# Custom files for run\_ner.py

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcustom-files-for-run-ner-py%2F9156%3E%5C
**Category:** Beginners\
**Created:** [August 11, 2021, 11:23am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcustom-files-for-run-ner-py%2F9156 "2021-08-11T11:23:51Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![IreneCrepax](https://avatars.discourse-cdn.com/v4/letter/i/54ee81/32.png) [@IreneCrepax](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FIreneCrepax)%5C
**Post date:** [August 11, 2021, 11:23am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcustom-files-for-run-ner-py%2F9156%2F1 "2021-08-11T11:23:51Z")

</div>

Hi, I am trying to train BERT using run\_ner.py with my own train and eval set  
python /rds/projects/d/divjakd-ooo-machines/Users/irene/BERT/transformers/examples/pytorch/token-classification/run\_ner.py   
–model\_name\_or\_path bert-base-uncased   
–train\_file ‘/rds/projects/d/divjakd-ooo-machines/Users/irene/articles\_old/Step1\_Extract\_sentences\_with\_articles\_BNC/Results/train\_set\_bert\_comma.csv’   
–validation\_file ‘/rds/projects/d/divjakd-ooo-machines/Users/irene/articles\_old/Step1\_Extract\_sentences\_with\_articles\_BNC/Results/eval\_set\_bert\_comma.csv’   
–output\_dir /tmp/test-ner   
–do\_train   
–do\_eval

but I am confused about the file format. I get an error saying  
File “/rds/projects/d/divjakd-ooo-machines/Users/irene/BERT/transformers/examples/pytorch/token-classification/run\_ner.py”, line 296, in main  
if isinstance(features[label\_column\_name].feature, ClassLabel):  
AttributeError: ‘Value’ object has no attribute ‘feature’

Any help? my files are csv comma separated, and have two columns, 0 and 1