110 lines
2.5 KiB
TOML
110 lines
2.5 KiB
TOML
[package]
|
|
name = "ruvector-onnx-embeddings"
|
|
version = "0.1.0"
|
|
edition = "2021"
|
|
authors = ["RuVector Team"]
|
|
description = "ONNX-based embedding generation for RuVector - Reimagined embedding pipeline in pure Rust"
|
|
license = "MIT"
|
|
repository = "https://github.com/ruvnet/ruvector"
|
|
keywords = ["onnx", "embeddings", "vector-database", "rust", "ml"]
|
|
categories = ["science", "algorithms"]
|
|
|
|
# Make this a standalone package, not part of the workspace
|
|
[workspace]
|
|
|
|
[dependencies]
|
|
# ONNX Runtime - Core inference engine
|
|
ort = { version = "2.0.0-rc.9", features = ["download-binaries", "half"] }
|
|
|
|
# Tokenization - HuggingFace tokenizers in Rust
|
|
tokenizers = { version = "0.20", default-features = false, features = ["progressbar", "onig"] }
|
|
|
|
# Tensor operations
|
|
ndarray = { version = "0.16", features = ["rayon"] }
|
|
half = "2.4"
|
|
|
|
# Async runtime
|
|
tokio = { version = "1.41", features = ["full"] }
|
|
|
|
# Serialization
|
|
serde = { version = "1.0", features = ["derive"] }
|
|
serde_json = "1.0"
|
|
|
|
# Error handling
|
|
thiserror = "2.0"
|
|
anyhow = "1.0"
|
|
|
|
# HTTP client for model downloads
|
|
reqwest = { version = "0.12", features = ["blocking", "stream"] }
|
|
futures-util = "0.3"
|
|
|
|
# Progress bars and CLI
|
|
indicatif = "0.17"
|
|
console = "0.15"
|
|
|
|
# Logging
|
|
tracing = "0.1"
|
|
tracing-subscriber = { version = "0.3", features = ["env-filter"] }
|
|
|
|
# File operations
|
|
sha2 = "0.10"
|
|
hex = "0.4"
|
|
tempfile = "3.14"
|
|
dirs = "5.0"
|
|
|
|
# Parallel processing
|
|
rayon = "1.10"
|
|
|
|
# Concurrency
|
|
parking_lot = "0.12"
|
|
|
|
# UUID for vector IDs
|
|
uuid = { version = "1.11", features = ["v4"] }
|
|
|
|
# GPU acceleration (optional)
|
|
wgpu = { version = "23.0", optional = true }
|
|
bytemuck = { version = "1.14", optional = true, features = ["derive"] }
|
|
|
|
[dev-dependencies]
|
|
criterion = { version = "0.5", features = ["html_reports"] }
|
|
approx = "0.5"
|
|
|
|
[[bench]]
|
|
name = "embedding_benchmark"
|
|
harness = false
|
|
|
|
[[bench]]
|
|
name = "gpu_benchmark"
|
|
harness = false
|
|
required-features = ["gpu"]
|
|
|
|
[[example]]
|
|
name = "basic_embedding"
|
|
path = "examples/basic.rs"
|
|
|
|
[[example]]
|
|
name = "batch_embedding"
|
|
path = "examples/batch.rs"
|
|
|
|
[[example]]
|
|
name = "semantic_search"
|
|
path = "examples/semantic_search.rs"
|
|
|
|
[features]
|
|
default = ["download-models"]
|
|
download-models = []
|
|
cuda = ["ort/cuda"]
|
|
tensorrt = ["ort/tensorrt"]
|
|
coreml = ["ort/coreml"]
|
|
simsimd = [] # Optional SIMD acceleration (not yet implemented)
|
|
|
|
# GPU acceleration features
|
|
gpu = ["dep:wgpu", "dep:bytemuck"]
|
|
cuda-wasm = ["gpu"] # CUDA-WASM transpilation (requires gpu)
|
|
webgpu = ["gpu"] # WebGPU backend alias
|
|
|
|
[profile.release]
|
|
opt-level = 3
|
|
lto = "thin"
|
|
codegen-units = 1
|