config(nyz): fix masac smac config multi_agent=True bug (#791)

Author: PaParaZz1

dizoo/smac/config/smac_10m11m_masac_config.py

@@ -26,7 +26,7 @@
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=132,
@@ -52,7 +52,6 @@
         collect=dict(
             env_num=collector_env_num,
             n_sample=1600,
-            unroll_len=1,
         ),
         command=dict(),
         eval=dict(
@@ -88,4 +87,4 @@
 if __name__ == '__main__':
 
     from ding.entry import serial_pipeline
-    serial_pipeline((main_config, create_config), seed=0)
+    serial_pipeline((main_config, create_config), seed=0, max_env_step=int(1e7))

dizoo/smac/config/smac_25m_masac_config.py

@@ -26,7 +26,7 @@
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=306,

dizoo/smac/config/smac_2c64zg_masac_config.py

@@ -26,7 +26,7 @@
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=404,

dizoo/smac/config/smac_3m_masac_config.py

@@ -26,6 +26,7 @@
     ),
     policy=dict(
         cuda=True,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=42,
@@ -41,7 +42,6 @@
             learning_rate_q=5e-4,
             learning_rate_policy=5e-4,
             learning_rate_alpha=5e-4,
-            ignore_done=False,
             target_theta=0.005,
             discount_factor=0.99,
             alpha=0.2,
@@ -63,7 +63,7 @@
                 start=1,
                 end=0.05,
                 decay=100000,
-            ),  # TODO(pu)
+            ),
             replay_buffer=dict(replay_buffer_size=1000000, ),
         ),
     ),

dizoo/smac/config/smac_3s5z_masac_config.py

@@ -26,6 +26,7 @@
     ),
     policy=dict(
         cuda=True,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=150,
@@ -41,7 +42,6 @@
             learning_rate_q=5e-4,
             learning_rate_policy=5e-4,
             learning_rate_alpha=5e-5,
-            ignore_done=False,
             target_theta=0.005,
             discount_factor=0.99,
             alpha=0.2,

dizoo/smac/config/smac_3s5zvs3s6z_masac_config.py

@@ -19,15 +19,14 @@
         stop_value=0.99,
         death_mask=True,
         special_global_state=special_global_state,
-        # save_replay_episodes = 1,
         manager=dict(
             shared_memory=False,
             reset_timeout=6000,
         ),
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=159,
@@ -89,4 +88,4 @@
 if __name__ == '__main__':
 
     from ding.entry import serial_pipeline
-    serial_pipeline((main_config, create_config), seed=0)
+    serial_pipeline((main_config, create_config), seed=0, max_env_step=int(1e7))

dizoo/smac/config/smac_5m6m_masac_config.py

@@ -26,6 +26,7 @@
     ),
     policy=dict(
         cuda=True,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=72,

dizoo/smac/config/smac_8m9m_masac_config.py

@@ -1,5 +1,4 @@
 from easydict import EasyDict
-from ding.entry import serial_pipeline
 
 agent_num = 8
 collector_env_num = 8
@@ -27,7 +26,7 @@
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=108,

dizoo/smac/config/smac_MMM2_masac_config.py

@@ -26,6 +26,7 @@
     ),
     policy=dict(
         cuda=True,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=204,

dizoo/smac/config/smac_MMM_masac_config.py

@@ -26,6 +26,7 @@
     ),
     policy=dict(
         cuda=True,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=186,

dizoo/smac/config/smac_corridor_masac_config.py

@@ -1,5 +1,4 @@
 from easydict import EasyDict
-from ding.entry import serial_pipeline
 
 agent_num = 6
 collector_env_num = 8
@@ -27,7 +26,7 @@
     ),
     policy=dict(
         cuda=True,
-        on_policy=False,
+        multi_agent=True,
         random_collect_size=0,
         model=dict(
             agent_obs_shape=192,
@@ -89,4 +88,4 @@
 if __name__ == '__main__':
 
     from ding.entry import serial_pipeline
-    serial_pipeline((main_config, create_config), seed=0)
+    serial_pipeline((main_config, create_config), seed=0, max_env_step=int(1e7))