Is there a way to split workload across 2 GPUs?

[FlippFuzz]

I've tried Google Colab, but moved to Kaggle because they provide more system RAM.
I ran out because I was trying to generate transcription for a 4h file.

Kaggle provides 2x T4 GPUs for their free tier.
Is there a easy way to parallelize the transcription across 2 different GPUs without loss of quality?
Currently only the first GPU does work, while the second is idle.