policy
MURA-RL-DeepEnsembleV4-83.20
blahmin · PyTorch
or hover any field below to flag it
Overview
Name
MURA-RL-DeepEnsembleV4-83.20
Author
blahmin
Framework
PyTorch
License
unknown
Skill type
other
Evidence level
untested
Task description
DeepEnsembleV4-83.20 builds on DeepSeek-R1's RL-based decision-making with CBAM attention, EfficientNet-B3 expert modeling, and transformer-enhanced gating. Trained from scratch, it set new records with 83.20% validation accuracy and 0.6608 Cohen’s Kappa, reversing the train/val accuracy dynamic for
Spaces
Action space
other · 0-dim · 0Hz
Observation space
- type: other
Links
HuggingFace repo
null
Paper (arXiv)
null
Compatible robots
3+17 mentioned but not in catalog yetCompatible environments
0No environments list MURA-RL-DeepEnsembleV4-83.20 yet.
Datasets that reference this policy
0No datasets reference MURA-RL-DeepEnsembleV4-83.20 yet.