test hub

README.md

@@ -0,0 +1,36 @@
+---
+env_name: Pendulum-v1
+tags:
+- Pendulum-v1
+- td3
+- reinforcement-learning
+- custom-implementation
+- TD3
+- DDPG
+- Pendulum
+model-index:
+- name: TD3-PendulumV1
+  results:
+  - task:
+      type: reinforcement-learning
+      name: reinforcement-learning
+    dataset:
+      name: Pendulum-v1
+      type: Pendulum-v1
+    metrics:
+    - type: mean_reward
+      value: -127.79 +/- 65.35
+      name: mean_reward
+      verified: false
+---
+
+# **TD3** Agent playing **Pendulum-v1**
+This is a trained model of a **TD3** agent playing **Pendulum-v1**.
+
+## Usage
+
+model = load_from_hub(repo_id="winkin119/TD3-PendulumV1", filename="td3_pendulum.pth")
+
+
+env = gym.make("Pendulum-v1")
+...

eval_result.json

@@ -0,0 +1 @@
+{"mean_reward": -127.78696345508857, "std_reward": 65.34958155884144, "datetime": "2025-07-24 23:20:12", "train_duration_min": "2.48"}

params.json

@@ -0,0 +1,52 @@
+{
+    "env_config": {
+        "env_id": "Pendulum-v1",
+        "env_kwargs": {},
+        "max_steps": null,
+        "use_image": false,
+        "vector_env_num": 6,
+        "use_multi_processing": true,
+        "image_shape": null,
+        "frame_stack": 1,
+        "frame_skip": 1,
+        "training_render_mode": null
+    },
+    "device": "cpu",
+    "learning_rate": 0.0003,
+    "gamma": 0.99,
+    "checkpoint_pathname": "",
+    "eval_episodes": 50,
+    "eval_random_seed": 42,
+    "eval_video_num": 10,
+    "total_steps": 240000,
+    "hidden_sizes": [
+        256,
+        256
+    ],
+    "critic_lr": 0.0003,
+    "replay_buffer_capacity": 240000,
+    "batch_size": 128,
+    "update_start_step": 20000,
+    "policy_delay": 3,
+    "policy_noise": 0.2,
+    "noise_clip": 0.5,
+    "exploration_noise": {
+        "_type": "LinearSchedule",
+        "_module": "practice.utils_for_coding.scheduler_utils",
+        "start_e": 0.3,
+        "end_e": 0.0,
+        "duration": 10000
+    },
+    "max_action": 2.0,
+    "tau": 0.005,
+    "max_grad_norm": 0.5,
+    "env_wrappers": [
+        "PendulumEnv",
+        "PassiveEnvChecker",
+        "OrderEnforcing",
+        "TimeLimit",
+        "CastObsFloat32Wrapper",
+        "RecordEpisodeStatistics",
+        "Autoreset"
+    ]
+}

td3_pendulum.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41931e48f2a454d63856f5fa02f997d62ed44406a5527ce8b2a1c629e5ea6b28
+size 273425

tensorboard/events.out.tfevents.1753370258.winkindeMacBook-Air.local.50207.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75d3019db394dc3c651f88dff4c3b8b01792c9b88cf79ff1a98087c724f5e742
+size 14868184