FashionBert errors

[StefanoSalvatori]

I tried to run the scripts provided for fashionbert and i'm getting this error:

Traceback (most recent call last):
  File "pretrain_main.py", line 194, in <module>
    main()
  File "pretrain_main.py", line 177, in main
    yield_single_examples=False):
  File "/usr/local/lib/python3.6/dist-packages/tensorflow/python/estimator/estimator.py", line 577, in predict
    features, None, model_fn_lib.ModeKeys.PREDICT, self.config)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow/python/estimator/estimator.py", line 1195, in _call_model_fn
     model_fn_results = self._model_fn(features=features, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/easytransfer/engines/model.py", line 578, in model_fn
    output = self.build_logits(features, mode=mode)
  File "pretrain_main.py", line 50, in build_logits
    input_sequence_length=_APP_FLAGS.input_sequence_length)
  File "/usr/local/lib/python3.6/dist-packages/easytransfer/model_zoo/__init__.py", line 38, in get_pretrained_model
     return ImageBertPreTrainedModel.get(pretrain_model_name_or_path, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/easytransfer/model_zoo/modeling_utils.py", line 99, in get
    model(model.dummy_inputs(kwargs.get('input_sequence_length', 512)), mode='eval', output_features=False)
  File "/usr/local/lib/python3.6/dist-packages/easytransfer/model_zoo/modeling_utils.py", line 69, in dummy_inputs
    input_ids = [[1]*seq_length]
  TypeError: can't multiply sequence by non-int of type 'NoneType'

After running:

python3 pretrain_main.py \
  --workerGPU=1 \
  --type=txt2img  \
  --mode=predict \
  --predict_input_fp=eval_txt2img_list.list_csv  \
  --predict_batch_size=64  \
  --output_dir=./fashionbert_out  \
  --pretrain_model_name_or_path=./pai-imagebert-base-en/model.ckpt  \
  --predict_checkpoint_path=./fashionbert_pretrain_model_fin/model.ckpt-54198  \
  --image_feature_size=131072  \
  --input_schema="image_feature:float:131072,image_mask:int:64,input_ids:int:64,input_mask:int:64,segment_ids:int:64,nx_sent_labels:int:1,prod_desc:str:1,text_prod_id:str:1,image_prod_id:str:1,prod_img_id:str:1"  \

I'm using easytransfer 0.1.2 (since the latest version causes another error as reported here #21) and tensorflow 1.12.3.

Can you give some advice on how to run the scripts? Are there other implementation of FashionBert (pytorch or tensorflow?)