dataflow_nlp 0.1.1

Dataflow is a data processing library, primarily for machine learning.
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

use std::collections::HashMap;

use super::{BasicVocab, TokenNotFoundError, Vocab};
use dataflow::prelude::Node;
use serde::{Deserialize, Serialize};

#[derive(Serialize, Deserialize, Default, Clone)]
pub struct BPEVocab {
    vocab: BasicVocab,
}

impl Vocab for BPEVocab {
    fn load() -> Self {
        use serde_json::Value;

        // Open vocab file
        let json: HashMap<String, Value> = serde_json::from_str(
            &include_str!("../resources/bpe_vocab.json")
                .replace('/', "")
                .replace('Ġ', ""),
        )
        .expect("Error parsing BPE vocab file!");
        // Build sorted vector of tokens from hashmap
        let mut token_vec: Vec<String> = vec![String::from(""); 50265]; // Happen to know the largest index in the json is 50264, this is a bad system
        for token in json.keys() {
            token_vec[json[token].as_u64().unwrap() as usize] = token.clone();
        }
        // Build vocab
        let mut vocab = BasicVocab::new();
        let mut temp_vec: Vec<String> = Vec::new();
        for token in token_vec {
            if !token.is_empty() {
                vocab.add_token(token.clone());
                temp_vec.push(token);
            }
        }
        BPEVocab { vocab }
    }

    fn len(&self) -> usize {
        self.vocab.len()
    }

    fn tokens_from_indexes(&self, indexes: &[usize]) -> Result<Vec<String>, TokenNotFoundError> {
        self.vocab.tokens_from_indexes(indexes)
    }

    fn batch_tokens_from_indexes(
        &self,
        indexes: &[Vec<usize>],
    ) -> Result<Vec<Vec<String>>, TokenNotFoundError> {
        self.vocab.batch_tokens_from_indexes(indexes)
    }

    fn indexes_from_tokens(&self, tokens: &[String]) -> Result<Vec<usize>, TokenNotFoundError> {
        self.vocab.indexes_from_tokens(tokens)
    }

    fn batch_indexes_from_tokens(
        &self,
        tokens: &[Vec<String>],
    ) -> Result<Vec<Vec<usize>>, TokenNotFoundError> {
        self.vocab.batch_indexes_from_tokens(tokens)
    }
}

impl Node<Vec<Vec<String>>> for BPEVocab {
    type Output = Vec<Vec<usize>>;

    fn process(&mut self, input: Vec<Vec<String>>) -> Self::Output {
        self.batch_indexes_from_tokens(&input).unwrap()
    }
}

impl Node<Vec<String>> for BPEVocab {
    type Output = Vec<usize>;

    fn process(&mut self, input: Vec<String>) -> Self::Output {
        self.indexes_from_tokens(&input).unwrap()
    }
}

impl Node<Vec<Vec<usize>>> for BPEVocab {
    type Output = Vec<Vec<String>>;

    fn process(&mut self, input: Vec<Vec<usize>>) -> Self::Output {
        self.batch_tokens_from_indexes(&input).unwrap()
    }
}

impl Node<Vec<usize>> for BPEVocab {
    type Output = Vec<String>;

    fn process(&mut self, input: Vec<usize>) -> Self::Output {
        self.tokens_from_indexes(&input).unwrap()
    }
}