🦦 Otter, a multi-modal model based on OpenFlamingo (open-sourced version of DeepMind's Flamingo), trained on MIMIC-IT and showcasing improved instruction-following and in-context learning ability.
I have a question about the batch size during fine-tuning.
I'm planning to fine-tune the "OTTER-Image-MPT7B" using my custom In-Context Learning dataset.
How much does the batch size affect the accuracy?
Could you recommend an appropriate batch size?
The dataset consists of 20,000 sets (2 context images and 1 query image each).
Hi, thanks for your great work!
I have a question about the batch size during fine-tuning.
I'm planning to fine-tune the "OTTER-Image-MPT7B" using my custom In-Context Learning dataset. How much does the batch size affect the accuracy? Could you recommend an appropriate batch size? The dataset consists of 20,000 sets (2 context images and 1 query image each).