ferrotorch 0.6.2

PyTorch in Rust — deep learning framework built on ferray
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

//! Pins every module path the umbrella crate's README documents.
//!
//! The umbrella crate's only job is re-exports. The README enumerates the
//! submodules a user can `use ferrotorch::*;` import. If a feature flag or
//! sub-crate dep is removed without updating the README, this test fails at
//! compile time instead of at the user's first build.

// Each test asserts a module path resolves; the `use … as _;` pattern is the
// minimum-friction way to do that without forcing every sub-crate to keep a
// specific named symbol stable.
//
// The `as _` rename suppresses `unused_imports` for value bindings, but it
// does NOT suppress it for module path imports (`use ferrotorch::nn as _;`
// — there is nothing nameable in a module path to "use"), so this file-level
// allow is load-bearing. Verified empirically: removing it produces 16
// `unused import` errors under `cargo clippy --all-targets -- -D warnings`.
// The audit's claim that this allow was redundant was mistaken.
#![allow(unused_imports)]

#[test]
fn always_on_modules_resolve() {
    use ferrotorch::data as _;
    use ferrotorch::nn as _;
    use ferrotorch::optim as _;
    use ferrotorch::vision as _;
}

#[cfg(feature = "train")]
#[test]
fn train_module_resolves() {
    use ferrotorch::train as _;
}

#[cfg(feature = "serialize")]
#[test]
fn serialize_module_resolves() {
    use ferrotorch::serialize as _;
}

#[cfg(feature = "jit")]
#[test]
fn jit_module_resolves() {
    use ferrotorch::jit as _;
}

#[cfg(feature = "jit-script")]
#[test]
fn jit_script_module_resolves() {
    use ferrotorch::jit_script as _;
}

#[cfg(feature = "distributions")]
#[test]
fn distributions_module_resolves() {
    use ferrotorch::distributions as _;
}

#[cfg(feature = "profiler")]
#[test]
fn profiler_module_resolves() {
    use ferrotorch::profiler as _;
}

#[cfg(feature = "hub")]
#[test]
fn hub_module_resolves() {
    use ferrotorch::hub as _;
}

#[cfg(feature = "tokenize")]
#[test]
fn tokenize_module_resolves() {
    use ferrotorch::tokenize as _;
}

#[cfg(feature = "gpu")]
#[test]
fn gpu_module_resolves() {
    use ferrotorch::gpu as _;
}

#[cfg(feature = "cubecl")]
#[test]
fn cubecl_module_resolves() {
    use ferrotorch::cubecl as _;
}

#[cfg(feature = "mps")]
#[test]
fn mps_module_resolves() {
    use ferrotorch::mps as _;
}

#[cfg(feature = "xpu")]
#[test]
fn xpu_module_resolves() {
    use ferrotorch::xpu as _;
}

#[cfg(feature = "distributed")]
#[test]
fn distributed_module_resolves() {
    use ferrotorch::distributed as _;
}

#[cfg(feature = "llama")]
#[test]
fn llama_module_resolves() {
    use ferrotorch::llama as _;
}

#[cfg(feature = "ml")]
#[test]
fn ml_module_resolves() {
    use ferrotorch::ml as _;
}