efficient-moe-agent-project
/

scienceworld

Reinforcement Learning

Model card Files Files and versions

Commit History

Add nothinking single-expert best ckpt (step200, val SR 0.379)

ce0b3dc
verified

Davidk1 commited on Aug 19

Add nothinking single-expert best ckpt (step200, val SR 0.379)

b75d194
verified

Davidk1 commited on Aug 19

add model card

88f09f5
verified

Davidk1 commited on Aug 5

add single_expert_warmstart/best_step140

02ab4f4
verified

Davidk1 commited on Aug 5

add single_expert_warmstart/final_step1000

98b34e7
verified

Davidk1 commited on Aug 5

initial commit

b986311
verified

Davidk1 commited on Aug 5