Upload A2C PandaReach (VecEnv + VecNormalize)
Browse files- README.md +1 -1
- a2c-panda-reach.zip +2 -2
- a2c-panda-reach/data +30 -24
- a2c-panda-reach/policy.optimizer.pth +1 -1
- a2c-panda-reach/policy.pth +1 -1
- config.json +1 -1
- results.json +1 -1
README.md
CHANGED
|
@@ -16,7 +16,7 @@ model-index:
|
|
| 16 |
type: PandaReachJointsDense-v3
|
| 17 |
metrics:
|
| 18 |
- type: mean_reward
|
| 19 |
-
value: -
|
| 20 |
name: mean_reward
|
| 21 |
verified: false
|
| 22 |
---
|
|
|
|
| 16 |
type: PandaReachJointsDense-v3
|
| 17 |
metrics:
|
| 18 |
- type: mean_reward
|
| 19 |
+
value: -7.73 +/- 5.33
|
| 20 |
name: mean_reward
|
| 21 |
verified: false
|
| 22 |
---
|
a2c-panda-reach.zip
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8468d15ed04e27145b2fc5c293479d65b5090bc196c1d89fa0b98f22468a5ce9
|
| 3 |
+
size 118558
|
a2c-panda-reach/data
CHANGED
|
@@ -5,10 +5,10 @@
|
|
| 5 |
"__module__": "stable_baselines3.common.policies",
|
| 6 |
"__firstlineno__": 839,
|
| 7 |
"__doc__": "\nMultiInputActorClass policy class for actor-critic algorithms (has both policy and value prediction).\nUsed by A2C, PPO and the likes.\n\n:param observation_space: Observation space (Tuple)\n:param action_space: Action space\n:param lr_schedule: Learning rate schedule (could be constant)\n:param net_arch: The specification of the policy and value networks.\n:param activation_fn: Activation function\n:param ortho_init: Whether to use or not orthogonal initialization\n:param use_sde: Whether to use State Dependent Exploration or not\n:param log_std_init: Initial value for the log standard deviation\n:param full_std: Whether to use (n_features x n_actions) parameters\n for the std instead of only (n_features,) when using gSDE\n:param use_expln: Use ``expln()`` function instead of ``exp()`` to ensure\n a positive standard deviation (cf paper). It allows to keep variance\n above zero and prevent it from growing too fast. In practice, ``exp()`` is usually enough.\n:param squash_output: Whether to squash the output using a tanh function,\n this allows to ensure boundaries when using gSDE.\n:param features_extractor_class: Uses the CombinedExtractor\n:param features_extractor_kwargs: Keyword arguments\n to pass to the features extractor.\n:param share_features_extractor: If True, the features extractor is shared between the policy and value networks.\n:param normalize_images: Whether to normalize images or not,\n dividing by 255.0 (True by default)\n:param optimizer_class: The optimizer to use,\n ``th.optim.Adam`` by default\n:param optimizer_kwargs: Additional keyword arguments,\n excluding the learning rate, to pass to the optimizer\n",
|
| 8 |
-
"__init__": "<function MultiInputActorCriticPolicy.__init__ at
|
| 9 |
"__static_attributes__": [],
|
| 10 |
"__abstractmethods__": "frozenset()",
|
| 11 |
-
"_abc_impl": "<_abc._abc_data object at
|
| 12 |
},
|
| 13 |
"verbose": 1,
|
| 14 |
"policy_kwargs": {
|
|
@@ -21,40 +21,46 @@
|
|
| 21 |
"weight_decay": 0
|
| 22 |
}
|
| 23 |
},
|
| 24 |
-
"num_timesteps":
|
| 25 |
-
"_total_timesteps":
|
| 26 |
"_num_timesteps_at_start": 0,
|
| 27 |
"seed": 0,
|
| 28 |
"action_noise": null,
|
| 29 |
-
"start_time":
|
| 30 |
"learning_rate": 0.0007,
|
| 31 |
"tensorboard_log": null,
|
| 32 |
"_last_obs": {
|
| 33 |
":type:": "<class 'collections.OrderedDict'>",
|
| 34 |
-
":serialized:": "
|
| 35 |
-
"achieved_goal": "[[ 0.
|
| 36 |
-
"desired_goal": "[[ 0.
|
| 37 |
-
"observation": "[[ 0.
|
| 38 |
},
|
| 39 |
"_last_episode_starts": {
|
| 40 |
":type:": "<class 'numpy.ndarray'>",
|
| 41 |
-
":serialized:": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
},
|
| 43 |
-
"_last_original_obs": null,
|
| 44 |
"_episode_num": 0,
|
| 45 |
"use_sde": false,
|
| 46 |
"sde_sample_freq": -1,
|
| 47 |
-
"_current_progress_remaining": 0.
|
| 48 |
"_stats_window_size": 100,
|
| 49 |
"ep_info_buffer": {
|
| 50 |
":type:": "<class 'collections.deque'>",
|
| 51 |
-
":serialized:": "
|
| 52 |
},
|
| 53 |
"ep_success_buffer": {
|
| 54 |
":type:": "<class 'collections.deque'>",
|
| 55 |
-
":serialized:": "
|
| 56 |
},
|
| 57 |
-
"_n_updates":
|
| 58 |
"observation_space": {
|
| 59 |
":type:": "<class 'gymnasium.spaces.dict.Dict'>",
|
| 60 |
":serialized:": "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",
|
|
@@ -78,10 +84,10 @@
|
|
| 78 |
"high_repr": "1.0",
|
| 79 |
"_np_random": "Generator(PCG64)"
|
| 80 |
},
|
| 81 |
-
"n_envs":
|
| 82 |
-
"n_steps":
|
| 83 |
"gamma": 0.99,
|
| 84 |
-
"gae_lambda":
|
| 85 |
"ent_coef": 0.0,
|
| 86 |
"vf_coef": 0.5,
|
| 87 |
"max_grad_norm": 0.5,
|
|
@@ -92,11 +98,11 @@
|
|
| 92 |
"__firstlineno__": 697,
|
| 93 |
"__annotations__": "{'observation_space': <class 'gymnasium.spaces.dict.Dict'>, 'obs_shape': dict[str, tuple[int, ...]], 'observations': dict[str, numpy.ndarray]}",
|
| 94 |
"__doc__": "\nDict Rollout buffer used in on-policy algorithms like A2C/PPO.\nExtends the RolloutBuffer to use dictionary observations\n\nIt corresponds to ``buffer_size`` transitions collected\nusing the current policy.\nThis experience will be discarded after the policy update.\nIn order to use PPO objective, we also store the current value of each state\nand the log probability of each taken action.\n\nThe term rollout here refers to the model-free notion and should not\nbe used with the concept of rollout used in model-based RL or planning.\nHence, it is only involved in policy and value function training but not action selection.\n\n:param buffer_size: Max number of element in the buffer\n:param observation_space: Observation space\n:param action_space: Action space\n:param device: PyTorch device\n:param gae_lambda: Factor for trade-off of bias vs variance for Generalized Advantage Estimator\n Equivalent to Monte-Carlo advantage estimate when set to 1.\n:param gamma: Discount factor\n:param n_envs: Number of parallel environments\n",
|
| 95 |
-
"__init__": "<function DictRolloutBuffer.__init__ at
|
| 96 |
-
"reset": "<function DictRolloutBuffer.reset at
|
| 97 |
-
"add": "<function DictRolloutBuffer.add at
|
| 98 |
-
"get": "<function DictRolloutBuffer.get at
|
| 99 |
-
"_get_samples": "<function DictRolloutBuffer._get_samples at
|
| 100 |
"__static_attributes__": [
|
| 101 |
"actions",
|
| 102 |
"advantages",
|
|
@@ -112,7 +118,7 @@
|
|
| 112 |
"values"
|
| 113 |
],
|
| 114 |
"__abstractmethods__": "frozenset()",
|
| 115 |
-
"_abc_impl": "<_abc._abc_data object at
|
| 116 |
},
|
| 117 |
"rollout_buffer_kwargs": {},
|
| 118 |
"normalize_advantage": false,
|
|
|
|
| 5 |
"__module__": "stable_baselines3.common.policies",
|
| 6 |
"__firstlineno__": 839,
|
| 7 |
"__doc__": "\nMultiInputActorClass policy class for actor-critic algorithms (has both policy and value prediction).\nUsed by A2C, PPO and the likes.\n\n:param observation_space: Observation space (Tuple)\n:param action_space: Action space\n:param lr_schedule: Learning rate schedule (could be constant)\n:param net_arch: The specification of the policy and value networks.\n:param activation_fn: Activation function\n:param ortho_init: Whether to use or not orthogonal initialization\n:param use_sde: Whether to use State Dependent Exploration or not\n:param log_std_init: Initial value for the log standard deviation\n:param full_std: Whether to use (n_features x n_actions) parameters\n for the std instead of only (n_features,) when using gSDE\n:param use_expln: Use ``expln()`` function instead of ``exp()`` to ensure\n a positive standard deviation (cf paper). It allows to keep variance\n above zero and prevent it from growing too fast. In practice, ``exp()`` is usually enough.\n:param squash_output: Whether to squash the output using a tanh function,\n this allows to ensure boundaries when using gSDE.\n:param features_extractor_class: Uses the CombinedExtractor\n:param features_extractor_kwargs: Keyword arguments\n to pass to the features extractor.\n:param share_features_extractor: If True, the features extractor is shared between the policy and value networks.\n:param normalize_images: Whether to normalize images or not,\n dividing by 255.0 (True by default)\n:param optimizer_class: The optimizer to use,\n ``th.optim.Adam`` by default\n:param optimizer_kwargs: Additional keyword arguments,\n excluding the learning rate, to pass to the optimizer\n",
|
| 8 |
+
"__init__": "<function MultiInputActorCriticPolicy.__init__ at 0x0000017EAFEE7740>",
|
| 9 |
"__static_attributes__": [],
|
| 10 |
"__abstractmethods__": "frozenset()",
|
| 11 |
+
"_abc_impl": "<_abc._abc_data object at 0x0000017EAFEE8880>"
|
| 12 |
},
|
| 13 |
"verbose": 1,
|
| 14 |
"policy_kwargs": {
|
|
|
|
| 21 |
"weight_decay": 0
|
| 22 |
}
|
| 23 |
},
|
| 24 |
+
"num_timesteps": 500736,
|
| 25 |
+
"_total_timesteps": 500000,
|
| 26 |
"_num_timesteps_at_start": 0,
|
| 27 |
"seed": 0,
|
| 28 |
"action_noise": null,
|
| 29 |
+
"start_time": 1772338710485410200,
|
| 30 |
"learning_rate": 0.0007,
|
| 31 |
"tensorboard_log": null,
|
| 32 |
"_last_obs": {
|
| 33 |
":type:": "<class 'collections.OrderedDict'>",
|
| 34 |
+
":serialized:": "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",
|
| 35 |
+
"achieved_goal": "[[-0.08335891 0.44960478 -0.02665034]\n [-0.58399945 -0.6678931 0.7820425 ]\n [ 0.40787783 0.01746196 0.50902754]\n [ 0.9704461 -1.2090456 0.3233709 ]\n [ 0.0489744 -0.4302222 0.54947084]\n [ 0.31267768 -0.5565364 -1.4349449 ]\n [ 0.704541 1.0311046 -0.07311366]\n [ 0.45444518 0.43149787 0.17831787]]",
|
| 36 |
+
"desired_goal": "[[-0.4154275 1.5640628 -0.2897496 ]\n [-1.2931356 -0.40794566 1.466245 ]\n [-0.19874583 0.65407354 0.6556445 ]\n [ 1.4411088 -1.1204636 0.07718875]\n [-1.2675456 -0.68667173 0.853017 ]\n [ 0.7859618 -0.26621374 -1.3615255 ]\n [ 0.9580833 0.5618754 -0.89246976]\n [ 1.4193127 1.0197951 -0.96095526]]",
|
| 37 |
+
"observation": "[[-0.08335891 0.44960478 -0.02665034 -2.1190984 1.8746047 -2.1056726 ]\n [-0.58399945 -0.6678931 0.7820425 -0.28948823 -1.7179918 0.78301406]\n [ 0.40787783 0.01746196 0.50902754 0.43464282 0.00255363 0.32012123]\n [ 0.9704461 -1.2090456 0.3233709 0.68215084 -1.1092457 0.17022473]\n [ 0.0489744 -0.4302222 0.54947084 -0.42583632 -1.9413825 0.65597755]\n [ 0.31267768 -0.5565364 -1.4349449 0.5830222 0.1626331 -0.5194508 ]\n [ 0.704541 1.0311046 -0.07311366 0.79510355 0.35738096 -0.19477238]\n [ 0.45444518 0.43149787 0.17831787 -0.2028764 1.8912816 -1.0641046 ]]"
|
| 38 |
},
|
| 39 |
"_last_episode_starts": {
|
| 40 |
":type:": "<class 'numpy.ndarray'>",
|
| 41 |
+
":serialized:": "gAWVfAAAAAAAAACME251bXB5Ll9jb3JlLm51bWVyaWOUjAtfZnJvbWJ1ZmZlcpSTlCiWCAAAAAAAAAAAAAEAAAAAAJSMBW51bXB5lIwFZHR5cGWUk5SMAmIxlImIh5RSlChLA4wBfJROTk5K/////0r/////SwB0lGJLCIWUjAFDlHSUUpQu"
|
| 42 |
+
},
|
| 43 |
+
"_last_original_obs": {
|
| 44 |
+
":type:": "<class 'collections.OrderedDict'>",
|
| 45 |
+
":serialized:": "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",
|
| 46 |
+
"achieved_goal": "[[ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]]",
|
| 47 |
+
"desired_goal": "[[ 0.04108851 -0.06906398 0.01229206]\n [ 0.00354649 0.13513911 0.04324788]\n [-0.07151636 -0.06045266 0.24426772]\n [-0.12430525 -0.07895685 0.24038234]\n [ 0.13291684 0.00339827 0.2928731 ]\n [ 0.09150088 0.09238224 0.15459767]\n [ 0.01144931 -0.04701874 0.11072017]\n [ 0.03752864 0.11916414 0.23270571]]",
|
| 48 |
+
"observation": "[[ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]]"
|
| 49 |
},
|
|
|
|
| 50 |
"_episode_num": 0,
|
| 51 |
"use_sde": false,
|
| 52 |
"sde_sample_freq": -1,
|
| 53 |
+
"_current_progress_remaining": -0.0014719999999999178,
|
| 54 |
"_stats_window_size": 100,
|
| 55 |
"ep_info_buffer": {
|
| 56 |
":type:": "<class 'collections.deque'>",
|
| 57 |
+
":serialized:": "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"
|
| 58 |
},
|
| 59 |
"ep_success_buffer": {
|
| 60 |
":type:": "<class 'collections.deque'>",
|
| 61 |
+
":serialized:": "gAWVhgAAAAAAAACMC2NvbGxlY3Rpb25zlIwFZGVxdWWUk5QpS2SGlFKUKIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIhlLg=="
|
| 62 |
},
|
| 63 |
+
"_n_updates": 489,
|
| 64 |
"observation_space": {
|
| 65 |
":type:": "<class 'gymnasium.spaces.dict.Dict'>",
|
| 66 |
":serialized:": "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",
|
|
|
|
| 84 |
"high_repr": "1.0",
|
| 85 |
"_np_random": "Generator(PCG64)"
|
| 86 |
},
|
| 87 |
+
"n_envs": 8,
|
| 88 |
+
"n_steps": 128,
|
| 89 |
"gamma": 0.99,
|
| 90 |
+
"gae_lambda": 0.95,
|
| 91 |
"ent_coef": 0.0,
|
| 92 |
"vf_coef": 0.5,
|
| 93 |
"max_grad_norm": 0.5,
|
|
|
|
| 98 |
"__firstlineno__": 697,
|
| 99 |
"__annotations__": "{'observation_space': <class 'gymnasium.spaces.dict.Dict'>, 'obs_shape': dict[str, tuple[int, ...]], 'observations': dict[str, numpy.ndarray]}",
|
| 100 |
"__doc__": "\nDict Rollout buffer used in on-policy algorithms like A2C/PPO.\nExtends the RolloutBuffer to use dictionary observations\n\nIt corresponds to ``buffer_size`` transitions collected\nusing the current policy.\nThis experience will be discarded after the policy update.\nIn order to use PPO objective, we also store the current value of each state\nand the log probability of each taken action.\n\nThe term rollout here refers to the model-free notion and should not\nbe used with the concept of rollout used in model-based RL or planning.\nHence, it is only involved in policy and value function training but not action selection.\n\n:param buffer_size: Max number of element in the buffer\n:param observation_space: Observation space\n:param action_space: Action space\n:param device: PyTorch device\n:param gae_lambda: Factor for trade-off of bias vs variance for Generalized Advantage Estimator\n Equivalent to Monte-Carlo advantage estimate when set to 1.\n:param gamma: Discount factor\n:param n_envs: Number of parallel environments\n",
|
| 101 |
+
"__init__": "<function DictRolloutBuffer.__init__ at 0x0000017EAFE36200>",
|
| 102 |
+
"reset": "<function DictRolloutBuffer.reset at 0x0000017EAFE362A0>",
|
| 103 |
+
"add": "<function DictRolloutBuffer.add at 0x0000017EAFE363E0>",
|
| 104 |
+
"get": "<function DictRolloutBuffer.get at 0x0000017EAFE36480>",
|
| 105 |
+
"_get_samples": "<function DictRolloutBuffer._get_samples at 0x0000017EAFE36520>",
|
| 106 |
"__static_attributes__": [
|
| 107 |
"actions",
|
| 108 |
"advantages",
|
|
|
|
| 118 |
"values"
|
| 119 |
],
|
| 120 |
"__abstractmethods__": "frozenset()",
|
| 121 |
+
"_abc_impl": "<_abc._abc_data object at 0x0000017EAFE25B40>"
|
| 122 |
},
|
| 123 |
"rollout_buffer_kwargs": {},
|
| 124 |
"normalize_advantage": false,
|
a2c-panda-reach/policy.optimizer.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 49685
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38993cbd1b88efb495fbad23fe4188ab5aa9b077453e16893e95ccaaf8a715d8
|
| 3 |
size 49685
|
a2c-panda-reach/policy.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 47676
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f780587eef1383ad06863c40c553738986c41ca96508d59c28ffa1d611b0b7f3
|
| 3 |
size 47676
|
config.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"policy_class": {":type:": "<class 'abc.ABCMeta'>", ":serialized:": "gAWVRQAAAAAAAACMIXN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi5wb2xpY2llc5SMG011bHRpSW5wdXRBY3RvckNyaXRpY1BvbGljeZSTlC4=", "__module__": "stable_baselines3.common.policies", "__firstlineno__": 839, "__doc__": "\nMultiInputActorClass policy class for actor-critic algorithms (has both policy and value prediction).\nUsed by A2C, PPO and the likes.\n\n:param observation_space: Observation space (Tuple)\n:param action_space: Action space\n:param lr_schedule: Learning rate schedule (could be constant)\n:param net_arch: The specification of the policy and value networks.\n:param activation_fn: Activation function\n:param ortho_init: Whether to use or not orthogonal initialization\n:param use_sde: Whether to use State Dependent Exploration or not\n:param log_std_init: Initial value for the log standard deviation\n:param full_std: Whether to use (n_features x n_actions) parameters\n for the std instead of only (n_features,) when using gSDE\n:param use_expln: Use ``expln()`` function instead of ``exp()`` to ensure\n a positive standard deviation (cf paper). It allows to keep variance\n above zero and prevent it from growing too fast. In practice, ``exp()`` is usually enough.\n:param squash_output: Whether to squash the output using a tanh function,\n this allows to ensure boundaries when using gSDE.\n:param features_extractor_class: Uses the CombinedExtractor\n:param features_extractor_kwargs: Keyword arguments\n to pass to the features extractor.\n:param share_features_extractor: If True, the features extractor is shared between the policy and value networks.\n:param normalize_images: Whether to normalize images or not,\n dividing by 255.0 (True by default)\n:param optimizer_class: The optimizer to use,\n ``th.optim.Adam`` by default\n:param optimizer_kwargs: Additional keyword arguments,\n excluding the learning rate, to pass to the optimizer\n", "__init__": "<function MultiInputActorCriticPolicy.__init__ at 0x000001DD1AF03740>", "__static_attributes__": [], "__abstractmethods__": "frozenset()", "_abc_impl": "<_abc._abc_data object at 0x000001DD1AF08280>"}, "verbose": 1, "policy_kwargs": {":type:": "<class 'dict'>", ":serialized:": "gAWVgQAAAAAAAAB9lCiMD29wdGltaXplcl9jbGFzc5SME3RvcmNoLm9wdGltLnJtc3Byb3CUjAdSTVNwcm9wlJOUjBBvcHRpbWl6ZXJfa3dhcmdzlH2UKIwFYWxwaGGURz/vrhR64UeujANlcHOURz7k+LWI42jxjAx3ZWlnaHRfZGVjYXmUSwB1dS4=", "optimizer_class": "<class 'torch.optim.rmsprop.RMSprop'>", "optimizer_kwargs": {"alpha": 0.99, "eps": 1e-05, "weight_decay": 0}}, "num_timesteps": 2000000, "_total_timesteps": 2000000, "_num_timesteps_at_start": 0, "seed": 0, "action_noise": null, "start_time": 1772318598024917700, "learning_rate": 0.0007, "tensorboard_log": null, "_last_obs": {":type:": "<class 'collections.OrderedDict'>", ":serialized:": "gAWVLAEAAAAAAACMC2NvbGxlY3Rpb25zlIwLT3JkZXJlZERpY3SUk5QpUpQojA1hY2hpZXZlZF9nb2FslIwTbnVtcHkuX2NvcmUubnVtZXJpY5SMC19mcm9tYnVmZmVylJOUKJYMAAAAAAAAANtwyj3ZY1++wpArPpSMBW51bXB5lIwFZHR5cGWUk5SMAmY0lImIh5RSlChLA4wBPJROTk5K/////0r/////SwB0lGJLAUsDhpSMAUOUdJRSlIwMZGVzaXJlZF9nb2FslGgHKJYMAAAAAAAAAHKqqD0tBVK8sTcTPJRoDksBSwOGlGgSdJRSlIwLb2JzZXJ2YXRpb26UaAcolhgAAAAAAAAA23DKPdljX77CkCs+AI+SPteKGD8XnYS+lGgOSwFLBoaUaBJ0lFKUdS4=", "achieved_goal": "[[ 0.09884807 -0.21815433 0.1675444 ]]", "desired_goal": "[[ 0.08235635 -0.01281862 0.00898545]]", "observation": "[[ 0.09884807 -0.21815433 0.1675444 0.28624725 0.5958685 -0.259011 ]]"}, "_last_episode_starts": {":type:": "<class 'numpy.ndarray'>", ":serialized:": "gAWVdQAAAAAAAACME251bXB5Ll9jb3JlLm51bWVyaWOUjAtfZnJvbWJ1ZmZlcpSTlCiWAQAAAAAAAAAAlIwFbnVtcHmUjAVkdHlwZZSTlIwCYjGUiYiHlFKUKEsDjAF8lE5OTkr/////Sv////9LAHSUYksBhZSMAUOUdJRSlC4="}, "_last_original_obs": null, "_episode_num": 0, "use_sde": false, "sde_sample_freq": -1, "_current_progress_remaining": 0.0, "_stats_window_size": 100, "ep_info_buffer": {":type:": "<class 'collections.deque'>", ":serialized:": "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"}, "ep_success_buffer": {":type:": "<class 'collections.deque'>", ":serialized:": "gAWVhgAAAAAAAACMC2NvbGxlY3Rpb25zlIwFZGVxdWWUk5QpS2SGlFKUKIiIiImIiImJiYmIiYmIiYiJiYiJiYiIiYiJiYiIiYiIiImIiIiIiIiIiIiIiIiIiIiJiIiIiYmJiYiIiImIiYmIiImJiIiIiIiIiYiIiYiIiImIiIiIiIiJiYiJiIiIiImJiIhlLg=="}, "_n_updates": 400000, "observation_space": {":type:": "<class 'gymnasium.spaces.dict.Dict'>", ":serialized:": "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", "spaces": "{'achieved_goal': Box(-10.0, 10.0, (3,), float32), 'desired_goal': Box(-10.0, 10.0, (3,), float32), 'observation': Box(-10.0, 10.0, (6,), float32)}", "_shape": null, "dtype": null, "_np_random": null}, "action_space": {":type:": "<class 'gymnasium.spaces.box.Box'>", ":serialized:": "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", "dtype": "float32", "_shape": [7], "low": "[-1. -1. -1. -1. -1. -1. -1.]", "bounded_below": "[ True True True True True True True]", "high": "[1. 1. 1. 1. 1. 1. 1.]", "bounded_above": "[ True True True True True True True]", "low_repr": "-1.0", "high_repr": "1.0", "_np_random": "Generator(PCG64)"}, "n_envs": 1, "n_steps": 5, "gamma": 0.99, "gae_lambda": 1.0, "ent_coef": 0.0, "vf_coef": 0.5, "max_grad_norm": 0.5, "rollout_buffer_class": {":type:": "<class 'abc.ABCMeta'>", ":serialized:": "gAWVOgAAAAAAAACMIHN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi5idWZmZXJzlIwRRGljdFJvbGxvdXRCdWZmZXKUk5Qu", "__module__": "stable_baselines3.common.buffers", "__firstlineno__": 697, "__annotations__": "{'observation_space': <class 'gymnasium.spaces.dict.Dict'>, 'obs_shape': dict[str, tuple[int, ...]], 'observations': dict[str, numpy.ndarray]}", "__doc__": "\nDict Rollout buffer used in on-policy algorithms like A2C/PPO.\nExtends the RolloutBuffer to use dictionary observations\n\nIt corresponds to ``buffer_size`` transitions collected\nusing the current policy.\nThis experience will be discarded after the policy update.\nIn order to use PPO objective, we also store the current value of each state\nand the log probability of each taken action.\n\nThe term rollout here refers to the model-free notion and should not\nbe used with the concept of rollout used in model-based RL or planning.\nHence, it is only involved in policy and value function training but not action selection.\n\n:param buffer_size: Max number of element in the buffer\n:param observation_space: Observation space\n:param action_space: Action space\n:param device: PyTorch device\n:param gae_lambda: Factor for trade-off of bias vs variance for Generalized Advantage Estimator\n Equivalent to Monte-Carlo advantage estimate when set to 1.\n:param gamma: Discount factor\n:param n_envs: Number of parallel environments\n", "__init__": "<function DictRolloutBuffer.__init__ at 0x000001DD1AE52200>", "reset": "<function DictRolloutBuffer.reset at 0x000001DD1AE522A0>", "add": "<function DictRolloutBuffer.add at 0x000001DD1AE523E0>", "get": "<function DictRolloutBuffer.get at 0x000001DD1AE52480>", "_get_samples": "<function DictRolloutBuffer._get_samples at 0x000001DD1AE52520>", "__static_attributes__": ["actions", "advantages", "episode_starts", "full", "gae_lambda", "gamma", "generator_ready", "log_probs", "observations", "returns", "rewards", "values"], "__abstractmethods__": "frozenset()", "_abc_impl": "<_abc._abc_data object at 0x000001DD1AE49580>"}, "rollout_buffer_kwargs": {}, "normalize_advantage": false, "lr_schedule": {":type:": "<class 'stable_baselines3.common.utils.FloatSchedule'>", ":serialized:": "gAWVeQAAAAAAAACMHnN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi51dGlsc5SMDUZsb2F0U2NoZWR1bGWUk5QpgZR9lIwOdmFsdWVfc2NoZWR1bGWUaACMEENvbnN0YW50U2NoZWR1bGWUk5QpgZR9lIwDdmFslEc/RvAGjbi6x3Nic2Iu", "value_schedule": "ConstantSchedule(val=0.0007)"}, "system_info": {"OS": "Windows-10-10.0.19045-SP0 10.0.19045", "Python": "3.13.7", "Stable-Baselines3": "2.7.1", "PyTorch": "2.9.0+cpu", "GPU Enabled": "False", "Numpy": "2.2.6", "Cloudpickle": "3.1.2", "Gymnasium": "1.2.3"}}
|
|
|
|
| 1 |
+
{"policy_class": {":type:": "<class 'abc.ABCMeta'>", ":serialized:": "gAWVRQAAAAAAAACMIXN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi5wb2xpY2llc5SMG011bHRpSW5wdXRBY3RvckNyaXRpY1BvbGljeZSTlC4=", "__module__": "stable_baselines3.common.policies", "__firstlineno__": 839, "__doc__": "\nMultiInputActorClass policy class for actor-critic algorithms (has both policy and value prediction).\nUsed by A2C, PPO and the likes.\n\n:param observation_space: Observation space (Tuple)\n:param action_space: Action space\n:param lr_schedule: Learning rate schedule (could be constant)\n:param net_arch: The specification of the policy and value networks.\n:param activation_fn: Activation function\n:param ortho_init: Whether to use or not orthogonal initialization\n:param use_sde: Whether to use State Dependent Exploration or not\n:param log_std_init: Initial value for the log standard deviation\n:param full_std: Whether to use (n_features x n_actions) parameters\n for the std instead of only (n_features,) when using gSDE\n:param use_expln: Use ``expln()`` function instead of ``exp()`` to ensure\n a positive standard deviation (cf paper). It allows to keep variance\n above zero and prevent it from growing too fast. In practice, ``exp()`` is usually enough.\n:param squash_output: Whether to squash the output using a tanh function,\n this allows to ensure boundaries when using gSDE.\n:param features_extractor_class: Uses the CombinedExtractor\n:param features_extractor_kwargs: Keyword arguments\n to pass to the features extractor.\n:param share_features_extractor: If True, the features extractor is shared between the policy and value networks.\n:param normalize_images: Whether to normalize images or not,\n dividing by 255.0 (True by default)\n:param optimizer_class: The optimizer to use,\n ``th.optim.Adam`` by default\n:param optimizer_kwargs: Additional keyword arguments,\n excluding the learning rate, to pass to the optimizer\n", "__init__": "<function MultiInputActorCriticPolicy.__init__ at 0x0000017EAFEE7740>", "__static_attributes__": [], "__abstractmethods__": "frozenset()", "_abc_impl": "<_abc._abc_data object at 0x0000017EAFEE8880>"}, "verbose": 1, "policy_kwargs": {":type:": "<class 'dict'>", ":serialized:": "gAWVgQAAAAAAAAB9lCiMD29wdGltaXplcl9jbGFzc5SME3RvcmNoLm9wdGltLnJtc3Byb3CUjAdSTVNwcm9wlJOUjBBvcHRpbWl6ZXJfa3dhcmdzlH2UKIwFYWxwaGGURz/vrhR64UeujANlcHOURz7k+LWI42jxjAx3ZWlnaHRfZGVjYXmUSwB1dS4=", "optimizer_class": "<class 'torch.optim.rmsprop.RMSprop'>", "optimizer_kwargs": {"alpha": 0.99, "eps": 1e-05, "weight_decay": 0}}, "num_timesteps": 500736, "_total_timesteps": 500000, "_num_timesteps_at_start": 0, "seed": 0, "action_noise": null, "start_time": 1772338710485410200, "learning_rate": 0.0007, "tensorboard_log": null, "_last_obs": {":type:": "<class 'collections.OrderedDict'>", ":serialized:": "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", "achieved_goal": "[[-0.08335891 0.44960478 -0.02665034]\n [-0.58399945 -0.6678931 0.7820425 ]\n [ 0.40787783 0.01746196 0.50902754]\n [ 0.9704461 -1.2090456 0.3233709 ]\n [ 0.0489744 -0.4302222 0.54947084]\n [ 0.31267768 -0.5565364 -1.4349449 ]\n [ 0.704541 1.0311046 -0.07311366]\n [ 0.45444518 0.43149787 0.17831787]]", "desired_goal": "[[-0.4154275 1.5640628 -0.2897496 ]\n [-1.2931356 -0.40794566 1.466245 ]\n [-0.19874583 0.65407354 0.6556445 ]\n [ 1.4411088 -1.1204636 0.07718875]\n [-1.2675456 -0.68667173 0.853017 ]\n [ 0.7859618 -0.26621374 -1.3615255 ]\n [ 0.9580833 0.5618754 -0.89246976]\n [ 1.4193127 1.0197951 -0.96095526]]", "observation": "[[-0.08335891 0.44960478 -0.02665034 -2.1190984 1.8746047 -2.1056726 ]\n [-0.58399945 -0.6678931 0.7820425 -0.28948823 -1.7179918 0.78301406]\n [ 0.40787783 0.01746196 0.50902754 0.43464282 0.00255363 0.32012123]\n [ 0.9704461 -1.2090456 0.3233709 0.68215084 -1.1092457 0.17022473]\n [ 0.0489744 -0.4302222 0.54947084 -0.42583632 -1.9413825 0.65597755]\n [ 0.31267768 -0.5565364 -1.4349449 0.5830222 0.1626331 -0.5194508 ]\n [ 0.704541 1.0311046 -0.07311366 0.79510355 0.35738096 -0.19477238]\n [ 0.45444518 0.43149787 0.17831787 -0.2028764 1.8912816 -1.0641046 ]]"}, "_last_episode_starts": {":type:": "<class 'numpy.ndarray'>", ":serialized:": "gAWVfAAAAAAAAACME251bXB5Ll9jb3JlLm51bWVyaWOUjAtfZnJvbWJ1ZmZlcpSTlCiWCAAAAAAAAAAAAAEAAAAAAJSMBW51bXB5lIwFZHR5cGWUk5SMAmIxlImIh5RSlChLA4wBfJROTk5K/////0r/////SwB0lGJLCIWUjAFDlHSUUpQu"}, "_last_original_obs": {":type:": "<class 'collections.OrderedDict'>", ":serialized:": "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", "achieved_goal": "[[ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01]]", "desired_goal": "[[ 0.04108851 -0.06906398 0.01229206]\n [ 0.00354649 0.13513911 0.04324788]\n [-0.07151636 -0.06045266 0.24426772]\n [-0.12430525 -0.07895685 0.24038234]\n [ 0.13291684 0.00339827 0.2928731 ]\n [ 0.09150088 0.09238224 0.15459767]\n [ 0.01144931 -0.04701874 0.11072017]\n [ 0.03752864 0.11916414 0.23270571]]", "observation": "[[ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]\n [ 3.8439669e-02 -2.1944723e-12 1.9740014e-01 0.0000000e+00\n -0.0000000e+00 0.0000000e+00]]"}, "_episode_num": 0, "use_sde": false, "sde_sample_freq": -1, "_current_progress_remaining": -0.0014719999999999178, "_stats_window_size": 100, "ep_info_buffer": {":type:": "<class 'collections.deque'>", ":serialized:": "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"}, "ep_success_buffer": {":type:": "<class 'collections.deque'>", ":serialized:": "gAWVhgAAAAAAAACMC2NvbGxlY3Rpb25zlIwFZGVxdWWUk5QpS2SGlFKUKIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIhlLg=="}, "_n_updates": 489, "observation_space": {":type:": "<class 'gymnasium.spaces.dict.Dict'>", ":serialized:": "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", "spaces": "{'achieved_goal': Box(-10.0, 10.0, (3,), float32), 'desired_goal': Box(-10.0, 10.0, (3,), float32), 'observation': Box(-10.0, 10.0, (6,), float32)}", "_shape": null, "dtype": null, "_np_random": null}, "action_space": {":type:": "<class 'gymnasium.spaces.box.Box'>", ":serialized:": "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", "dtype": "float32", "_shape": [7], "low": "[-1. -1. -1. -1. -1. -1. -1.]", "bounded_below": "[ True True True True True True True]", "high": "[1. 1. 1. 1. 1. 1. 1.]", "bounded_above": "[ True True True True True True True]", "low_repr": "-1.0", "high_repr": "1.0", "_np_random": "Generator(PCG64)"}, "n_envs": 8, "n_steps": 128, "gamma": 0.99, "gae_lambda": 0.95, "ent_coef": 0.0, "vf_coef": 0.5, "max_grad_norm": 0.5, "rollout_buffer_class": {":type:": "<class 'abc.ABCMeta'>", ":serialized:": "gAWVOgAAAAAAAACMIHN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi5idWZmZXJzlIwRRGljdFJvbGxvdXRCdWZmZXKUk5Qu", "__module__": "stable_baselines3.common.buffers", "__firstlineno__": 697, "__annotations__": "{'observation_space': <class 'gymnasium.spaces.dict.Dict'>, 'obs_shape': dict[str, tuple[int, ...]], 'observations': dict[str, numpy.ndarray]}", "__doc__": "\nDict Rollout buffer used in on-policy algorithms like A2C/PPO.\nExtends the RolloutBuffer to use dictionary observations\n\nIt corresponds to ``buffer_size`` transitions collected\nusing the current policy.\nThis experience will be discarded after the policy update.\nIn order to use PPO objective, we also store the current value of each state\nand the log probability of each taken action.\n\nThe term rollout here refers to the model-free notion and should not\nbe used with the concept of rollout used in model-based RL or planning.\nHence, it is only involved in policy and value function training but not action selection.\n\n:param buffer_size: Max number of element in the buffer\n:param observation_space: Observation space\n:param action_space: Action space\n:param device: PyTorch device\n:param gae_lambda: Factor for trade-off of bias vs variance for Generalized Advantage Estimator\n Equivalent to Monte-Carlo advantage estimate when set to 1.\n:param gamma: Discount factor\n:param n_envs: Number of parallel environments\n", "__init__": "<function DictRolloutBuffer.__init__ at 0x0000017EAFE36200>", "reset": "<function DictRolloutBuffer.reset at 0x0000017EAFE362A0>", "add": "<function DictRolloutBuffer.add at 0x0000017EAFE363E0>", "get": "<function DictRolloutBuffer.get at 0x0000017EAFE36480>", "_get_samples": "<function DictRolloutBuffer._get_samples at 0x0000017EAFE36520>", "__static_attributes__": ["actions", "advantages", "episode_starts", "full", "gae_lambda", "gamma", "generator_ready", "log_probs", "observations", "returns", "rewards", "values"], "__abstractmethods__": "frozenset()", "_abc_impl": "<_abc._abc_data object at 0x0000017EAFE25B40>"}, "rollout_buffer_kwargs": {}, "normalize_advantage": false, "lr_schedule": {":type:": "<class 'stable_baselines3.common.utils.FloatSchedule'>", ":serialized:": "gAWVeQAAAAAAAACMHnN0YWJsZV9iYXNlbGluZXMzLmNvbW1vbi51dGlsc5SMDUZsb2F0U2NoZWR1bGWUk5QpgZR9lIwOdmFsdWVfc2NoZWR1bGWUaACMEENvbnN0YW50U2NoZWR1bGWUk5QpgZR9lIwDdmFslEc/RvAGjbi6x3Nic2Iu", "value_schedule": "ConstantSchedule(val=0.0007)"}, "system_info": {"OS": "Windows-10-10.0.19045-SP0 10.0.19045", "Python": "3.13.7", "Stable-Baselines3": "2.7.1", "PyTorch": "2.9.0+cpu", "GPU Enabled": "False", "Numpy": "2.2.6", "Cloudpickle": "3.1.2", "Gymnasium": "1.2.3"}}
|
results.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"mean_reward": -
|
|
|
|
| 1 |
+
{"mean_reward": -7.727921200000002, "std_reward": 5.32886649071387, "is_deterministic": true, "n_eval_episodes": 10, "eval_datetime": "2026-03-01T05:36:47.634036"}
|