download our pre-trained checkpoint and then fine-tune it on your own NER training data. For example, here is a script of fine-tuning msra_ner

python run_token_level_classification.py \
    --task_name msra \
    --do_train \
    --do_eval \
    --do_lower_case \
    --data_dir /path/to/dataset \
    --bert_model /path/to/zen_model \
    --max_seq_length 128 \
    --do_train  \
    --do_eval \
    --train_batch_size 128 \
    --num_train_epochs 30 \
    --warmup_proportion 0.1