furnace-server 0.14.0

Fast, unified, Rust-native, Axum-compatible inference endpoints and nervous system for tensor models and LLMs: KServe / Triton and OpenAI APIs from one pure-Rust server
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

// The gRPC code for the Open Inference Protocol (feature `grpc`). The built-in ONNX models are
// generated by furnace-burn's build script.

fn main() {
    // Without it, Cargo would run this script again after any change in the package
    println!("cargo:rerun-if-changed=build.rs");
    #[cfg(feature = "grpc")]
    compile_grpc();
}

/// Rust types and the service trait for the Open Inference Protocol's gRPC API, from
/// proto/open_inference_grpc.proto. protox parses the file in pure Rust, so no `protoc` is needed.
#[cfg(feature = "grpc")]
fn compile_grpc() {
    println!("cargo:rerun-if-changed=proto/open_inference_grpc.proto");
    let descriptors = protox::compile(["open_inference_grpc.proto"], ["proto"])
        .expect("failed to parse proto/open_inference_grpc.proto");
    tonic_prost_build::configure()
        .build_server(true)
        .build_client(true)
        .compile_fds(descriptors)
        .expect("failed to generate the gRPC code");
}