verl
verl copied to clipboard
Could not override 'actor_rollout_ref.actor.dtype'. To append to your config use +actor_rollout_ref.actor.dtype=float16
System Info
Information
- [ ] The official example scripts
- [ ] My own modified scripts
Tasks
- [ ] An officially supported task in the
examplesfolder (such as GLUE/SQuAD, ...) - [ ] My own task or dataset (give details below)
Reproduction
recipe/flowrl/run_flowrl_qwen2.5_7b_fp16.sh脚本有问题
Expected behavior
recipe/flowrl/run_flowrl_qwen2.5_7b_fp16.sh脚本有问题
我也遇到相同的错误,尝试使用+actor_rollout_ref.actor.fsdp_config.mixed_precision.param_dtype=${dtype}
但是发现训练很不稳定