evolutionary-policy-optimization 0.0.41__py3-none-any.whl → 0.0.43__py3-none-any.whl

evolutionary_policy_optimization/epo.py

@@ -875,25 +875,32 @@ def create_agent(
     actor_num_actions,
     actor_dim_hiddens: int | tuple[int, ...],
     critic_dim_hiddens: int | tuple[int, ...],
+    latent_gene_pool_kwargs: dict = dict(),
+    actor_kwargs: dict = dict(),
+    critic_kwargs: dict = dict(),
 ) -> Agent:
 
     latent_gene_pool = LatentGenePool(
         num_latents = num_latents,
-        dim_latent = dim_latent
+        dim_latent = dim_latent,
+        **latent_gene_pool_kwargs
     )
 
     actor = Actor(
         num_actions = actor_num_actions,
         dim_state = dim_state,
         dim_latent = dim_latent,
-        dim_hiddens = actor_dim_hiddens
+        dim_hiddens = actor_dim_hiddens,
+        **actor_kwargs
     )
 
     critic = Critic(
         dim_state = dim_state,
         dim_latent = dim_latent,
-        dim_hiddens = critic_dim_hiddens
-    )                
+        dim_hiddens = critic_dim_hiddens,
+        **critic_kwargs
+    )
+
     return Agent(actor = actor, critic = critic, latent_gene_pool = latent_gene_pool)
 
 # EPO - which is just PPO with natural selection of a population of latent variables conditioning the agent

evolutionary_policy_optimization/mock_env.py

@@ -7,7 +7,7 @@ from torch.nn import Module
 # functions
 
 def cast_tuple(v):
-    return v if isinstance(v, tuple) else v\
+    return v if isinstance(v, tuple) else (v,)
 
 # mock env
 

{evolutionary_policy_optimization-0.0.41.dist-info → evolutionary_policy_optimization-0.0.43.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: evolutionary-policy-optimization
-Version: 0.0.41
+Version: 0.0.43
 Summary: EPO - Pytorch
 Project-URL: Homepage, https://pypi.org/project/evolutionary-policy-optimization/
 Project-URL: Repository, https://github.com/lucidrains/evolutionary-policy-optimization

evolutionary_policy_optimization-0.0.43.dist-info/RECORD

@@ -0,0 +1,8 @@
+evolutionary_policy_optimization/__init__.py,sha256=0q0aBuFgWi06MLMD8FiHzBYQ3_W4LYWrwmCtF3u5H2A,201
+evolutionary_policy_optimization/epo.py,sha256=Yf-iw1gqmAUEVzg6_PwYy-q4005eroZKUYGxNgwCsKk,30440
+evolutionary_policy_optimization/experimental.py,sha256=9FrJGviLESlYysHI3i83efT9g2ZB9ha4u3K9HXN98_w,1100
+evolutionary_policy_optimization/mock_env.py,sha256=6AIc4mwL_C6JkAxwESJgCLxXHMzCAu2FcffVg3HkSm0,920
+evolutionary_policy_optimization-0.0.43.dist-info/METADATA,sha256=pMVLppijepjmI1A9wVVhdX2IXo4BNPsOozpMAAsS6Lo,6213
+evolutionary_policy_optimization-0.0.43.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+evolutionary_policy_optimization-0.0.43.dist-info/licenses/LICENSE,sha256=1yCiA9b5nhslTavxPjsQAO-wpOnwJR9-l8LTVi7GJuk,1066
+evolutionary_policy_optimization-0.0.43.dist-info/RECORD,,

evolutionary_policy_optimization-0.0.41.dist-info/RECORD

@@ -1,8 +0,0 @@
-evolutionary_policy_optimization/__init__.py,sha256=0q0aBuFgWi06MLMD8FiHzBYQ3_W4LYWrwmCtF3u5H2A,201
-evolutionary_policy_optimization/epo.py,sha256=GL3nH5crOj4y_Amu2BY0s95MJL7F2t-X085y40SgUK0,30260
-evolutionary_policy_optimization/experimental.py,sha256=9FrJGviLESlYysHI3i83efT9g2ZB9ha4u3K9HXN98_w,1100
-evolutionary_policy_optimization/mock_env.py,sha256=QqVPZVJtrvQmSDcnYDTob_A5sDwiUzGj6_tmo6BII5c,918
-evolutionary_policy_optimization-0.0.41.dist-info/METADATA,sha256=TFKI2B2PeyU6pHwqmCu130k-U2Li_QmUkvVB39-4uDw,6213
-evolutionary_policy_optimization-0.0.41.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-evolutionary_policy_optimization-0.0.41.dist-info/licenses/LICENSE,sha256=1yCiA9b5nhslTavxPjsQAO-wpOnwJR9-l8LTVi7GJuk,1066
-evolutionary_policy_optimization-0.0.41.dist-info/RECORD,,