salesforce / decaNLP

The Natural Language Decathlon: A Multitask Challenge for NLP
BSD 3-Clause "New" or "Revised" License
2.34k stars 474 forks source link

Error while validating models. #35

Closed ashleyyy94 closed 5 years ago

ashleyyy94 commented 5 years ago

Hi,

Whenever I try to validate my models or the pretrained models provided, I run into this error:

Example: nvidia-docker run -it --rm -v pwd:/decaNLP/ decanlp bash -c "python /decaNLP/predict.py --evaluate validation --path /decaNLP/mqan_decanlp_qa_first --checkpoint_name model.pth --gpu 0"

. . . Loading from /decaNLP/mqan_decanlp_qa_first/model.pth Initializing Model Traceback (most recent call last): File "/decaNLP/predict.py", line 262, in model = Model(field, args) File "/decaNLP/models/multitask_question_answering_network.py", line 32, in init if self.args.cove or self.args.intermediate_cove: AttributeError: 'Namespace' object has no attribute 'intermediate_cove'

Is there a way to resolve this? Training the decaNLP Model runs win but validation throws this error.

Thank you!

dfenglei commented 5 years ago

yes, You can look the below for fixing this error. https://github.com/salesforce/decaNLP/pull/33

ashleyyy94 commented 5 years ago

Thank you. That helped to fix it!