[fsdp,vllm,trainer,algo] feat: On-Policy Distillation #5910
e2e_one_step_off_policy.yml
on: pull_request
setup
7s
e2e_one_step_off_policy_fsdp2
4m 13s
e2e_one_step_off_policy_megatron
4m 9s
cleanup
7s