ledmands
commited on
Commit
·
98cdb04
1
Parent(s):
f7c1ad0
Updated README metadata with eval metrics. Updated README with note on replay and eval metrics
Browse files
README.md
CHANGED
|
@@ -19,14 +19,15 @@ model-index:
|
|
| 19 |
type: ALE/Pacman-v5
|
| 20 |
metrics:
|
| 21 |
- type: mean_reward
|
| 22 |
-
value:
|
| 23 |
name: mean_reward
|
| 24 |
verified: false
|
| 25 |
---
|
| 26 |
|
| 27 |
# Agent using DQN to play ALE/Pacman-v5
|
| 28 |
|
| 29 |
-
|
|
|
|
| 30 |
|
| 31 |
This is an agent that is trained using Stable Baselines3 as part of the capstone project for South Hills School in Spring 2024.
|
| 32 |
The goal of this project is to gain familiarity with reinforcement learning concepts and tools, and to train an agent to score up into the 400-500 point range in Pacman.
|
|
|
|
| 19 |
type: ALE/Pacman-v5
|
| 20 |
metrics:
|
| 21 |
- type: mean_reward
|
| 22 |
+
value: 455.60 +/- 40.10
|
| 23 |
name: mean_reward
|
| 24 |
verified: false
|
| 25 |
---
|
| 26 |
|
| 27 |
# Agent using DQN to play ALE/Pacman-v5
|
| 28 |
|
| 29 |
+
# Update 20 May 2024: Latest DQN model is version 2.8
|
| 30 |
+
# NOTE: Video preview is version 2.8, best model playing for 10,000 steps. Evaluation metrics are self-reported based on 10 episodes of evaluation. Can be found in agents/dqn_v2-8/evals.txt
|
| 31 |
|
| 32 |
This is an agent that is trained using Stable Baselines3 as part of the capstone project for South Hills School in Spring 2024.
|
| 33 |
The goal of this project is to gain familiarity with reinforcement learning concepts and tools, and to train an agent to score up into the 400-500 point range in Pacman.
|