agilerl 2.4.2.dev1__tar.gz → 2.4.3.dev0__tar.gz

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/.pre-commit-config.yaml

@@ -24,7 +24,7 @@ repos:
           - id: mixed-line-ending
             args: [--fix=lf]
     - repo: https://github.com/psf/black-pre-commit-mirror
-      rev: 25.12.0
+      rev: 26.1.0
       hooks:
           - id: black
     - repo: https://github.com/codespell-project/codespell
@@ -35,7 +35,7 @@ repos:
                 - --skip=*.css,*.js,*.map,*.scss,*.svg
                 - --ignore-words-list=magent,pres,roate
     - repo: https://github.com/astral-sh/ruff-pre-commit
-      rev: v0.14.9
+      rev: v0.14.14
       hooks:
           - id: ruff-check
             args:
@@ -53,6 +53,6 @@ repos:
           - id: yamlfmt
     - repo: https://github.com/astral-sh/uv-pre-commit
       # uv version.
-      rev: 0.9.27
+      rev: 0.9.28
       hooks:
           - id: uv-lock

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agilerl
-Version: 2.4.2.dev1
+Version: 2.4.3.dev0
 Summary: AgileRL is a deep reinforcement learning library focused on improving RL development through RLOps.
 Author-email: Nick Ustaran-Anderegg <dev@agilerl.com>
 License-Expression: Apache-2.0

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/agilerl/algorithms/core/base.py

@@ -2066,8 +2066,7 @@ class LLMAlgorithm(EvolvableAlgorithm, ABC):
         accelerator: Optional[Accelerator] = None,
     ) -> None:
         raise NotImplementedError(
-            "The load class method is not supported for this algorithm class."
-            """
+            "The load class method is not supported for this algorithm class." """
             To load a saved LLM, please load the model as follows, and then re-instantiate the GRPO
             class, using the pre-trained model.
 

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/agilerl/algorithms/ippo.py

@@ -671,7 +671,7 @@ class IPPO(MultiAgentRLAlgorithm):
         :param action_space: Action space for the agent
         :type action_space: gymnasium.spaces
         """
-        (states, actions, log_probs, rewards, dones, values, next_state, next_done) = (
+        states, actions, log_probs, rewards, dones, values, next_state, next_done = (
             experiences
         )
 

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/agilerl/wrappers/agent.py

@@ -597,8 +597,8 @@ class AsyncAgentsWrapper(AgentWrapper[MultiAgentRLAlgorithm]):
         :return: Learning information
         :rtype: Any
         """
-        (states, actions, log_probs, rewards, dones, values, next_state, next_done) = (
-            map(self.stack_experiences, experiences)
+        states, actions, log_probs, rewards, dones, values, next_state, next_done = map(
+            self.stack_experiences, experiences
         )
 
         # Handle case where we haven't collected a next state for each sub-agent

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "agilerl"
-version = "2.4.2.dev1"
+version = "2.4.3.dev0"
 description = "AgileRL is a deep reinforcement learning library focused on improving RL development through RLOps."
 authors = [{ name = "Nick Ustaran-Anderegg", email = "dev@agilerl.com" }]
 license = "Apache-2.0"

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/tests/test_modules/test_make_evolvable.py

@@ -194,9 +194,7 @@ def test_instantiation_with_rainbow():
         network, input_tensor, support=support, rainbow=True
     )
     assert isinstance(evolvable_network, MakeEvolvable)
-    assert (
-        str(evolvable_network)
-        == """MakeEvolvable(
+    assert str(evolvable_network) == """MakeEvolvable(
   (feature_net): Sequential(
     (feature_linear_layer_0): Linear(in_features=3, out_features=128, bias=True)
     (feature_activation_0): ReLU()
@@ -212,7 +210,6 @@ def test_instantiation_with_rainbow():
     (advantage_linear_layer_output): NoisyLinear(in_features=8, out_features=102)
   )
 )"""
-    )
     del network, evolvable_network
 
 

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/tests/test_train/test_train.py

@@ -1498,8 +1498,10 @@ def test_train_off_policy_agent_calls_made_rainbow(
 def test_train_off_policy_save_elite_warning(
     env, population_off_policy, tournament, mutations, memory
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitnesses = train_off_policy(
             env,
@@ -2137,8 +2139,10 @@ def test_train_on_policy_save_elite_warning(
     tournament,
     mutations,
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitnesses = train_on_policy(
             env,
@@ -2703,8 +2707,10 @@ def test_train_multi_agent_on_policy_rgb_vectorized(
 def test_train_multi_save_elite_warning(
     multi_env, population_multi_agent, on_policy, multi_memory, tournament, mutations
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitnesses = train_multi_agent_off_policy(
             multi_env,
@@ -2730,8 +2736,10 @@ def test_train_multi_save_elite_warning(
 def test_train_multi_save_elite_warning_on_policy(
     multi_env, population_multi_agent, on_policy, multi_memory, tournament, mutations
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitnesses = train_multi_agent_on_policy(
             multi_env,
@@ -3567,8 +3575,10 @@ def test_train_offline_save_elite_warning(
     offline_init_hp,
     dummy_h5py_data,
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitness = train_offline(
             env,
@@ -4057,8 +4067,10 @@ def test_train_bandit_agent_calls_made(
 def test_train_bandit_save_elite_warning(
     bandit_env, population_bandit, tournament, mutations, bandit_memory
 ):
-    warning_string = "'save_elite' set to False but 'elite_path' has been defined, elite will not\
+    warning_string = (
+        "'save_elite' set to False but 'elite_path' has been defined, elite will not\
                       be saved unless 'save_elite' is set to True."
+    )
     with pytest.warns(match=warning_string):
         pop, pop_fitnesses = train_bandits(
             bandit_env,

{agilerl-2.4.2.dev1 → agilerl-2.4.3.dev0}/uv.lock

@@ -53,7 +53,7 @@ wheels = [
 
 [[package]]
 name = "agilerl"
-version = "2.4.2.dev1"
+version = "2.4.3.dev0"
 source = { editable = "." }
 dependencies = [
     { name = "accelerate" },