geoore
/

SHOREKEEPER

mixture-of-experts

Model card Files Files and versions

SHOREKEEPER / src /training

6.61 kB

Ctrl+K

Ctrl+K

2 contributors

History: 1 commit

geoore's picture

Restructure to src/ layout with attention, per-layer MoE, and working chat

73400c8 6 days ago

grpo.py

6.61 kB
Restructure to src/ layout with attention, per-layer MoE, and working chat 6 days ago