Upload folder using huggingface_hub

README.md

@@ -1,6 +1,6 @@
 ---
 tags:
-- Pixelcopter
+- Pixelcopter-PLE-v0
 - reinforce
 - reinforcement-learning
 - custom-implementation
@@ -12,15 +12,16 @@ model-index:
       type: reinforcement-learning
       name: reinforcement-learning
     dataset:
-      name: Pixelcopter
-      type: Pixelcopter
+      name: Pixelcopter-PLE-v0
+      type: Pixelcopter-PLE-v0
     metrics:
     - type: mean_reward
-      value: 15.00 +/- 13.20
+      value: 19.92 +/- 21.58
       name: mean_reward
       verified: false
 ---
-# **REINFORCE** Agent playing **Pixelcopter**
+# **REINFORCE** Agent playing **Pixelcopter-PLE-v0**
 
-Mean reward: 15.00 +/- 13.20
-Mean episode length: 125.7 +/- 75.9
+This is a trained **REINFORCE** agent playing **Pixelcopter-PLE-v0**.
+
+Trained following Unit 4 of the Hugging Face Deep Reinforcement Learning Course.

hyperparameters.json

@@ -1,6 +1,7 @@
 {
+  "env_id": "Pixelcopter-PLE-v0",
   "h_size": 128,
-  "lr": 0.003,
+  "lr": 0.00025,
   "gamma": 0.99,
   "max_t": 5000,
   "n_training_episodes": 20000,
@@ -8,5 +9,8 @@
   "obs_scale": 55.0,
   "clip_obs": 10.0,
   "use_entropy_bonus": false,
-  "entropy_coef": 0.01
+  "entropy_coef": 0.01,
+  "use_gap_shaping": true,
+  "gap_shaping_k": 0.5,
+  "n_evaluation_episodes": 50
 }

model.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af30224aad8da0fdde8fdbcffe2ea45a9bc9baeda340483f959164f28f9fc9ca
+oid sha256:bdc2104a7449dda62ded6ae2a5dc76165f2d4f9d79d7a32f6b7e274ae97042da
 size 8265

results.json

@@ -1,9 +1,9 @@
 {
-  "env_id": "Pixelcopter",
-  "mean_reward": 15.0,
-  "std_reward": 13.196969349058897,
-  "mean_episode_length": 125.7,
-  "std_episode_length": 75.91659897545463,
+  "env_id": "Pixelcopter-PLE-v0",
+  "mean_reward": 19.92,
+  "std_reward": 21.582251967762772,
+  "mean_episode_length": 153.18,
+  "std_episode_length": 122.8784260966912,
   "n_evaluation_episodes": 50,
-  "eval_datetime": "2026-02-17T01:22:11.133074"
+  "eval_datetime": "2026-02-17T01:48:08.114790"
 }