Add RL examples

examples/ppo/README.md

@@ -0,0 +1,45 @@
+# Training PPO with decentralized averaging
+
+This tutorial will walk you through the steps to set up collaborative training of an on-policy reinforcement learning algorighm [PPO](https://arxiv.org/pdf/1707.06347.pdf) to play Atari Breakout. It uses [stable-baselines3 implementation of PPO](https://stable-baselines3.readthedocs.io/en/master/modules/ppo.html), hyperparameters for the algorithm are taken from [rl-baselines3-zoo](https://github.com/DLR-RM/rl-baselines3-zoo/blob/master/hyperparams/ppo.yml), collaborative training is built on `hivemind.Optimizer` to exchange information between peers.
+
+## Preparation
+
+* Install hivemind: `pip install git+https://github.com/learning-at-home/hivemind.git`
+* Dependencies: `pip install -r requirements.txt`
+
+## Running an experiment
+
+### First peer
+Run the first DHT peer to welcome trainers and record training statistics (e.g., loss and performance):
+- In this example, we use [tensorboard](https://www.tensorflow.org/tensorboard) to plot training metrics. If you're unfamiliar with Tensorboard, here's a [quickstart tutorial](https://www.tensorflow.org/tensorboard/get_started).
+- Run `python3 ppo.py`
+
+```
+$ python3 ppo.py
+To connect other peers to this one, use --initial_peers /ip4/127.0.0.1/tcp/41926/p2p/QmUmiebP4BxdEPEpQb28cqyhaheDugFRn7M
+CoLJr556xYt
+A.L.E: Arcade Learning Environment (version 0.7.4+069f8bd)
+[Powered by Stella]
+Using cuda device
+Wrapping the env in a VecTransposeImage.
+[W NNPACK.cpp:51] Could not initialize NNPACK! Reason: Unsupported hardware.
+Jun 20 13:23:20.515 [INFO] Found no active peers: None
+Jun 20 13:23:20.533 [INFO] Initializing optimizer manually since it has no tensors in state dict. To override this, prov
+ide initialize_optimizer=False
+Logging to logs/bs-256.target_bs-32768.n_envs-8.n_steps-128.n_epochs-1_1
+---------------------------------
+| rollout/           |          |
+|    ep_len_mean     | 521      |
+|    ep_rew_mean     | 0        |
+| time/              |          |
+|    fps             | 582      |
+|    iterations      | 1        |
+|    time_elapsed    | 1        |
+|    total_timesteps | 1024     |
+| train/             |          |
+|    timesteps       | 1024     |
+---------------------------------
+Jun 20 13:23:23.525 [INFO] ppo_hivemind accumulated 1024 samples for epoch #0 from 1 peers. ETA 52.20 sec (refresh in 1$
+.00 sec)
+
+```

examples/ppo/ppo.py

@@ -0,0 +1,117 @@
+import argparse
+from importlib.resources import path
+import pathlib
+import torch
+
+import hivemind
+from hivemind import Float16Compression, SizeAdaptiveCompression, Uniform8BitQuantization
+
+from stable_baselines3 import PPO
+from stable_baselines3.common.env_util import make_atari_env
+from stable_baselines3.common.vec_env import VecFrameStack
+
+def parse_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--n-steps', type=int, default=128, help='Number of rollout steps per each agent')
+    parser.add_argument('--n-envs', type=int, default=8, help='Number of training envs')
+    parser.add_argument('--batch-size', type=int, default=256)
+    parser.add_argument('--target-batch-size', type=int, default=32768)
+    parser.add_argument('--n-epochs', type=int, default=1, help='Number of training epochs per each rollout')
+    parser.add_argument('--learning-rate', type=float, default=2.5e-4)
+    parser.add_argument('--tb-logs-path', type=pathlib.Path, default='./logs', help='Path to tensorboard logs folder')
+    parser.add_argument('--experiment-prefix', type=str, help='Experiment prefix for tensorboard logs')
+    parser.add_argument('--initial-peers', nargs='+', default=[])
+    parser.add_argument('--averaging-compression', action='store_true')
+    args = parser.parse_args()
+    return args
+
+def generate_experiment_name(args):
+    exp_name_dict = {
+        'bs': args.batch_size,
+        'target_bs': args.target_batch_size,
+        'n_envs': args.n_envs,
+        'n_steps': args.n_steps,
+        'n_epochs': args.n_epochs,
+    }
+
+    exp_name = [f'{key}-{value}' for key, value in exp_name_dict.items()]
+    exp_name = '.'.join(exp_name)
+
+    if args.experiment_prefix:
+        exp_name = f'{args.experiment_prefix}.{exp_name}'
+    exp_name = exp_name.replace('000.', 'k.')
+    return exp_name
+
+
+class AdamWithClipping(torch.optim.Adam):
+    def __init__(self, *args, max_grad_norm: float, **kwargs):
+        self.max_grad_norm = max_grad_norm
+        super().__init__(*args, **kwargs)
+
+    def step(self, *args, **kwargs):
+        iter_params = (param for group in self.param_groups for param in group["params"])
+        torch.nn.utils.clip_grad_norm_(iter_params, self.max_grad_norm)
+        return super().step(*args, **kwargs)
+
+
+def configure_dht_opts(args):
+    opts = {
+        'start': True,
+    }
+    if args.initial_peers:
+        opts['initial_peers'] = args.initial_peers
+
+    return opts
+
+
+if __name__ == "__main__":
+    args = parse_args()
+
+    dht_opts = configure_dht_opts(args)
+    dht = hivemind.DHT(**dht_opts)
+    print("To connect other peers to this one, use --initial_peers", *[str(addr) for addr in dht.get_visible_maddrs()])
+
+    env = make_atari_env('BreakoutNoFrameskip-v4', n_envs=args.n_envs)
+    env = VecFrameStack(env, n_stack=4)
+
+    model = PPO(
+        'CnnPolicy', env,
+        verbose=1,
+        batch_size=args.batch_size,
+        n_steps=args.n_steps,
+        n_epochs=args.n_epochs,
+        learning_rate=args.learning_rate,
+        clip_range=0.1,
+        vf_coef=0.5,
+        ent_coef=0.01,
+        tensorboard_log=args.tb_logs_path,
+        max_grad_norm=10000.0,
+        policy_kwargs={'optimizer_class': AdamWithClipping, 'optimizer_kwargs': {'max_grad_norm': 0.5}}
+    )
+
+    compression_opts = {}
+    if args.averaging_compression:
+        averaging_compression = SizeAdaptiveCompression(
+            threshold=2 ** 10 + 1, less=Float16Compression(), greater_equal=Uniform8BitQuantization()
+        )
+        compression_opts.update({
+            'grad_compression': averaging_compression,
+            'state_averaging_compression': averaging_compression
+        })
+
+    model.policy.optimizer_class = hivemind.Optimizer
+    model.policy.optimizer = hivemind.Optimizer(
+        dht=dht,
+        optimizer=model.policy.optimizer,
+        run_id='ppo_hivemind',
+        batch_size_per_step=args.batch_size,
+        target_batch_size=args.target_batch_size,
+        offload_optimizer=False,
+        verbose=True,
+        use_local_updates=False,
+        matchmaking_time=4,
+        averaging_timeout=15,
+        **compression_opts,
+    )
+    model.policy.optimizer.load_state_from_peers()
+    model.learn(total_timesteps=int(5e11), tb_log_name=generate_experiment_name(args))

examples/ppo/requirements.txt

@@ -0,0 +1 @@
+stable-baselines3[extra]==1.5.0