In your article, I noticed a mention of "condition tokens," but I couldn’t find the corresponding code in the model repository. May I ask whether the code you’ve released supports pre-training? If not, what modifications would be needed to adapt this codebase for pre-training purposes?
The image below shows the input definition in the source code. Based on my understanding, src refers to Gene tokens and values refers to Expression values. It seems there are no Condition tokens defined here:
In your article, I noticed a mention of "condition tokens," but I couldn’t find the corresponding code in the model repository. May I ask whether the code you’ve released supports pre-training? If not, what modifications would be needed to adapt this codebase for pre-training purposes?
The image below shows the input definition in the source code. Based on my understanding, src refers to Gene tokens and values refers to Expression values. It seems there are no Condition tokens defined here: