Interleave ablation study
Created on February 25|Last edited on February 25
Comment
Larger models do benefit from interleaving.
act thpt: ppo_atari_envpool_jax_3_buffer
1
act thpt: ppo_atari_envpool_jax_3_buffer_interleave
1
act thpt: ppo_atari_envpool_jax_3_buffer (large nn)
1
act thpt: ppo_atari_envpool_jax_3_buffer_interleave (large nn)
1
Training:
ppo_atari_envpool_jax_3_buffer (large nn)
1
ppo_atari_envpool_jax_3_buffer_interleave (large nn)
1
Add a comment