PPO Agent playing MicrortsDefeatCoacAIShaped-v3

This is a trained model of a PPO agent playing MicrortsDefeatCoacAIShaped-v3 using the /sgoodfriend/rl-algo-impls repo.

All models trained at this commit can be found at https://api.wandb.ai/links/sgoodfriend/sjo3qukl.

Training Results

This model was trained from 3 trainings of PPO agents using different initial seeds. These agents were trained by checking out 9ba0ab5. The best and last models were kept from each training. This submission has loaded the best models from each training, reevaluates them, and selects the best model from these latest evaluations (mean - std).

algo	env	seed	reward_mean	reward_std	eval_episodes	best	wandb_url
ppo	MicrortsDefeatCoacAIShaped-v3	1	0.769231	0.638971	26	*	wandb
ppo	MicrortsDefeatCoacAIShaped-v3	2	0.692308	0.721602	26		wandb
ppo	MicrortsDefeatCoacAIShaped-v3	3	0.423077	0.884615	26		wandb

Prerequisites: Weights & Biases (WandB)

Training and benchmarking assumes you have a Weights & Biases project to upload runs to. By default training goes to a rl-algo-impls project while benchmarks go to rl-algo-impls-benchmarks. During training and benchmarking runs, videos of the best models and the model weights are uploaded to WandB.

Before doing anything below, you'll need to create a wandb account and run wandb login.

Usage

/sgoodfriend/rl-algo-impls: https://github.com/sgoodfriend/rl-algo-impls

Note: While the model state dictionary and hyperaparameters are saved, the latest implementation could be sufficiently different to not be able to reproduce similar results. You might need to checkout the commit the agent was trained on: 9ba0ab5.

# Downloads the model, sets hyperparameters, and runs agent for 3 episodes
python enjoy.py --wandb-run-path=sgoodfriend/rl-algo-impls-benchmarks/a0smxvhw

Setup hasn't been completely worked out yet, so you might be best served by using Google Colab starting from the colab_enjoy.ipynb notebook.

Training

If you want the highest chance to reproduce these results, you'll want to checkout the commit the agent was trained on: 9ba0ab5. While training is deterministic, different hardware will give different results.

python train.py --algo ppo --env MicrortsDefeatCoacAIShaped-v3 --seed 1

Setup hasn't been completely worked out yet, so you might be best served by using Google Colab starting from the colab_train.ipynb notebook.

Benchmarking (with Lambda Labs instance)

This and other models from https://api.wandb.ai/links/sgoodfriend/sjo3qukl were generated by running a script on a Lambda Labs instance. In a Lambda Labs instance terminal:

git clone git@github.com:sgoodfriend/rl-algo-impls.git
cd rl-algo-impls
bash ./lambda_labs/setup.sh
wandb login
bash ./lambda_labs/benchmark.sh [-a {"ppo a2c dqn vpg"}] [-e ENVS] [-j {6}] [-p {rl-algo-impls-benchmarks}] [-s {"1 2 3"}]

Alternative: Google Colab Pro+

As an alternative, colab_benchmark.ipynb, can be used. However, this requires a Google Colab Pro+ subscription and running across 4 separate instances because otherwise running all jobs will exceed the 24-hour limit.

Hyperparameters

This isn't exactly the format of hyperparams in hyperparams/ppo.yml, but instead the Wandb Run Config. However, it's very close and has some additional data:

additional_keys_to_log:
- microrts_stats
- microrts_results
algo: ppo
algo_hyperparams:
  batch_size: 3072
  clip_range: 0.1
  clip_range_decay: none
  clip_range_vf: 0.1
  ent_coef: 0.01
  gamma_end: 0.999
  learning_rate: 0.00025
  learning_rate_decay: spike
  max_grad_norm: 0.5
  n_epochs: 4
  n_steps: 512
  ppo2_vf_coef_halving: true
  vf_coef: 0.5
device: auto
env: Microrts-selfplay-unet-decay
env_hyperparams:
  env_type: microrts
  make_kwargs:
    map_paths:
    - maps/16x16/basesWorkers16x16.xml
    max_steps: 4000
    num_selfplay_envs: 36
    render_theme: 2
    reward_weight:
    - 10
    - 1
    - 1
    - 0.2
    - 1
    - 4
  n_envs: 24
  self_play_kwargs:
    num_old_policies: 12
    save_steps: 300000
    swap_steps: 6000
    swap_window_size: 4
    window: 33
env_id: MicrortsDefeatCoacAIShaped-v3
eval_hyperparams:
  deterministic: false
  env_overrides:
    bots:
      coacAI: 2
      droplet: 2
      guidedRojoA3N: 2
      izanagi: 2
      lightRushAI: 2
      mixedBot: 2
      naiveMCTSAI: 2
      passiveAI: 2
      randomAI: 2
      randomBiasedAI: 2
      rojo: 2
      tiamat: 2
      workerRushAI: 2
    make_kwargs:
      map_paths:
      - maps/16x16/basesWorkers16x16.xml
      max_steps: 4000
      num_selfplay_envs: 0
      render_theme: 2
      reward_weight:
      - 1
      - 0
      - 0
      - 0
      - 0
      - 0
    n_envs: 26
    self_play_kwargs: {}
  max_video_length: 4000
  n_episodes: 26
  score_function: mean
  step_freq: 1000000
microrts_reward_decay_callback: true
n_timesteps: 300000000
policy_hyperparams:
  activation_fn: relu
  actor_head_style: unet
  cnn_flatten_dim: 256
  cnn_style: microrts
  v_hidden_sizes:
  - 256
  - 128
seed: 1
use_deterministic_algorithms: true
wandb_entity: null
wandb_group: null
wandb_project_name: rl-algo-impls-benchmarks
wandb_tags:
- benchmark_9ba0ab5
- host_192-9-155-233
- branch_main
- v0.0.9

sgoodfriend
/

ppo-Microrts-selfplay-unet-decay