Cargo.toml 2.17 KB
Newer Older
1
[package]
2
name = "sglang_router_rs"
3
4
5
version = "0.0.0"
edition = "2021"

6
7
8
[features]
default = ["huggingface"]
huggingface = ["tokenizers"]
9
tiktoken = ["tiktoken-rs"]
10

11
[lib]
12
name = "sglang_router_rs"
13
14
15
# Pure Rust library: Just omit crate-type (defaults to rlib)
# Python/C binding + Rust library: Use ["cdylib", "rlib"]
crate-type = ["cdylib", "rlib"]
16

17
18
19
20
[[bin]]
name = "sglang-router"
path = "src/main.rs"

21
[dependencies]
22
clap = { version = "4", features = ["derive"] }
23
24
25
axum = { version = "0.8.4", features = ["macros", "ws", "tracing"] }
tower = { version = "0.5", features = ["full"] }
tower-http = { version = "0.6", features = ["trace", "compression-gzip", "cors", "timeout", "limit", "request-id", "util"] }
26
serde = { version = "1.0", features = ["derive"] }
27
serde_json = "1.0"
28
bytes = "1.8.0"
29
rand = "0.9.2"
30
reqwest = { version = "0.12.8", features = ["stream", "blocking", "json"] }
31
futures-util = "0.3"
32
futures = "0.3"
33
pyo3 = { version = "0.25.1", features = ["extension-module"] }
34
35
dashmap = "6.1.0"
http = "1.1.0"
36
37
tokio = { version = "1.42.0", features = ["full"] }
async-trait = "0.1"
38
39
40
41
tracing = "0.1"
tracing-subscriber = { version = "0.3", features = ["env-filter", "json", "chrono"] }
tracing-log = "0.2"
tracing-appender = "0.2.3"
42
chrono = "0.4"
43
44
kube = { version = "1.1.0", features = ["runtime", "derive"] }
k8s-openapi = { version = "0.25.0", features = ["v1_33"] }
45
46
metrics = "0.24.2"
metrics-exporter-prometheus = "0.17.0"
47
uuid = { version = "1.10", features = ["v4", "serde"] }
48
thiserror = "2.0.12"
49
url = "2.5.4"
50
tokio-stream = { version = "0.1", features = ["sync"] }
51
anyhow = "1.0"
52
tokenizers = { version = "0.21.4", optional = true }
53
tiktoken-rs = { version = "0.5", optional = true }
54
55
56

[dev-dependencies]
criterion = { version = "0.5", features = ["html_reports"] }
57
58
59
tower = { version = "0.5", features = ["util"] }
http-body-util = "0.1"
portpicker = "0.1"
60
61
62
63
64
65

[[bench]]
name = "request_processing"
harness = false
path = "benches/request_processing.rs"

66
67
68
[profile.release]
lto = "thin"
codegen-units = 1
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83

[profile.dev]
opt-level = 0
debug = true
split-debuginfo = "unpacked"
incremental = true


[profile.dev.build-override]
opt-level = 3
codegen-units = 1

[profile.dev-opt]
inherits = "dev"
opt-level = 1