说明文档

PPO 智能体游玩 SnowballTarget

这是一个使用 Unity ML-Agents 库训练的 PPO 智能体，用于游玩 SnowballTarget。

使用方法（使用 ML-Agents）

文档：https://github.com/huggingface/ml-agents#get-started 我们编写了一个完整的教程，教你如何使用 ML-Agents 训练你的第一个智能体并将其发布到 Hub：

继续训练

mlagents-learn ./config/ppo/SnowballTarget.yaml --run-id=\"SnowballTarget-v1\" --resume

训练超参数

behaviors:
SnowballTarget:
  trainer_type: ppo
  summary_freq: 10000
  keep_checkpoints: 5
  checkpoint_interval: 50000
  max_steps: 900000
  time_horizon: 128
  threaded: true
  hyperparameters:
    learning_rate: 0.0001
    learning_rate_schedule: linear
    batch_size: 128
    buffer_size: 4096
    beta: 0.005
    epsilon: 0.2
    lambd: 0.95
    num_epoch: 5
  network_settings:
    normalize: false
    hidden_units: 256
    num_layers: 3
    vis_encode_type: simple
  reward_signals:
    extrinsic:
      gamma: 0.99
      strength: 1.0

观看你的智能体游玩

你可以直接在浏览器中观看你的智能体游玩：.

访问 https://huggingface.co/spaces/unity/ML-Agents-SnowballTarget
第一步：输入你的 model_id：kinkpunk/PPO-SnowballTarget
第二步：选择你的 .nn /.onnx 文件
点击"观看智能体游玩"👀

kinkpunk/PPO-SnowballTarget

作者 kinkpunk

reinforcement-learning ml-agents

↓ 5 ♥ 0

创建时间: 2023-01-17 12:46:50+00:00

更新时间: 2023-01-17 14:08:18+00:00

在 Hugging Face 上查看

文件 (20)

.gitattributes

README.md

SnowballTarget.onnx ONNX

SnowballTarget/SnowballTarget-749984.onnx ONNX

SnowballTarget/SnowballTarget-749984.pt

SnowballTarget/SnowballTarget-799880.onnx ONNX

SnowballTarget/SnowballTarget-799880.pt

SnowballTarget/SnowballTarget-849968.onnx ONNX

SnowballTarget/SnowballTarget-849968.pt

SnowballTarget/SnowballTarget-899928.onnx ONNX

SnowballTarget/SnowballTarget-899928.pt

SnowballTarget/SnowballTarget-901208.onnx ONNX

SnowballTarget/SnowballTarget-901208.pt

SnowballTarget/checkpoint.pt

SnowballTarget/events.out.tfevents.1673956828.4e8d1bd07e66.782.0

config.json

configuration.yaml

run_logs/Player-0.log

run_logs/timers.json

run_logs/training_status.json