Closed makevin23 closed 11 months ago
Hi, there was an error in the training and dataset loading script for which the training script was throwing an error. Now that issue is resolved. Unfortunately, I do not have access to multiple GPU for this task unfortunately. Could you please try again with a single GPU (python train.py --dataset qald9
)?
Thanks for your understanding
Hi, I have some problem while training this model. When I run
on my server with 2 GPUs, it didn't load the parameters from the
params.json
file and train 70 epochs, but it just didn't start training. I assumetrain.py
is not able to readparamas.json
and take the default epoch -1. Can you fix this problem?