Struct RustAI 

pub struct RustAI { /* private fields */ }

Unified facade for the rust-ai ecosystem.

Provides high-level APIs that orchestrate multiple ecosystem crates to accomplish common AI engineering tasks.

Capabilities

Workflow	Description
finetune()	LoRA, DoRA, AdaLoRA adapter creation
quantize()	4-bit (NF4/FP4) and 1.58-bit (BitNet) quantization
vsa()	VSA-based operations and optimization
train()	YAML-driven training pipelines

Example

use rust_ai_core::{RustAI, RustAIConfig};

let config = RustAIConfig::new()
    .with_verbose(true);

let ai = RustAI::new(config).unwrap();
println!("Device: {:?}", ai.device());
println!("Ecosystem: {:?}", ai.ecosystem());

Implementations

impl RustAI

pub fn new(config: RustAIConfig) -> Result<Self>

Create a new RustAI instance.

Initializes the device and ecosystem information.

Arguments

• config - Configuration options

Returns

A configured RustAI instance.

Errors

Returns CoreError::DeviceNotAvailable if device initialization fails.

Example

use rust_ai_core::{RustAI, RustAIConfig};

let ai = RustAI::new(RustAIConfig::default())?;

pub fn device(&self) -> &Device

Get the active device.

pub fn ecosystem(&self) -> &EcosystemInfo

Get ecosystem information.

pub fn config(&self) -> &RustAIConfig

Get the configuration.

pub fn is_cuda(&self) -> bool

Check if CUDA is available and active.

pub fn finetune(&self) -> FinetuneBuilder<'_>

Start a fine-tuning workflow.

Returns a builder for configuring and running fine-tuning.

Example

let config = ai.finetune()
    .model("meta-llama/Llama-2-7b")
    .adapter(AdapterType::Lora)
    .rank(64)
    .build()?;

pub fn quantize(&self) -> QuantizeBuilder<'_>

Start a quantization workflow.

Returns a builder for configuring and running quantization.

Example

let config = ai.quantize()
    .method(QuantizeMethod::Nf4)
    .bits(4)
    .build();

pub fn vsa(&self) -> VsaBuilder<'_>

Start a VSA workflow.

Returns a builder for VSA-based operations.

Example

let config = ai.vsa()
    .dimension(10000)
    .build();

pub fn train(&self) -> TrainBuilder<'_>

Start an Axolotl training pipeline.

Returns a builder for YAML-driven training configuration.

Example

let config = ai.train()
    .config_file("config.yaml")
    .build()?;

pub fn info(&self) -> RustAIInfo

Get information about the RustAI environment.

Returns a struct containing version info, ecosystem, and device details.