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
use Progress;
/// Trait for logging training progress at each step and end of epoch.
///
/// # Call sequence
///
/// Implementors can expect the following sequence of calls for a complete training run:
///
/// ```text
/// start(total_epochs, total_items)
/// for each epoch:
/// start_split("train", total_items_train)
/// update_split(items_processed) // called once per batch
/// ...
/// end_split()
/// start_split("valid", total_items_valid)
/// update_split(items_processed) // called once per batch
/// ...
/// end_split()
/// update_epoch(epoch)
/// end()
/// ```
///
/// `end()` is called whether training completes normally or is interrupted early.
///
/// Implementors are responsible for tracking `total_items` and epoch state in order
/// to reconstruct the full progress picture when `update_split` is called.
/// Trait for logging evaluation progress at each step and end of evaluation.
///
/// # Call sequence
///
/// Implementors can expect the following sequence of calls for a complete evaluation run:
///
/// ```text
/// start_global_progress(total_tests)
/// for each test split:
/// start_test(name, total_items)
/// update_test_progress(items_processed) // called once per batch
/// ...
/// end_test()
/// end_global_progress()
/// ```
///
/// `end_global_progress()` is called whether evaluation completes normally or is interrupted early.
///
/// Implementors are responsible for tracking `total_tests` and `total_items` (stored from
/// `start_global_progress` and `start_test`) to reconstruct the full progress picture.
/// Two-level progress snapshot combining run-level and phase-level tracking.
///
/// `global_progress` spans the full training run (e.g., epochs completed out of total),
/// while `split_progress` tracks the current phase (e.g., batches within the current epoch).