Trouble with dataset_benchmark
and tensors_benchmark
. Continual Learning plugins do not work.
#1577
Unanswered
Vincent710129
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
🐛 Describe the bug
I prepare 2 sets of training and test Pytorch datasets from 2 different domains with the EWC plugin to perform domain-incremental CL. However, the evaluation results show no CL is performed. The evaluation results are the same as performing naive fine-tuning. This bug happens for both
dataset_benchmark
andtensors_benchmark
.🐜 To Reproduce
Due to the confidentiality of my dataset, I cannot share the dataset here. Below is my working manuscript for your information for debugging.
🐝 Expected behavior
I expect the EWC plugin to perform domain-incremental CL with EWC on my self-defined datasets from 2 different domains with
dataset_benchmark
.🐞 Screenshots
Results from EWC:
Results from Naive Fine-tuning:
Results from my own CL strategy:
Beta Was this translation helpful? Give feedback.
All reactions