border-core 0.0.5

Reinforcement learning library
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
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137

//! Configuration of [Trainer](super::Trainer).
use anyhow::Result;
use serde::{Deserialize, Serialize};
use std::{
    fs::File,
    io::{BufReader, Write},
    path::Path,
};

/// Configuration of [Trainer](super::Trainer).
#[derive(Debug, Deserialize, Serialize, PartialEq, Clone)]
pub struct TrainerConfig {
    pub(super) max_opts: usize,
    pub(super) eval_episodes: usize,
    pub(super) eval_threshold: Option<f32>,
    pub(super) model_dir: Option<String>,
    pub(super) opt_interval: usize,
    pub(super) eval_interval: usize,
    pub(super) record_interval: usize,
    pub(super) save_interval: usize,
}

impl Default for TrainerConfig {
    fn default() -> Self {
        Self {
            max_opts: 0,
            eval_interval: 0,
            eval_episodes: 0,
            eval_threshold: None,
            model_dir: None,
            opt_interval: 1,
            record_interval: usize::MAX,
            save_interval: usize::MAX,
        }
    }
}

impl TrainerConfig {
    /// Sets the number of optimization steps.
    pub fn max_opts(mut self, v: usize) -> Self {
        self.max_opts = v;
        self
    }

    /// Sets the interval of evaluation in optimization steps.
    pub fn eval_interval(mut self, v: usize) -> Self {
        self.eval_interval = v;
        self
    }

    /// Sets the number of episodes for evaluation.
    pub fn eval_episodes(mut self, v: usize) -> Self {
        self.eval_episodes = v;
        self
    }

    /// Sets the evaluation threshold.
    pub fn eval_threshold(mut self, v: f32) -> Self {
        self.eval_threshold = Some(v);
        self
    }

    /// Sets the directory the trained model being saved.
    pub fn model_dir<T: Into<String>>(mut self, model_dir: T) -> Self {
        self.model_dir = Some(model_dir.into());
        self
    }

    /// Sets the interval of optimization in environment steps.
    pub fn opt_interval(mut self, opt_interval: usize) -> Self {
        self.opt_interval = opt_interval;
        self
    }

    /// Sets the interval of recording in optimization steps.
    pub fn record_interval(mut self, record_interval: usize) -> Self {
        self.record_interval = record_interval;
        self
    }

    /// Sets the interval of saving in optimization steps.
    pub fn save_interval(mut self, save_interval: usize) -> Self {
        self.save_interval = save_interval;
        self
    }

    /// Constructs [TrainerConfig] from YAML file.
    pub fn load(path: impl AsRef<Path>) -> Result<Self> {
        let file = File::open(path)?;
        let rdr = BufReader::new(file);
        let b = serde_yaml::from_reader(rdr)?;
        Ok(b)
    }

    /// Saves [TrainerConfig].
    pub fn save(&self, path: impl AsRef<Path>) -> Result<()> {
        let mut file = File::create(path)?;
        file.write_all(serde_yaml::to_string(&self)?.as_bytes())?;
        Ok(())
    }
}

// #[cfg(test)]
// mod tests {
//     use super::*;
//     use tempdir::TempDir;

//     #[test]
//     fn test_serde_trainer_builder() -> Result<()> {
//         let builder = TrainerBuilder::default()
//             .max_opts(100)
//             .eval_interval(10000)
//             .n_episodes_per_eval(5)
//             .model_dir("some/directory");

//         let dir = TempDir::new("trainer_builder")?;
//         let path = dir.path().join("trainer_builder.yaml");
//         println!("{:?}", path);

//         builder.save(&path)?;
//         let builder_ = TrainerBuilder::load(&path)?;
//         assert_eq!(builder, builder_);
//         // let yaml = serde_yaml::to_string(&trainer)?;
//         // println!("{}", yaml);
//         // assert_eq!(
//         //     yaml,
//         //     "---\n\
//         //      max_opts: 100\n\
//         //      eval_interval: 10000\n\
//         //      n_episodes_per_eval: 5\n\
//         //      eval_threshold: ~\n\
//         //      model_dir: some/directory\n\
//         // "
//         // );
//         Ok(())
//     }
// }