File size: 2,572 Bytes
0584798
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from __future__ import annotations

import json
import sys
from pathlib import Path

import numpy as np
import torch
from datasets import Dataset
from sklearn.metrics import accuracy_score, f1_score

BASE_DIR = Path(__file__).resolve().parent.parent
if str(BASE_DIR) not in sys.path:
    sys.path.insert(0, str(BASE_DIR))


def load_labeled_rows(path: Path, label_field: str, label2id: dict[str, int]) -> list[dict]:
    rows = []
    with path.open("r", encoding="utf-8") as handle:
        for line in handle:
            item = json.loads(line)
            rows.append({"text": item["text"], "label": label2id[item[label_field]]})
    return rows


def load_labeled_rows_from_paths(paths: list[Path], label_field: str, label2id: dict[str, int]) -> list[dict]:
    rows = []
    for path in paths:
        if not path.exists():
            continue
        rows.extend(load_labeled_rows(path, label_field, label2id))
    return rows


def prepare_dataset(rows: list[dict], tokenizer, max_length: int) -> Dataset:
    dataset = Dataset.from_list(rows)

    def tokenize(batch):
        return tokenizer(batch["text"], truncation=True, padding="max_length", max_length=max_length)

    dataset = dataset.map(tokenize, batched=True)
    dataset = dataset.remove_columns(["text"])
    dataset.set_format("torch")
    return dataset


def build_balanced_class_weights(rows: list[dict], num_labels: int) -> torch.Tensor:
    counts = np.zeros(num_labels, dtype=np.float32)
    for row in rows:
        counts[row["label"]] += 1.0

    nonzero = counts > 0
    if not np.any(nonzero):
        return torch.ones(num_labels, dtype=torch.float32)

    total = float(counts.sum())
    active_labels = float(np.count_nonzero(nonzero))
    weights = np.ones(num_labels, dtype=np.float32)
    weights[nonzero] = total / (active_labels * counts[nonzero])
    return torch.tensor(weights, dtype=torch.float32)


def build_label_weight_tensor(labels: tuple[str, ...], weight_map: dict[str, float]) -> torch.Tensor:
    return torch.tensor(
        [float(weight_map.get(label, 1.0)) for label in labels],
        dtype=torch.float32,
    )


def compute_classification_metrics(eval_pred):
    logits, labels = eval_pred
    preds = np.argmax(logits, axis=-1)
    return {
        "accuracy": accuracy_score(labels, preds),
        "macro_f1": f1_score(labels, preds, average="macro"),
    }


def write_json(path: Path, payload: dict) -> None:
    path.parent.mkdir(parents=True, exist_ok=True)
    path.write_text(json.dumps(payload, indent=2, sort_keys=True) + "\n", encoding="utf-8")