No module names 'bert'

[allenyummy]

I download pytorch, allennlp, and git clone your repository, and run the script below.
I got the error like this.
What's wrong I didn't notice ?
Thanks your reply.

--

#!bin/bash

####local path
DATA_DIR=raw_data/
BERT_DIR=MTMSN/bert_pretrain_model/bert-base-uncased/

SEQ=512
TBS=12
PBS=24
LR=3e-5
EPOCH=10.0

EXPERIMENT_DIR=MTMSN/experiment/bert-base-uncased_S-${SEQ}_B-${TBS}_LR-${LR}_E-${EPOCH}/

CUDA_VISIBLE_DEVICES=0,1,2 python3 -m MTMSN/mtmsn/bert.run_mtmsn \
  --vocab_file $BERT_DIR/vocab.txt \
  --bert_config_file $BERT_DIR/config.json \
  --init_checkpoint $BERT_DIR/pytorch_model.bin \
  --do_train \
  --do_predict \
  --do_lower_case \
  --train_file $DATA_DIR/drop_dataset_train.json \
  --predict_file $DATA_DIR/drop_dataset_dev.json \
  --train_batch_size $TBS \
  --predict_batch_size $PBS \
  --num_train_epochs $EPOCH \
  --learning_rate $LR \
  --max_seq_length $SEQ \
  --span_extraction \
  --addition_subtraction \
  --counting \
  --negation \
  --gradient_accumulation_steps 2 \
  --output_dir $EXPERIMENT_DIR/mtmsn_base

[huminghao16]

Please use the exact command described in the README

[allenyummy]

I can run this program. But I encountered another question.

Is there any solution for this ?