Multi-trial dataset for training

[aloejhb]

Hi,
Thanks for developing CEBRA!

I have a question regarding the multi-trial data.
In my experiments, I recorded the neuronal activities of a same animal in multiple trials with different stimuli.
Since the inter-trial-interval is long, I did not record the activities between the trials.

Now if I would like to train a single CEBRA model with data from all trials using time, is there a way to inform CEBRA that the trials are seperated in time, so CEBRA does not connect the trials as if they were recorded immediately after each other?

[aloejhb]

This seem to be a duplicate to #49 . Thanks for the answer and looking forward to the update!