Push agent to the Hub

README.md

@@ -1,31 +1,31 @@
 ---
 tags:
-- Pixelcopter-PLE-v0
+- PandaReachDense-v3
 - ppo
 - deep-reinforcement-learning
 - reinforcement-learning
 - custom-implementation
 - deep-rl-course
 model-index:
-- name: SARSA_QLEARNING
+- name: Actor-Critic
   results:
   - task:
       type: reinforcement-learning
       name: reinforcement-learning
     dataset:
-      name: Pixelcopter-PLE-v0
-      type: Pixelcopter-PLE-v0
+      name: PandaReachDense-v3
+      type: PandaReachDense-v3
     metrics:
     - type: mean_reward
-      value: 46.00 +/- 43.67
+      value: -0.15 +/- 0.03
       name: mean_reward
       verified: false
 ---
 
-  # REINFORCE Agent Playing Pixelcopter-PLE-v0
+  # Actor-Critic Agent Playing PandaReachDense-v3
 
-  This is a trained model of a REINFORCE agent playing Pixelcopter-PLE-v0.
+  This is a trained model of an AC agent playing PandaReachDense-v3.
 
   # Hyperparameters
-  hp_seed: 1<br />hp_torch_deterministic: True<br />hp_nb_frames: 8<br />hp_total_timesteps: 1005000<br />hp_learning_t: 500<br />hp_num_envs: 12<br />hp_learning_rate: 0.0001<br />hp_gamma: 0.99<br />hp_buffer_size: 10000<br />hp_batch_size: 32
+  hp_seed: 2444<br />hp_torch_deterministic: True<br />hp_total_timesteps: 20500<br />hp_critic_nstep: 1<br />hp_num_envs: 12<br />hp_learning_rate_actor: 0.001<br />hp_learning_rate_critic: 0.005<br />hp_minlr_actor: 2e-06<br />hp_minlr_critic: 1e-05<br />hp_gamma: 0.99<br />hp_reg_term: 3<br />hp_batch_size: 64
   

model_scripted.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8af84a2d5fb993b2d8b56d648a394caef04d58b68bb4a4cf3f2e8927ae141369
-size 1186859
+oid sha256:7197c2dc571e18719d98046e282a9b5f3f8c5dd6ea89be308c89a43ea257ea7a
+size 34816

results.json

@@ -1 +1 @@
-{"env_id": "Pixelcopter-PLE-v0", "mean_reward": 46.0, "std_reward": 43.669211121796096, "n_evaluation_episodes": 10, "eval_datetime": "2024-04-01T09:00:24.807934"}
+{"env_id": "PandaReachDense-v3", "mean_reward": "-0.14658388", "std_reward": "0.03386605", "n_evaluation_episodes": 10, "eval_datetime": "2024-04-01T13:22:24.586948"}