main
toml 181 lines 5.81 KB
Raw
1 [workspace]
2 members = [
3 "proto",
4 "src/bench",
5 "src/build-info",
6 "src/chat",
7 "src/cmd",
8 "src/engine-core-client",
9 "src/llm",
10 "src/managed-engine",
11 "src/metrics",
12 "src/mock-engine",
13 "src/parser",
14 "src/parser/python",
15 "src/server",
16 "src/text",
17 "src/tokenizer",
18 "src/tracing",
19 ]
20 resolver = "3"
21
22 [workspace.package]
23 version = "0.1.0"
24 edition = "2024"
25 license = "Apache-2.0"
26
27 [workspace.dependencies]
28 anyhow = "1.0.100"
29 arc-swap = "1.9.0"
30 async-openai = { version = "0.33.1", default-features = false, features = ["native-tls"] }
31 async-trait = "0.1.89"
32 asynk-strim-attr = "0.1.0"
33 auto_enums = { version = "0.8.9", features = ["tokio1"] }
34 auto_impl = "1.3.0"
35 axum = "0.8.8"
36 base64 = "0.22.1"
37 bytemuck = { version = "1.25.0", features = ["extern_crate_alloc"] }
38 byteorder = "1.5.0"
39 bytes = "1.12.0"
40 chrono = "0.4.42"
41 clap = { version = "4.5.38", features = ["derive", "env"] }
42 criterion = "0.5.1"
43 dirs = "6.0.0"
44 easy-ext = "1.0.3"
45 educe = "0.6.0"
46 enum-as-inner = "0.7.0"
47 expect-test = "1.5.1"
48 fastokens = { version = "0.2.1", default-features = false }
49 futures = "0.3.31"
50 half = { version = "2.7.1", features = ["bytemuck"] }
51 hex = "0.4.3"
52 hf-hub = { version = "0.5.0", default-features = false, features = ["tokio"] }
53 http-body = "1.0.1"
54 hyper = { version = "1.10.1", features = ["http1", "server"] }
55 hyper-util = { version = "0.1.20", features = [
56 "server-graceful",
57 "service",
58 "tokio",
59 ] }
60 image = { version = "0.25.9", default-features = false, features = ["jpeg"] }
61 indexmap = "2.13.0"
62 indicatif = "0.18.4"
63 itertools = "0.14.0"
64 libc = "0.2.177"
65 llm-multimodal = { git = "https://github.com/smg-project/llm-multimodal", rev = "f0985ef65967615db2c79279aa07818499301bfd", default-features = false, features = ["native-tls"] }
66 mimalloc = "0.1.52"
67 minijinja = { version = "2.24", features = ["unstable_machinery", "json", "builtins", "loader", "loop_controls", "preserve_order", "fuel"] }
68 minijinja-contrib = { version = "2.24", features = ["pycompat"] }
69 native-tls-vendored = { package = "native-tls", version = "0.2.18", features = ["vendored"] }
70 ndarray = { version = "0.17", features = ["serde"] }
71 openai-harmony = { package = "oss-harmony", git = "https://github.com/oss-harmony/harmony", tag = "v0.0.11", default-features = false }
72 openai-protocol = "1.6.0"
73 openssl = "0.10"
74 parking_lot = "0.12.5"
75 parquet = { version = "59.2.0", default-features = false, features = ["json", "snap", "zstd", "flate2", "flate2-rust_backend", "lz4", "brotli"] }
76 paste = "1.0.15"
77 prometheus-client = "0.24.0"
78 prometheus-client-derive-encode = "0.5.0"
79 prost = "0.14.3"
80 prost-types = "0.14.3"
81 protox = "0.9.1"
82 pyo3 = "0.29.0"
83 pythonize = "0.29.0"
84 rand = "0.9.3"
85 rand_distr = "0.5.1"
86 rayon = "1.11.0"
87 reasoning-parser = "1.2.2"
88 regex-automata = "0.4.14"
89 regex-syntax = "0.8.10"
90 reqwest = { version = "0.12.8", default-features = false, features = ["native-tls"] }
91 reqwest-0-13 = { package = "reqwest", version = "0.13.4", default-features = false, features = ["native-tls"] }
92 riptoken = { version = "0.3.0", default-features = false }
93 rlimit = "0.11.0"
94 rmp-serde = "1.3.1"
95 rmpv = { version = "1.3.1", features = ["with-serde"] }
96 rustc-hash = "1.1.0"
97 serde = { version = "1.0.228", features = ["derive", "rc"] }
98 serde-json-fmt = "0.1.0"
99 serde_default = "0.2.0"
100 serde_json = { version = "1.0.145", features = ["preserve_order"] }
101 serde_path_to_error = "0.1.20"
102 serde_repr = "0.1.20"
103 serde_tuple = "1.1.3"
104 serde_with = "3.21.0"
105 serial_test = { version = "3.2.0", features = ["file_locks"] }
106 sha2 = "0.10.9"
107 smallvec = "1.15.1"
108 socket2 = "0.6.3"
109 strum = { version = "0.27.2", features = ["derive"] }
110 subenum = "1.1.3"
111 subtle = "2.6"
112 task-local = "0.1.1"
113 tekken = { version = "0.2.0", default-features = false }
114 tempfile = "3.23.0"
115 thiserror = "2.0.16"
116 thiserror-ext = "0.3.0"
117 tiktoken-rs = "0.9.1"
118 time = { version = "0.3.47", features = ["formatting", "local-offset", "macros"] }
119 tls-listener = { version = "0.11.2", default-features = false, features = ["openssl", "tokio-net", "axum"] }
120 tokenizers = "0.22.0"
121 tokio = { version = "1.47.1", features = [
122 "macros",
123 "net",
124 "rt-multi-thread",
125 "sync",
126 "time",
127 ] }
128 tokio-openssl = "0.6"
129 tokio-stream = "0.1"
130 tokio-util = { version = "0.7.18", features = ["rt"] }
131 tonic = "0.14.6"
132 tonic-build = "0.14.6"
133 tonic-health = "0.14.6"
134 tonic-prost = "0.14.6"
135 tonic-prost-build = "0.14.6"
136 tool-parser = "1.2.0"
137 tower = { version = "0.5.3", features = ["util"] }
138 tower-http = { version = "0.6.8", features = ["cors", "trace"] }
139 tracing = { version = "0.1.44", features = ["release_max_level_debug"] }
140 tracing-futures = { version = "0.2.5", features = ["futures-03"] }
141 tracing-subscriber = { version = "0.3.20", features = ["env-filter", "fmt"] }
142 trait-set = "0.3.0"
143 url = "2.5.7"
144 uuid = { version = "1.22.0", features = ["v4"] }
145 validator = { version = "0.20.0", features = ["derive"] }
146 vllm-bench = { path = "src/bench" }
147 vllm-build-info = { path = "src/build-info" }
148 vllm-chat = { path = "src/chat" }
149 vllm-engine-core-client = { path = "src/engine-core-client" }
150 vllm-llm = { path = "src/llm" }
151 vllm-managed-engine = { path = "src/managed-engine" }
152 vllm-metrics = { path = "src/metrics" }
153 vllm-parser = { path = "src/parser" }
154 vllm-proto = { path = "proto" }
155 vllm-server = { path = "src/server" }
156 vllm-text = { path = "src/text" }
157 vllm-tokenizer = { path = "src/tokenizer" }
158 vllm-tracing = { path = "src/tracing" }
159 winnow = { version = "1.0.2", features = ["simd"] }
160 xgrammar-structural-tag = "0.4.1"
161 zeromq = { version = "0.6.0", default-features = false, features = [
162 "tokio-runtime",
163 "all-transport",
164 ] }
165
166 [workspace.lints.clippy]
167 too_many_arguments = "allow"
168
169 [profile.dev]
170 panic = "abort"
171
172 # Speed up cold tokenizer construction in tests.
173 [profile.dev.package]
174 fastokens = { opt-level = 3 }
175 regex-automata = { opt-level = 3 }
176 serde_json = { opt-level = 3 }
177 tokenizers = { opt-level = 3 }
178
179 [profile.release]
180 lto = "thin"
181 panic = "abort"