fix(pu): fix typo

opendilab · Jan 23, 2025 · 11ab64d · 11ab64d
1 parent d717529
commit 11ab64d
Show file tree

Hide file tree

Showing 2 changed files with 1 addition and 2 deletions.
diff --git a/lzero/mcts/buffer/game_buffer_unizero.py b/lzero/mcts/buffer/game_buffer_unizero.py
@@ -522,7 +522,6 @@ def _compute_target_reward_value(self, reward_value_context: List[Any], model: A
             m_output = model.initial_inference(batch_obs, batch_action)
             # ======================================================================
 
-            # if not model.training:
             # if not in training, obtain the scalars of the value/reward
             [m_output.latent_state, m_output.value, m_output.policy_logits] = to_detach_cpu_numpy(
                 [

diff --git a/zoo/atari/config/atari_efficientzero_ddp_config.py b/zoo/atari/config/atari_efficientzero_ddp_config.py
@@ -84,7 +84,7 @@
     Overview:
         This script should be executed with <nproc_per_node> GPUs.
         Run the following command to launch the script:
-        torchrun--nproc_per_node=2 ./LightZero/zoo/atari/config/atari_efficientzero_ddp_config.py
+        torchrun --nproc_per_node=2 ./LightZero/zoo/atari/config/atari_efficientzero_ddp_config.py
     """
     from ding.utils import DDPContext
     from lzero.entry import train_muzero