Hierarchy Summary

• AdaptiveDraftGateSettings
  • AdaptiveDraftGateOptions

• AneInstallProgram
  • AneInstallResult

• ArtifactIdentityStore
  • FileArtifactIdentityStore

• AssistantDrafterModel
  • GemmaAssistantDrafter

• Cache
  • AttentionCache
    • BatchedRotatingCache
    • TrainingAttentionCache
      • TrainingCache
    • CommittedAttentionCache
      • BatchedKVCache
      • BatchedQuantizedKVCache
      • BatchedTurboQuantKVCache
      • Bf16FirstQuantizedKVCache
        • Bf16FirstKv4Head256Cache
      • Bf16FirstRotatingQuantizedKVCache
      • Bf16FirstTurboQuantKVCache
      • DelayedQuantizedKVCache
      • DelayedRotatingQuantizedKVCache
      • DelayedTurboQuantKVCache
      • KVCache
      • PagedKVCache
      • QuantizedKVCache
      • RotatingKVCache
      • RotatingQuantizedKVCache
        • BatchedRotatingQuantCache
      • SpeculativeRotatingKVCache
      • TurboQuantKVCache
    • BidirectionalAttentionCache
      • Bf16FirstQuantizedKVCache
      • Bf16FirstRotatingQuantizedKVCache
      • Bf16FirstTurboQuantKVCache
      • KVCache
      • PagedKVCache
      • RotatingKVCache
    • EncoderAttentionCache
      • Bf16FirstQuantizedKVCache
      • Bf16FirstRotatingQuantizedKVCache
      • Bf16FirstTurboQuantKVCache
      • KVCache
      • PagedKVCache
      • RotatingKVCache
    • BlockAttentionCache
      • Bf16FirstQuantizedKVCache
      • Bf16FirstRotatingQuantizedKVCache
      • Bf16FirstTurboQuantKVCache
      • KVCache
      • PagedKVCache
      • QuantizedKVCache
      • RotatingKVCache
      • RotatingQuantizedKVCache
      • TurboQuantKVCache
  • GatedDeltaCache
    • SSMCache
      • BatchedSSMCache
    • TrainingSSMCache
  • LatentAttentionCache
    • MLACache
  • RowBatchCache
  • BatchableCache
    • BatchedKVCache
    • BatchedQuantizedKVCache
    • BatchedRotatingCache
    • BatchedRotatingQuantCache
    • BatchedSSMCache
    • BatchedTurboQuantKVCache
    • MLACache
    • MlxRowLayout
    • SpeculativeRowLayout
      • SpeculativeRotatingKVCache
  • PaddedPrefillCache
    • BatchedKVCache
    • BatchedQuantizedKVCache
    • BatchedRotatingCache
    • BatchedRotatingQuantCache
    • BatchedSSMCache
    • BatchedTurboQuantKVCache
    • DelayedRotatingQuantizedKVCache
  • DecodeTrace
  • CompiledDecodeCache
    • Bf16FirstQuantizedKVCache
    • Bf16FirstRotatingQuantizedKVCache
    • Bf16FirstTurboQuantKVCache
    • KVCache
    • PagedKVCache
    • QuantizedKVCache
    • RotatingKVCache
    • RotatingQuantizedKVCache
    • TurboQuantKVCache

• Cancellation
  • CancellationSource

• CatalogError
  • CatalogFailure

• ContinuationRequest
  • OrdinaryContinuation

• DeepspecDrafterModel
  • DeepspecDrafter

• DisposableResource
  • PreparedStateChange

• DraftProvider
  • AssistantProvider
  • DeepspecProvider
  • Dflash2Provider
  • DflashProvider
  • Glm52NativeMtpProvider
  • NgramProvider
  • QwenMtpProvider
  • TwoModelProvider

• DsparkDrafterModel
  • DflashDrafter

• DurabilitySnapshotStats
  • DurabilityFlushResult

• EventBus
  • EventHub

• FixedPerLayerInputs
  • TracePerLayerInputs

• GatedDeltaNet
  • QwenGatedDeltaNet
    • QwenFusedGatedDeltaNet
      • QwenFusedHeadsDeltaNet

• GegluMLP
  • GemmaMlpWeights

• GemmaLayerWeights
  • GemmaInputLayerWeights
  • GemmaMoeLayerWeights

• GenerationSession
  • EngineSession

• Glm52MlaWeightSource
  • Glm52WeightSource
    • ColibriGlm52ResidentWeights

• Glm52RuntimeOpenOptions
  • ModelOpenOptions

• KvAttentionView
  • KvDonorAttention

• KvCodec
  • TurboQuantCodec

• LogitsProcessorOptions
  • StepSamplerOptions
  • GenerateOptions

• MemoryPlan
  • Glm52MemoryPlan
  • FitReport

• MlxCompiledDecodeStep
  • CompiledDecode

• MlxDeclaredGraph
  • DiffusionGemmaModel
  • Gemma4E4b
  • Gemma4Model
  • Gemma4Moe
  • Gemma4TwelveB
  • Glm52Model
  • MiniCPM5Model
  • Qwen35Model
  • Qwen38TrellisM1Max
  • Qwen38TrellisM4Pro
  • Qwen3Model
  • Qwen3MoeModel
  • UniversalDenseModel
  • MlxTokenGraph

• MlxDenoisingOperations
  • DiffusionGemmaModel

• MlxGroupPreparation
  • MlxPrefillRows
    • MlxPrefillCohort

• MlxModelMemory
  • MlxTokenGraph

• MlxPrefillState
  • PrefillState

• MlxTokenPhases
  • Gemma4E4b
  • Gemma4Model
  • Gemma4Moe
  • Gemma4TwelveB
  • Glm52Model
  • MiniCPM5Model
  • Qwen35Model
  • Qwen38TrellisM1Max
  • Qwen38TrellisM4Pro
  • Qwen3Model
  • Qwen3MoeModel
  • UniversalDenseModel

• ModelHost
  • WhisperModelHost

• ModelHostError
  • ModelHostFailure

• ModelImplementationProvider
  • ModelImplementationRegistry

• ModelProfile
  • ArtifactModelProfile

• NativeMtpHead
  • Glm52MtpGraph

• Overlay
  • Palette
  • ShortcutSheet

• PackedTrellisTensor
  • LdlqEncoding

• PrefillStep
  • MlxPrefillStep

• PrefixCache
  • PromptCache
    • TieredPromptCache

• PrefixCacheHit
  • PromptCacheEntry

• ProposalSource
  • StrictRowSource

• Qwen3Attention
  • Qwen3TextAttention
  • Qwen3MediaAttention

• QwenAttentionLayer
  • QwenResidualAttentionLayer

• QwenDeltaLayer
  • QwenResidualDeltaLayer

• RecurrentMtpModule
  • MtpModule

• ReduceMixedGateUp
  • ReduceSplitMixedGateUp

• RequestStateCache
  • KvCacheService

• RetentionPolicy
  • CostSizeRetention
  • LruRetention

• SamplerOptions
  • StepSamplerOptions
  • GenerateOptions

• StorageEntrySpec
  • StorageRegistration

• StrictFillContext
  • ToolCallFillContext

• UniversalRope
  • NoRope

• WeightTransformPlan
  • LlamaWeightTransformPlan
  • QwenWeightTransformPlan

• WhisperDecoderCache
  • WhisperKvCache
  • WhisperFastCache

• WhisperDecodingOptions
  • WhisperTranscribeOptions

• WhisperGraph
  • WhisperModel