relayrl_algorithms 0.3.0

A collection of Multi-Agent Deep Reinforcement Learning Algorithms (IPPO, MAPPO, etc.)
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42

use crate::templates::base_replay_buffer::{Batch, GenericReplayBuffer, ReplayBufferError};

use async_trait::async_trait;
use relayrl_types::prelude::trajectory::RelayRLTrajectory;

use std::any::Any;

type SharedReplayBuffer = super::super::replay_buffer::TD3ReplayBuffer;

pub struct MultiagentTD3ReplayBuffer {
    inner: SharedReplayBuffer,
}

pub type MATD3ReplayBuffer = MultiagentTD3ReplayBuffer;

impl Default for MultiagentTD3ReplayBuffer {
    fn default() -> Self {
        Self::new(1_000_000, 128)
    }
}

impl MultiagentTD3ReplayBuffer {
    pub fn new(buffer_size: usize, batch_size: usize) -> Self {
        Self {
            inner: SharedReplayBuffer::new(buffer_size, batch_size),
        }
    }
}

#[async_trait]
impl GenericReplayBuffer for MultiagentTD3ReplayBuffer {
    async fn insert_trajectory(
        &self,
        trajectory: RelayRLTrajectory,
    ) -> Result<Box<dyn Any>, ReplayBufferError> {
        self.inner.insert_trajectory(trajectory).await
    }

    async fn sample_buffer(&self) -> Result<Batch, ReplayBufferError> {
        self.inner.sample_buffer().await
    }
}