entrenar 0.7.8

Training & Optimization library with autograd, LoRA, quantization, and model merging
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

//! Cell types for Jupyter notebooks.

use serde::{Deserialize, Serialize};

use super::split_source;

/// Cell type in a notebook
#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "snake_case")]
pub enum CellType {
    /// Executable code cell
    Code,
    /// Markdown documentation cell
    Markdown,
    /// Raw text cell
    Raw,
}

impl std::fmt::Display for CellType {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        match self {
            Self::Code => write!(f, "code"),
            Self::Markdown => write!(f, "markdown"),
            Self::Raw => write!(f, "raw"),
        }
    }
}

/// A single cell in a notebook
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
pub struct Cell {
    /// Type of cell
    pub cell_type: CellType,
    /// Source content (lines)
    pub source: Vec<String>,
    /// Cell outputs (for code cells)
    pub outputs: Vec<CellOutput>,
    /// Execution count (for code cells)
    pub execution_count: Option<u32>,
    /// Cell metadata
    pub metadata: CellMetadata,
}

impl Cell {
    /// Create a new code cell
    pub fn code(source: impl Into<String>) -> Self {
        Self {
            cell_type: CellType::Code,
            source: split_source(source.into()),
            outputs: Vec::new(),
            execution_count: None,
            metadata: CellMetadata::default(),
        }
    }

    /// Create a new markdown cell
    pub fn markdown(source: impl Into<String>) -> Self {
        Self {
            cell_type: CellType::Markdown,
            source: split_source(source.into()),
            outputs: Vec::new(),
            execution_count: None,
            metadata: CellMetadata::default(),
        }
    }

    /// Create a new raw cell
    pub fn raw(source: impl Into<String>) -> Self {
        Self {
            cell_type: CellType::Raw,
            source: split_source(source.into()),
            outputs: Vec::new(),
            execution_count: None,
            metadata: CellMetadata::default(),
        }
    }

    /// Add an output to the cell
    pub fn with_output(mut self, output: CellOutput) -> Self {
        self.outputs.push(output);
        self
    }

    /// Set execution count
    pub fn with_execution_count(mut self, count: u32) -> Self {
        self.execution_count = Some(count);
        self
    }

    /// Get the source as a single string
    pub fn source_text(&self) -> String {
        self.source.join("")
    }
}

/// Cell output
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
pub struct CellOutput {
    /// Output type
    pub output_type: String,
    /// Output data
    pub data: Option<serde_json::Value>,
    /// Text output
    pub text: Option<Vec<String>>,
}

impl CellOutput {
    /// Create a stream output (stdout/stderr)
    pub fn stream(_name: &str, text: impl Into<String>) -> Self {
        Self {
            output_type: "stream".to_string(),
            data: None,
            text: Some(split_source(text.into())),
        }
    }

    /// Create an execute result output
    pub fn execute_result(data: serde_json::Value) -> Self {
        Self { output_type: "execute_result".to_string(), data: Some(data), text: None }
    }
}

/// Cell metadata
#[derive(Debug, Clone, Default, PartialEq, Eq, Serialize, Deserialize)]
pub struct CellMetadata {
    /// Whether the cell is collapsed
    #[serde(skip_serializing_if = "Option::is_none")]
    pub collapsed: Option<bool>,
    /// Whether the cell is scrolled
    #[serde(skip_serializing_if = "Option::is_none")]
    pub scrolled: Option<bool>,
}