This repository has been archived by the owner on Oct 31, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 68
TaBERT wit MAPO #22
Comments
Hi Soumik, I'm having the same problem, did you find any solution to it? Thank you and have a great day! |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Hi
I have run successfully BERT with MAPO. Everything is running fine. This is the command I used : python experiments.py train --seed 0 --cuda --work-dir=runs/demo_run --config=data/config/config.vanilla_bert.json --extra-config='{"actor_use_table_bert_proxy": true}'. This is the MAPO code https://github.com/pcyin/pytorch_neural_symbolic_machines.
Now I am trying to reproduce TaBERT with MAPO. But when I do this the code is not running. The Learner exits and that is where it ends. Same command as BERT.
For running TaBERT with MAPO I have downloaded the model from 'https://drive.google.com/uc?id=17NTNIqxqYexAzaH_TgEfK42-KmjIRC-g'. Then I just have changed the config file 'table_bert_model_or_config' param with TaBERT model path.
I am trying to reproduce Table 1 in https://www.aclweb.org/anthology/2020.acl-main.745.pdf : w/ TABERTLarge (K = 3) 52.2 ±0.7 , 53.0 , 51.8 ±0.6 , 52.3 .
Can you please let me know where I am doing error?
Thanks.
The text was updated successfully, but these errors were encountered: