Sequential layer-wise model merging for Vision Transformers using reinforcement learning.
reinforcement-learning computer-vision pytorch multi-task-learning vision-transformer model-merging grpo ties-merging task-vectors
-
Updated
Jun 6, 2026 - Python