Struct MLConfig 

pub struct MLConfig {
    pub enabled: bool,
    pub model_variant: ModelVariant,
    pub threshold: f32,
    pub fallback_to_heuristic: bool,
    pub cache_enabled: bool,
    pub cache_config: CacheSettings,
    pub extra: HashMap<String, Value>,
}

ML detection configuration for scanners

Fields

• enabled: Whether ML detection is enabled
• model_variant: Model precision (FP32, FP16, INT8)
• threshold: Detection threshold (0.0 to 1.0)
• fallback_to_heuristic: Use heuristic if ML fails
• cache_enabled: Enable result caching
• cache_config: Cache configuration

Recommended Configurations

Production (balanced):

MLConfig {
    enabled: true,
    model_variant: ModelVariant::FP16,
    threshold: 0.5,
    fallback_to_heuristic: true,
    cache_enabled: true,
    cache_config: CacheSettings::default(),
}

High accuracy:

MLConfig {
    enabled: true,
    model_variant: ModelVariant::FP32,
    threshold: 0.3,  // Lower threshold = more sensitive
    fallback_to_heuristic: false,  // Require ML
    cache_enabled: true,
    cache_config: CacheSettings::default(),
}

Edge/Mobile:

MLConfig {
    enabled: true,
    model_variant: ModelVariant::INT8,
    threshold: 0.6,
    fallback_to_heuristic: true,
    cache_enabled: true,
    cache_config: CacheSettings {
        max_size: 100,  // Smaller cache
        ttl: Duration::from_secs(600),
    },
}

Fields

enabled: bool

Whether ML detection is enabled

model_variant: ModelVariant

Model variant to use (FP32, FP16, INT8)

threshold: f32

Detection threshold (0.0 to 1.0)

• Higher = fewer false positives, more false negatives
• Lower = fewer false negatives, more false positives
• Recommended: 0.5 for balanced results

fallback_to_heuristic: bool

Use heuristic detection if ML fails or is unavailable

cache_enabled: bool

Enable result caching for repeated inputs

cache_config: CacheSettings

Cache settings

extra: HashMap<String, Value>

Additional model-specific configuration

Implementations

impl MLConfig

pub fn new(model_variant: ModelVariant, threshold: f32) -> Self

Create a new ML configuration

pub fn production() -> Self

Create ML configuration for production use

• FP16 model (balanced speed/accuracy)
• 0.5 threshold (balanced sensitivity)
• Heuristic fallback enabled
• Caching enabled with 1000 entries, 1 hour TTL

pub fn edge() -> Self

Create ML configuration for edge/mobile deployment

• INT8 model (smallest size)
• 0.6 threshold (fewer false positives)
• Heuristic fallback enabled
• Smaller cache (100 entries, 10 minutes TTL)

pub fn high_accuracy() -> Self

Create ML configuration for high accuracy

• FP32 model (highest accuracy)
• 0.3 threshold (very sensitive)
• No heuristic fallback
• Aggressive caching

pub fn disabled() -> Self

Disable ML detection (heuristic-only mode)

pub fn validate(&self) -> Result<(), String>

Validate configuration

Trait Implementations

impl Clone for MLConfig

fn clone(&self) -> MLConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MLConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for MLConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for MLConfig

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for MLConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.