issues
search
zuoxingdong
/
lagom
lagom: A PyTorch infrastructure for rapid prototyping of reinforcement learning algorithms.
MIT License
373
stars
30
forks
source link
update ExperimentWorker: set num_thread=1 for CPU mode to avoid getti…
#131
Closed
zuoxingdong
closed
5 years ago
zuoxingdong
commented
5 years ago
…ng stuck, particularly with LayerNorm
…ng stuck, particularly with LayerNorm