PaliGemma 1
Google DeepMind — Gemma
rich detail 19 method fields published
Evaluation
Transfer with limited examples study
fine-tuning PaliGemma with limited numbers of examples (64, 256, 1024, 4096), sweeping learning rate, epochs and batch size, with 5 seeds per setting; relative regret vs full-dataset results
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.