LMRuntime.com / Generated reference
UAIX.LmRuntime API Reference
Search the documented public types and members generated from public package XML documentation.
A searchable reference generated from public package XML documentation: 376 documented types and 2,044 documented members across the package family.
All documented types are shown.
UAIX.LmRuntime.Abstractions
99 documented types and 631 documented members.
IClaimBoundaryPolicyUAIX.LmRuntime.Abstractions
1 member
Evaluates generated text against bounded claim-boundary rules.
EvaluateAsync(string,System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.Governance.ClaimBoundaryRule>,System.Threading.CancellationToken)
Evaluates the async against the supplied policy and bounded observation state.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
rules- The claim-boundary rules. Implementations may use defaults when this collection is empty.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<ClaimBoundaryDecision> that completes with the result of IClaimBoundaryPolicy.EvaluateAsync: Evaluates the async against the supplied policy and bounded observation state. Fault and cancellation states are propagated without a successful partial result.
IClaimLifecyclePolicyUAIX.LmRuntime.Abstractions
1 member
Enforces explicit evidence, boundary, and human-authority transitions for runtime claims.
Evaluate(UAIX.LmRuntime.Governance.ClaimTransitionRequest)
Evaluates the supplied request against the supplied policy and bounded observation state.
request- The requested transition and its bounded evidence signals.
Returns: The ClaimTransitionDecision result produced by IClaimLifecyclePolicy.Evaluate for this contract: Evaluates the supplied request against the supplied policy and bounded observation state. It is published only after all documented validation and ownership transitions succeed.
IConstraintClosureAnalyzerUAIX.LmRuntime.Abstractions
1 member
Analyzes a bounded work-constraint registry for closed maintenance cycles and retirement candidates.
Analyze(UAIX.LmRuntime.Governance.ConstraintRegistrySnapshot)
Computes deterministic graph closure information for the supplied registry snapshot.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
Returns: The closure report, including validation diagnostics and retirement candidates.
IConstraintPolicyUAIX.LmRuntime.Abstractions
1 member
Evaluates request-side constraints before budgeted execution is allowed.
EvaluateAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.Governance.ConstraintRule>,System.Threading.CancellationToken)
Evaluates the request against configured constraint rules.
request- The InferenceRequest containing the complete caller-owned inputs for EvaluateAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
rules- The rules sequence used by this operation; its required length, ordering, and element bounds are validated before access.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<ConstraintDecision> that completes with the result of IConstraintPolicy.EvaluateAsync: Evaluates the request against configured constraint rules. Fault and cancellation states are propagated without a successful partial result.
IEvidenceLedgerUAIX.LmRuntime.Abstractions
1 member
Appends immutable governance evidence receipts.
AppendAsync(UAIX.LmRuntime.Governance.GovernanceDecisionReceipt,System.Threading.CancellationToken)
Appends the async to the current IEvidenceLedger state after validating capacity, ordering, and ownership constraints.
receipt- The immutable governance decision receipt appended to the caller-owned evidence ledger in the supplied order.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: A task that represents completion of the asynchronous operation.
IInferenceRuntimeUAIX.LmRuntime.Abstractions
2 members
Defines the public runtime orchestration API.
GenerateAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Threading.CancellationToken)
Generates the async through the deterministic execution path owned by IInferenceRuntime.
request- The InferenceRequest containing the complete caller-owned inputs for GenerateAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous Task<InferenceResponse> that completes with the result of IInferenceRuntime.GenerateAsync: Generates the async through the deterministic execution path owned by IInferenceRuntime. Fault and cancellation states are propagated without a successful partial result.
StreamAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Threading.CancellationToken)
Streams the async in observable sequence order while honoring caller cancellation.
request- The InferenceRequest containing the complete caller-owned inputs for StreamAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The normalized streaming events, enumerated in source order with caller cancellation and failure propagation governed by StreamAsync.
IInferenceSessionUAIX.LmRuntime.Abstractions
3 members
Defines a stateful inference session with explicit prefill and decode phases.
SessionId
Gets the session identifier.
DecodeNextAsync(System.Threading.CancellationToken)
Decodes the next token or text chunk from the session.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<StreamingDelta> that completes with the result of IInferenceSession.DecodeNextAsync: Decodes the next token or text chunk from the session. Fault and cancellation states are propagated without a successful partial result.
PrefillAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Threading.CancellationToken)
Prefills the async into the current model state after validating token and cache bounds.
request- The InferenceRequest containing the complete caller-owned inputs for PrefillAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The number of prompt tokens accepted by the session.
IMemoryFirewallUAIX.LmRuntime.Abstractions
1 member
Evaluates source-routed memory packets before they can enter an active retrieval tier.
Evaluate(UAIX.LmRuntime.Governance.MemoryFirewallRequest)
Evaluates packet provenance, integrity, freshness, entropy, trust, contradiction, evidence, and review state.
request- The MemoryFirewallRequest containing the complete caller-owned inputs for Evaluate; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
Returns: The memory region containing the supplied request, bounded to the validated range owned by the result.
IModelAdapterUAIX.LmRuntime.Abstractions
4 members
Defines the execution boundary for provider-hosted or local models.
ProviderName
Gets the provider or backend name.
GenerateAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Threading.CancellationToken)
Generates the async through the deterministic execution path owned by IModelAdapter.
request- The InferenceRequest containing the complete caller-owned inputs for GenerateAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous Task<InferenceResponse> that completes with the result of IModelAdapter.GenerateAsync: Generates the async through the deterministic execution path owned by IModelAdapter. Fault and cancellation states are propagated without a successful partial result.
GetModelAsync(string,System.Threading.CancellationToken)
Gets model metadata for the requested model.
model- The model whose validated metadata, tensors, or runtime state are consumed by this operation.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<ModelDescriptor> that completes with the result of IModelAdapter.GetModelAsync: Gets model metadata for the requested model. Fault and cancellation states are propagated without a successful partial result.
StreamAsync(UAIX.LmRuntime.Contracts.InferenceRequest,System.Threading.CancellationToken)
Streams the async in observable sequence order while honoring caller cancellation.
request- The InferenceRequest containing the complete caller-owned inputs for StreamAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The normalized streaming event sequence, enumerated in source order with caller cancellation and failure propagation governed by StreamAsync.
IQuarantineLedgerUAIX.LmRuntime.Abstractions
1 member
Writes append-only records for generated or unreviewed artifacts that require quarantine.
AppendAsync(UAIX.LmRuntime.Governance.QuarantineRecord,System.Threading.CancellationToken)
Appends the async to the current IQuarantineLedger state after validating capacity, ordering, and ownership constraints.
record- The immutable quarantine record appended to the caller-owned ledger without altering its evidence fields.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: A task that represents completion of the asynchronous operation.
IResourceEconomyEngineUAIX.LmRuntime.Abstractions
1 member
Advances the structural-control resource economy using explicit gain and cost inputs.
Evaluate(UAIX.LmRuntime.Governance.ResourceEconomyState,UAIX.LmRuntime.Governance.ResourceEconomyInput,UAIX.LmRuntime.Governance.ResourceEconomyPolicy)
Evaluates the supplied state against the supplied policy and bounded observation state.
state- The immutable state observed before the candidate action.
input- The gain and burden estimates for the candidate action.
policy- The policy that define validation limits and execution behavior; required values are checked before use.
Returns: The ResourceEconomyTransition result produced by IResourceEconomyEngine.Evaluate for this contract: Evaluates the supplied state against the supplied policy and bounded observation state. It is published only after all documented validation and ownership transitions succeed.
IReviewGatePolicyUAIX.LmRuntime.Abstractions
1 member
Evaluates slow-loop review gates before consequential runtime work proceeds.
EvaluateAsync(UAIX.LmRuntime.Governance.ReviewGateRequest,System.Threading.CancellationToken)
Evaluates whether a request can continue in the automatic lane.
request- The ReviewGateRequest containing the complete caller-owned inputs for EvaluateAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<ReviewGateDecision> that completes with the result of IReviewGatePolicy.EvaluateAsync: Evaluates whether a request can continue in the automatic lane. Fault and cancellation states are propagated without a successful partial result.
IRuntimeBudgetGovernorUAIX.LmRuntime.Abstractions
1 member
Evaluates whether runtime work can proceed under the configured resource economy.
EvaluateAsync(UAIX.LmRuntime.Governance.RuntimeBudgetRequest,UAIX.LmRuntime.Governance.RuntimeBudget,System.Threading.CancellationToken)
Evaluates the requested runtime work against a resource budget.
request- The RuntimeBudgetRequest containing the complete caller-owned inputs for EvaluateAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
budget- The immutable runtime budget whose quantitative limits constrain the evaluated operation or proposal.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<RuntimeBudgetDecision> that completes with the result of IRuntimeBudgetGovernor.EvaluateAsync: Evaluates the requested runtime work against a resource budget. Fault and cancellation states are propagated without a successful partial result.
IStructuralChangePlannerUAIX.LmRuntime.Abstractions
1 member
Evaluates proposed structural mutations through budget, no-op, evidence, and review gates.
PlanAsync(UAIX.LmRuntime.Governance.StructuralChangeRequest,UAIX.LmRuntime.Governance.RuntimeBudget,System.Threading.CancellationToken)
Plans the disposition for a proposed structural change.
request- The StructuralChangeRequest containing the complete caller-owned inputs for PlanAsync; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
budget- The immutable runtime budget whose quantitative limits constrain the evaluated operation or proposal.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<StructuralChangeDecision> that completes with the result of IStructuralChangePlanner.PlanAsync: Plans the disposition for a proposed structural change. Fault and cancellation states are propagated without a successful partial result.
IStructuralOperatorEngineUAIX.LmRuntime.Abstractions
1 member
Converts resource and claim-boundary decisions into an auditable structural operator receipt.
DecideAsync(UAIX.LmRuntime.Governance.RuntimeBudgetDecision,UAIX.LmRuntime.Governance.ClaimBoundaryDecision,System.Threading.CancellationToken)
Selects the structural operator for the current decision context.
budgetDecision- The prior budget decision that must authorize the structural operation before any proposal can advance.
claimBoundaryDecision- The claim-boundary decision used to constrain the resulting action, or null when no claim evaluation applies.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<GovernanceDecisionReceipt> that completes with the result of IStructuralOperatorEngine.DecideAsync: Selects the structural operator for the current decision context. Fault and cancellation states are propagated without a successful partial result.
IStructuralPhaseDetectorUAIX.LmRuntime.Abstractions
1 member
Classifies the current structural regime from bounded fast-loop measurements.
Detect(UAIX.LmRuntime.Governance.StructuralObservationWindow,UAIX.LmRuntime.Governance.StructuralPhasePolicy)
Classifies one observation window using deterministic threshold precedence.
window- The immutable observation window whose ordered measurements are evaluated to determine the structural phase.
policy- The policy that define validation limits and execution behavior; required values are checked before use.
Returns: The StructuralPhaseAssessment result produced by IStructuralPhaseDetector.Detect for this contract: Classifies one observation window using deterministic threshold precedence. It is published only after all documented validation and ownership transitions succeed.
ITeleodynamicControlPlaneUAIX.LmRuntime.Abstractions
1 member
Coordinates the resource, proposal, closure, phase, claim, and trace components on a slow timescale.
Evaluate(UAIX.LmRuntime.Governance.TeleodynamicControlRequest)
Evaluates one complete control cycle without mutating model weights, tokenizer state, or generated output.
request- The TeleodynamicControlRequest containing the complete caller-owned inputs for Evaluate; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
Returns: The phase, closure, structural decision, claim decision when present, and tamper-evident trace entry.
ITeleodynamicDecisionEngineUAIX.LmRuntime.Abstractions
1 member
Selects a bounded structural proposal or explicit no-op using resource and evidence constraints.
Decide(string,System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.Governance.StructuralProposal>,UAIX.LmRuntime.Governance.ResourceEconomyState,UAIX.LmRuntime.Governance.ResourceEconomyPolicy,UAIX.LmRuntime.Governance.TeleodynamicDecisionPolicy,System.DateTimeOffset)
Evaluates the proposal set and returns one deterministic, no-op-aware decision.
decisionId- The stable identifier assigned by the caller to the decision cycle.
proposals- The proposals sequence used by this operation; its required length, ordering, and element bounds are validated before access.
state- The validated state value consumed by the operation; mutations, when applicable, are limited to the explicitly documented state owner.
resourcePolicy- The immutable resource-economy policy that defines allowed transitions and quantitative thresholds for the decision.
decisionPolicy- The immutable decision policy that constrains proposal selection, no-op behavior, and review requirements.
createdUtc- The caller-supplied UTC timestamp recorded in the deterministic decision receipt; non-UTC offsets must be normalized by the caller.
Returns: The TeleodynamicDecision result produced by ITeleodynamicDecisionEngine.Decide for this contract: Evaluates the proposal set and returns one deterministic, no-op-aware decision. It is published only after all documented validation and ownership transitions succeed.
ITeleodynamicTraceChainUAIX.LmRuntime.Abstractions
3 members
Maintains an append-only, SHA-256-linked trace of bounded structural-control decisions.
Append(UAIX.LmRuntime.Governance.TeleodynamicTraceRequest)
Appends one canonical event to the trace.
request- The TeleodynamicTraceRequest containing the complete caller-owned inputs for Append; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
Returns: The immutable trace entry assigned to the event.
Snapshot
Gets an immutable snapshot of the current trace entries.
Returns: The current trace entries in ascending sequence order.
Verify(System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.Governance.TeleodynamicTraceEntry>)
Verifies an arbitrary trace snapshot without mutating the current chain.
entries- The entries to verify in their supplied order.
Returns: The first detected chain error or a valid result.
ITokenizerUAIX.LmRuntime.Abstractions
6 members
Defines tokenization behavior for runtime token budgeting and model parity work.
Name
Gets the tokenizer name.
CountTokens(string)
Counts tokens in a single text value.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: The int value computed by ITokenizer.CountTokens for this contract: Counts tokens in a single text value. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts tokens across a set of model messages.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by ITokenizer.CountTokens for this contract: Counts tokens across a set of model messages. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes token identifiers into text when the tokenizer has a vocabulary.
tokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes text into token identifiers when the tokenizer has a vocabulary.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
IUaiMemoryStoreUAIX.LmRuntime.Abstractions
2 members
Defines append and query behavior for.uai-backed runtime memory.
AppendAsync(UAIX.LmRuntime.Contracts.UaiMemoryEntry,System.Threading.CancellationToken)
Appends the async to the current IUaiMemoryStore state after validating capacity, ordering, and ownership constraints.
entry- The entry examined or transformed by this operation; it must satisfy the declared type and range constraints.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: A task that completes when the entry has been written.
ReadAsync(UAIX.LmRuntime.Contracts.MemoryQuery,System.Threading.CancellationToken)
Reads the async from the current IUaiMemoryStore state using the component's validated representation.
query- The bounded memory query defining the caller-authorized selection criteria and result ceiling.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<IReadOnlyList<UaiMemoryEntry>> that completes with the result of IUaiMemoryStore.ReadAsync: Reads the async from the current IUaiMemoryStore state using the component's validated representation. Fault and cancellation states are propagated without a successful partial result.
RuntimePackageVersionUAIX.LmRuntime.Abstractions
2 members
Provides the runtime package version emitted by the centrally configured assembly metadata.
Package projects obtain their version from Directory.Build.props. Runtime components should use this type instead of repeating a version literal, which keeps evidence, session, CLI, and governance records aligned with the assembly that produced them.
Current
Gets the normalized three-part package version of the active runtime distribution.
Resolve(System.Reflection.Assembly)
Resolves a normalized three-part package version from an assembly identity.
assembly- The assembly whose centrally generated version metadata is authoritative.
Returns: A major.minor.build version string, or 0.0.0 when the assembly has no version metadata.
AdapterStatusUAIX.LmRuntime.Contracts
3 members
Represents normalized adapter status information.
Message
Gets a normalized warning or diagnostic message.
RequestId
Gets the provider request identifier, if supplied by the backend.
StatusCode
Gets the provider status code, if applicable.
FinishReasonUAIX.LmRuntime.Contracts
8 members
Identifies why an inference response stopped.
Cancelled
The request was cancelled.
ContentFilter
The runtime stopped generation because content was filtered.
Error
The backend reported an execution error.
Length
The model hit the configured maximum output token budget.
PolicyDenied
The runtime selected no-op because policy, budget, or claim-boundary rules blocked automatic execution.
Stop
The model naturally stopped.
ToolCall
The model emitted or requested a tool call.
Unknown
The provider or backend did not return a reason.
InferenceRequestUAIX.LmRuntime.Contracts
14 members
Represents a provider-neutral inference request.
ConversationId
Gets the conversation identifier when one is present.
MaxOutputTokens
Gets the maximum output token budget.
Messages
Gets the message sequence.
Metadata
Gets caller-supplied metadata propagated to adapters and diagnostics.
Model
Gets the requested model identifier.
ResponseFormat
Gets the requested response format.
Seed
Gets the deterministic sampler seed when one is present.
StopSequences
Gets stop sequences used to terminate generation.
Temperature
Gets the sampling temperature.
ToolChoice
Gets tool selection guidance.
Tools
Gets tool definitions available to the model.
TopK
Gets the top-k sampling cutoff when one is present.
TopP
Gets the nucleus sampling probability cutoff.
UseMemory
Gets a value indicating whether.uai memory should be injected before execution.
InferenceResponseUAIX.LmRuntime.Contracts
11 members
Represents a normalized inference response.
AdapterStatus
Gets normalized adapter status information.
ConversationId
Gets the conversation identifier when one is present.
CreatedUtc
Gets the UTC creation timestamp.
FinishReason
Gets the legacy textual finish reason.
FinishReasonKind
Gets the strongly typed finish reason.
GovernanceReceipt
Gets the governance receipt emitted by budget or claim-boundary policy.
Model
Gets the resolved model identifier.
OutputText
Gets the output text.
Provider
Gets the provider or backend name.
ResponseId
Gets the response identifier.
Usage
Gets normalized usage data.
InferenceUsageUAIX.LmRuntime.Contracts
4 members
Represents normalized model usage data.
CachedInputTokens
Gets the provider-cached input token count when available.
EstimatedCostMicros
Gets the estimated cost in one-millionth currency units, if known.
InputTokens
Gets the input token count.
OutputTokens
Gets the output token count.
LlmMessageUAIX.LmRuntime.Contracts
8 members
Represents a canonical chat or completion message.
Content
Gets the text content for the message.
Role
Gets the message role.
ToolCallId
Gets the tool call identifier associated with a tool message when one is present.
Assistant(string)
Creates an immutable assistant-role LLM message from caller-supplied content.
content- The content processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: A new assistant-role message whose content is never null.
Developer(string)
Creates an immutable developer-role LLM message from caller-supplied content.
content- The content processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: A new developer-role message whose content is never null.
System(string)
Creates an immutable system-role LLM message from caller-supplied content.
content- The content processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: A new system-role message whose content is never null.
Tool(string,string)
Creates an immutable tool-role LLM message correlated to the supplied tool call.
toolCallId- The caller-owned tool call id used for deterministic correlation by Tool; it must satisfy the documented range and grammar and grants no additional authority.
content- The content processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: A new tool-role message containing the supplied correlation identifier and non-null content.
User(string)
Creates an immutable user-role LLM message from caller-supplied content.
content- The content processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
Returns: A new user-role message whose content is never null.
LlmRoleUAIX.LmRuntime.Contracts
5 members
Identifies the role of a message in a canonical inference request.
Assistant
Assistant message content.
Developer
Developer-level instruction context.
System
System-level instruction context.
Tool
Tool input or output content.
User
End-user message content.
MemoryQueryUAIX.LmRuntime.Contracts
3 members
Represents a query against.uai memory entries.
ConversationId
Gets the conversation identifier filter when one is present.
MaxEntries
Gets the maximum number of entries to return.
Text
Gets full-text query text when one is present.
ModelDescriptorUAIX.LmRuntime.Contracts
6 members
Describes a model visible to the runtime.
Capabilities
Gets capability names exposed by the model.
ContextLength
Gets the maximum context length in tokens.
IsLocal
Gets a value indicating whether the model executes locally.
ModelId
Gets the model identifier.
Provider
Gets the provider or backend name.
SupportsStreaming
Gets a value indicating whether streaming is supported.
ProviderErrorUAIX.LmRuntime.Contracts
4 members
Represents a normalized adapter or provider error.
Code
Gets the normalized error code.
Message
Gets the error message.
Retriable
Gets a value indicating whether retry may be safe.
RetryAfter
Gets the suggested retry delay when available.
ResponseFormatUAIX.LmRuntime.Contracts
3 members
Describes structured output requirements for a response.
JsonSchema
Gets the JSON schema document used when UAIX.LmRuntime.Contracts.ResponseFormat.Kind is UAIX.LmRuntime.Contracts.ResponseFormatKind.JsonSchema.
Kind
Gets the response format kind.
Strict
Gets a value indicating whether the backend should enforce strict schema adherence when supported.
ResponseFormatKindUAIX.LmRuntime.Contracts
3 members
Identifies the canonical output format mode.
JsonObject
JSON object response.
JsonSchema
JSON schema-constrained response.
Text
Free-form text response.
RuntimeOptionsUAIX.LmRuntime.Contracts
24 members
Defines runtime orchestration settings.
AvailableResourceBudget
Gets the available resource budget used by the default governor.
ClaimBoundaryRules
Gets additional claim-boundary rules used by the default claim policy.
ConstraintRules
Gets additional request-side constraint rules used by the default constraint policy.
DefaultModel
Gets the default model identifier.
EnableClaimBoundaryPolicy
Gets a value indicating whether generated text is evaluated against claim-boundary rules.
EnableConstraintPolicy
Gets a value indicating whether request-side constraint rules are evaluated before adapter execution.
EnableReviewGatePolicy
Gets a value indicating whether slow-loop review gates are evaluated before execution.
EnableTeleodynamicGovernance
Gets a value indicating whether Teleodynamic governance gates are evaluated before execution.
FailOnEvidenceLedgerError
Gets a value indicating whether ledger append errors should fail inference.
MaxContextTokens
Gets the maximum context tokens accepted by the orchestrator before adapter execution.
MaxMemoryCharacters
Gets the maximum memory characters injected into the system context.
MaxMemoryEntries
Gets the maximum number of memory entries injected into a request.
MaxToolDefinitions
Gets the maximum number of tool definitions exposed to a request.
MaxUncertaintyScore
Gets the maximum uncertainty score accepted in the automatic lane.
MemoryEntryCost
Gets the resource weight assigned to each injected memory entry.
QuarantineGeneratedNeedsHumanReview
Gets a value indicating whether generated items needing human review are written to the quarantine ledger.
ReturnNoOpResponseOnClaimBoundaryViolation
Gets a value indicating whether claim-boundary violations return a no-op response.
ReturnNoOpResponseOnGovernanceDenial
Gets a value indicating whether budget-denied requests return a no-op response instead of throwing.
ReturnNoOpResponseOnReviewGateRequired
Gets a value indicating whether review-gated requests return a no-op response instead of throwing.
ReviewMinuteCost
Gets the resource weight assigned to each declared review minute.
TokenCostWeight
Gets the resource weight assigned to each token.
ToolDefinitionCost
Gets the resource weight assigned to each exposed tool definition.
UncertaintyCost
Gets the resource weight assigned to normalized uncertainty.
ViabilityFloor
Gets the minimum resource reserve below which automatic actions are blocked.
StreamingDeltaUAIX.LmRuntime.Contracts
8 members
Represents a normalized streaming inference event.
CreatedUtc
Gets the UTC event timestamp.
Error
Gets an error message for error events.
ResponseId
Gets the response identifier.
Text
Gets the text delta for text events.
ToolArgumentsDelta
Gets a tool-call argument delta.
ToolCallId
Gets the tool-call identifier for tool deltas.
Type
Gets the event type.
Usage
Gets the usage payload when available.
StreamingEventTypeUAIX.LmRuntime.Contracts
6 members
Identifies the type of a streaming inference event.
Completed
The stream completed successfully.
Delta
The event contains text delta content.
Error
The stream completed with an error.
Start
The stream has started.
ToolCallDelta
The event contains a tool-call delta.
Usage
The event contains usage information.
TokenCountResultUAIX.LmRuntime.Contracts
2 members
Represents tokenizer count output.
TokenCount
Gets the token count.
Tokenizer
Gets the tokenizer name.
ToolChoiceUAIX.LmRuntime.Contracts
2 members
Defines runtime guidance for model tool selection.
Automatic
Gets a value indicating whether tool selection is automatic.
RequiredToolName
Gets a required tool name when a specific tool must be used.
ToolDefinitionUAIX.LmRuntime.Contracts
3 members
Defines a callable tool exposed through the canonical inference contract.
Description
Gets the tool description.
JsonSchema
Gets the JSON schema document used to validate tool arguments.
Name
Gets the tool name.
UaiFileMemoryOptionsUAIX.LmRuntime.Contracts
6 members
Defines settings for the.uai file memory store.
IncludeShortTermMemoryFiles
Gets a value indicating whether short-term.sui memory files are included when reading memory.
MemoryFileName
Gets the memory file name.
RootDirectory
Gets the root directory for.uai memory files.
ShortTermMemoryDirectoryName
Gets the directory name under the.uai root that contains short-term.sui memory files.
ShortTermMemoryFilePattern
Gets the file search pattern used to discover short-term memory units.
SkipInvalidEntries
Gets a value indicating whether invalid entries should be skipped instead of throwing.
UaiMemoryEntryUAIX.LmRuntime.Contracts
6 members
Represents a persisted.uai memory entry.
Content
Gets the persisted memory content.
ContentSha256
Gets the SHA-256 hash of normalized content.
ConversationId
Gets the associated conversation identifier.
CreatedUtc
Gets the UTC creation timestamp.
EntryId
Gets the memory entry identifier.
Role
Gets the memory role.
RuntimeTelemetryNamesUAIX.LmRuntime.Diagnostics
16 members
Defines stable telemetry names emitted by the runtime core.
BlockedActionCounter
Teleodynamic blocked action counter metric name.
BudgetDecisionCounter
Teleodynamic budget decision counter metric name.
ClaimBoundaryViolationCounter
Claim-boundary violation counter metric name.
ConstraintDecisionCounter
Request-side constraint decision counter metric name.
ConstraintViolationCounter
Request-side constraint violation counter metric name.
EvidenceReceiptCounter
Evidence receipt counter metric name.
FailureCounter
Failure counter metric name.
MemoryFirewallDecisionCounter
Memory-firewall decision counter metric name.
MemoryQuarantineCounter
Memory-firewall quarantine counter metric name.
QuarantineRecordCounter
Quarantine record counter metric name.
RequestCounter
Request counter metric name.
RequestDurationMs
Request duration histogram metric name.
ReviewGateDecisionCounter
Slow-loop review-gate decision counter metric name.
SourceName
The ActivitySource and Meter name.
TeleodynamicControlDecisionCounter
Explicit teleodynamic control-cycle decision counter metric name.
TeleodynamicNoOpCounter
Explicit teleodynamic control-cycle no-op counter metric name.
BudgetDecisionStatusUAIX.LmRuntime.Governance
4 members
Identifies the outcome of a runtime budget evaluation.
Approved
The requested action is affordable under the configured viability budget.
Blocked
The requested action is blocked and should not execute automatically.
NoOpSelected
No-op was selected as the dominant safe action.
Unknown
The decision has not been evaluated.
ClaimBoundaryDecisionUAIX.LmRuntime.Governance
6 members
Represents the result of applying claim-boundary rules to text.
Allowed
Gets a value indicating whether the text stayed within claim boundaries.
CreatedUtc
Gets the UTC decision timestamp.
Message
Gets a bounded decision message.
NoOpReason
Gets the selected no-op reason when the decision blocks output.
SafeReplacementText
Gets a replacement text that can be emitted in automatic lanes.
ViolatedRuleIds
Gets violated rule identifiers.
ClaimBoundaryRuleUAIX.LmRuntime.Governance
5 members
Defines a bounded claim-boundary rule applied to generated text or runtime claims.
Message
Gets the human-readable rule message.
NoOpReason
Gets the no-op reason associated with this rule.
Pattern
Gets the case-insensitive text pattern that triggers the rule.
RuleId
Gets the stable rule identifier.
Severity
Gets the rule severity.
ClaimBoundarySeverityUAIX.LmRuntime.Governance
3 members
Identifies how strongly a claim-boundary rule should affect runtime behavior.
Advisory
The rule adds advisory context only.
Block
The rule should block or replace output in automatic lanes.
Warning
The rule should be logged and surfaced for review.
ClaimLifecycleStatusUAIX.LmRuntime.Governance
6 members
Identifies the bounded evidence lifecycle assigned to a runtime or release claim.
Bounded
The claim is constrained to an explicit domain, digest, environment, and evidence scope.
Promoted
The bounded claim has received explicit human approval for its intended publication lane.
Raw
The claim has been captured but has not received an evidence review.
Rejected
The claim has been rejected and cannot be promoted without a new evidence cycle.
Restricted
The claim remains usable only under narrower restrictions than originally requested.
Reviewed
The claim and cited evidence have received an initial review.
ClaimStatusUAIX.LmRuntime.Governance
6 members
Identifies the evidence status assigned to a claim, generated artifact, or externally visible output.
ApprovedPublicOutcome
The artifact is an approved public outcome with reviewed evidence.
GeneratedNeedsHumanReview
The artifact was generated or transformed by automation and requires human review before promotion.
MachineReadableEvidence
The artifact is a machine-readable evidence payload intended for agents and due-diligence workflows.
PublicReadyTemplate
The artifact is a buyer-safe methodology template and does not claim a specific client outcome.
Quarantined
The artifact is quarantined and cannot be promoted automatically.
Unknown
No claim status was supplied.
ClaimTransitionDecisionUAIX.LmRuntime.Governance
5 members
Reports whether a requested claim-lifecycle transition is permitted by evidence and authority rules.
Allowed
Gets a value indicating whether the requested transition is allowed.
ClaimId
Gets the stable claim identifier.
EffectiveStatus
Gets the status that remains effective after policy evaluation.
Message
Gets the bounded policy explanation.
NoOpReason
Gets the no-op reason when the requested transition is denied.
ClaimTransitionRequestUAIX.LmRuntime.Governance
11 members
Describes one requested transition in the explicit claim-evidence lifecycle.
AuditTracePresent
Gets a value indicating whether an auditable decision trace supports the claim.
BoundaryViolation
Gets a value indicating whether a claim-boundary violation was identified.
ClaimId
Gets the stable claim identifier.
CurrentStatus
Gets the currently recorded claim lifecycle status.
DomainBounded
Gets a value indicating whether the claim has an explicit domain and applicability boundary.
HumanApproved
Gets a value indicating whether an authorized human reviewer approved the requested promotion.
IndependentEvidenceCount
Gets the number of independent evidence references attached to the claim.
ProximityOnlyEvidence
Gets a value indicating whether the transition is justified only by association with a nearby approved claim.
Rationale
Gets the bounded reviewer rationale attached to restrictive or terminal transitions.
RequestedStatus
Gets the requested claim lifecycle status.
ResourceTracePresent
Gets a value indicating whether a candidate-bound resource trace supports the claim.
ConstraintClosureReportUAIX.LmRuntime.Governance
6 members
Reports whether active work and constraint nodes participate in closed maintenance cycles.
Closed
Gets a value indicating whether every active work and constraint node participates in a closed cycle.
Diagnostics
Gets validation diagnostics for duplicate identifiers, missing endpoints, or invalid values.
OpenNodeIds
Gets active node identifiers that do not participate in a closed directed cycle.
RegistryId
Gets the analyzed registry identifier.
RetirementCandidateIds
Gets active nodes marked for bounded retirement because maintenance burden exceeds evidence strength.
StronglyConnectedComponents
Gets strongly connected components in deterministic node-identifier order.
ConstraintDecisionUAIX.LmRuntime.Governance
8 members
Represents the result of request-side constraint evaluation.
Allowed
Gets a value indicating whether automatic execution may proceed.
CreatedUtc
Gets the UTC decision timestamp.
EvidenceReferences
Gets evidence references associated with this decision.
MatchedRuleIds
Gets matched rule identifiers.
Message
Gets a bounded explanation suitable for logs and receipts.
NoOpReason
Gets the selected no-op reason when automatic execution is blocked.
RequiresReview
Gets a value indicating whether the resulting artifact requires review before promotion.
SafeReplacementText
Gets the safe replacement text when no-op is selected.
ConstraintEdgeUAIX.LmRuntime.Governance
4 members
Defines one directed maintenance or channeling relationship in a work-constraint graph.
Relationship
Gets the bounded relationship label, such as maintains, channels, depends-on, or verifies.
SourceNodeId
Gets the source node identifier.
Strength
Gets the normalized relationship strength in the inclusive range from zero through one.
TargetNodeId
Gets the target node identifier.
ConstraintNodeUAIX.LmRuntime.Governance
6 members
Defines one bounded node in the work-constraint closure graph.
EvidenceStrength
Gets the non-negative evidence strength associated with the node.
Kind
Gets the semantic role of the node.
MaintenanceBurden
Gets the non-negative recurring maintenance burden attributed to the node.
Metadata
Gets bounded metadata attached to the node.
NodeId
Gets the stable node identifier.
Status
Gets the lifecycle state of the node.
ConstraintNodeKindUAIX.LmRuntime.Governance
4 members
Identifies the role a node plays in a work-constraint registry.
Constraint
The node represents a constraint that channels or limits work.
Evidence
The node represents evidence that supports a work or constraint decision.
Unknown
No node role was supplied.
Work
The node represents work that consumes resources and maintains constraints.
ConstraintNodeStatusUAIX.LmRuntime.Governance
3 members
Identifies the current lifecycle state of a work-constraint registry node.
Active
The node is active and participates in closure analysis.
Frozen
The node is frozen and remains auditable but cannot be expanded automatically.
Retired
The node is retired and excluded from active closure analysis.
ConstraintRegistrySnapshotUAIX.LmRuntime.Governance
4 members
Captures one immutable work-constraint graph supplied to closure analysis.
CapturedUtc
Gets the UTC time at which the registry snapshot was captured.
Edges
Gets the directed graph edges.
Nodes
Gets the graph nodes.
RegistryId
Gets the stable registry identifier.
ConstraintRuleUAIX.LmRuntime.Governance
5 members
Defines an evidence-bounded runtime constraint rule.
Message
Gets the bounded explanation for the decision receipt.
Pattern
Gets the case-insensitive substring pattern matched by the default policy.
RuleId
Gets the stable rule identifier.
Scope
Gets the inspected request surface.
Severity
Gets the severity applied when the pattern is matched.
ConstraintScopeUAIX.LmRuntime.Governance
5 members
Identifies the request surface inspected by a constraint rule.
All
Inspect all supported request surfaces.
Messages
Inspect message content.
Metadata
Inspect metadata keys and values.
ResponseFormat
Inspect response-format hints and schemas.
Tools
Inspect tool names, descriptions, and schemas.
ConstraintSeverityUAIX.LmRuntime.Governance
3 members
Identifies how strongly a runtime constraint rule affects automatic execution.
Block
The rule blocks automatic execution and selects no-op.
Information
The rule only records an informational observation.
ReviewRequired
The rule permits execution but marks the artifact as needing review before promotion.
EvidenceReferenceUAIX.LmRuntime.Governance
5 members
Identifies a source used to justify a governance decision.
Note
Gets a bounded note describing why the evidence is relevant.
Path
Gets the evidence path or stable identifier.
Sha256
Gets the SHA-256 hash when the evidence is file-backed.
Source
Gets the evidence source name.
Span
Gets the line, byte, or section span when one is present.
GovernanceActionKindUAIX.LmRuntime.Governance
8 members
Identifies a Teleodynamic structural operator selected by the runtime control plane.
Add
Add a bounded structure such as a tool, memory edge, adapter, or prompt template.
Freeze
Freeze a structure so it remains auditable while automatic expansion and promotion are disabled.
Merge
Merge overlapping structures after evidence shows lower maintenance burden.
NoOp
Select no mutation because evidence, budget, or claim boundaries do not justify action.
Reactivate
Reactivate a previously frozen structure after fresh evidence repays its maintenance burden.
Retire
Retire a structure whose maintenance burden is no longer justified.
Split
Split a structure into narrower lanes when evidence shows ambiguous or overloaded behavior.
Unknown
No operator was specified.
GovernanceDecisionReceiptUAIX.LmRuntime.Governance
14 members
Represents an immutable evidence-bearing receipt for a runtime governance decision.
Action
Gets the selected structural operator.
BudgetDecision
Gets the budget decision that contributed to the receipt.
ClaimBoundaryDecision
Gets the claim-boundary decision that contributed to the receipt.
ConstraintDecision
Gets the request-side constraint decision that contributed to the receipt.
ConversationId
Gets the conversation identifier when one is present.
CreatedUtc
Gets the UTC receipt timestamp.
EvidenceReferences
Gets evidence references associated with the receipt.
Metadata
Gets bounded metadata for downstream audit and telemetry correlation.
Model
Gets the associated model identifier.
NoOpReason
Gets the selected no-op reason when applicable.
PackageVersion
Gets the package version that emitted the receipt.
QuarantineRecord
Gets the quarantine record emitted for the receipt, when one was written.
ReceiptId
Gets the stable receipt identifier.
ReviewGateDecision
Gets the slow-loop review-gate decision that contributed to the receipt.
MemoryFirewallDecisionUAIX.LmRuntime.Governance
7 members
Represents the memory-firewall disposition for one packet.
CreatedUtc
Gets the UTC time at which the firewall decision was produced.
Message
Gets a bounded explanation of the disposition.
PacketId
Gets the packet identifier.
PromotionAllowed
Gets a value indicating whether the requested promotion was approved.
Reason
Gets the primary firewall reason.
Status
Gets the resulting firewall status.
TargetTier
Gets the target memory tier when promotion is allowed.
MemoryFirewallPolicyUAIX.LmRuntime.Governance
8 members
Defines source, freshness, contradiction, entropy, and review boundaries for memory promotion.
MaximumEntropyScore
Gets the maximum entropy accepted in the automatic memory lane.
MaximumLongTermAge
Gets the maximum age of long-term packets before re-review is required.
MaximumMediumTermAge
Gets the maximum age of medium-term packets before retirement.
MaximumShortTermAge
Gets the maximum age of short-term packets before retirement.
MinimumLongTermEvidenceReferences
Gets the minimum evidence count required for durable long-term promotion.
MinimumTrustScore
Gets the minimum source trust accepted in the automatic memory lane.
RequireHumanReviewForLongTerm
Gets a value indicating whether long-term promotion requires explicit human review.
ReviewContradictions
Gets a value indicating whether contradiction always routes a packet to review.
MemoryFirewallReasonUAIX.LmRuntime.Governance
11 members
Identifies the primary reason for a memory-firewall disposition.
Contradiction
The packet conflicts with one or more active memory references.
CorruptPacket
The packet was explicitly marked corrupt.
DigestMismatch
The observed content digest does not match the packet declaration.
EvidenceRequired
The requested promotion does not carry sufficient evidence references.
ExcessiveEntropy
The packet entropy or uncertainty exceeds the automatic-lane threshold.
InvalidDigest
A declared or observed content digest is not a valid SHA-256 value.
LowTrust
The packet source trust score is below the configured threshold.
MissingSource
The packet does not identify a bounded source.
None
No firewall restriction was required.
ReviewRequired
The requested tier or disposition requires explicit review.
StalePacket
The packet exceeded its freshness or expiry boundary.
MemoryFirewallRequestUAIX.LmRuntime.Governance
4 members
Describes the requested disposition and review proof evaluated by the memory firewall.
HumanReviewed
Gets a value indicating whether explicit human review was completed.
Packet
Gets the packet to evaluate.
RequestedStatus
Gets the requested packet status.
ReviewReference
Gets a bounded review reference when human review was completed.
MemoryPacketUAIX.LmRuntime.Governance
14 members
Describes a privacy-preserving memory packet using provenance and integrity metadata rather than raw content.
ContradictionReferences
Gets bounded references to memories or evidence involved in a contradiction.
CreatedUtc
Gets the UTC packet creation time.
DeclaredContentSha256
Gets the content SHA-256 declared by the producer.
EntropyScore
Gets the normalized entropy or unresolved uncertainty score.
EvidenceReferences
Gets evidence references supporting provenance, trust, or contradiction analysis.
ExpiresUtc
Gets the explicit UTC expiry time when one is present.
HasContradiction
Gets a value indicating whether the packet contradicts active memory.
IsCorrupt
Gets a value indicating whether an upstream integrity check marked the packet corrupt.
Metadata
Gets bounded metadata that excludes raw memory content.
ObservedContentSha256
Gets the independently observed content SHA-256 supplied to the firewall.
PacketId
Gets the stable packet identifier.
Source
Gets the bounded provenance source.
Tier
Gets the requested source-routed memory tier.
TrustScore
Gets the normalized trust score assigned to the packet source.
MemoryPacketStatusUAIX.LmRuntime.Governance
6 members
Identifies the current firewall disposition of a memory packet.
Candidate
The packet has not yet passed a firewall decision.
Promoted
The packet may be persisted in the approved target tier.
Quarantined
The packet is isolated from active retrieval while evidence or integrity is unresolved.
Rejected
The packet is invalid or unsafe to retain as active memory.
Retired
The packet is no longer eligible for active retrieval because it expired or was superseded.
ReviewRequired
The packet is structurally valid but requires an explicit human or policy review.
MemoryTierUAIX.LmRuntime.Governance
3 members
Identifies a source-routed memory tier managed by the memory firewall.
LongTerm
Durable memory that requires provenance, contradiction checks, and governed promotion.
MediumTerm
Reviewable working memory retained across a bounded sequence of interactions.
ShortTerm
Ephemeral context retained only for the immediate execution window.
NoOpReasonUAIX.LmRuntime.Governance
16 members
Identifies why the runtime selected no-op.
ClaimBoundaryViolation
The request or output crosses a claim boundary.
ConstraintViolation
The request crossed a configured runtime constraint.
ExcessiveUncertainty
The uncertainty score is too high for an automatic action.
HardLimitExceeded
The request exceeds a hard token, memory, or tool-count limit.
InsufficientBudget
The requested work would exceed the available runtime budget.
IrreversibleAction
The proposed action lacks a documented rollback path required for automatic execution.
MaintenanceCycleOpen
The work-constraint registry contains active structure outside a closed maintenance cycle.
NoFeasibleImprovement
No feasible proposal produced enough bounded benefit to dominate explicit no-op.
None
No no-op reason applies.
PhaseUnstable
The current structural phase is unstable and does not permit the requested automatic action.
PromotionEvidenceMissing
The requested claim promotion lacks candidate-bound evidence required by the lifecycle policy.
ProximityOnlyEvidence
The requested claim promotion relies only on proximity to another approved claim.
ReviewRequired
The requested work requires human review before promotion.
UnsupportedAction
The requested operator is not supported in the current lane.
ViabilityFloor
The requested work would cross the configured viability floor.
WeakEvidence
The evidence packet is too weak to justify a structural mutation.
QuarantineRecordUAIX.LmRuntime.Governance
11 members
Represents an append-only quarantine ledger record for generated or unreviewed runtime artifacts.
ArtifactKind
Gets the artifact kind associated with the quarantine record.
ClaimStatus
Gets the evidence status assigned to the quarantined item.
ConversationId
Gets the conversation identifier when one is present.
CreatedUtc
Gets the UTC record timestamp.
EvidenceReferences
Gets evidence references associated with the quarantined item.
Metadata
Gets bounded metadata for downstream review tools.
Model
Gets the associated model identifier.
PackageVersion
Gets the package version that emitted the quarantine record.
Reason
Gets the reason the item was quarantined or review-gated.
RecordId
Gets the stable quarantine record identifier.
ReviewGateStatus
Gets the review-gate status that caused the record.
ResourceEconomyInputUAIX.LmRuntime.Governance
13 members
Describes the measured benefits and burdens of one observation or proposed structural action.
ActionCost
Gets the estimated one-time cost of validating and applying the action.
CorrelationId
Gets the stable correlation identifier copied into the transition record.
CountAsObservation
Gets a value indicating whether the committed transition represents a fast-loop observation.
CountAsStructuralReservation
Gets a value indicating whether the committed transition represents a structural reservation.
EnergyCost
Gets the estimated energy or compute cost associated with the action.
EvidenceReferences
Gets evidence references supporting the gain and cost measurements.
MaintenanceCost
Gets the estimated recurring cost of retaining and reviewing the resulting structure.
MemoryCost
Gets the estimated memory cost associated with the action.
PredictiveGain
Gets observed predictive-loss reduction expressed as a non-negative raw gain.
ReviewCost
Gets the estimated human-review cost associated with the action.
TransitionUtc
Gets the zero-offset UTC timestamp assigned to the transition.
UncertaintyCost
Gets an explicit uncertainty cost already expressed in resource units.
UncertaintyScore
Gets normalized proposal uncertainty in the inclusive range from zero through one.
ResourceEconomyPolicyUAIX.LmRuntime.Governance
13 members
Defines the bounded resource-economy policy used when structural proposals and observations are evaluated.
The policy makes predictive gain, action cost, maintenance burden, compute, memory, review, uncertainty, viability, and capacity explicit so a structural action cannot be justified by accuracy alone.
ActionCostWeight
Gets the multiplier applied to one-time action cost.
Capacity
Gets the maximum resource balance retained after a transition.
DecayRate
Gets the fraction of the current resource balance removed as natural decay during each observation.
EnergyCostWeight
Gets the multiplier applied to estimated compute or energy cost.
InitialResource
Gets the resource balance assigned to a newly created stateful economy.
MaintenanceCostWeight
Gets the multiplier applied to recurring maintenance cost.
MemoryCostWeight
Gets the multiplier applied to memory cost.
MinimumNetGain
Gets the minimum positive net resource change required before growth is preferred over no-op.
PredictiveGainWeight
Gets the multiplier applied to measured predictive gain.
ReviewCostWeight
Gets the multiplier applied to human-review cost.
UncertaintyReserve
Gets the resource reserve charged for a fully uncertain proposal.
UncertaintyReserveRatio
Gets the fraction of the current balance protected in addition to the viability floor.
ViabilityFloor
Gets the minimum resource balance required for an automatically approved action.
ResourceEconomyStateUAIX.LmRuntime.Governance
16 members
Captures the endogenous resource state carried between fast observations and structural-control cycles.
This state is independent from model weights and inference tensors. It records bounded accounting context used by the slow-loop control plane and cannot alter deterministic parity mode without an explicit external actuator.
Capacity
Gets the configured capacity used to normalize resource-retention metrics.
CumulativeActionCost
Gets the cumulative weighted one-time action cost charged by the resource economy.
CumulativeDecay
Gets the cumulative endogenous decay charged by the resource economy.
CumulativeEnergyCost
Gets the cumulative weighted energy or compute cost charged by the resource economy.
CumulativeMaintenanceCost
Gets the cumulative weighted maintenance cost charged by the resource economy.
CumulativeMemoryCost
Gets the cumulative weighted memory cost charged by the resource economy.
CumulativePredictiveGain
Gets the cumulative weighted predictive gain accepted by the resource economy.
CumulativeReviewCost
Gets the cumulative weighted human-review cost charged by the resource economy.
CumulativeUncertaintyReserve
Gets the cumulative uncertainty reserve charged by the resource economy.
CurrentResource
Gets the current resource balance available to sustain runtime structure.
Cycle
Gets the monotonically increasing committed transition number.
ObservationCount
Gets the number of accepted fast-loop observations.
StructuralReservationCount
Gets the number of approved structural reservations.
UpdatedUtc
Gets the UTC time at which the state was produced.
ViabilityFloor
Gets the configured viability floor used by the current economy.
ViabilityMargin
Gets current resource minus the configured viability floor.
ResourceEconomyTransitionUAIX.LmRuntime.Governance
20 members
Reports the deterministic resource transition associated with an observation or structural proposal.
CorrelationId
Gets the observation or reservation identifier associated with the transition.
CreatedUtc
Gets the UTC time at which the transition was produced.
DecayCost
Gets the natural decay charged to the transition.
EffectiveViabilityFloor
Gets the effective viability boundary after the protected reserve is added to the policy floor.
EvidenceReferences
Gets the evidence references supporting the transition inputs.
Message
Gets the bounded diagnostic explaining the transition outcome.
NetChange
Gets the unconstrained net resource change before capacity clamping.
NextState
Gets the candidate state after gain, decay, costs, and capacity bounds are applied.
NoOpReason
Gets the no-op reason when the transition is not viable or does not repay its burden.
PreviousState
Gets the state supplied before the transition.
ProtectedReserve
Gets the resource reserve protected in proportion to the previous balance.
Viable
Gets a value indicating whether the transition remains above the configured viability floor.
WeightedActionCost
Gets the weighted one-time action cost.
WeightedCost
Gets the total weighted action, maintenance, compute, memory, review, and uncertainty cost.
WeightedEnergyCost
Gets the weighted compute or energy cost.
WeightedGain
Gets the weighted gain credited to the transition.
WeightedMaintenanceCost
Gets the weighted recurring maintenance cost.
WeightedMemoryCost
Gets the weighted memory cost.
WeightedReviewCost
Gets the weighted human-review cost.
WeightedUncertaintyCost
Gets the uncertainty cost and reserve charged to the transition.
ReviewGateDecisionUAIX.LmRuntime.Governance
9 members
Represents the result of slow-loop review gate evaluation.
Allowed
Gets a value indicating whether automatic execution may proceed.
CreatedUtc
Gets the UTC review-gate timestamp.
EvidenceReferences
Gets evidence references used by the review decision.
Message
Gets the bounded review-gate message.
Metadata
Gets bounded metadata for review and audit systems.
NoOpReason
Gets the no-op reason when automatic execution is blocked.
ReviewState
Gets the review-state label assigned to the work.
Status
Gets the review-gate status.
TriggeredRuleIds
Gets triggered review rule identifiers.
ReviewGateRequestUAIX.LmRuntime.Governance
9 members
Describes runtime work that may require slow-loop review before execution or promotion.
Action
Gets the requested structural operator.
ConversationId
Gets the conversation identifier when one is present.
CreatedUtc
Gets the UTC timestamp when the review request was created.
EstimatedReviewMinutes
Gets the estimated review effort in minutes.
EvidenceReferences
Gets evidence references used to justify the action.
Metadata
Gets bounded request metadata used by the policy.
Model
Gets the model identifier associated with the request.
ToolDefinitions
Gets the number of tool definitions exposed to the request.
UncertaintyScore
Gets the normalized uncertainty score declared by the caller or derived by the orchestrator.
ReviewGateStatusUAIX.LmRuntime.Governance
5 members
Identifies the slow-loop review gate disposition.
Approved
The work is already approved for the selected lane.
Blocked
The work is blocked by review policy and must select no-op.
NotRequired
No additional review is required for the selected lane.
ReviewRequired
Human review is required before promotion or mutation.
Unknown
No review decision has been made.
RuntimeBudgetUAIX.LmRuntime.Governance
12 members
Defines resource-economy limits used by the runtime budget governor.
AvailableResource
Gets the available resource budget for the evaluated action.
MaxInputTokens
Gets the maximum accepted input-token count before a hard limit block.
MaxMemoryEntries
Gets the maximum number of injected memory entries accepted before a hard limit block.
MaxOutputTokens
Gets the maximum accepted output-token count before a hard limit block.
MaxToolDefinitions
Gets the maximum number of tool definitions accepted before a hard limit block.
MaxUncertaintyScore
Gets the maximum uncertainty score accepted in an automatic lane.
MemoryEntryCost
Gets the resource cost assigned to each injected memory entry.
ReviewMinuteCost
Gets the resource cost assigned to each declared review minute.
TokenCostWeight
Gets the resource cost assigned to each token considered by the request.
ToolDefinitionCost
Gets the resource cost assigned to each exposed tool definition.
UncertaintyCost
Gets the resource cost assigned to uncertainty after normalization.
ViabilityFloor
Gets the minimum reserve below which automatic structural growth is blocked.
RuntimeBudgetDecisionUAIX.LmRuntime.Governance
10 members
Represents a resource-economy decision for a runtime action.
Action
Gets the selected or requested structural operator.
ActionCost
Gets the computed action cost.
Approved
Gets a value indicating whether the action may continue automatically.
AvailableResource
Gets the configured available resource value used during evaluation.
CreatedUtc
Gets the UTC decision timestamp.
EvidenceReferences
Gets evidence references used by the decision.
Message
Gets a bounded decision message.
NoOpReason
Gets the no-op reason when no-op was selected.
Status
Gets the decision status.
ViabilityFloor
Gets the configured viability floor used during evaluation.
RuntimeBudgetRequestUAIX.LmRuntime.Governance
12 members
Describes a request or structural action being evaluated by the runtime budget governor.
Action
Gets the requested structural operator.
ConversationId
Gets the conversation identifier associated with the action when one is present.
EstimatedReviewMinutes
Gets the expected human review burden in minutes.
EvidenceReferences
Gets evidence references associated with the decision.
InputTokens
Gets the input-token count after prompt and memory preparation.
MemoryEntries
Gets the number of memory entries injected into the prompt.
Model
Gets the model identifier associated with the action.
OutputTokens
Gets the requested output-token count.
RetrievalFanOut
Gets the retrieval fan-out declared by the caller or retriever.
Source
Gets a bounded source label for the evaluated action.
ToolDefinitions
Gets the number of tool definitions exposed to the model.
UncertaintyScore
Gets the normalized uncertainty score in the range zero to one.
StructuralChangeDecisionUAIX.LmRuntime.Governance
9 members
Represents the budget, review, and evidence disposition for a proposed structural change.
Action
Gets the selected structural operator.
BudgetDecision
Gets the resource-budget decision.
CreatedUtc
Gets the UTC decision timestamp.
GovernanceReceipt
Gets the governance receipt emitted for the decision.
Message
Gets the bounded decision message.
NoOpReason
Gets the no-op reason when no mutation may occur.
QuarantineRecord
Gets the quarantine record when the decision is held for review.
ReviewGateDecision
Gets the review-gate decision.
Status
Gets the decision status.
StructuralChangeDecisionStatusUAIX.LmRuntime.Governance
3 members
Identifies the disposition of a proposed structural change.
Approved
The proposal was approved for the caller's mutation lane.
NoOpSelected
No mutation is allowed.
RequiresReview
The proposal was converted to slow-loop review.
StructuralChangeRequestUAIX.LmRuntime.Governance
16 members
Describes a slow-loop structural change proposed for runtime configuration, memory, tools, prompts, or backend routing.
Action
Gets the requested structural operator.
ChangeId
Gets the caller-provided change identifier.
ClaimText
Gets claim text associated with the change, if any.
ConversationId
Gets the conversation affected by the change when one is present.
EstimatedOutputTokens
Gets the estimated generated-token cost of validation or rollback evidence.
EstimatedReviewMinutes
Gets the estimated human review minutes required before promotion.
EstimatedTokens
Gets the estimated token cost of validating or applying the change.
EvidenceReferences
Gets evidence references supporting the requested change.
MemoryEntries
Gets the number of memory entries affected by the change.
Metadata
Gets bounded metadata used for receipts and audit correlation.
Model
Gets the model affected by the change when one is present.
RetrievalFanOut
Gets the retrieval fan-out required to validate the change.
TargetKind
Gets the type of structure being changed.
TargetName
Gets the structure name or stable key.
ToolDefinitions
Gets the number of tool definitions affected by the change.
UncertaintyScore
Gets the normalized uncertainty score for the change.
StructuralObservationWindowUAIX.LmRuntime.Governance
10 members
Captures bounded fast-loop measurements used by the slow-loop structural phase detector.
ActionRate
Gets the normalized rate of structural actions in the current window.
CurrentComplexity
Gets the current normalized structural complexity.
CurrentLoss
Gets the current normalized predictive loss.
DriftScore
Gets the normalized cross-context or cross-window behavioral drift score.
MaintenancePressure
Gets the normalized recurring maintenance pressure.
PreviousComplexity
Gets the normalized structural complexity from the preceding window.
PreviousLoss
Gets the normalized predictive loss from the preceding window.
ReviewPressure
Gets the normalized human-review pressure associated with recent changes.
SampleCount
Gets the number of observations represented by the window.
ViabilityMargin
Gets the resource balance minus the configured viability floor.
StructuralPhaseUAIX.LmRuntime.Governance
7 members
Identifies the bounded structural regime inferred from a recent observation window.
Drifting
Behavior has changed materially across contexts or observation periods.
Growth
Loss is improving while bounded structural change remains economically viable.
OverStructured
Complexity and maintenance burden are growing without sufficient predictive improvement.
PhaseLocked
Loss, complexity, and action rate are stable within configured tolerances.
ResourceConstrained
The resource margin is too small to support additional automatic structure.
UnderStructured
Loss remains high while complexity is low and viable growth remains available.
Unknown
The observation window is too small or invalid for classification.
StructuralPhaseAssessmentUAIX.LmRuntime.Governance
5 members
Reports the structural phase and bounded reasoning derived from an observation window.
ComplexityDelta
Gets the current-minus-previous complexity delta.
Confidence
Gets the normalized confidence in the classification.
LossDelta
Gets the current-minus-previous loss delta.
Message
Gets the bounded explanation for the classification.
Phase
Gets the classified structural phase.
StructuralPhasePolicyUAIX.LmRuntime.Governance
11 members
Defines deterministic thresholds for structural phase classification.
ComplexityGrowthThreshold
Gets the complexity-growth threshold used to identify over-structured behavior.
DriftThreshold
Gets the drift score at or above which the phase is classified as drifting.
GrowthImprovementThreshold
Gets the minimum loss improvement required to classify a window as bounded growth.
HighLossThreshold
Gets the loss level above which a low-complexity system is considered under-structured.
LowComplexityThreshold
Gets the complexity level below which a high-loss system may be considered under-structured.
MaintenancePressureThreshold
Gets the maintenance-pressure threshold used to identify over-structured behavior.
MinimumSampleCount
Gets the minimum number of observations required for classification.
ReviewPressureThreshold
Gets the review-pressure threshold used to identify over-structured behavior.
StableActionRate
Gets the maximum action rate accepted as phase-locked stability.
StableComplexityDelta
Gets the maximum absolute complexity change accepted as phase-locked stability.
StableLossDelta
Gets the maximum absolute loss change accepted as phase-locked stability.
StructuralProposalUAIX.LmRuntime.Governance
15 members
Describes one reversible candidate considered by the slow structural-control loop.
Action
Gets the proposed structural operator.
ActionCost
Gets the one-time cost of validating and applying the proposal.
ComplexityDelta
Gets the signed structural-complexity change; negative values reduce complexity.
EnergyCost
Gets the estimated energy or compute cost of the proposal.
EvidenceReferences
Gets the evidence references that justify the proposal estimates.
ExpectedPredictiveGain
Gets the expected predictive-loss reduction produced by the proposal.
MaintenanceCost
Gets the recurring cost of retaining the resulting structure.
MemoryCost
Gets the estimated memory cost of the proposal.
Metadata
Gets bounded metadata attached to the proposal.
ProposalId
Gets the stable proposal identifier used by evidence and trace records.
RequiresHumanReview
Gets a value indicating whether the proposal requires explicit human approval before application.
Reversible
Gets a value indicating whether the proposal can be rolled back through a documented inverse action.
ReviewCost
Gets the expected human-review cost of the proposal.
Target
Gets the stable name of the affected structure.
UncertaintyScore
Gets the normalized proposal uncertainty in the inclusive range from zero through one.
StructuralProposalEvaluationUAIX.LmRuntime.Governance
6 members
Reports the viability and local objective for one structural proposal.
Feasible
Gets a value indicating whether all automatic-action gates were satisfied.
Message
Gets the bounded diagnostic explaining the evaluation.
NoOpReason
Gets the reason the proposal was excluded when it was not feasible.
Objective
Gets the local objective, where lower values represent a more favorable bounded action.
Proposal
Gets the evaluated proposal.
ResourceTransition
Gets the resource transition predicted for the proposal.
TeleodynamicControlRequestUAIX.LmRuntime.Governance
11 members
Aggregates one opt-in slow-loop evaluation without exposing or mutating model inference state.
ClaimTransition
Gets a claim-lifecycle transition evaluated during the cycle when one is present.
ConstraintRegistry
Gets the work-constraint registry snapshot.
CreatedUtc
Gets the zero-offset UTC time assigned to the control cycle and trace entry.
CycleId
Gets the stable control-cycle identifier.
DecisionPolicy
Gets the proposal-scoring and safety policy.
Metadata
Gets bounded metadata copied into the decision trace.
ObservationWindow
Gets the recent fast-loop observation window.
PhasePolicy
Gets the structural-phase threshold policy.
Proposals
Gets the structural proposals considered during the cycle.
ResourcePolicy
Gets the resource-economy policy.
ResourceState
Gets the resource state observed before the cycle.
TeleodynamicControlResultUAIX.LmRuntime.Governance
5 members
Reports the complete bounded output of one opt-in teleodynamic control cycle.
ClaimDecision
Gets the claim-lifecycle decision when one is present.
ConstraintClosure
Gets the work-constraint closure analysis.
Decision
Gets the selected structural proposal or explicit no-op decision.
PhaseAssessment
Gets the structural phase inferred from the observation window.
TraceEntry
Gets the tamper-evident trace entry appended for the cycle.
TeleodynamicDecisionUAIX.LmRuntime.Governance
8 members
Reports the deterministic selection made across structural proposals and the explicit no-op candidate.
CreatedUtc
Gets the UTC time at which the decision was produced.
DecisionId
Gets the stable decision identifier.
Evaluations
Gets all bounded proposal evaluations in stable proposal-identifier order.
Message
Gets the bounded explanation for the selected action.
NoOpReason
Gets the reason no-op was selected, or UAIX.LmRuntime.Governance.NoOpReason.None for an actionable proposal.
SelectedAction
Gets the selected action, including explicit no-op.
SelectedProposalId
Gets the selected proposal identifier, or when no-op wins.
SelectedTransition
Gets the resource transition associated with the selected proposal or the unchanged no-op state.
TeleodynamicDecisionPolicyUAIX.LmRuntime.Governance
8 members
Defines deterministic scoring and safety rules for choosing among structural proposals and no-op.
ComplexityWeight
Gets the weight applied to absolute structural-complexity growth in the local objective.
DeferHumanReviewProposals
Gets a value indicating whether proposals requiring human review are excluded from automatic selection.
MaximumProposals
Gets the maximum number of proposals accepted in one bounded evaluation.
MaximumUncertainty
Gets the maximum uncertainty accepted for an automatically actionable proposal.
MinimumEvidenceCount
Gets the minimum evidence-reference count required for an automatically actionable proposal.
NoOpAdvantage
Gets the non-negative advantage a proposal must have over no-op before it can be selected.
RequireReversibleAutomaticAction
Gets a value indicating whether automatic proposal selection requires a documented rollback path.
UncertaintyWeight
Gets the weight applied to uncertainty in the local objective.
TeleodynamicTraceEntryUAIX.LmRuntime.Governance
4 members
Represents one immutable entry in the SHA-256-linked teleodynamic decision trace.
ContentHash
Gets the lowercase SHA-256 digest covering the sequence, previous hash, and canonical request fields.
PreviousHash
Gets the preceding entry hash, or 64 zero characters for the first entry.
Request
Gets the canonical event request stored by the entry.
Sequence
Gets the one-based sequence number assigned by the trace chain.
TeleodynamicTraceRequestUAIX.LmRuntime.Governance
8 members
Describes the canonical fields appended to the tamper-evident teleodynamic decision trace.
Action
Gets the selected governance action.
CreatedUtc
Gets the UTC event time supplied by the orchestrator.
EventId
Gets the stable decision or event identifier.
EventKind
Gets the bounded event kind.
Metadata
Gets bounded metadata serialized in ordinal key order.
NoOpReason
Gets the no-op reason when no structural action was selected.
ResourceAfter
Gets the resource balance selected after the event.
ResourceBefore
Gets the resource balance observed before the event.
TraceChainVerificationResultUAIX.LmRuntime.Governance
3 members
Reports whether a teleodynamic trace chain is structurally and cryptographically intact.
FirstInvalidSequence
Gets the one-based sequence number of the first invalid entry, or zero when the chain is valid.
Message
Gets the bounded verification diagnostic.
Valid
Gets a value indicating whether every sequence, predecessor hash, and content hash is valid.
UAIX.LmRuntime.Tensors
13 documented types and 76 documented members.
GgmlTensorTypeUAIX.LmRuntime.Tensors
21 members
Identifies GGML tensor storage types as encoded in GGUF tensor descriptors.
BF16
16-bit brain floating point.
F16
16-bit floating point.
F32
32-bit floating point.
F64
64-bit floating point.
I16
16-bit signed integer.
I32
32-bit signed integer.
I64
64-bit signed integer.
I8
8-bit signed integer.
IQ4_NL
IQ4_NL block quantization.
Q2_K
Q2_K block quantization.
Q3_K
Q3_K block quantization.
Q4_0
Q4_0 block quantization.
Q4_1
Q4_1 block quantization.
Q4_K
Q4_K block quantization.
Q5_0
Q5_0 block quantization.
Q5_1
Q5_1 block quantization.
Q5_K
Q5_K block quantization.
Q6_K
Q6_K block quantization.
Q8_0
Q8_0 block quantization.
Q8_1
Q8_1 block quantization.
Q8_K
Q8_K block quantization.
ITensorUAIX.LmRuntime.Tensors
2 members
Defines tensor metadata common to all backend placements.
DataType
Gets the tensor element representation.
Shape
Gets the tensor shape.
Q4_1BlockUAIX.LmRuntime.Tensors
3 members
Represents a Q4_1 quantized block descriptor.
Minimum
Gets the block minimum.
PackedValues
Gets the packed values.
Scale
Gets the block scale.
Q5_0BlockUAIX.LmRuntime.Tensors
3 members
Represents a Q5_0 quantized block descriptor.
HighBits
Gets the high-bit metadata.
PackedValues
Gets the packed low-bit values.
Scale
Gets the block scale.
Q5_1BlockUAIX.LmRuntime.Tensors
4 members
Represents a Q5_1 quantized block descriptor.
HighBits
Gets the high-bit metadata.
Minimum
Gets the block minimum.
PackedValues
Gets the packed low-bit values.
Scale
Gets the block scale.
Q8_1BlockUAIX.LmRuntime.Tensors
3 members
Represents a Q8_1 quantized block descriptor.
Scale
Gets the block scale.
Sum
Gets the block sum metadata.
Values
Gets the quantized values.
QuantizedBlockTraitUAIX.LmRuntime.Tensors
3 members
Describes a quantized block layout.
BlockByteCount
Gets the physical bytes in one block.
BlockElementCount
Gets the logical elements in one block.
GgmlType
Gets the GGML tensor type.
QuantizedBlockTraitsUAIX.LmRuntime.Tensors
1 member
Provides quantized block trait lookup.
Get(UAIX.LmRuntime.Tensors.GgmlTensorType)
Gets quantized block layout information for a GGML tensor type.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
Returns: The QuantizedBlockTrait result produced by QuantizedBlockTraits.Get for this contract: Gets quantized block layout information for a GGML tensor type. It is published only after all documented validation and ownership transitions succeed.
TensorDataTypeUAIX.LmRuntime.Tensors
24 members
Identifies supported tensor element representations.
BFloat16
16-bit brain floating point.
Float16
16-bit IEEE floating point.
Float32
32-bit IEEE floating point.
Int16
16-bit signed integer.
Int32
32-bit signed integer.
Int64
64-bit signed integer.
Int8
8-bit signed integer.
IQ4_NL
GGML IQ4_NL block quantization.
MXFP4
MXFP4 packed floating-point storage.
NVFP4
NVFP4 packed floating-point storage.
Q2_K
GGML Q2_K block quantization.
Q3_K
GGML Q3_K block quantization.
Q4_0
GGML Q4_0 block quantization.
Q4_1
GGML Q4_1 block quantization.
Q4_K
GGML Q4_K block quantization.
Q5_0
GGML Q5_0 block quantization.
Q5_1
GGML Q5_1 block quantization.
Q5_K
GGML Q5_K block quantization.
Q6_K
GGML Q6_K block quantization.
Q8_0
GGML Q8_0 block quantization.
Q8_1
GGML Q8_1 block quantization.
TQ1_0
TQ1_0 ternary quantized storage.
TQ2_0
TQ2_0 ternary quantized storage.
Unknown
Unknown or unsupported tensor representation.
TensorShapeUAIX.LmRuntime.Tensors
3 members
Represents immutable tensor shape metadata.
Dimensions
Gets the tensor dimensions.
ElementCount
Gets the number of tensor elements.
From(long[])
Creates the result from the tensor shape after validating the supplied representation.
Returns: The TensorShape result produced by TensorShape.From for this contract: Creates the result from the tensor shape after validating the supplied representation. It is published only after all documented validation and ownership transitions succeed.
TensorTypeTraitsUAIX.LmRuntime.Tensors
5 members
Describes storage traits for a GGML tensor type.
BlockByteCount
Gets the number of physical bytes in one storage block.
BlockElementCount
Gets the number of logical elements in one physical storage block.
DataType
Gets the runtime tensor data type.
GgmlType
Gets the GGML tensor type.
IsQuantized
Gets a value indicating whether the type is block-quantized.
TensorTypeTraitsCatalogUAIX.LmRuntime.Tensors
2 members
Provides GGML tensor type trait lookup and byte-length validation.
ComputeByteLength(UAIX.LmRuntime.Tensors.GgmlTensorType,ulong)
Computes the physical byte length required for a tensor type and element count.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
elementCount- The element count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The ulong value computed by TensorTypeTraitsCatalog.ComputeByteLength for this contract: Computes the physical byte length required for a tensor type and element count. Range, finite-value, and overflow checks are completed before the value is returned.
Get(UAIX.LmRuntime.Tensors.GgmlTensorType)
Gets traits for the specified tensor type.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
Returns: The TensorTypeTraits result produced by TensorTypeTraitsCatalog.Get for this contract: Gets traits for the specified tensor type. It is published only after all documented validation and ownership transitions succeed.
VectorMathUAIX.LmRuntime.Tensors
2 members
Provides allocation-free vector math kernels used by tests and CPU fallback paths.
Dot(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>)
Computes the dot product of two equal-length vectors.
left- The left sequence used by this operation; its required length, ordering, and element bounds are validated before access.
right- The right sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by VectorMath.Dot for this contract: Computes the dot product of two equal-length vectors. Range, finite-value, and overflow checks are completed before the value is returned.
RmsNorm(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>,float)
Applies RMS normalization using an explicit weight vector.
input- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
weight- The weight sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
epsilon- The positive normalization epsilon added to the mean-square term to avoid division by zero while preserving deterministic numerical behavior.
UAIX.LmRuntime.Acceleration
13 documented types and 80 documented members.
DiagnosticRuntimeBackendUAIX.LmRuntime.Acceleration
5 members
Provides shared diagnostic behavior for backend packages that declare compatibility before native execution is proven.
The class keeps package-visible backend declarations DRY. It performs no native loads, subprocess execution, provider calls, network access, or model downloads; concrete packages remain responsible only for naming their compatibility API, runtime identifiers, and package-specific diagnostics.
Capabilities
Gets the backend capability declaration.
DisplayName
Gets the human-readable backend name.
Id
Gets the stable backend identifier.
Kind
Gets the backend kind.
ProbeAsync(UAIX.LmRuntime.Acceleration.RuntimeBackendOptions,System.Threading.CancellationToken)
Probes a diagnostic backend and reports unavailable until native assets, runtime libraries, and devices are proven.
options- The caller-supplied probing and selection options.
cancellationToken- A token that cancels the probe.
Returns: The backend probe result.
IRuntimeBackendUAIX.LmRuntime.Acceleration
5 members
Defines a local runtime backend that can report capabilities and probe available devices.
Backends are package-visible compatibility surfaces. Implementations must not perform hidden downloads, provider calls, subprocess execution, or remote inference during probing or selection.
Capabilities
Gets the backend capability declaration.
DisplayName
Gets the human-readable backend name.
Id
Gets the stable backend identifier.
Kind
Gets the backend kind.
ProbeAsync(UAIX.LmRuntime.Acceleration.RuntimeBackendOptions,System.Threading.CancellationToken)
Probes the backend for local execution availability and devices.
options- The caller-supplied probing and selection options.
cancellationToken- A token that cancels the probe.
Returns: The backend probe result.
IRuntimeBackendRegistryUAIX.LmRuntime.Acceleration
3 members
Defines a registry of runtime backends available to a host application.
The registry is explicit and local. Merely installing a package does not grant hidden runtime authority unless the host registers the backend instance.
FindById(string)
Finds a backend by identifier.
backendId- The backend identifier to find.
Returns: The matching backend, or when no backend is registered for the ID.
GetBackends
Gets the currently registered runtime backends.
Returns: The ordered registered backend list.
Register(UAIX.LmRuntime.Acceleration.IRuntimeBackend)
Registers one runtime backend.
backend- The backend to register.
IRuntimeBackendSelectorUAIX.LmRuntime.Acceleration
1 member
Defines backend selection over a registered set of local runtime backends.
Selection returns explicit backend and device identity so CPU fallback is never confused with GPU execution.
SelectAsync(UAIX.LmRuntime.Acceleration.RuntimeBackendOptions,System.Threading.CancellationToken)
Selects a backend according to the supplied policy and options.
options- The backend selection options.
cancellationToken- A token that cancels selection.
Returns: The backend selection result.
RuntimeBackendCapabilitiesUAIX.LmRuntime.Acceleration
13 members
Reports the execution and compatibility capabilities declared by a runtime backend.
Capabilities are package-visible declarations. Availability is still decided by probing native assets, runtime libraries, drivers, and devices before a backend is selected for execution.
BackendApiName
Gets or sets the backend API name, such as CUDA, DirectML, Vulkan, ROCm, Metal, or managed CPU.
Diagnostics
Gets diagnostic messages describing compatibility, package state, and probe status.
NativeAssetState
Gets or sets the backend-reported native asset state.
RuntimeIdentifiers
Gets runtime identifiers supported or observed by the backend package.
SupportsCpuOnlyExecution
Gets or sets a value indicating whether the backend can execute without GPU hardware.
SupportsDeviceSelection
Gets or sets a value indicating whether callers can select among backend devices.
SupportsGguf
Gets or sets a value indicating whether the backend supports GGUF model inputs.
SupportsGpuExecution
Gets or sets a value indicating whether the backend can execute on a GPU device.
SupportsLlama
Gets or sets a value indicating whether the backend supports LLaMA-family model binding.
SupportsMemoryQuery
Gets or sets a value indicating whether the backend can query device or execution memory.
SupportsStreaming
Gets or sets a value indicating whether the backend supports streaming token output.
UsesGpuAcceleration
Gets or sets a value indicating whether the backend uses GPU acceleration when it executes successfully.
UsesNativeInference
Gets or sets a value indicating whether the backend uses native inference components.
RuntimeBackendKindUAIX.LmRuntime.Acceleration
8 members
Identifies the accelerator API family or execution class represented by a runtime backend.
The value describes the backend package family, not a claim that a device, driver, or native library is available on the current machine.
CpuManaged
The managed CPU reference backend.
Cuda
An NVIDIA CUDA backend.
DirectML
A Windows DirectML backend.
ExternalNative
A backend supplied by an external native package or host adapter.
Metal
An Apple Metal backend.
Rocm
An AMD ROCm backend.
Unknown
An unknown or not-yet-classified backend.
Vulkan
A Vulkan compute backend.
RuntimeBackendOptionsUAIX.LmRuntime.Acceleration
7 members
Provides caller-supplied backend selection and probing options.
These options never authorize hidden downloads, provider fallback, subprocess execution, or remote inference. They only describe how registered local backends should be probed and selected.
AllowCpuFallback
Gets or sets a value indicating whether prefer policies may fall back to CPU execution.
NativeAssetDirectory
Gets or sets an optional directory that a backend may inspect for native assets.
Policy
Gets or sets the backend selection policy.
PreferredBackendId
Gets or sets the preferred or required backend identifier for backend-id policies.
PreferredDeviceId
Gets or sets the preferred backend-local device identifier when device selection is supported.
RequestedRuntimeIdentifier
Gets or sets the runtime identifier requested by the caller, when different from the current process.
RequireNativeAssets
Gets or sets a value indicating whether a selected backend must have native assets available.
RuntimeBackendProbeResultUAIX.LmRuntime.Acceleration
6 members
Reports the result of probing one registered runtime backend.
A backend may declare capabilities but still probe unavailable when native assets, runtime libraries, drivers, or devices are absent.
BackendId
Gets or sets the backend identifier.
BackendKind
Gets or sets the backend kind.
Capabilities
Gets or sets the backend capabilities.
Devices
Gets discovered devices reported by the backend.
Diagnostics
Gets probe diagnostics.
IsAvailable
Gets or sets a value indicating whether the backend is available for execution.
RuntimeBackendRegistryUAIX.LmRuntime.Acceleration
3 members
Provides an in-memory explicit registry of local runtime backends.
Registration order is preserved so callers can make deterministic selection decisions.
FindById(string)
Finds a backend by identifier.
backendId- The backend identifier to find.
Returns: The matching backend, or when no backend is registered for the ID.
GetBackends
Gets the currently registered runtime backends.
Returns: The ordered registered backend list.
Register(UAIX.LmRuntime.Acceleration.IRuntimeBackend)
Registers one runtime backend.
backend- The backend to register.
RuntimeBackendSelectorUAIX.LmRuntime.Acceleration
2 members
Selects a local runtime backend from an explicit registry.
The selector probes registered backends and reports fallback identity instead of silently substituting one execution class for another.
RuntimeBackendSelector(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Initializes a new instance of the UAIX.LmRuntime.Acceleration.RuntimeBackendSelector class.
registry- The backend registry to select from.
SelectAsync(UAIX.LmRuntime.Acceleration.RuntimeBackendOptions,System.Threading.CancellationToken)
Selects a backend according to the supplied policy and options.
options- The backend selection options.
cancellationToken- A token that cancels selection.
Returns: The backend selection result.
RuntimeDeviceDescriptorUAIX.LmRuntime.Acceleration
12 members
Describes one CPU or accelerator device discovered by a runtime backend probe.
Device descriptors are diagnostic evidence. They should identify the device that would execute work without exposing private file paths, credentials, prompt text, or generated text.
BackendApiName
Gets or sets the acceleration API name associated with the device.
BackendId
Gets or sets the identifier of the backend that reported the device.
BackendKind
Gets or sets the backend kind that reported the device.
ComputeCapability
Gets or sets the backend-reported compute capability, feature level, or API version.
DeviceId
Gets or sets the stable backend-local device identifier.
Diagnostics
Gets diagnostic messages associated with the device descriptor.
DisplayName
Gets or sets the human-readable device name.
IsCpu
Gets or sets a value indicating whether the descriptor represents CPU execution.
IsGpu
Gets or sets a value indicating whether the descriptor represents GPU execution.
MemoryBytes
Gets or sets the backend-reported device memory in bytes when known.
RuntimeIdentifier
Gets or sets the runtime identifier associated with the probed process.
Vendor
Gets or sets the vendor or implementation owner reported by the backend.
RuntimeSelectionPolicyUAIX.LmRuntime.Acceleration
6 members
Defines how the selector chooses among registered runtime backends.
Policies that prefer a backend may fall back when the options allow it. Policies that require a backend fail clearly when the required capability is unavailable.
PreferBackendId
Prefer a specific backend identifier and fall back only when fallback is permitted.
PreferCpu
Prefer a CPU-capable backend but allow a different available backend when CPU is absent.
PreferGpu
Prefer an available GPU backend and fall back to CPU only when fallback is permitted.
RequireBackendId
Select only a specific backend identifier.
RequireCpu
Select only a CPU-capable backend.
RequireGpu
Select only an available GPU backend.
RuntimeSelectionResultUAIX.LmRuntime.Acceleration
9 members
Reports the selected backend, selected device, and any fallback used by a selection policy.
Selection results make CPU fallback explicit so callers never mistake fallback execution for GPU execution.
Diagnostics
Gets selection diagnostics.
FailureReason
Gets or sets the selection failure reason when selection did not succeed.
Policy
Gets or sets the policy used for selection.
SelectedBackendId
Gets or sets the selected backend identifier.
SelectedBackendKind
Gets or sets the selected backend kind.
SelectedCapabilities
Gets or sets the selected backend capabilities.
SelectedDevice
Gets or sets the selected device descriptor.
Succeeded
Gets or sets a value indicating whether selection succeeded.
UsedCpuFallback
Gets or sets a value indicating whether a GPU-prefer policy fell back to CPU execution.
UAIX.LmRuntime.Gguf
32 documented types and 148 documented members.
Float32TensorReaderUAIX.LmRuntime.Gguf
5 members
Reads float32 values from a validated tensor view with explicit GGUF byte-order handling.
ElementCount
Gets the logical float32 element count.
CopyTo(System.Span<float>)
Decodes every element into a caller-provided destination.
destination- The destination with room for every tensor element.
Float32TensorReader(UAIX.LmRuntime.Gguf.IReadOnlyTensorView)
Initializes a new Float32TensorReader instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
ReadElement(int)
Reads one element by flat storage index.
index- The zero-based index; it must identify an existing position within the relevant validated range.
Returns: The float value computed by Float32TensorReader.ReadElement for this contract: Reads one element by flat storage index. Range, finite-value, and overflow checks are completed before the value is returned.
ToArray
Creates an explicit bounded managed copy for scalar reference execution.
Returns: A newly allocated float[] containing the ordered result of Float32TensorReader.ToArray: Creates an explicit bounded managed copy for scalar reference execution. The caller owns the returned array and later mutation cannot alter the source object.
GgufArtifactClassifierUAIX.LmRuntime.Gguf
1 member
Classifies GGUF artifacts from metadata.
Classify(UAIX.LmRuntime.Gguf.GgufModel)
Classifies the GGUF artifact kind from validated metadata and tensor-layout evidence.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The GgufArtifactKind result produced by GgufArtifactClassifier.Classify for this contract: Classifies the GGUF artifact kind from validated metadata and tensor-layout evidence. It is published only after all documented validation and ownership transitions succeed.
GgufArtifactKindUAIX.LmRuntime.Gguf
5 members
Identifies the likely role of a GGUF artifact.
BaseModel
Base model artifact.
LoraSidecar
LoRA sidecar artifact.
MultimodalProjectorSidecar
Multimodal projector sidecar artifact.
MultiTokenPredictionSidecar
Multi-token-prediction sidecar artifact.
VocabOnly
Vocabulary-only artifact.
GgufByteOrderUAIX.LmRuntime.Gguf
3 members
Identifies the byte order used by a GGUF artifact.
Auto
Detect the byte order from the version field and reject ambiguous headers.
BigEndian
Interpret multi-byte values as big-endian.
LittleEndian
Interpret multi-byte values as little-endian.
GgufDiagnosticUAIX.LmRuntime.Gguf
3 members
Represents a structured GGUF parser diagnostic.
ByteOffset
Gets the byte offset associated with the diagnostic when known.
Code
Gets the stable diagnostic code.
Message
Gets the diagnostic message.
GgufDumpWriterUAIX.LmRuntime.Gguf
1 member
Writes human-readable GGUF metadata and tensor catalog dumps.
Write(UAIX.LmRuntime.Gguf.GgufModel)
Creates a textual dump for a parsed GGUF model.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The text produced by GgufDumpWriter.Write for this contract: Creates a textual dump for a parsed GGUF model. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufFormatExceptionUAIX.LmRuntime.Gguf
3 members
Represents a GGUF format violation with a stable code and byte offset.
ByteOffset
Gets the byte offset associated with the violation when known.
Code
Gets the stable diagnostic code.
GgufFormatException(string,string,System.Nullable<ulong>,System.Exception)
Initializes a new GgufFormatException instance with validated dependencies and operational bounds.
code- The stable machine-readable diagnostic code used to classify the failure without relying on localized message text.
message- The display-safe diagnostic message describing the failure without embedding prompt text, generated text, credentials, or private file contents.
byteOffset- The zero-based byte offset into the relevant source or destination; range validation occurs before access.
innerException- The underlying exception preserved for diagnostic chaining, or null when no lower-level failure is available.
GgufHashingReaderUAIX.LmRuntime.Gguf
1 member
Computes integrity hashes for GGUF and related model artifacts.
ComputeSha256(string)
Computes the SHA-256 hash of a file.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
Returns: The text produced by GgufHashingReader.ComputeSha256 for this contract: Computes the SHA-256 hash of a file. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufJsonDumpWriterUAIX.LmRuntime.Gguf
1 member
Writes machine-readable GGUF dump JSON.
Write(UAIX.LmRuntime.Gguf.GgufModel)
Writes a stable JSON dump for a parsed GGUF model.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The text produced by GgufJsonDumpWriter.Write for this contract: Writes a stable JSON dump for a parsed GGUF model. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufMetadataAccessorsUAIX.LmRuntime.Gguf
6 members
Provides strongly typed accessors for GGUF metadata values.
TryGetBoolean(UAIX.LmRuntime.Gguf.GgufModel,string,bool@)
Tries to read a Boolean metadata value.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
value- When the method returns, contains the value produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains a Boolean value.
TryGetByteArray(UAIX.LmRuntime.Gguf.GgufModel,string,System.Collections.Generic.IReadOnlyList<byte>@)
Tries to read a metadata byte array.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
values- When the method returns, contains the values produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains integral elements convertible to bytes.
TryGetInt32(UAIX.LmRuntime.Gguf.GgufModel,string,int@)
Tries to read a metadata 32-bit signed integer scalar.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
value- When the method returns, contains the value produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains an integral value convertible to int.
TryGetInt32Array(UAIX.LmRuntime.Gguf.GgufModel,string,System.Collections.Generic.IReadOnlyList<int>@)
Tries to read a metadata 32-bit signed integer array.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
values- When the method returns, contains the values produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains integral elements convertible to int.
TryGetSingleArray(UAIX.LmRuntime.Gguf.GgufModel,string,System.Collections.Generic.IReadOnlyList<float>@)
Tries to read a metadata single-precision floating point array.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
values- When the method returns, contains the values produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains numeric elements convertible to float.
TryGetStringArray(UAIX.LmRuntime.Gguf.GgufModel,string,System.Collections.Generic.IReadOnlyList<string>@)
Tries to read a metadata string array.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
values- When the method returns, contains the values produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains only string elements.
GgufMetadataArrayUAIX.LmRuntime.Gguf
3 members
Represents a typed GGUF metadata array.
ElementType
Gets the element type.
Items
Gets the array values.
ToString
Returns a concise description of this array.
Returns: The text produced by GgufMetadataArray.ToString for this contract: Returns a concise description of this array. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufMetadataKeyPolicyUAIX.LmRuntime.Gguf
1 member
Validates GGUF metadata key syntax.
Validate(string)
Validates the supplied key against the invariants required by GgufMetadataKeyPolicy.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
Returns: The GgufValidationError? result produced by GgufMetadataKeyPolicy.Validate for this contract: Validates the supplied key against the invariants required by GgufMetadataKeyPolicy. It is published only after all documented validation and ownership transitions succeed.
GgufMetadataValueUAIX.LmRuntime.Gguf
4 members
Represents a typed GGUF metadata value.
Type
Gets the metadata value type.
Value
Gets the scalar value or a UAIX.LmRuntime.Gguf.GgufMetadataArray for arrays.
Create(UAIX.LmRuntime.Gguf.GgufMetadataValueType,object)
Creates the GGUF metadata value from the validated inputs required by GgufMetadataValue.
type- The type containing validated format or tokenizer metadata required by this operation.
value- The value input of type object? read by GgufMetadataValue.Create; it must satisfy the member-specific nullability, identity, range, and ownership rules before dependent work begins.
Returns: The metadata value, with ownership and disposal obligations defined by the returned type and the Create contract.
ToString
Returns the string representation of this value.
Returns: The text produced by GgufMetadataValue.ToString for this contract: Returns the string representation of this value. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufMetadataValueTypeUAIX.LmRuntime.Gguf
13 members
Identifies GGUF metadata value types.
Array
Typed array.
Bool
Boolean.
Float32
32-bit floating point.
Float64
64-bit floating point.
Int16
16-bit signed integer.
Int32
32-bit signed integer.
Int64
64-bit signed integer.
Int8
8-bit signed integer.
String
UTF-8 string.
UInt16
16-bit unsigned integer.
UInt32
32-bit unsigned integer.
UInt64
64-bit unsigned integer.
UInt8
8-bit unsigned integer.
GgufModelUAIX.LmRuntime.Gguf
15 members
Represents a parsed GGUF artifact catalog and metadata dictionary.
Alignment
Gets the tensor-data alignment.
ArtifactKind
Gets the inferred artifact kind.
ByteOrder
Gets the resolved byte order used by the artifact.
FileLength
Gets the source file length captured during parsing.
Metadata
Gets the parsed metadata dictionary.
Path
Gets the source file path.
Shard
Gets shard metadata.
TensorDataOffset
Gets the absolute tensor-data section start offset.
Tensors
Gets the tensor catalog.
Version
Gets the GGUF format version.
Load(string,UAIX.LmRuntime.Gguf.GgufParseOptions)
Loads the GGUF model from a verified local source into GgufModel.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional GgufParseOptions controlling Load; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The parsed model catalog, with ownership and disposal obligations defined by the returned type and the Load contract.
TryGetSingle(string,float@)
Tries to get a single-precision metadata value.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
value- When the method returns, contains the value produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and can be converted to a single-precision value.
TryGetString(string,string@)
Tries to get a string metadata value.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
value- When the method returns, contains the value produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and contains a string.
TryGetTensor(string,UAIX.LmRuntime.Gguf.GgufTensorDescriptor@)
Tries to resolve a tensor descriptor by its exact GGUF name.
name- The exact ordinal name used for catalog lookup, canonical hashing, or diagnostic labeling as defined by the containing member.
tensor- When the method returns, contains the tensor produced by the operation when successful; otherwise contains the type's default value.
Returns: True when try get tensor succeeds for the supplied values; otherwise, false.
TryGetUInt32(string,uint@)
Tries to get an unsigned 32-bit metadata value.
key- The exact ordinal metadata or catalog key used for lookup; the operation does not normalize or retain a mutable alias of the key.
value- When the method returns, contains the value produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the key exists and can be converted to an unsigned 32-bit integer.
GgufParseOptionsUAIX.LmRuntime.Gguf
7 members
Defines safety limits for GGUF parsing.
ByteOrder
Gets the requested GGUF byte order. Auto uses the version-field heuristic.
MaxArrayDepth
Gets the maximum recursive array depth.
MaxArrayLength
Gets the maximum metadata array length.
MaxDimensionCount
Gets the maximum supported dimension count per tensor.
MaxMetadataCount
Gets the maximum supported metadata key-value count.
MaxStringBytes
Gets the maximum metadata string byte length.
MaxTensorCount
Gets the maximum supported tensor count.
GgufParseResultUAIX.LmRuntime.Gguf
3 members
Represents a non-throwing GGUF parse result.
Diagnostics
Gets parser diagnostics.
Model
Gets the parsed model when parsing succeeded.
Succeeded
Gets a value indicating whether parsing succeeded without diagnostics.
GgufReaderUAIX.LmRuntime.Gguf
2 members
Reads and validates GGUF model artifacts.
Read(string,UAIX.LmRuntime.Gguf.GgufParseOptions)
Reads a GGUF artifact from disk without copying tensor payloads to managed memory.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional GgufParseOptions controlling Read; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The GgufModel result produced by GgufReader.Read for this contract: Reads a GGUF artifact from disk without copying tensor payloads to managed memory. It is published only after all documented validation and ownership transitions succeed.
TryRead(string,UAIX.LmRuntime.Gguf.GgufParseOptions)
Parses a GGUF artifact and returns structured diagnostics instead of throwing for format failures.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional GgufParseOptions controlling TryRead; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: True when the GGUF parse result is produced successfully; otherwise, false and no successful result is published.
GgufShardInfoUAIX.LmRuntime.Gguf
3 members
Describes sharding metadata found in a GGUF artifact.
IsSharded
Gets a value indicating whether sharding metadata was found.
ShardCount
Gets the shard count when present.
ShardIndex
Gets the shard index when present.
GgufShardResolverUAIX.LmRuntime.Gguf
1 member
Resolves GGUF shard files adjacent to a root artifact.
ResolveShards(string)
Resolves likely shard paths for a GGUF model path.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
Returns: An ordered read-only IReadOnlyList<string> result from GgufShardResolver.ResolveShards: Resolves likely shard paths for a GGUF model path. Mutable internal collection aliases are not exposed through the returned contract.
GgufStrictValidatorUAIX.LmRuntime.Gguf
1 member
Validates GGUF artifacts without claiming execution parity.
Validate(string,UAIX.LmRuntime.Gguf.GgufParseOptions)
Validates the supplied path and the supplied options against the invariants required by GgufStrictValidator.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional GgufParseOptions controlling Validate; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The GgufValidationReport result produced by GgufStrictValidator.Validate for this contract: Validates the supplied path and the supplied options against the invariants required by GgufStrictValidator. It is published only after all documented validation and ownership transitions succeed.
GgufTensorDescriptorUAIX.LmRuntime.Gguf
7 members
Describes a tensor stored in a GGUF artifact without copying its payload.
AbsoluteOffset
Gets the absolute file offset for this tensor.
ByteLength
Gets the physical tensor byte length.
Dimensions
Gets tensor dimensions in GGUF order.
ElementCount
Gets the logical element count.
GgmlType
Gets the GGML storage type.
Name
Gets the tensor name.
RelativeOffset
Gets the relative tensor offset from the tensor data section.
GgufTensorPayloadViewUAIX.LmRuntime.Gguf
3 members
Represents a zero-copy tensor payload range inside a GGUF file.
AbsoluteOffset
Gets the absolute file offset.
LengthBytes
Gets the payload byte length.
FromTensor(UAIX.LmRuntime.Gguf.GgufTensorDescriptor)
Creates a payload view from a tensor descriptor.
tensor- The validated GGUF tensor descriptor whose dimensions, offsets, storage type, and byte length drive the operation.
Returns: The GgufTensorPayloadView result produced by GgufTensorPayloadView.FromTensor for this contract: Creates a payload view from a tensor descriptor. It is published only after all documented validation and ownership transitions succeed.
GgufValidationErrorUAIX.LmRuntime.Gguf
3 members
Represents a GGUF validation error.
ByteOffset
Gets the byte offset associated with the error, if known.
Code
Gets the diagnostic code.
Message
Gets the diagnostic message.
GgufValidationReportUAIX.LmRuntime.Gguf
3 members
Represents the result of strict GGUF validation.
Errors
Gets validation errors with byte offsets when known.
IsValid
Gets a value indicating whether validation succeeded.
Model
Gets the parsed model when validation succeeded.
IMappedModelFileUAIX.LmRuntime.Gguf
6 members
Defines bounded, zero-copy access to tensor payloads in a mapped model artifact.
FileLength
Gets the mapped file length in bytes.
Model
Gets the parsed GGUF model catalog associated with the mapping.
GetMemorySegments(ulong,ulong,int)
Creates bounded read-only memory windows for a validated mapped-file range.
absoluteOffset- The zero-based absolute offset into the relevant source or destination; range validation occurs before access.
byteLength- The bounded payload length in bytes used to validate offsets and prevent arithmetic overflow before slicing mapped storage.
maximumSegmentByteCount- The positive maximum byte count per returned segment.
Returns: Ordered borrowed segments whose combined length equals byteLength.
GetTensorBytes(UAIX.LmRuntime.Gguf.GgufTensorDescriptor)
Gets a synchronous read-only span over one tensor payload.
tensor- The validated GGUF tensor descriptor whose dimensions, offsets, storage type, and byte length drive the operation.
Returns: A read-only span valid until this mapping is disposed.
GetTensorMemory(UAIX.LmRuntime.Gguf.GgufTensorDescriptor)
Gets a read-only memory view over one tensor payload.
tensor- The validated GGUF tensor descriptor whose dimensions, offsets, storage type, and byte length drive the operation.
Returns: A read-only memory view whose owner remains the mapped file.
TryGetTensorMemory(string,System.ReadOnlyMemory<byte>@)
Attempts to get a read-only memory view by tensor name.
tensorName- The exact ordinal GGUF tensor catalog name used for lookup and diagnostics.
memory- When the method returns, contains the memory produced by the operation when successful; otherwise contains the type's default value.
Returns: True when try get tensor memory succeeds for the supplied values; otherwise, false.
IReadOnlyTensorViewUAIX.LmRuntime.Gguf
8 members
Defines a read-only tensor view backed by validated model storage.
ByteOrder
Gets the GGUF byte order.
DataType
Gets the runtime data type.
Descriptor
Gets the authoritative GGUF tensor descriptor.
IsDisposed
Gets whether the backing owner has been disposed.
LogicalDimensions
Gets normalized logical row-major dimensions.
StorageDimensions
Gets dimensions in GGUF storage order.
GetMemory
Gets a borrowed read-only memory view over the tensor payload.
Returns: The memory region containing the memory, bounded to the validated range owned by the result.
GetSpan
Gets a synchronous borrowed span over the tensor payload.
Returns: The bounded ReadOnlySpan<byte> view produced by IReadOnlyTensorView.GetSpan: Gets a synchronous borrowed span over the tensor payload. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
MappedGgufFileUAIX.LmRuntime.Gguf
14 members
Owns a read-only operating-system mapping of a GGUF artifact and exposes bounded tensor payload views.
The mapping is the sole owner of the acquired unmanaged pointer. Returned spans and memory views borrow that pointer and become invalid immediately after UAIX.LmRuntime.Gguf.MappedGgufFile.Dispose. Public entry points verify that descriptors belong to the parsed catalog before pointer arithmetic occurs.
FileLength
Gets the mapped file length captured from the validated catalog.
IsDisposed
Gets a value indicating whether the mapping has been disposed.
Model
Gets the immutable GGUF catalog that was validated before the operating-system mapping was opened.
CopyBytes(ulong,System.Span<byte>)
Copies a validated mapped-file range into a caller-owned destination.
absoluteOffset- The zero-based absolute offset into the relevant source or destination; range validation occurs before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
CopyTensorBytes(UAIX.LmRuntime.Gguf.GgufTensorDescriptor,System.Span<byte>)
Copies a complete tensor payload into a caller-owned destination.
tensor- The validated GGUF tensor descriptor whose dimensions, offsets, storage type, and byte length drive the operation.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
CreateTensorView(UAIX.LmRuntime.Gguf.GgufTensorDescriptor,System.Collections.Generic.IReadOnlyList<ulong>)
Creates a read-only typed tensor view whose lifetime is owned by this mapping.
tensor- The validated GGUF tensor descriptor whose dimensions, offsets, storage type, and byte length drive the operation.
logicalDimensions- Optional logical row-major dimensions; storage dimensions are used when omitted.
Returns: The borrowed mapped tensor view, with ownership and disposal obligations defined by the returned type and the CreateTensorView contract.
Dispose
Releases the acquired view pointer, operating-system mapping handles, and source file handle.
GetBytes(ulong,ulong)
Gets a read-only span over a validated file range.
absoluteOffset- The zero-based absolute offset into the relevant source or destination; range validation occurs before access.
byteLength- The bounded payload length in bytes used to validate offsets and prevent arithmetic overflow before slicing mapped storage.
Returns: The bounded ReadOnlySpan<byte> view produced by MappedGgufFile.GetBytes: Gets a read-only span over a validated file range. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
GetMemorySegments(ulong,ulong,int)
Splits a validated mapped-file range into bounded borrowed memory segments.
absoluteOffset- The absolute file offset at which the first segment begins.
byteLength- The total number of bytes represented by all returned segments.
maximumSegmentByteCount- The positive upper bound for each managed segment descriptor.
Returns: Ordered borrowed memory segments that exactly cover the requested range.
GetTensorBytes(UAIX.LmRuntime.Gguf.GgufTensorDescriptor)
Gets a synchronous borrowed span over one catalog tensor payload.
tensor- The descriptor that must match the mapping's authoritative tensor catalog.
Returns: A read-only span whose lifetime cannot exceed the current synchronous call chain.
GetTensorMemory(UAIX.LmRuntime.Gguf.GgufTensorDescriptor)
Gets a borrowed memory object over one catalog tensor payload.
tensor- The descriptor that must match the mapping's authoritative tensor catalog.
Returns: A read-only memory view backed directly by the operating-system mapping.
MappedGgufFile(string,UAIX.LmRuntime.Gguf.GgufParseOptions)
Opens, validates, and maps a GGUF file read-only.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional GgufParseOptions controlling MappedGgufFile; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
MappedGgufFile(UAIX.LmRuntime.Gguf.GgufModel)
Maps a GGUF file using an already validated catalog.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
TryGetTensorMemory(string,System.ReadOnlyMemory<byte>@)
Tries to resolve a named tensor and create a borrowed mapped-memory view.
tensorName- The exact ordinal GGUF tensor catalog name used for lookup and diagnostics.
memory- Receives the borrowed tensor memory when the tensor exists; otherwise receives empty memory.
Returns: True when the tensor exists and a view was created; otherwise false.
MappedTensorViewUAIX.LmRuntime.Gguf
9 members
Carries validated tensor geometry and a borrowed read-only view into a mapped GGUF file.
ByteOrder
DataType
Descriptor
IsDisposed
LogicalDimensions
StorageDimensions
GetMemory
Retrieves the memory from the current MappedTensorView state after validating the requested access.
Returns: The memory region containing the memory, bounded to the validated range owned by the result.
GetSpan
Retrieves the span from the current MappedTensorView state after validating the requested access.
Returns: The bounded ReadOnlySpan<byte> view produced by MappedTensorView.GetSpan: Retrieves the span from the current MappedTensorView state after validating the requested access. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
MappedTensorView(UAIX.LmRuntime.Gguf.MappedGgufFile,UAIX.LmRuntime.Gguf.GgufTensorDescriptor,System.Collections.Generic.IReadOnlyList<ulong>)
Initializes a new MappedTensorView instance with validated dependencies and operational bounds.
owner- The mapped model-file owner that keeps the tensor payload alive for the lifetime of the created view.
descriptor- The validated tensor descriptor retained by the view; its offsets and lengths remain bounded by the mapped file.
logicalDimensions- The logical dimensions sequence used by this operation; its required length, ordering, and element bounds are validated before access.
ModelFileSegmentUAIX.LmRuntime.Gguf
5 members
Owns one pooled, bounded model-file segment returned by UAIX.LmRuntime.Gguf.SegmentedModelFileReader.
The segment owns an System.Buffers.ArrayPool`1 lease. Consumers may retain UAIX.LmRuntime.Gguf.ModelFileSegment.Memory only while this instance is alive and must dispose the segment exactly once when the bytes are no longer needed.
IsDisposed
Gets whether the pooled segment has been released.
Length
Gets the number of valid bytes in the segment.
Memory
Gets the read-only segment bytes while this owner remains alive.
Offset
Gets the unsigned logical file offset represented by this segment.
Dispose
Returns the rented byte array to the shared pool and invalidates Memory.
SegmentedModelFileOptionsUAIX.LmRuntime.Gguf
1 member
Configures bounded segmented reads from a model file whose logical offsets remain unsigned 64-bit values.
MaximumSegmentByteCount
Gets the maximum bytes returned by one owned segment.
SegmentedModelFileReaderUAIX.LmRuntime.Gguf
7 members
Reads bounded windows from a local model file without representing the complete file as one managed span.
Logical positions remain unsigned 64-bit values until the final checked conversion required by the operating-system random-access API. Returned owned segments are copies in pooled memory; caller-provided spans remain caller owned. This class performs no network access and never follows URLs.
IsDisposed
Gets whether the reader has released its file handle.
Length
Gets the file length as an unsigned 64-bit value.
MaximumSegmentByteCount
Gets the maximum owned segment size.
Dispose
Releases resources owned by SegmentedModelFileReader and transitions it to the disposed state.
ReadExactly(ulong,System.Span<byte>)
Reads exactly into a caller-owned bounded destination.
offset- The zero-based offset into the relevant source or destination; range validation occurs before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
ReadSegment(ulong,int)
Reads the segment from the current binary source using the component's validated representation.
offset- The zero-based offset into the relevant source or destination; range validation occurs before access.
length- The length used to bound this operation; it must be nonnegative and within the supported range.
Returns: The ModelFileSegment result produced by SegmentedModelFileReader.ReadSegment for this contract: Reads the segment from the current binary source using the component's validated representation. It is published only after all documented validation and ownership transitions succeed.
SegmentedModelFileReader(string,UAIX.LmRuntime.Gguf.SegmentedModelFileOptions)
Opens a local model file for deterministic random access.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional SegmentedModelFileOptions controlling SegmentedModelFileReader; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
UAIX.LmRuntime.Sampling
16 documented types and 71 documented members.
GenerationControllerUAIX.LmRuntime.Sampling
9 members
Enforces stop, limit, cancellation, usage, and output-publication boundaries for one generation.
FinishReason
Gets the first decisive finish reason, or UAIX.LmRuntime.Sampling.GenerationFinishReason.None while active.
OutputTokenIds
Gets generated token identifiers retained under stop-token emission policy.
State
Gets the session-local sampling state used by the same generation.
Usage
Gets exact tokenizer-ID usage without deriving counts from text or bytes.
AcceptToken(int,System.ReadOnlySpan<byte>,System.Threading.CancellationToken)
Attempts to accept and publish one generated token at a bounded cancellation point.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
decodedBytes- The decoded bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
cancellationToken- Cancellation observed before the token is recorded or published.
Returns: The token/output/finish transition produced by this step.
CompleteVisibleBytes
Completes an otherwise active stream and flushes any bytes retained as a possible stop prefix.
Returns: Remaining visible bytes; the finish reason remains GenerationFinishReason.None.
Fail(System.Exception)
Records an execution failure without exposing exception details through the stable finish reason.
exception- The non-null execution failure observed by the caller.
Returns: The GenerationFinishReason result produced by GenerationController.Fail for this contract: Records an execution failure without exposing exception details through the stable finish reason. It is published only after all documented validation and ownership transitions succeed.
GenerationController(int,int,UAIX.LmRuntime.Sampling.SamplingOptions)
Initializes a controller from validated vocabulary and prompt-token bounds.
vocabularySize- The positive tokenizer vocabulary size used to validate token identifiers and size bounded result buffers.
promptTokenCount- The prompt token count used to bound this operation; it must be nonnegative and within the supported range.
options- The optional SamplingOptions controlling GenerationController; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
ObserveCancellation(System.Threading.CancellationToken)
Observes cancellation between decode steps without publishing another token.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The GenerationFinishReason result produced by GenerationController.ObserveCancellation for this contract: Observes cancellation between decode steps without publishing another token. It is published only after all documented validation and ownership transitions succeed.
GenerationFinishReasonUAIX.LmRuntime.Sampling
7 members
Identifies the first decisive condition that ended a generation.
Cancelled
Cancellation was observed before publishing another token.
ContextLimit
The configured prompt-plus-generation context bound was reached.
ExecutionError
An execution error ended generation.
None
Generation remains active.
StopText
A configured decoded UTF-8 stop sequence ended generation.
StopToken
An exact configured token identifier ended generation.
TokenLimit
The configured maximum generated-token count was reached.
GenerationStepResultUAIX.LmRuntime.Sampling
3 members
Represents the observable result of attempting to publish one generated token.
FinishReason
Gets the stable finish reason after this step.
TokenAccepted
Gets whether the token identifier was retained in generated-token output.
VisibleBytes
Gets bytes newly safe to publish after stop-prefix matching.
GenerationUsageUAIX.LmRuntime.Sampling
3 members
Records tokenizer-ID-based usage accounting.
CompletionTokens
Gets the exact number of generated token identifiers accepted by the controller.
PromptTokens
Gets the exact number of prompt token identifiers consumed.
TotalTokens
Gets the checked sum of prompt and completion token counts.
GreedySamplerUAIX.LmRuntime.Sampling
1 member
Provides deterministic greedy token selection with explicit non-finite input policy.
Select(System.ReadOnlySpan<float>)
Selects the highest logit index with deterministic lower-index tie-breaking.
logits- The non-empty source logits. NaN is rejected; infinities compare normally.
Returns: The int value computed by GreedySampler.Select for this contract: Selects the highest logit index with deterministic lower-index tie-breaking. Range, finite-value, and overflow checks are completed before the value is returned.
LogitProcessorUAIX.LmRuntime.Sampling
4 members
Applies validated, deterministic token-history and bias policies to logits.
ApplyBias(System.Span<float>,System.Collections.Generic.IReadOnlyDictionary<int,float>)
Adds all validated per-token biases exactly once.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
biases- The ordered biases collection of type IReadOnlyDictionary<int, float>; LogitProcessor.ApplyBias validates nullability, count, and element constraints before consuming or snapshotting it and does not retain a mutable caller alias.
ApplyHistoryPenalties(System.Span<float>,UAIX.LmRuntime.Sampling.SamplingOptions,System.Collections.Generic.IReadOnlyDictionary<int,int>)
Applies sign-aware repetition, frequency, and presence penalties in one deterministic pass.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
options- The optional SamplingOptions controlling ApplyHistoryPenalties; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
tokenCounts- The ordered token counts collection of type IReadOnlyDictionary<int, int>; LogitProcessor.ApplyHistoryPenalties validates nullability, count, and element constraints before consuming or snapshotting it and does not retain a mutable caller alias.
Process(System.ReadOnlySpan<float>,UAIX.LmRuntime.Sampling.SamplingOptions,UAIX.LmRuntime.Sampling.SamplingState)
Produces a processed copy of the source logits without exposing partially mutated caller data on validation failure.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
options- The optional SamplingOptions controlling Process; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
state- The validated state value consumed by the operation; mutations, when applicable, are limited to the explicitly documented state owner.
Returns: A newly allocated float[] containing the ordered result of LogitProcessor.Process: Produces a processed copy of the source logits without exposing partially mutated caller data on validation failure. The caller owns the returned array and later mutation cannot alter the source object.
SuppressEarlyStopTokens(System.Span<float>,UAIX.LmRuntime.Sampling.SamplingOptions,int)
Marks configured stop tokens ineligible before the exact minimum-token boundary.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
options- The optional SamplingOptions controlling SuppressEarlyStopTokens; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
generatedTokenCount- The generated token count used to bound this operation; it must be nonnegative and within the supported range.
LogitScoreUAIX.LmRuntime.Sampling
2 members
Represents a scored token candidate.
Score
Gets the logit score.
TokenIndex
Gets the token index.
ProbabilitySamplerUAIX.LmRuntime.Sampling
2 members
Builds stable normalized distributions and samples them with session-local deterministic state.
BuildDistribution(System.ReadOnlySpan<float>,UAIX.LmRuntime.Sampling.SamplingOptions)
Builds a stable, filtered, and normalized probability distribution without consuming random state.
processedLogits- The processed logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
options- The optional SamplingOptions controlling BuildDistribution; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: Candidates ordered by descending probability and ascending token identifier for ties.
Select(System.ReadOnlySpan<float>,UAIX.LmRuntime.Sampling.SamplingOptions,UAIX.LmRuntime.Sampling.SamplingState)
Processes logits and selects one token under the supplied deterministic policy.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
options- The optional SamplingOptions controlling Select; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
state- The validated state value consumed by the operation; mutations, when applicable, are limited to the explicitly documented state owner.
Returns: The SamplingDecision result produced by ProbabilitySampler.Select for this contract: Processes logits and selects one token under the supplied deterministic policy. It is published only after all documented validation and ownership transitions succeed.
SamplingCandidateUAIX.LmRuntime.Sampling
2 members
Represents one normalized token candidate retained after all filters.
Probability
Gets the normalized candidate probability.
TokenId
Gets the token identifier.
SamplingDecisionUAIX.LmRuntime.Sampling
3 members
Describes one deterministic or stochastic sampling decision.
Candidates
Gets the candidate distribution used for selection, ordered by probability and token identifier.
IsGreedy
Gets whether the zero-temperature greedy path made the decision.
TokenId
Gets the selected token identifier.
SamplingOptionsUAIX.LmRuntime.Sampling
17 members
Defines deterministic logit-processing, candidate-selection, and generation-stop policies for one sampling session.
Instances are treated as immutable configuration after a UAIX.LmRuntime.Sampling.SamplingState is created. Validation is intentionally performed before a logit buffer is modified so an invalid option cannot leave partially adjusted data.
FrequencyPenalty
Gets the amount subtracted for each prior occurrence of a token.
IncludeStopSequence
Gets whether matched stop-sequence bytes are included in visible output.
IncludeStopToken
Gets whether a matched stop token is retained in emitted token identifiers.
LogitBias
Gets finite per-token additive logit biases.
MaximumContextTokens
Gets the maximum prompt-plus-generation token count.
MaximumGeneratedTokens
Gets the maximum number of generated tokens. Zero permits no generated tokens.
MinimumGeneratedTokens
Gets the minimum generated-token count before stop-token or stop-text policies become eligible.
MinimumP
Gets the minimum probability relative to the highest candidate probability.
PresencePenalty
Gets the amount subtracted once from any token that has previously appeared.
RepetitionPenalty
Gets the positive sign-aware repetition penalty.
Seed
Gets the deterministic per-session pseudo-random generator seed.
StopSequences
Gets UTF-8 stop strings matched across decoded token boundaries.
StopTokenIds
Gets exact token identifiers that terminate generation after the minimum-token boundary.
Temperature
Gets the non-negative temperature. Zero selects the deterministic greedy path.
TopK
Gets the maximum candidate count. Zero disables top-k filtering.
TopP
Gets the normalized nucleus probability threshold in the inclusive range zero through one.
Validate(int)
Validates every option and token-indexed policy against a vocabulary size without mutating caller data.
vocabularySize- The positive number of logits accepted by the model.
SamplingStateUAIX.LmRuntime.Sampling
6 members
Stores token history and deterministic pseudo-random state for exactly one generation session.
A state instance must not be shared by independent requests. Keeping history and random state together makes session isolation explicit and prevents interleaved requests from consuming one another's random sequence.
GeneratedTokenCount
Gets the number of generated token identifiers recorded by this session.
Random
Gets the session-local deterministic random generator.
TokenCounts
Gets the session-owned prior-token counts through a read-only interface.
GetTokenCount(int)
Gets the prior count for one token without adding it to history.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
Returns: The recorded count, or zero when the token has not appeared.
RecordToken(int)
Records one generated token for repetition, frequency, presence, and usage policies.
tokenId- The non-negative tokenizer identifier generated by the model.
SamplingState(UAIX.LmRuntime.Sampling.SamplingOptions)
Initializes isolated state from the immutable session options.
options- The options whose seed initializes the session generator.
StopSequenceMatcherUAIX.LmRuntime.Sampling
5 members
Matches UTF-8 stop sequences across arbitrary decoded-byte boundaries while retaining only a bounded possible prefix.
MaximumRetainedBytes
Gets the maximum retained prefix bytes, bounded by the longest configured stop sequence.
RetainedByteCount
Gets the current possible stop-prefix byte count.
Append(System.ReadOnlySpan<byte>)
Appends one decoded byte chunk and releases bytes that can no longer participate in a stop match.
bytes- The next decoded UTF-8 bytes; chunks may split a code point or stop sequence.
Returns: The newly visible bytes and optional terminal match.
Complete
Completes matching and releases any retained non-matching prefix bytes.
Returns: A newly allocated byte[] containing the ordered result of StopSequenceMatcher.Complete: Completes matching and releases any retained non-matching prefix bytes. The caller owns the returned array and later mutation cannot alter the source object.
StopSequenceMatcher(System.Collections.Generic.IEnumerable<string>,bool)
Initializes a matcher from non-empty stop strings.
stopSequences- The exact Unicode strings encoded as UTF-8 for byte matching.
includeMatchedBytes- Whether a terminal match is included in visible output.
StopSequenceMatchResultUAIX.LmRuntime.Sampling
3 members
Represents visible bytes released by one bounded stop-sequence matching step.
Matched
Gets whether a terminal stop sequence has matched.
MatchedStopSequence
Gets the exact stop string that completed during this step, if any.
VisibleBytes
Gets newly visible bytes that cannot participate in a future stop match.
TopKSelectorUAIX.LmRuntime.Sampling
1 member
Provides partial top-k selection for logit arrays.
SelectTopK(System.ReadOnlySpan<float>,int)
Selects the highest scoring token candidates without sorting the full input.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
k- The numeric k consumed by SelectTopK; it must satisfy the member's documented range, geometry, and finite-value requirements.
Returns: The selected candidates in descending score order with deterministic index tie-breaking.
Xoshiro256StarStarUAIX.LmRuntime.Sampling
3 members
Implements the xoshiro256** generator with SplitMix64 seed expansion and deterministic unsigned arithmetic.
The generator is intentionally session-local and not thread-safe. Its transition follows the published xoshiro256** reference algorithm; seed expansion prevents the prohibited all-zero state for a zero seed.
NextUInt64
Returns the next 64-bit output and advances the generator exactly once.
Returns: The ulong value computed by Xoshiro256StarStar.NextUInt64 for this contract: Returns the next 64-bit output and advances the generator exactly once. Range, finite-value, and overflow checks are completed before the value is returned.
NextUnitDouble
Returns a uniformly distributed value in the half-open interval [0, 1).
Returns: The double value computed by Xoshiro256StarStar.NextUnitDouble for this contract: Returns a uniformly distributed value in the half-open interval [0, 1). Range, finite-value, and overflow checks are completed before the value is returned.
Xoshiro256StarStar(ulong)
Initializes the generator from one deterministic 64-bit seed.
seed- The numeric seed consumed by Xoshiro256StarStar; it must satisfy the member's documented range, geometry, and finite-value requirements.
UAIX.LmRuntime.Kernels.Cpu
21 documented types and 87 documented members.
Avx2Float32KernelsUAIX.LmRuntime.Kernels.Cpu
2 members
Provides dedicated AVX2 float32 correctness kernels with scalar tails.
These kernels are selected only when AVX2 is explicitly requested and supported. Scalar implementations remain the numerical authority, and no throughput claim is implied until executed benchmark evidence exists.
Dot(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>)
Computes a float32 dot product with AVX/FMA vector arithmetic and a scalar tail.
left- The left sequence used by this operation; its required length, ordering, and element bounds are validated before access.
right- The right sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by Avx2Float32Kernels.Dot for this contract: Computes a float32 dot product with AVX/FMA vector arithmetic and a scalar tail. Range, finite-value, and overflow checks are completed before the value is returned.
MatVec(System.ReadOnlySpan<float>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a row-major float32 matrix-vector product by reusing the dedicated AVX2 dot kernel.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
CpuKernelDispatcherUAIX.LmRuntime.Kernels.Cpu
6 members
Dispatches correctness-first CPU kernels through explicitly selectable implementation tiers.
Scalar implementations remain the numerical authority. Portable and architecture-specific paths are additive and can always be bypassed by requesting UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier.Scalar.
DotFloat32(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier,UAIX.LmRuntime.Kernels.Cpu.CpuKernelSelection@)
Computes a float32 dot product through the selected implementation tier.
left- The left sequence used by this operation; its required length, ordering, and element bounds are validated before access.
right- The right sequence used by this operation; its required length, ordering, and element bounds are validated before access.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
selection- When the method returns, contains the selection produced by the operation when successful; otherwise contains the type's default value.
Returns: The float value computed by CpuKernelDispatcher.DotFloat32 for this contract: Computes a float32 dot product through the selected implementation tier. Range, finite-value, and overflow checks are completed before the value is returned.
DotQ4_0(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>,UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier,UAIX.LmRuntime.Kernels.Cpu.CpuKernelSelection@)
Computes a Q4_0 block dot product through a scalar or portable-vector correctness path.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
selection- When the method returns, contains the selection produced by the operation when successful; otherwise contains the type's default value.
Returns: The float value computed by CpuKernelDispatcher.DotQ4_0 for this contract: Computes a Q4_0 block dot product through a scalar or portable-vector correctness path. Range, finite-value, and overflow checks are completed before the value is returned.
DotQ8_0(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>,UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier,UAIX.LmRuntime.Kernels.Cpu.CpuKernelSelection@)
Computes a Q8_0 block dot product through a scalar or portable-vector correctness path.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
selection- When the method returns, contains the selection produced by the operation when successful; otherwise contains the type's default value.
Returns: The float value computed by CpuKernelDispatcher.DotQ8_0 for this contract: Computes a Q8_0 block dot product through a scalar or portable-vector correctness path. Range, finite-value, and overflow checks are completed before the value is returned.
MatVecFloat32(System.ReadOnlySpan<float>,int,int,System.ReadOnlySpan<float>,System.Span<float>,UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier,UAIX.LmRuntime.Kernels.Cpu.CpuKernelSelection@)
Computes a row-major float32 matrix-vector product through the selected implementation tier.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
selection- When the method returns, contains the selection produced by the operation when successful; otherwise contains the type's default value.
RmsNorm(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>,float,UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier,UAIX.LmRuntime.Kernels.Cpu.CpuKernelSelection@)
Applies RMS normalization using the selected scalar or portable-vector accumulation tier.
input- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
weight- The weight sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
epsilon- The positive normalization epsilon added to the mean-square term to avoid division by zero while preserving deterministic numerical behavior.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
selection- When the method returns, contains the selection produced by the operation when successful; otherwise contains the type's default value.
SelectFloat32DotTier(UAIX.LmRuntime.Kernels.Cpu.CpuKernelTier)
Selects the implemented tier for a float32 dot product.
requestedTier- The requested tier containing caller-supplied values for this operation; all required fields are validated before processing.
Returns: The CpuKernelSelection result produced by CpuKernelDispatcher.SelectFloat32DotTier for this contract: Selects the implemented tier for a float32 dot product. It is published only after all documented validation and ownership transitions succeed.
CpuKernelSelectionUAIX.LmRuntime.Kernels.Cpu
4 members
Describes the requested and selected CPU kernel tier for one operation.
Operation
Gets the operation name.
Reason
Gets the stable selection rationale.
RequestedTier
Gets the requested tier.
SelectedTier
Gets the selected tier.
CpuKernelTierUAIX.LmRuntime.Kernels.Cpu
6 members
Identifies a managed CPU kernel implementation tier.
AdvSimd
Uses an ARM64 AdvSimd implementation when the operation provides one.
Auto
Selects the highest supported tier implemented for the requested operation.
Avx2
Uses an AVX2 implementation when the operation provides one.
Avx512
Uses an AVX-512 implementation when the operation provides one.
PortableVector
Uses portable System.Numerics.Vector`1 operations.
Scalar
Uses the scalar correctness implementation.
FusedQuantizedDotProductUAIX.LmRuntime.Kernels.Cpu
1 member
Provides fused dequantize-and-dot reference kernels.
Dot(UAIX.LmRuntime.Tensors.GgmlTensorType,System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>)
Computes a dot product between a quantized block and float activations.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by FusedQuantizedDotProduct.Dot for this contract: Computes a dot product between a quantized block and float activations. Range, finite-value, and overflow checks are completed before the value is returned.
KQuantizedBlockLayoutUAIX.LmRuntime.Kernels.Cpu
4 members
Describes one audited K-quantized block layout.
ByteCount
Gets the physical byte count.
ElementCount
Gets the logical element count.
Format
Gets the format name.
LayoutDescription
Gets the audited layout statement.
KQuantizedCpuKernelsUAIX.LmRuntime.Kernels.Cpu
17 members
Provides correctness-first scalar GGML Q4_K and Q6_K block kernels.
These methods operate on one exact 256-element block and never materialize a complete model matrix. All scale fields are interpreted as little-endian IEEE half values because current direct K-quant execution is limited to little-endian GGUF storage.
BlockElementCount
Gets the number of logical values in one K-quant block.
Q4_KBlockBytes
Gets the historical Q4_K block-byte constant retained for source compatibility.
Q4KBlockByteCount
Gets the exact Q4_K block byte count.
Q6_KBlockBytes
Gets the historical Q6_K block-byte constant retained for source compatibility.
Q6KBlockByteCount
Gets the exact Q6_K block byte count.
Q4KLayout
Gets the audited Q4_K block layout.
Q6KLayout
Gets the audited Q6_K block layout.
DequantizeQ4_K(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one Q4_K block using the historical method name retained for source compatibility.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
DequantizeQ4K(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one exact Q4_K block into a caller-owned destination.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
DequantizeQ6_K(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one Q6_K block using the historical method name retained for source compatibility.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
DequantizeQ6K(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one exact Q6_K block into a caller-owned destination.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
DotQ4K(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>)
Computes an allocation-free dot product for one exact Q4_K block.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by KQuantizedCpuKernels.DotQ4K for this contract: Computes an allocation-free dot product for one exact Q4_K block. Range, finite-value, and overflow checks are completed before the value is returned.
DotQ6K(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>)
Computes an allocation-free dot product for one exact Q6_K block.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by KQuantizedCpuKernels.DotQ6K for this contract: Computes an allocation-free dot product for one exact Q6_K block. Range, finite-value, and overflow checks are completed before the value is returned.
MatVecQ4_K(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a Q4_K matrix-vector product using the historical method name retained for source compatibility.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MatVecQ4K(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a row-major Q4_K matrix-vector product without whole-matrix dequantization.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MatVecQ6_K(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a Q6_K matrix-vector product using the historical method name retained for source compatibility.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MatVecQ6K(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a row-major Q6_K matrix-vector product without whole-matrix dequantization.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
Q4_0BlockUAIX.LmRuntime.Kernels.Cpu
2 members
Defines the exact packed Q4_0 block layout used by GGML storage.
QuantizedValues
Stores 32 signed 4-bit values in 16 packed bytes.
Scale
Gets or sets the little-endian IEEE half scale field.
Q4_1DequantizerUAIX.LmRuntime.Kernels.Cpu
1 member
Dequantizes Q4_1 blocks for scalar reference parity.
DequantizeBlock(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one Q4_1 block into destination floats.
source- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
Q4KBlockUAIX.LmRuntime.Kernels.Cpu
4 members
Defines the audited packed GGML Q4_K block layout for 256 logical values.
MinimumScale
Gets or sets the common little-endian IEEE half minimum scale.
QuantizedValues
Stores 256 four-bit quants in 128 bytes.
Scale
Gets or sets the common little-endian IEEE half scale.
ScaleMinimums
Stores eight packed 6-bit scales and eight packed 6-bit minimum factors.
Q5_0DequantizerUAIX.LmRuntime.Kernels.Cpu
1 member
Dequantizes Q5_0 blocks for scalar reference parity.
DequantizeBlock(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one Q5_0 block into destination floats.
source- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
Q6_KDequantizerUAIX.LmRuntime.Kernels.Cpu
1 member
Dequantizes Q6_K blocks for scalar reference parity.
DequantizeBlock(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one exact GGML Q6_K block into destination floats.
source- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
Q6KBlockUAIX.LmRuntime.Kernels.Cpu
4 members
Defines the audited packed GGML Q6_K block layout for 256 logical values.
HighBits
Stores the upper two bits for 256 quants.
LowBits
Stores the lower four bits for 256 quants.
Scale
Gets or sets the common little-endian IEEE half scale.
Scales
Stores sixteen signed sub-block scales.
Q8_0BlockUAIX.LmRuntime.Kernels.Cpu
2 members
Defines the exact packed Q8_0 block layout used by GGML storage.
QuantizedValues
Stores 32 signed 8-bit values.
Scale
Gets or sets the little-endian IEEE half scale field.
QuantizedCpuKernelsUAIX.LmRuntime.Kernels.Cpu
9 members
Provides allocation-free scalar correctness kernels for high-value GGML quantization formats.
BlockElementCount
Gets the logical element count in a Q4_0 or Q8_0 block.
Q4_0BlockBytes
Gets the byte length of a Q4_0 block.
Q8_0BlockBytes
Gets the byte length of a Q8_0 block.
DequantizeQ4_0(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one little-endian Q4_0 block into float32 values.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The destination buffer with room for 32 values.
DequantizeQ8_0(System.ReadOnlySpan<byte>,System.Span<float>)
Dequantizes one little-endian Q8_0 block into float32 values.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
destination- The destination buffer with room for 32 values.
DotQ4_0(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>)
Computes an allocation-free dequantize-and-dot operation for one Q4_0 block.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by QuantizedCpuKernels.DotQ4_0 for this contract: Computes an allocation-free dequantize-and-dot operation for one Q4_0 block. Range, finite-value, and overflow checks are completed before the value is returned.
DotQ8_0(System.ReadOnlySpan<byte>,System.ReadOnlySpan<float>)
Computes an allocation-free dequantize-and-dot operation for one Q8_0 block.
block- The block sequence used by this operation; its required length, ordering, and element bounds are validated before access.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The float value computed by QuantizedCpuKernels.DotQ8_0 for this contract: Computes an allocation-free dequantize-and-dot operation for one Q8_0 block. Range, finite-value, and overflow checks are completed before the value is returned.
MatVecQ4_0(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a row-major Q4_0 matrix-vector product without materializing full-precision rows.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The number of logical columns; it must be divisible by 32.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MatVecQ8_0(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a row-major Q8_0 matrix-vector product without materializing full-precision rows.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The number of logical columns; it must be divisible by 32.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
QuantizedKernelParityReportUAIX.LmRuntime.Kernels.Cpu
2 members
Represents the result of comparing an optimized quantized kernel to a reference kernel.
MaxAbsoluteError
Gets the maximum absolute error observed.
Passed
Gets a value indicating whether outputs are within tolerance.
QuantizedKernelParityRunnerUAIX.LmRuntime.Kernels.Cpu
1 member
Compares quantized kernels against scalar references.
CompareAgainstReference(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,float)
Compares two output vectors with an absolute tolerance.
reference- The reference sequence used by this operation; its required length, ordering, and element bounds are validated before access.
actual- The actual sequence used by this operation; its required length, ordering, and element bounds are validated before access.
tolerance- The numeric tolerance consumed by CompareAgainstReference; it must satisfy the member's documented range, geometry, and finite-value requirements.
Returns: The QuantizedKernelParityReport result produced by QuantizedKernelParityRunner.CompareAgainstReference for this contract: Compares two output vectors with an absolute tolerance. It is published only after all documented validation and ownership transitions succeed.
ReferenceCpuKernelsUAIX.LmRuntime.Kernels.Cpu
6 members
Provides scalar and portable CPU reference kernels for correctness anchoring.
ApplyRope(System.Span<float>,int,float)
Applies RoPE rotation using generated trigonometric tables for the supplied position.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
theta- The rotary angle in radians applied to the paired vector components at the addressed position.
ApplyRopeInPlace(System.Span<float>,System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,int)
Applies RoPE rotation to one query or key vector in place using precomputed sine and cosine values.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
cos- The cos sequence used by this operation; its required length, ordering, and element bounds are validated before access.
sin- The sin sequence used by this operation; its required length, ordering, and element bounds are validated before access.
ropeDimensions- The even number of leading head dimensions transformed by rotary positional encoding.
MatVec(System.ReadOnlySpan<float>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a matrix-vector product for row-major float32 weights.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
RmsNorm(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>,float)
Applies RMS normalization using the shared vector math implementation.
input- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
weight- The weight sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
epsilon- The positive normalization epsilon added to the mean-square term to avoid division by zero while preserving deterministic numerical behavior.
Softmax(System.Span<float>)
Computes softmax probabilities for the supplied values using numerically stable normalization.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
SoftmaxInPlace(System.Span<float>)
Computes softmax probabilities for the in place using numerically stable normalization.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
ReferenceMatrixRowDispatcherUAIX.LmRuntime.Kernels.Cpu
2 members
Dispatches correctness-first matrix-vector operations for supported mapped scalar and quantized rows.
Describe(UAIX.LmRuntime.Tensors.GgmlTensorType,int,int)
Computes the exact storage byte count for a supported matrix.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The ReferenceMatrixStorageDescriptor result produced by ReferenceMatrixRowDispatcher.Describe for this contract: Computes the exact storage byte count for a supported matrix. It is published only after all documented validation and ownership transitions succeed.
MatVec(UAIX.LmRuntime.Tensors.GgmlTensorType,System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>)
Computes a little-endian matrix-vector product without materializing a complete dequantized matrix.
type- The declared GGML tensor or metadata type used to select the corresponding decoding and validation rules.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
activations- The activations sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
ReferenceMatrixStorageDescriptorUAIX.LmRuntime.Kernels.Cpu
4 members
Describes one supported scalar matrix storage layout.
ColumnCount
Gets the logical column count.
GgmlType
Gets the GGML tensor type.
RequiredByteCount
Gets the exact required storage byte count.
RowCount
Gets the logical row count.
Scalar16CpuKernelsUAIX.LmRuntime.Kernels.Cpu
8 members
Provides correctness-first scalar F16 and BF16 decoding and matrix-vector kernels.
CopyBFloat16(System.ReadOnlySpan<byte>,System.Span<float>,bool)
Copies BF16 values into a caller-owned float32 destination.
source- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
destination- The destination with one element per BF16 value.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
CopyFloat16(System.ReadOnlySpan<byte>,System.Span<float>,bool)
Copies F16 values into a caller-owned float32 destination.
source- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
destination- The destination with one element per F16 value.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
DecodeBFloat16(System.ReadOnlySpan<byte>,bool)
Decodes one bfloat16 value with explicit byte order.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
Returns: The float value computed by Scalar16CpuKernels.DecodeBFloat16 for this contract: Decodes one bfloat16 value with explicit byte order. Range, finite-value, and overflow checks are completed before the value is returned.
DecodeBFloat16(System.ReadOnlySpan<byte>,bool,bool)
Decodes one bfloat16 value with explicit byte order and non-finite policy.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
rejectNonFinite- Whether NaN and infinity are rejected as invalid model weights.
Returns: The float value computed by Scalar16CpuKernels.DecodeBFloat16 for this contract: Decodes one bfloat16 value with explicit byte order and non-finite policy. Range, finite-value, and overflow checks are completed before the value is returned.
DecodeFloat16(System.ReadOnlySpan<byte>,bool)
Decodes one IEEE binary16 value with explicit byte order.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
Returns: The float value computed by Scalar16CpuKernels.DecodeFloat16 for this contract: Decodes one IEEE binary16 value with explicit byte order. Range, finite-value, and overflow checks are completed before the value is returned.
DecodeFloat16(System.ReadOnlySpan<byte>,bool,bool)
Decodes one IEEE binary16 value with explicit byte order and non-finite policy.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
rejectNonFinite- Whether NaN and infinity are rejected as invalid model weights.
Returns: The float value computed by Scalar16CpuKernels.DecodeFloat16 for this contract: Decodes one IEEE binary16 value with explicit byte order and non-finite policy. Range, finite-value, and overflow checks are completed before the value is returned.
MatVecBFloat16(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>,bool)
Computes a row-major BF16 matrix-vector product without whole-matrix conversion.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
MatVecFloat16(System.ReadOnlySpan<byte>,int,int,System.ReadOnlySpan<float>,System.Span<float>,bool)
Computes a row-major F16 matrix-vector product without whole-matrix conversion.
matrix- The matrix sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
bigEndian- True to decode source bytes in big-endian order; false to decode the little-endian representation used by the local GGUF artifact.
UAIX.LmRuntime.Tokenization
45 documented types and 160 documented members.
BpeMergeRuleUAIX.LmRuntime.Tokenization
3 members
Represents one parsed BPE merge rule.
Left
Gets the left symbol.
Right
Gets the right symbol.
TryParse(string,UAIX.LmRuntime.Tokenization.BpeMergeRule@)
Attempts to parse the Boolean result while reporting invalid input without a successful result.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
rule- When the method returns, contains the rule produced by the operation when successful; otherwise contains the type's default value.
Returns: True when the rule contains two non-empty symbols.
ChatTemplateConformanceSuiteUAIX.LmRuntime.Tokenization
1 member
Runs chat-template conformance checks against rendered message sequences.
RenderAndValidate(string,System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.Contracts.LlmMessage>)
Renders and validates a chat template against a message sequence.
template- The template text. The current safe subset ignores arbitrary code.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenizerParityReport result produced by ChatTemplateConformanceSuite.RenderAndValidate for this contract: Renders and validates a chat template against a message sequence. It is published only after all documented validation and ownership transitions succeed.
ChatTemplateRendererUAIX.LmRuntime.Tokenization
1 member
Renders a minimal safe chat template suitable for deterministic tests and initial GGUF tokenizer work.
Render(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Renders messages using a small role/content template rather than a general Jinja interpreter.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The text produced by ChatTemplateRenderer.Render for this contract: Renders messages using a small role/content template rather than a general Jinja interpreter. The returned string is detached from mutable caller storage and is not persisted by the operation.
DetokenizationOptionsUAIX.LmRuntime.Tokenization
3 members
Describes detokenization behavior for one decode operation.
CleanSpaces
Gets whether tokenizer-specific space cleanup should be applied.
RemoveSpecialTokens
Gets whether special tokens should be removed from decoded text.
UnparseSpecialTokens
Gets whether special tokens should be emitted as their raw token text.
GgufPreTokenizerRegistryUAIX.LmRuntime.Tokenization
2 members
Provides a conservative allow-list for tokenizer.ggml.pre values implemented by this build.
GetSupportedNames
Retrieves the supported names from the configured tokenizer after validating the requested access.
Returns: An ordered read-only IReadOnlyList<string> result from GgufPreTokenizerRegistry.GetSupportedNames: Retrieves the supported names from the configured tokenizer after validating the requested access. Mutable internal collection aliases are not exposed through the returned contract.
IsSupported(string)
Determines whether a pre-tokenizer identifier is supported.
name- The exact ordinal name used for catalog lookup, canonical hashing, or diagnostic labeling as defined by the containing member.
Returns: True when the identifier is absent or explicitly supported.
GgufSpecialTokenMapUAIX.LmRuntime.Tokenization
6 members
Represents special token identifiers resolved from GGUF metadata.
BosTokenId
Gets the beginning-of-sequence token identifier.
EosTokenId
Gets the end-of-sequence token identifier.
PaddingTokenId
Gets the padding token identifier.
SeparatorTokenId
Gets the separator token identifier.
UnknownTokenId
Gets the unknown token identifier.
EnumerateKnownTokenIds
Enumerates the known token identifiers in stable source order without exposing mutable internal storage.
Returns: An ordered sequence containing the known token identifiers as produced by the validated operation.
GgufTokenUAIX.LmRuntime.Tokenization
4 members
Represents one GGUF vocabulary token.
Score
Gets the tokenizer score associated with the token.
Text
Gets the raw token text from the GGUF vocabulary.
TokenId
Gets the token identifier used by model embedding rows.
Type
Gets the token type associated with the token.
GgufTokenizerEngineFactoryUAIX.LmRuntime.Tokenization
1 member
Selects a concrete tokenizer engine from validated GGUF tokenizer metadata.
Create(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Creates the GGUF tokenizer engine from the validated inputs required by GgufTokenizerEngineFactory.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Returns: The concrete tokenizer engine, with ownership and disposal obligations defined by the returned type and the Create contract.
GgufTokenizerFactoryUAIX.LmRuntime.Tokenization
2 members
Creates strict, metadata-routed tokenizers for parsed GGUF artifacts.
Create(UAIX.LmRuntime.Gguf.GgufModel)
Creates the tokenizer from the validated inputs required by GgufTokenizerFactory.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The ITokenizer result produced by GgufTokenizerFactory.Create for this contract: Creates the tokenizer from the validated inputs required by GgufTokenizerFactory. It is published only after all documented validation and ownership transitions succeed.
CreateStrict(UAIX.LmRuntime.Gguf.GgufModel)
Creates a tokenizer after strict GGUF tokenizer metadata validation.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The selected tokenizer, with ownership and disposal obligations defined by the returned type and the CreateStrict contract.
GgufTokenizerFingerprintUAIX.LmRuntime.Tokenization
1 member
Computes a deterministic SHA-256 identity for model-facing GGUF tokenizer metadata.
Create(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Computes a canonical tokenizer fingerprint without treating decoded text as token-ID authority.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Returns: The text produced by GgufTokenizerFingerprint.Create for this contract: Computes a canonical tokenizer fingerprint without treating decoded text as token-ID authority. The returned string is detached from mutable caller storage and is not persisted by the operation.
GgufTokenizerMetadataUAIX.LmRuntime.Tokenization
21 members
Captures tokenizer metadata loaded from a GGUF model.
AddBos
Gets whether model-defined BOS insertion is enabled.
AddedTokens
Gets the added tokens from GGUF metadata.
AddEos
Gets whether model-defined EOS insertion is enabled.
AddSeparator
Gets whether model-defined separator insertion is enabled.
AddSpacePrefix
Gets whether a leading space prefix is added before text fragments.
ChatTemplate
Gets the chat template from GGUF metadata, when present.
CleanSpaces
Gets whether detokenization should clean spaces around punctuation.
EscapeWhitespaces
Gets whether whitespace characters are escaped using SentencePiece whitespace notation.
HuggingFaceTokenizerJson
Gets the embedded Hugging Face tokenizer JSON, when present.
Merges
Gets the BPE merge rules from GGUF metadata.
PrecompiledCharsMap
Gets the optional precompiled SentencePiece normalization character map.
PreTokenizer
Gets the tokenizer pre-tokenizer name, when present.
RemoveExtraWhitespaces
Gets whether tokenizer-specific extra whitespace removal is enabled.
ScoresPresent
Gets a value indicating whether tokenizer.ggml.scores was present.
SourceScoreCount
Gets the source score-array length, or zero when the metadata key was absent.
SourceTokenTypeCount
Gets the source token-type-array length, or zero when the metadata key was absent.
SpecialTokens
Gets the special token identifiers.
TokenizerModel
Gets the tokenizer model name from GGUF metadata.
Tokens
Gets the vocabulary tokens indexed by token identifier.
TokenTypesPresent
Gets a value indicating whether tokenizer.ggml.token_type was present.
VocabularySize
Gets the effective vocabulary size from the token array.
GgufTokenizerMetadataReaderUAIX.LmRuntime.Tokenization
2 members
Builds tokenizer metadata from a parsed GGUF artifact.
Read(UAIX.LmRuntime.Gguf.GgufModel)
Reads tokenizer metadata without throwing for validation failures.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The GgufTokenizerMetadata result produced by GgufTokenizerMetadataReader.Read for this contract: Reads tokenizer metadata without throwing for validation failures. It is published only after all documented validation and ownership transitions succeed.
ReadStrict(UAIX.LmRuntime.Gguf.GgufModel)
Reads tokenizer metadata and throws when validation fails.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The GgufTokenizerMetadata result produced by GgufTokenizerMetadataReader.ReadStrict for this contract: Reads tokenizer metadata and throws when validation fails. It is published only after all documented validation and ownership transitions succeed.
GgufTokenizerMetadataValidationResultUAIX.LmRuntime.Tokenization
2 members
Represents tokenizer metadata validation output.
Diagnostics
Gets validation diagnostics.
IsValid
Gets a value indicating whether no diagnostics were emitted.
GgufTokenizerMetadataValidatorUAIX.LmRuntime.Tokenization
1 member
Validates GGUF tokenizer metadata before runtime use.
Validate(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Validates the supplied metadata against the invariants required by GgufTokenizerMetadataValidator.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Returns: The GgufTokenizerMetadataValidationResult result produced by GgufTokenizerMetadataValidator.Validate for this contract: Validates the supplied metadata against the invariants required by GgufTokenizerMetadataValidator. It is published only after all documented validation and ownership transitions succeed.
GgufTokenTypeUAIX.LmRuntime.Tokenization
6 members
Identifies the tokenizer token type stored in GGUF metadata.
Byte
Byte-fallback token.
Control
Control token.
Normal
Normal token text.
Unknown
Unknown token.
Unused
Unused token slot.
UserDefined
User-defined token.
Gpt2BpeTokenizerUAIX.LmRuntime.Tokenization
10 members
Executes the GPT-2 byte-level BPE tokenizer path from GGUF vocabulary and merge metadata.
Metadata
Name
CountTokens(string)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
text- The text to process using the configured encoding and normalization rules.
Returns: The int value computed by Gpt2BpeTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by Gpt2BpeTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Encode(string,UAIX.LmRuntime.Tokenization.TokenizationOptions)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
options- The optional TokenizationOptions controlling Encode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The TokenizationResult result produced by Gpt2BpeTokenizer.Encode for this contract: Encodes the supplied text with the configured tokenizer and validated special-token policy. It is published only after all documented validation and ownership transitions succeed.
Gpt2BpeTokenizer(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Initializes the tokenizer from validated GGUF metadata.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text to process using the configured encoding and normalization rules.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
Gpt2BpeTokenizerEngineUAIX.LmRuntime.Tokenization
4 members
Implements GPT-2 byte-level BPE from GGUF vocabulary and merge metadata.
Name
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
EncodeRaw(string,UAIX.LmRuntime.Tokenization.TokenizerFragmentContext,System.Collections.Generic.IList<int>,System.Collections.Generic.IList<string>)
Encodes the raw with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
context- The context that supplies session-scoped identity and boundary state; it is validated before dependent work begins.
destination- The destination buffer that receives the produced values.
trace- The trace sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Gpt2BpeTokenizerEngine(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Initializes a GPT-2 BPE engine from validated GGUF tokenizer metadata.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
HuggingFaceTokenizerJsonAdapterUAIX.LmRuntime.Tokenization
1 member
Provides an optional seam for embedded Hugging Face tokenizer JSON metadata.
Create(string)
Creates a tokenizer from embedded tokenizer JSON metadata when supported.
json- The json text consumed by HuggingFaceTokenizerJsonAdapter.Create; null, emptiness, length, encoding, identifier, or path rules are enforced as documented, and the value is not persisted by this operation.
Returns: A tokenizer instance, with ownership and disposal obligations defined by the returned type and the Create contract.
IGgufTokenizerUAIX.LmRuntime.Tokenization
3 members
Encodes and decodes text for a GGUF-backed model.
Metadata
Gets the tokenizer metadata used by this tokenizer.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,UAIX.LmRuntime.Tokenization.TokenizationOptions)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
options- The optional TokenizationOptions controlling Encode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The TokenizationResult result produced by IGgufTokenizer.Encode for this contract: Encodes the supplied text with the configured tokenizer and validated special-token policy. It is published only after all documented validation and ownership transitions succeed.
IGgufTokenizerEngineUAIX.LmRuntime.Tokenization
3 members
Defines a family-specific tokenizer engine that operates after special-token partitioning.
Name
Gets the stable engine name.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
EncodeRaw(string,UAIX.LmRuntime.Tokenization.TokenizerFragmentContext,System.Collections.Generic.IList<int>,System.Collections.Generic.IList<string>)
Encodes a raw-text fragment into model token identifiers.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
context- The context that supplies session-scoped identity and boundary state; it is validated before dependent work begins.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
trace- The trace sequence used by this operation; its required length, ordering, and element bounds are validated before access.
IGgufTokenizerFactoryUAIX.LmRuntime.Tokenization
1 member
Creates tokenizer instances from GGUF tokenizer metadata.
Create(UAIX.LmRuntime.Gguf.GgufModel)
Creates a tokenizer for a parsed GGUF model.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The tokenizer selected from metadata, with ownership and disposal obligations defined by the returned type and the Create contract.
InvalidGgufTokenizerExceptionUAIX.LmRuntime.Tokenization
1 member
Thrown when GGUF tokenizer metadata is invalid.
InvalidGgufTokenizerException(string)
Initializes a new InvalidGgufTokenizerException instance with validated dependencies and operational bounds.
message- The display-safe diagnostic message describing the failure without embedding prompt text, generated text, credentials, or private file contents.
InvalidUtf16PolicyUAIX.LmRuntime.Tokenization
2 members
Defines how tokenizer entry points handle invalid UTF-16 surrogate sequences.
Reject
Rejects invalid UTF-16 before tokenizer-specific normalization or segmentation.
Replace
Replaces each invalid surrogate code unit with the Unicode replacement character.
MetadataDrivenGgufTokenizerUAIX.LmRuntime.Tokenization
12 members
Executes a GGUF tokenizer by combining special-token partitioning with a family-specific tokenizer engine.
Real GGUF execution never falls back to whitespace tokenization. Unsupported tokenizer families fail during construction so token identifiers cannot silently diverge from the model embedding table.
Metadata
Name
CountTokens(string)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
text- The text to process using the configured encoding and normalization rules.
Returns: The int value computed by MetadataDrivenGgufTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by MetadataDrivenGgufTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.MetadataDrivenGgufTokenizerDetokenizationOptions)
Decodes model token identifiers using the stable metadata-driven compatibility options contract.
tokenIds- The token identifiers to process in sequence order.
options- The optional MetadataDrivenGgufTokenizerDetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Encode(string,UAIX.LmRuntime.Tokenization.TokenizationOptions)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
options- The optional TokenizationOptions controlling Encode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The TokenizationResult result produced by MetadataDrivenGgufTokenizer.Encode for this contract: Encodes the supplied text with the configured tokenizer and validated special-token policy. It is published only after all documented validation and ownership transitions succeed.
MetadataDrivenGgufTokenizer(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Initializes a tokenizer from validated GGUF metadata.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
MetadataDrivenGgufTokenizer(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata,UAIX.LmRuntime.Tokenization.IGgufTokenizerEngine)
Initializes a tokenizer with an explicitly selected family engine.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
engine- The validated IGgufTokenizerEngine dependency consumed by MetadataDrivenGgufTokenizer; ownership and lifetime remain with the caller unless this member explicitly documents a transfer.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text to process using the configured encoding and normalization rules.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
MetadataDrivenGgufTokenizerDetokenizationOptionsUAIX.LmRuntime.Tokenization
4 members
Provides a stable LocalEndpoint-facing name for metadata-driven GGUF detokenization controls.
This compatibility type mirrors UAIX.LmRuntime.Tokenization.DetokenizationOptions without inheritance because the canonical options type is sealed. It allows integration code to use a descriptive contract while the tokenizer retains one canonical internal representation.
CleanSpaces
Gets whether tokenizer-specific space cleanup should be applied.
RemoveSpecialTokens
Gets whether special tokens should be removed from decoded text.
UnparseSpecialTokens
Gets whether special tokens should be emitted as their raw token text.
ToDetokenizationOptions
Creates the canonical detokenization options consumed by the tokenizer engine.
Returns: A new canonical options instance with the same behavior flags.
RwkvWorldTokenizerUAIX.LmRuntime.Tokenization
6 members
Marks the RWKV tokenizer family as an explicit unsupported boundary until a dedicated engine is implemented.
Name
CountTokens(string)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
text- The text to process using the configured encoding and normalization rules.
Returns: The int value computed by RwkvWorldTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by RwkvWorldTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text to process using the configured encoding and normalization rules.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
SentencePieceBpeTokenizerEngineUAIX.LmRuntime.Tokenization
4 members
Implements the SentencePiece-BPE execution path used by LLaMA-style GGUF vocabularies.
Name
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
EncodeRaw(string,UAIX.LmRuntime.Tokenization.TokenizerFragmentContext,System.Collections.Generic.IList<int>,System.Collections.Generic.IList<string>)
Encodes the raw with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
context- The context that supplies session-scoped identity and boundary state; it is validated before dependent work begins.
destination- The destination buffer that receives the produced values.
trace- The trace sequence used by this operation; its required length, ordering, and element bounds are validated before access.
SentencePieceBpeTokenizerEngine(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Initializes the engine from validated GGUF tokenizer metadata.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
SentencePieceGgufTokenizerUAIX.LmRuntime.Tokenization
10 members
Executes the SentencePiece-BPE tokenizer path used by LLaMA-style GGUF artifacts.
Metadata
Name
CountTokens(string)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
text- The text to process using the configured encoding and normalization rules.
Returns: The int value computed by SentencePieceGgufTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by SentencePieceGgufTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.DetokenizationOptions)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
options- The optional DetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Encode(string,UAIX.LmRuntime.Tokenization.TokenizationOptions)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
options- The optional TokenizationOptions controlling Encode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The TokenizationResult result produced by SentencePieceGgufTokenizer.Encode for this contract: Encodes the supplied text with the configured tokenizer and validated special-token policy. It is published only after all documented validation and ownership transitions succeed.
SentencePieceGgufTokenizer(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Initializes the tokenizer from validated GGUF metadata.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text to process using the configured encoding and normalization rules.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
SpecialTokenFragmentUAIX.LmRuntime.Tokenization
7 members
Represents one fragment emitted by special-token partitioning.
Kind
Gets the fragment kind.
Length
Gets the fragment length in source text characters.
Offset
Gets the character offset in the source text.
Text
Gets the raw text fragment.
TokenId
Gets the token identifier for token fragments.
Raw(string,int)
Creates a raw-text fragment representing an unmodified source slice at the supplied offset.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
offset- The zero-based offset into the relevant source or destination; range validation occurs before access.
Returns: The SpecialTokenFragment result produced by SpecialTokenFragment.Raw for this contract: Creates a raw-text fragment representing an unmodified source slice at the supplied offset. It is published only after all documented validation and ownership transitions succeed.
Token(int,string,int)
Creates a special-token fragment at the supplied source-text offset.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
offset- The zero-based offset into the relevant source or destination; range validation occurs before access.
Returns: The SpecialTokenFragment result produced by SpecialTokenFragment.Token for this contract: Creates a special-token fragment at the supplied source-text offset. It is published only after all documented validation and ownership transitions succeed.
SpecialTokenFragmentKindUAIX.LmRuntime.Tokenization
2 members
Identifies the type of fragment emitted by special-token partitioning.
RawText
A raw text fragment that must be processed by the tokenizer engine.
Token
A pre-resolved token identifier fragment.
SpecialTokenMapUAIX.LmRuntime.Tokenization
4 members
Represents model special-token identities.
BeginningOfSequence
Gets the beginning-of-sequence token identifier.
EndOfSequence
Gets the end-of-sequence token identifier.
Padding
Gets the padding token identifier.
Unknown
Gets the unknown token identifier.
SpecialTokenPartitionerUAIX.LmRuntime.Tokenization
1 member
Partitions raw text around tokenizer special tokens before normal tokenization.
Partition(string,UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata,bool)
Partitions text around known special tokens using longest-token-first matching.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
parseSpecial- Whether control and unknown tokens should be parsed as special tokens.
Returns: An ordered read-only IReadOnlyList<SpecialTokenFragment> result from SpecialTokenPartitioner.Partition: Partitions text around known special tokens using longest-token-first matching. Mutable internal collection aliases are not exposed through the returned contract.
StreamingUtf8TokenDecoderUAIX.LmRuntime.Tokenization
2 members
Incrementally decodes byte-token payloads without corrupting UTF-8 sequences split across token boundaries.
Decode(System.ReadOnlySpan<byte>,bool)
Decodes one byte fragment and retains incomplete UTF-8 state for the next call.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
flush- The flush flag selecting the documented branch of Decode; it does not grant authority beyond this in-memory operation.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Reset
Resets the requested state to its validated initial state without publishing partial state.
TokenBudgetTruncatorUAIX.LmRuntime.Tokenization
1 member
Truncates message sequences by token budget.
TruncateMessages(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>,UAIX.LmRuntime.Abstractions.ITokenizer,int)
Truncates messages so the total token count does not exceed the budget.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
tokenizer- The validated ITokenizer dependency consumed by TruncateMessages; ownership and lifetime remain with the caller unless this member explicitly documents a transfer.
maxTokens- The numeric max tokens consumed by TruncateMessages; it must satisfy the member's documented range, geometry, and finite-value requirements.
Returns: An ordered read-only IReadOnlyList<LlmMessage> result from TokenBudgetTruncator.TruncateMessages: Truncates messages so the total token count does not exceed the budget. Mutable internal collection aliases are not exposed through the returned contract.
TokenizationOptionsUAIX.LmRuntime.Tokenization
6 members
Describes tokenization behavior for one encode operation.
AddSpecialTokens
Gets whether model-defined special tokens should be added.
EmitTrace
Gets whether content-minimized trace data should be emitted for parity diagnostics.
InvalidUtf16Policy
Gets the policy for invalid UTF-16 surrogate sequences.
OverrideAddBos
Gets an optional override for BOS insertion.
OverrideAddEos
Gets an optional override for EOS insertion.
ParseSpecialTokens
Gets whether raw special-token text should be parsed as special tokens.
TokenizationResultUAIX.LmRuntime.Tokenization
2 members
Represents the output of a tokenizer encode operation.
TokenIds
Gets the emitted token identifiers.
Trace
Gets optional content-minimized events used for tokenizer parity diagnostics.
TokenizerFragmentContextUAIX.LmRuntime.Tokenization
2 members
Describes the position of one raw-text fragment within special-token partitioning.
IsFirstFragment
Gets a value indicating whether this is the first raw-text fragment in the input.
PreviousFragmentWasSpecial
Gets a value indicating whether the immediately preceding fragment was a special token.
TokenizerGoldenCorpusUAIX.LmRuntime.Tokenization
1 member
Loads tokenizer golden corpora.
Load(string)
Loads ordered tokenizer golden record collection from a verified local source into TokenizerGoldenCorpus.
json- The json text consumed by TokenizerGoldenCorpus.Load; null, emptiness, length, encoding, identifier, or path rules are enforced as documented, and the value is not persisted by this operation.
Returns: An ordered read-only IReadOnlyList<TokenizerGoldenRecord> result from TokenizerGoldenCorpus.Load: Loads ordered tokenizer golden record collection from a verified local source into TokenizerGoldenCorpus. Mutable internal collection aliases are not exposed through the returned contract.
TokenizerGoldenRecordUAIX.LmRuntime.Tokenization
2 members
Represents one tokenizer golden record.
ExpectedTokenIds
Gets expected token identifiers.
Text
Gets the source text.
TokenizerParityReportUAIX.LmRuntime.Tokenization
1 member
Represents tokenizer parity diagnostics.
Mismatches
Gets tokenizer mismatches.
TokenizerTextSafetyUAIX.LmRuntime.Tokenization
1 member
Validates and normalizes managed strings before tokenizer-specific processing.
NormalizeUtf16(string,UAIX.LmRuntime.Tokenization.InvalidUtf16Policy)
Validates a managed string and optionally replaces unpaired surrogate code units.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
policy- The policy that define validation limits and execution behavior; required values are checked before use.
Returns: The original string when valid, or a normalized replacement string when requested.
TokenizerVocabularyReconcilerUAIX.LmRuntime.Tokenization
1 member
Reconciles embedded Hugging Face vocabulary identifiers against authoritative GGUF token-array order.
Reconcile(UAIX.LmRuntime.Tokenization.GgufTokenizerMetadata)
Validates embedded tokenizer JSON without allowing it to reorder GGUF token identifiers.
metadata- The metadata containing validated format or tokenizer metadata required by this operation.
Returns: The TokenizerVocabularyReconciliationResult result produced by TokenizerVocabularyReconciler.Reconcile for this contract: Validates embedded tokenizer JSON without allowing it to reorder GGUF token identifiers. It is published only after all documented validation and ownership transitions succeed.
TokenizerVocabularyReconciliationResultUAIX.LmRuntime.Tokenization
3 members
Describes a consistency check between GGUF vocabulary order and embedded Hugging Face tokenizer JSON.
Diagnostics
Gets bounded deterministic diagnostics.
EmbeddedJsonPresent
Gets whether embedded tokenizer JSON was present.
IsConsistent
Gets whether the embedded tokenizer JSON is absent or consistent with GGUF token identifiers.
UnsupportedTokenizerExceptionUAIX.LmRuntime.Tokenization
1 member
Thrown when a GGUF tokenizer family is not supported.
UnsupportedTokenizerException(string)
Initializes a new UnsupportedTokenizerException instance with validated dependencies and operational bounds.
message- The display-safe diagnostic message describing the failure without embedding prompt text, generated text, credentials, or private file contents.
WhitespaceTokenizerUAIX.LmRuntime.Tokenization
6 members
Provides a deterministic tokenizer for tests, examples, and fallback token budgeting.
Name
CountTokens(string)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
text- The text to process using the configured encoding and normalization rules.
Returns: The int value computed by WhitespaceTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. Range, finite-value, and overflow checks are completed before the value is returned.
CountTokens(System.Collections.Generic.IEnumerable<UAIX.LmRuntime.Contracts.LlmMessage>)
Counts the tokens using the same deterministic rules as the corresponding processing operation.
messages- The messages sequence used by this operation; its required length, ordering, and element bounds are validated before access.
Returns: The TokenCountResult result produced by WhitespaceTokenizer.CountTokens for this contract: Counts the tokens using the same deterministic rules as the corresponding processing operation. It is published only after all documented validation and ownership transitions succeed.
Decode(System.Collections.Generic.IEnumerable<int>)
Decodes the supplied token sequence with the configured tokenizer while preserving sequence order.
tokenIds- The token identifiers to process in sequence order.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Encode(string,bool,bool)
Encodes the supplied text with the configured tokenizer and validated special-token policy.
text- The text to process using the configured encoding and normalization rules.
addBos- A value indicating whether add BOS applies to this operation.
addEos- A value indicating whether add EOS applies to this operation.
Returns: An ordered read-only collection of token identifiers produced by the configured tokenizer.
Tokenize(string)
Tokenizes the supplied text with the configured metadata and preserves deterministic token order.
text- The text to process using the configured encoding and normalization rules.
Returns: An ordered read-only collection of token text values produced by the configured tokenizer.
UAIX.LmRuntime.Models.Llama
102 documented types and 605 documented members.
ArrayLlamaLayerWeightSourceUAIX.LmRuntime.Models.Llama
10 members
Provides one array-backed LLaMA layer weight source.
AttentionKey
AttentionNorm
AttentionOutput
AttentionQuery
AttentionValue
FeedForwardDown
FeedForwardGate
FeedForwardNorm
FeedForwardUp
ArrayLlamaLayerWeightSource(int,UAIX.LmRuntime.Models.Llama.LlamaReferenceLayerWeights,UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Initializes a new ArrayLlamaLayerWeightSource instance with validated dependencies and operational bounds.
blockIndex- The zero-based block index; it must identify an existing position within the relevant validated range.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
ArrayLlamaModelWeightSourceUAIX.LmRuntime.Models.Llama
10 members
Adapts the v1.8.0 float-array model to the storage-neutral v1.9.0 execution contracts.
Layers
ManagedCopiedByteCount
OutputNorm
OutputProjection
StorageDiagnostics
StorageSummary
TokenEmbeddings
UsesTiedOutputProjection
ArrayLlamaModelWeightSource(UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.LlamaReferenceModelWeights)
Initializes a new ArrayLlamaModelWeightSource instance with validated dependencies and operational bounds.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
Create(UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.LlamaReferenceModelWeights)
Creates an array-backed source after validating its complete model contract.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
Returns: The array-backed source, with ownership and disposal obligations defined by the returned type and the Create contract.
ArrayMatrixSourceUAIX.LmRuntime.Models.Llama
9 members
Provides an immutable row-major array-backed matrix adapter for compatibility and deterministic fixtures.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
TensorName
Gets the semantic tensor name.
ArrayMatrixSource(string,float[],int,int)
Initializes a new ArrayMatrixSource instance with validated dependencies and operational bounds.
tensorName- The exact ordinal GGUF tensor catalog name used for lookup and diagnostics.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
rowCount- The row count used to bound this operation; it must be nonnegative and within the supported range.
columnCount- The column count used to bound this operation; it must be nonnegative and within the supported range.
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
ArrayVectorSourceUAIX.LmRuntime.Models.Llama
7 members
Provides an immutable array-backed vector adapter for compatibility and deterministic fixtures.
DataType
Length
StorageDiagnostics
StorageType
TensorName
Gets the semantic tensor name.
ArrayVectorSource(string,float[])
Initializes a new ArrayVectorSource instance with validated dependencies and operational bounds.
tensorName- The exact ordinal GGUF tensor catalog name used for lookup and diagnostics.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
CopyTo(System.Span<float>)
Copies the to into caller-owned storage after validating the requested range and capacity.
destination- The destination buffer that receives the produced values.
FixtureDirectoryVerifierUAIX.LmRuntime.Models.Llama
1 member
Verifies fixture manifests, artifact paths, digests, and basic loadability without network access.
Verify(string)
Verifies the supplied fixture directory and returns bounded evidence only after every required check succeeds.
fixtureDirectory- The directory containing a fixture manifest and GGUF artifact.
Returns: The FixtureVerificationResult result produced by FixtureDirectoryVerifier.Verify for this contract: Verifies the supplied fixture directory and returns bounded evidence only after every required check succeeds. It is published only after all documented validation and ownership transitions succeed.
FixtureVerificationDiagnosticUAIX.LmRuntime.Models.Llama
2 members
Represents one diagnostic emitted while verifying a checked-in GGUF fixture directory.
Code
Gets the stable diagnostic code.
Message
Gets the diagnostic message.
FixtureVerificationResultUAIX.LmRuntime.Models.Llama
5 members
Represents the result of bounded, offline fixture directory verification.
ArtifactPath
Gets the normalized GGUF artifact path.
ArtifactSha256
Gets the verified SHA-256 digest.
Diagnostics
Gets verification diagnostics.
FixtureDirectory
Gets the normalized fixture directory.
IsValid
Gets whether no verification diagnostics were emitted.
GroupedQueryAttentionMapUAIX.LmRuntime.Models.Llama
1 member
Maps query heads to grouped key/value heads.
MapHead(int,int,int)
Maps an attention query head to the corresponding KV head.
queryHead- The zero-based query-head index mapped deterministically to its corresponding key/value head.
queryHeadCount- The query head count used to bound this operation; it must be nonnegative and within the supported range.
keyValueHeadCount- The key value head count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The int value computed by GroupedQueryAttentionMap.MapHead for this contract: Maps an attention query head to the corresponding KV head. Range, finite-value, and overflow checks are completed before the value is returned.
ILlamaLayerWeightSourceUAIX.LmRuntime.Models.Llama
9 members
Exposes immutable weights required by one LLaMA transformer block.
AttentionKey
Gets the key projection matrix.
AttentionNorm
Gets the attention normalization vector.
AttentionOutput
Gets the attention output projection matrix.
AttentionQuery
Gets the query projection matrix.
AttentionValue
Gets the value projection matrix.
FeedForwardDown
Gets the feed-forward down projection matrix.
FeedForwardGate
Gets the feed-forward gate projection matrix.
FeedForwardNorm
Gets the feed-forward normalization vector.
FeedForwardUp
Gets the feed-forward up projection matrix.
ILlamaModelWeightSourceUAIX.LmRuntime.Models.Llama
8 members
Exposes immutable model weights required by the deterministic LLaMA reference session.
Layers
Gets transformer-block weights in execution order.
ManagedCopiedByteCount
Gets persistent managed model-weight bytes represented by this source.
OutputNorm
Gets the final output normalization vector.
OutputProjection
Gets the output projection matrix.
StorageDiagnostics
Gets storage diagnostics for every distinct semantic source.
StorageSummary
Gets a stable summary of physical storage types used by the model.
TokenEmbeddings
Gets the token embedding table.
UsesTiedOutputProjection
Gets a value indicating whether output projection aliases token embeddings.
ILlamaSessionUAIX.LmRuntime.Models.Llama
1 member
Defines the lifecycle for a LLaMA-family inference session.
DecodeAsync(int,System.Threading.CancellationToken)
Decodes the next token for the active sequence.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: An asynchronous ValueTask<int> that completes with the result of ILlamaSession.DecodeAsync: Decodes the next token for the active sequence. Fault and cancellation states are propagated without a successful partial result.
IReadOnlyMatrixSourceUAIX.LmRuntime.Models.Llama
7 members
Exposes an immutable logical row-major matrix without requiring a particular storage representation.
ColumnCount
Gets the logical column count.
DataType
Gets the logical runtime data type.
RowCount
Gets the logical row count.
StorageDiagnostics
Gets immutable storage diagnostics.
StorageType
Gets the physical GGML storage type.
CopyRowTo(int,System.Span<float>)
Copies and, when required, dequantizes one logical row into a caller-owned float32 destination.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies this matrix by a float32 vector without materializing a complete float32 matrix.
vector- The input vector with at least ColumnCount values.
output- The output buffer with at least RowCount values.
IReadOnlyVectorSourceUAIX.LmRuntime.Models.Llama
5 members
Exposes an immutable logical vector without requiring a particular storage representation.
DataType
Gets the logical runtime data type.
Length
Gets the logical vector length.
StorageDiagnostics
Gets immutable storage diagnostics.
StorageType
Gets the physical GGML storage type.
CopyTo(System.Span<float>)
Copies every vector value into a caller-owned float32 destination.
destination- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
IReferenceKvCacheUAIX.LmRuntime.Models.Llama
13 members
Defines a typed, deterministic key/value cache contract for the scalar LLaMA reference runtime.
ConfigurationFingerprint
Gets the configuration fingerprint required by compatible snapshots.
ContextLength
Gets the maximum sequence capacity.
HeadWidth
Gets the float width of one key/value head.
KeyValueHeadCount
Gets the number of key/value heads per layer.
LayerCount
Gets the number of transformer layers.
UsedTokenCount
Gets the highest contiguous token position written plus one.
WriteBehavior
Gets the deterministic append-versus-overwrite behavior.
CreateSnapshot
Creates a bounded snapshot for tiny-fixture testing and replay.
Returns: The immutable cache snapshot, with ownership and disposal obligations defined by the returned type and the CreateSnapshot contract.
GetKey(int,int,int)
Retrieves the key from the current cache state after validating the requested access.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
headIndex- The zero-based head index; it must identify an existing position within the relevant validated range.
Returns: The bounded ReadOnlySpan<float> view produced by IReferenceKvCache.GetKey: Retrieves the key from the current cache state after validating the requested access. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
GetValue(int,int,int)
Retrieves the value from the current cache state after validating the requested access.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
headIndex- The zero-based head index; it must identify an existing position within the relevant validated range.
Returns: The bounded ReadOnlySpan<float> view produced by IReferenceKvCache.GetValue: Retrieves the value from the current cache state after validating the requested access. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
Reset
Resets the requested state to its validated initial state without publishing partial state.
Restore(UAIX.LmRuntime.Models.Llama.ReferenceKvCacheSnapshot)
Restores the supplied snapshot from a validated persisted representation.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
Write(int,int,System.ReadOnlySpan<float>,System.ReadOnlySpan<float>)
Appends or replaces one layer's key and value vectors at a sequence position.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
key- The flattened key vector for all key/value heads.
value- The flattened value vector for all key/value heads.
LlamaBoundLayerWeightSetUAIX.LmRuntime.Models.Llama
10 members
Represents the mapped tensors required by one LLaMA transformer block.
AttentionKey
Gets the key projection tensor.
AttentionNorm
Gets the attention normalization tensor.
AttentionOutput
Gets the attention output projection tensor.
AttentionQuery
Gets the query projection tensor.
AttentionValue
Gets the value projection tensor.
BlockIndex
Gets the zero-based transformer block index.
FeedForwardDown
Gets the feed-forward down projection tensor.
FeedForwardGate
Gets the feed-forward gate projection tensor.
FeedForwardNorm
Gets the feed-forward normalization tensor.
FeedForwardUp
Gets the feed-forward up projection tensor.
LlamaBoundTensorUAIX.LmRuntime.Models.Llama
5 members
Represents one semantic LLaMA weight bound to mapped model storage.
Binding
Gets the validated binding manifest entry.
BlockIndex
Gets the optional transformer block index.
Role
Gets the semantic tensor role.
StorageMode
Gets the storage mode represented by this binding.
View
Gets the borrowed mapped tensor view.
LlamaBoundWeightSetUAIX.LmRuntime.Models.Llama
11 members
Resolves a complete LLaMA binding manifest into stable mapped tensor views.
This object does not own the operating-system mapping. Every view borrows storage from the supplied UAIX.LmRuntime.Gguf.MappedGgufFile and becomes invalid when that mapping is disposed.
Bindings
Gets all semantic mapped tensor bindings.
Configuration
Gets the validated model configuration.
Layers
Gets the block-local mapped weight sets.
Manifest
Gets the complete tensor binding manifest.
Mapping
Gets the mapping that owns all borrowed tensor bytes.
Output
Gets the output projection tensor or tied embedding alias.
OutputNorm
Gets the final output normalization tensor.
TokenEmbeddings
Gets the token embedding tensor.
Get(UAIX.LmRuntime.Models.Llama.LlamaTensorRole,System.Nullable<int>)
Retrieves the llama bound tensor from the current LlamaBoundWeightSet state after validating the requested access.
role- The semantic LLaMA tensor role used to select the required bound tensor from the validated manifest.
blockIndex- The zero-based block index; it must identify an existing position within the relevant validated range.
Returns: The LlamaBoundTensor result produced by LlamaBoundWeightSet.Get for this contract: Retrieves the llama bound tensor from the current LlamaBoundWeightSet state after validating the requested access. It is published only after all documented validation and ownership transitions succeed.
LlamaBoundWeightSet(UAIX.LmRuntime.Gguf.MappedGgufFile,UAIX.LmRuntime.Models.Llama.TensorBindingManifest,UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Initializes a mapped LLaMA weight set from a complete binding manifest.
mapping- The mapped GGUF file that owns tensor storage.
manifest- The validated manifest that binds tensor requirements, model identity, and storage diagnostics used by the operation.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
MaterializeFloat32ReferenceWeights(int)
Materializes bounded float32 arrays for the scalar correctness runtime.
maximumCopiedBytes- The maximum total bytes that may be copied from mapped storage.
Returns: The immutable reference weights and explicit copy ledger.
LlamaGeneratedTokenUAIX.LmRuntime.Models.Llama
4 members
Describes one token selected during deterministic greedy generation.
The value contains only a zero-based sequence number, token identifier, and selected logit. It does not contain prompt text, decoded output, model bytes, file paths, persistent state, or provider information.
SelectedLogit
Gets the selected token's deterministic argmax logit.
Sequence
Gets the zero-based token-selection sequence.
TokenId
Gets the selected model vocabulary identifier.
LlamaGeneratedToken(int,int,float)
Initializes a new LlamaGeneratedToken instance with validated dependencies and operational bounds.
sequence- The zero-based selection sequence within the current generation operation.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
selectedLogit- The deterministic selected logit at this generation sequence position. NaN is rejected; infinities remain valid because UAIX.LmRuntime.Sampling.GreedySampler compares them using the same deterministic ordering as finite values.
LlamaGenerationStopReasonUAIX.LmRuntime.Models.Llama
5 members
Identifies why deterministic greedy generation stopped.
Cancelled
Cooperative cancellation was observed between committed inference steps.
ContextCapacity
The model context window could not accept another evaluated token.
EndOfSequence
The configured end-of-sequence token was selected.
MaximumTokens
The requested maximum number of tokens was produced.
StopToken
A caller-configured stop token was selected.
LlamaGreedyGenerationOptionsUAIX.LmRuntime.Models.Llama
4 members
Defines allocation-bounded deterministic greedy generation controls.
EndOfSequenceTokenId
Gets the optional end-of-sequence token identifier.
MaximumTokens
Gets the maximum number of generated tokens.
ResetSession
Gets whether the session is reset before prompt prefill.
StopTokenIds
Gets additional token identifiers that terminate generation after being emitted.
LlamaGreedyGenerationResultUAIX.LmRuntime.Models.Llama
5 members
Describes an allocation-bounded greedy generation operation.
FinalSelectedLogit
Gets the selected logit of the final generated token, or negative infinity when none was generated.
GeneratedTokenCount
Gets the number of generated token identifiers written to the caller buffer.
Position
Gets the next sequence position maintained by the session.
PromptTokenCount
Gets the number of prompt tokens evaluated for this operation.
StopReason
Gets the deterministic stop reason.
LlamaGreedyTokenResultUAIX.LmRuntime.Models.Llama
5 members
Represents exactly one greedily selected token produced by the reference runtime.
Position
Gets the zero-based position whose logits selected this token.
PromptTokenCount
Gets the number of prompt tokens evaluated.
SelectedLogit
Gets the selected token logit.
TokenId
Gets the selected token identifier.
TokenText
Gets the selected token text when a tokenizer is attached.
LlamaLayerWeightsUAIX.LmRuntime.Models.Llama
3 members
Represents one transformer block's reference weights.
AttentionKey
Gets the attention key projection matrix.
AttentionQuery
Gets the attention query projection matrix.
AttentionValue
Gets the attention value projection matrix.
LlamaLogitComparatorUAIX.LmRuntime.Models.Llama
1 member
Compares deterministic next-token vectors without widening caller-provided tolerances.
Compare(System.Collections.Generic.IReadOnlyList<float>,System.Collections.Generic.IReadOnlyList<float>,UAIX.LmRuntime.Models.Llama.LlamaParityTolerance)
Compares two logit vectors using absolute-or-relative error acceptance.
reference- The reference sequence used by this operation; its required length, ordering, and element bounds are validated before access.
candidate- The candidate sequence used by this operation; its required length, ordering, and element bounds are validated before access.
tolerance- The tolerance input of type LlamaParityTolerance read by LlamaLogitComparator.Compare; it must satisfy the member-specific nullability, identity, range, and ownership rules before dependent work begins.
Returns: The LlamaLogitComparison result produced by LlamaLogitComparator.Compare for this contract: Compares two logit vectors using absolute-or-relative error acceptance. It is published only after all documented validation and ownership transitions succeed.
LlamaLogitComparisonUAIX.LmRuntime.Models.Llama
6 members
Summarizes a deterministic comparison of two next-token logit vectors.
FirstFailingCandidateValue
Gets the candidate value at the first failing index.
FirstFailingIndex
Gets the first failing logit index, or when none failed.
FirstFailingReferenceValue
Gets the reference value at the first failing index.
IsWithinTolerance
Gets whether every compared logit satisfies the configured tolerance.
MaximumAbsoluteError
Gets the largest absolute logit difference.
MeanAbsoluteError
Gets the arithmetic mean absolute logit difference.
LlamaLogitComputerUAIX.LmRuntime.Models.Llama
1 member
Computes reference logits from a hidden state and output projection.
ComputeLogits(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>)
Computes logits from a hidden vector and a row-major projection matrix.
hiddenState- The hidden state sequence used by this operation; its required length, ordering, and element bounds are validated before access.
projection- The projection sequence used by this operation; its required length, ordering, and element bounds are validated before access.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
LlamaMappedGreedyTokenResultUAIX.LmRuntime.Models.Llama
20 members
Represents an end-to-end prompt-to-one-token result from a mapped GGUF model.
Architecture
Gets the model architecture identifier.
Evidence
Gets evidence statements for the deterministic one-token operation.
FinishReason
Gets the deterministic finish reason.
KeyValueCacheTokenCount
Gets the resulting key/value cache token count.
Logits
Gets the complete next-token logits for parity diagnostics.
ManagedAllocatedByteCount
Gets managed bytes allocated on the current thread during the measured operation.
ManagedModelWeightCopiedByteCount
Gets the managed model-weight bytes copied by the session path.
ModelName
Gets the model display name declared by GGUF metadata.
ModelPath
Gets the normalized GGUF model path used for the operation.
ModelSha256
Gets the optional complete-file model digest computed during load.
Position
Gets the sequence position that produced the logits.
Prompt
Gets the input prompt.
PromptTokenIds
Gets the exact prompt token identifiers.
RuntimeMode
Gets the runtime mode.
SelectedLogit
Gets the selected token logit.
StorageSummary
Gets the mapped storage-type summary.
Timings
Gets measured operation timings.
TokenId
Gets the selected token identifier.
TokenizerTrace
Gets tokenizer trace events when requested.
TokenText
Gets the selected token text.
LlamaMappedModelUAIX.LmRuntime.Models.Llama
14 members
Owns a mapped GGUF artifact and immutable LLaMA runtime composition.
BindingManifest
Gets the tensor binding manifest.
Configuration
Gets the validated LLaMA configuration.
IsDisposed
Gets whether the model has been disposed.
Manifest
Gets the immutable load evidence manifest.
Mapping
Gets the mapped model storage owner.
Options
Gets the load options retained for deterministic session creation.
Tokenizer
Gets the exact metadata-driven tokenizer.
TokenizerMetadata
Gets validated GGUF tokenizer metadata.
Weights
Gets the mapped semantic weight set.
WeightSource
Gets the direct mapped execution weight source.
CreateMaterializedReferenceSession
Creates an independent compatibility session over explicitly materialized float32 arrays.
Returns: The materialized compatibility session, with ownership and disposal obligations defined by the returned type and the CreateMaterializedReferenceSession contract.
CreateReferenceSession
Creates an independent scalar reference session with its own key/value state.
Returns: The new mapped reference session, with ownership and disposal obligations defined by the returned type and the CreateReferenceSession contract.
Dispose
Releases resources owned by LlamaMappedModel and transitions it to the disposed state.
GetReferenceMaterialization
Gets the bounded reference materialization evidence, creating it on first use.
Returns: The LlamaReferenceWeightMaterialization result produced by LlamaMappedModel.GetReferenceMaterialization for this contract: Gets the bounded reference materialization evidence, creating it on first use. It is published only after all documented validation and ownership transitions succeed.
LlamaMappedModelLoaderUAIX.LmRuntime.Models.Llama
1 member
Loads a local GGUF artifact into a mapped, tokenizer-aware LLaMA model composition.
Load(string,UAIX.LmRuntime.Models.Llama.LlamaMappedModelLoadOptions)
Loads and validates one mapped local model.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional LlamaMappedModelLoadOptions controlling Load; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The owned mapped model, with ownership and disposal obligations defined by the returned type and the Load contract.
LlamaMappedModelLoadOptionsUAIX.LmRuntime.Models.Llama
5 members
Configures loading of a mapped LLaMA GGUF artifact.
BindingOptions
Gets semantic tensor binding validation options.
ComputeModelSha256
Gets whether a SHA-256 digest of the complete artifact should be computed during load.
MaximumReferenceMaterializationBytes
Gets the maximum bytes that scalar reference sessions may copy from mapped F32 weights.
ParseOptions
Gets GGUF parser safety limits.
RuntimeMode
Gets the runtime mode.
LlamaMappedModelLoadTimingsUAIX.LmRuntime.Models.Llama
5 members
Records measured stages of mapped model loading.
CompositionDuration
Gets architecture, tokenizer, and binding composition duration.
HashDuration
Gets optional complete-file digest duration.
MapDuration
Gets operating-system memory-map creation duration.
ParseDuration
Gets metadata and tensor catalog parse duration.
TotalDuration
Gets total load duration.
LlamaMappedModelManifestUAIX.LmRuntime.Models.Llama
13 members
Describes the immutable evidence produced while loading a mapped LLaMA model.
Architecture
Gets the architecture identifier.
BoundTensorCount
Gets the bound tensor count.
Evidence
Gets load evidence messages.
GgufVersion
Gets the GGUF version.
ManagedModelWeightCopiedByteCount
Gets the managed model-weight byte count copied by the default execution path.
ModelByteCount
Gets the exact mapped GGUF file length observed during parsing.
ModelName
Gets the model display name.
ModelPath
Gets the normalized model path.
ModelSha256
Gets the optional complete-file SHA-256 digest.
RuntimeMode
Gets the selected execution mode.
StorageSummary
Gets the physical tensor storage summary used by direct mapped execution.
Timings
Gets load-stage timings.
Tokenizer
Gets the tokenizer implementation name.
LlamaMappedReferenceSessionUAIX.LmRuntime.Models.Llama
12 members
Combines exact GGUF tokenization with an independent scalar reference session.
IsDisposed
Gets whether this session has released its state.
KvCache
Gets the typed session-local key/value cache.
Position
Gets the current next-token sequence position.
DecodeOneGreedy(string,UAIX.LmRuntime.Models.Llama.LlamaOneTokenOptions)
Tokenizes a prompt, executes prefill, selects argmax, and decodes exactly one token.
prompt- The prompt processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
options- The optional LlamaOneTokenOptions controlling DecodeOneGreedy; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The LlamaMappedGreedyTokenResult result produced by LlamaMappedReferenceSession.DecodeOneGreedy for this contract: Tokenizes a prompt, executes prefill, selects argmax, and decodes exactly one token. It is published only after all documented validation and ownership transitions succeed.
Dispose
Releases resources owned by LlamaMappedReferenceSession and transitions it to the disposed state.
ExportState(UAIX.LmRuntime.Models.Llama.LlamaSessionPersistenceOptions)
Exports complete deterministic state with model, configuration, tokenizer, and cache-layout identities.
options- Optional persistence metadata. Empty identity fields are resolved from the mapped model.
Returns: A newly allocated byte[] containing the ordered result of LlamaMappedReferenceSession.ExportState: Exports complete deterministic state with model, configuration, tokenizer, and cache-layout identities. The caller owns the returned array and later mutation cannot alter the source object.
GenerateGreedy(string,System.Span<int>,System.Span<float>,UAIX.LmRuntime.Models.Llama.LlamaGreedyGenerationOptions,UAIX.LmRuntime.Tokenization.TokenizationOptions,System.Action<UAIX.LmRuntime.Models.Llama.LlamaGeneratedToken>,System.Threading.CancellationToken)
Tokenizes a prompt, generates greedy token identifiers, and reports each selected token synchronously.
prompt- The prompt processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
generatedTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
finalLogits- The final logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
generationOptions- The generation options that define validation limits and execution behavior; required values are checked before use.
tokenizationOptions- The tokenization options that define validation limits and execution behavior; required values are checked before use.
tokenObserver- The optional observer invoked once for each selected token.
cancellationToken- A token observed before selection and before the next committed model step.
Returns: The LlamaGreedyGenerationResult result produced by LlamaMappedReferenceSession.GenerateGreedy for this contract: Tokenizes a prompt, generates greedy token identifiers, and reports each selected token synchronously. It is published only after all documented validation and ownership transitions succeed.
GenerateGreedy(string,System.Span<int>,System.Span<float>,UAIX.LmRuntime.Models.Llama.LlamaGreedyGenerationOptions,UAIX.LmRuntime.Tokenization.TokenizationOptions,System.Threading.CancellationToken)
Tokenizes a prompt and generates greedy token identifiers into caller-owned buffers.
prompt- The prompt processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
generatedTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
finalLogits- The final logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
generationOptions- The generation options that define validation limits and execution behavior; required values are checked before use.
tokenizationOptions- The tokenization options that define validation limits and execution behavior; required values are checked before use.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The LlamaGreedyGenerationResult result produced by LlamaMappedReferenceSession.GenerateGreedy for this contract: Tokenizes a prompt and generates greedy token identifiers into caller-owned buffers. It is published only after all documented validation and ownership transitions succeed.
LoadState(string,int)
Loads and restores complete deterministic state from a local artifact.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The verified artifact, with ownership and disposal obligations defined by the returned type and the LoadState contract.
Reset
Resets this session's sequence and key/value state.
RestoreState(System.ReadOnlySpan<byte>,int)
Restores verified complete state after enforcing mapped model and tokenizer identities.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The LlamaSessionArtifact result produced by LlamaMappedReferenceSession.RestoreState for this contract: Restores verified complete state after enforcing mapped model and tokenizer identities. It is published only after all documented validation and ownership transitions succeed.
SaveState(string,UAIX.LmRuntime.Models.Llama.LlamaSessionPersistenceOptions)
Saves complete deterministic state to a local artifact.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
options- The optional LlamaSessionPersistenceOptions controlling SaveState; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The LlamaSessionArtifact result produced by LlamaMappedReferenceSession.SaveState for this contract: Saves complete deterministic state to a local artifact. It is published only after all documented validation and ownership transitions succeed.
LlamaModelConfigUAIX.LmRuntime.Models.Llama
17 members
Represents LLaMA-family transformer configuration reconstructed from GGUF metadata.
Architecture
Gets the architecture name.
AttentionHeadCount
Gets the attention head count.
AttentionKeyValueHeadCount
Gets the attention key/value head count.
BlockCount
Gets the transformer block count.
ContextLength
Gets the training context length.
EmbeddingLength
Gets the embedding length.
FeedForwardLength
Gets the feed-forward hidden length.
HeadDimension
Gets the dimension of one query attention head.
KeyValueDimension
Gets the flattened key/value projection dimension.
ModelName
Gets the optional model display name.
RmsNormEpsilon
Gets the RMSNorm epsilon.
RopeDimensionCount
Gets the RoPE dimension count per attention head.
RopeFrequencyBase
Gets the RoPE frequency base.
SupportsTiedOutputProjection
Gets whether the loader may use token embeddings as the output projection when output.weight is absent.
VocabularySize
Gets the vocabulary size.
FromGguf(UAIX.LmRuntime.Gguf.GgufModel)
Creates a LLaMA-family configuration from GGUF metadata.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
Returns: The LlamaModelConfig result produced by LlamaModelConfig.FromGguf for this contract: Creates a LLaMA-family configuration from GGUF metadata. It is published only after all documented validation and ownership transitions succeed.
Validate
Validates architectural invariants required by the scalar LLaMA runtime.
LlamaOneTokenFinishReasonUAIX.LmRuntime.Models.Llama
1 member
Identifies why a bounded one-token generation operation ended.
OneTokenCompleted
Exactly one greedy token was selected as requested.
LlamaOneTokenOptionsUAIX.LmRuntime.Models.Llama
4 members
Configures one deterministic mapped-model greedy-token operation.
AddSpecialTokens
Gets whether model-defined BOS/EOS behavior should be applied.
EmitTokenizerTrace
Gets whether tokenizer trace events should be captured.
ParseSpecialTokens
Gets whether raw special-token text should be recognized.
ResetSession
Gets whether the session should reset before prompt evaluation.
LlamaOneTokenTimingsUAIX.LmRuntime.Models.Llama
4 members
Records measured stages of exactly one mapped-model greedy decode operation.
PrefillDuration
Gets prompt prefill duration.
SelectionDuration
Gets greedy selection and token decode duration.
TokenizationDuration
Gets prompt tokenization duration.
TotalDuration
Gets total operation duration.
LlamaParityToleranceUAIX.LmRuntime.Models.Llama
3 members
Configures exact token and explicit floating-point tolerance checks for cross-storage parity.
AbsoluteTolerance
Gets the absolute per-logit tolerance.
RelativeTolerance
Gets the relative per-logit tolerance.
Validate
Validates the absolute and relative parity tolerances used for numerical comparison.
LlamaReferenceAttentionUAIX.LmRuntime.Models.Llama
1 member
Provides reference causal attention behavior.
ApplyCausal(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,int,System.Span<float>)
Applies a minimal causal attention score computation.
query- The query sequence used by this operation; its required length, ordering, and element bounds are validated before access.
keys- The keys sequence used by this operation; its required length, ordering, and element bounds are validated before access.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
headSize- The numeric head size consumed by ApplyCausal; it must satisfy the member's documented range, geometry, and finite-value requirements.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
LlamaReferenceExecutorUAIX.LmRuntime.Models.Llama
1 member
Provides scalar/reference execution anchors for LLaMA-family graphs.
Forward(System.ReadOnlySpan<float>,UAIX.LmRuntime.Models.Llama.LlamaWeights,System.Span<float>)
Executes a minimal reference forward pass over hidden-state logits.
hiddenState- The hidden state sequence used by this operation; its required length, ordering, and element bounds are validated before access.
weights- The LLaMA weights used by the reference path.
logits- The logits sequence used by this operation; its required length, ordering, and element bounds are validated before access.
LlamaReferenceFixtureUAIX.LmRuntime.Models.Llama
5 members
Represents a deterministic tiny reference fixture with one transformer block.
Configuration
Gets the fixture model configuration.
PromptTokenIds
Gets the canonical fixture prompt tokens.
Tokenizer
Gets the fixture tokenizer.
Weights
Gets the fixture model weights.
CreateSession
Creates the session from the validated inputs required by LlamaReferenceFixture.
Returns: A session with empty key/value cache state.
LlamaReferenceFixtureFactoryUAIX.LmRuntime.Models.Llama
1 member
Creates deterministic tiny fixtures used by reference-runtime tests and examples.
CreateDeterministic
Creates a one-block, five-token deterministic LLaMA fixture.
Returns: The fixture configuration, weights, tokenizer, and prompt, with ownership and disposal obligations defined by the returned type and the CreateDeterministic contract.
LlamaReferenceForwardPassUAIX.LmRuntime.Models.Llama
2 members
Provides tiny reference building blocks for LLaMA-family correctness tests.
ApplyRope(System.Span<float>,System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,int)
Applies LLaMA-style RoPE to a query or key vector in place.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
cos- The cos sequence used by this operation; its required length, ordering, and element bounds are validated before access.
sin- The sin sequence used by this operation; its required length, ordering, and element bounds are validated before access.
ropeDimensions- The even number of leading head dimensions transformed by rotary positional encoding.
RmsNorm(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>,float)
Applies the LLaMA RMSNorm operation through the CPU reference kernel.
input- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
weight- The weight sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
epsilon- The positive normalization epsilon added to the mean-square term to avoid division by zero while preserving deterministic numerical behavior.
LlamaReferenceLayerWeightsUAIX.LmRuntime.Models.Llama
9 members
Stores immutable float32 weights for one scalar/reference LLaMA transformer block.
AttentionKey
Gets the key projection matrix in row-major logical order.
AttentionNorm
Gets the attention RMSNorm scale.
AttentionOutput
Gets the attention output projection matrix in row-major logical order.
AttentionQuery
Gets the query projection matrix in row-major logical order.
AttentionValue
Gets the value projection matrix in row-major logical order.
FeedForwardDown
Gets the feed-forward down projection matrix in row-major logical order.
FeedForwardGate
Gets the feed-forward gate projection matrix in row-major logical order.
FeedForwardNorm
Gets the feed-forward RMSNorm scale.
FeedForwardUp
Gets the feed-forward up projection matrix in row-major logical order.
LlamaReferenceMaterializationRecordUAIX.LmRuntime.Models.Llama
5 members
Records one explicit managed copy made for the bounded scalar reference runtime.
BlockIndex
Gets the optional transformer block index.
CopiedByteCount
Gets the copied byte count.
Role
Gets the semantic tensor role.
StorageMode
Gets the resulting storage mode.
TensorName
Gets the source tensor name.
LlamaReferenceModelWeightsUAIX.LmRuntime.Models.Llama
5 members
Stores immutable float32 weights for the deterministic LLaMA reference runtime.
Layers
Gets transformer block weights in execution order.
OutputNorm
Gets the final RMSNorm scale.
OutputProjection
Gets the output projection matrix in row-major logical order. An empty value means tied embeddings.
TokenEmbeddings
Gets the token embedding table in row-major logical order.
Validate(UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Validates all reference-weight shapes against a LLaMA configuration.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
LlamaReferenceRmsNormUAIX.LmRuntime.Models.Llama
1 member
Provides reference RMSNorm behavior.
Apply(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>,float)
Applies the supplied input to the supplied values while preserving the operation's numeric and shape invariants.
input- The source data consumed by the operation; caller-owned storage is not retained after the method returns.
weights- The weights sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
epsilon- The positive normalization epsilon added to the mean-square term to avoid division by zero while preserving deterministic numerical behavior.
LlamaReferenceRopeUAIX.LmRuntime.Models.Llama
1 member
Provides reference RoPE behavior.
Apply(System.Span<float>,int,float)
Applies rotary position embedding to adjacent hidden-state pairs.
values- The values sequence used by this operation; its required length, ordering, and element bounds are validated before access.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
theta- The rotary angle in radians applied to the paired vector components at the addressed position.
LlamaReferenceSessionUAIX.LmRuntime.Models.Llama
17 members
Executes a deterministic, scalar-first LLaMA forward path for tiny correctness fixtures.
This class is the numerical correctness anchor for later optimized kernels. It is intentionally limited to batch size one and F32, Q8_0, or Q4_0 mapped or array-backed weights. It performs no governance or adaptive policy operations and therefore belongs exclusively to deterministic parity mode.
ContextCapacity
Gets the configured sequence capacity.
KvCache
Gets the typed key/value cache owned by this session.
Position
Gets the next sequence position to be evaluated.
VocabularySize
Gets the configured vocabulary size.
WeightSource
Gets the immutable model weight source used by this session.
CaptureState
Captures complete deterministic session state without serializing live model pointers.
Returns: The LlamaReferenceSessionSnapshot result produced by LlamaReferenceSession.CaptureState for this contract: Captures complete deterministic session state without serializing live model pointers. It is published only after all documented validation and ownership transitions succeed.
CopyLastLogitsTo(System.Span<float>)
Copies the most recently computed logits to a caller-provided destination.
destination- The destination with room for the configured vocabulary.
DecodeOneGreedy(System.Collections.Generic.IReadOnlyList<int>,bool)
Evaluates a prompt and returns exactly one greedily selected next token.
promptTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
resetSession- Whether existing key/value state should be cleared first.
Returns: The LlamaGreedyTokenResult result produced by LlamaReferenceSession.DecodeOneGreedy for this contract: Evaluates a prompt and returns exactly one greedily selected next token. It is published only after all documented validation and ownership transitions succeed.
GenerateGreedy(System.Collections.Generic.IReadOnlyList<int>,System.Span<int>,System.Span<float>,UAIX.LmRuntime.Models.Llama.LlamaGreedyGenerationOptions,System.Action<UAIX.LmRuntime.Models.Llama.LlamaGeneratedToken>,System.Threading.CancellationToken)
Generates deterministic greedy token identifiers and reports each selection to a synchronous observer.
promptTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
generatedTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
finalLogits- The caller-owned destination for the final available logits.
options- The optional LlamaGreedyGenerationOptions controlling GenerateGreedy; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
tokenObserver- The optional synchronous observer invoked once after each token is selected.
cancellationToken- A token observed before selection and before the next committed model step.
Returns: The LlamaGreedyGenerationResult result produced by LlamaReferenceSession.GenerateGreedy for this contract: Generates deterministic greedy token identifiers and reports each selection to a synchronous observer. It is published only after all documented validation and ownership transitions succeed.
GenerateGreedy(System.Collections.Generic.IReadOnlyList<int>,System.Span<int>,System.Span<float>,UAIX.LmRuntime.Models.Llama.LlamaGreedyGenerationOptions,System.Threading.CancellationToken)
Generates deterministic greedy token identifiers into caller-owned buffers.
promptTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
generatedTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
finalLogits- The caller-owned destination for the final available logits.
options- The optional LlamaGreedyGenerationOptions controlling GenerateGreedy; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
cancellationToken- The caller-provided token used to cancel the operation before additional work or results are published.
Returns: The LlamaGreedyGenerationResult result produced by LlamaReferenceSession.GenerateGreedy for this contract: Generates deterministic greedy token identifiers into caller-owned buffers. It is published only after all documented validation and ownership transitions succeed.
LlamaReferenceSession(UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.ILlamaModelWeightSource,UAIX.LmRuntime.Tokenization.IGgufTokenizer)
Initializes a reference session over immutable array-backed or direct mapped weight sources.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
tokenizer- The optional tokenizer used only to decode the selected token text.
LlamaReferenceSession(UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.LlamaReferenceModelWeights,UAIX.LmRuntime.Tokenization.IGgufTokenizer)
Initializes a reference session through the v1.8.0 array-backed compatibility path.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
tokenizer- The optional tokenizer used only to decode the selected token text.
Prefill(System.Collections.Generic.IReadOnlyList<int>,bool)
Evaluates every prompt token and leaves the final logits available for deterministic selection.
promptTokenIds- The ordered token identifiers to process; sequence order is preserved and each identifier is validated where required.
resetSession- Whether existing sequence and cache state should be cleared first.
Reset
Clears sequence state and all key/value cache contents.
RestoreState(UAIX.LmRuntime.Models.Llama.LlamaReferenceSessionSnapshot)
Restores complete deterministic state after validating sequence, vocabulary, and cache identities.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
RunStep(int,System.Span<float>)
Evaluates one input token and writes next-token logits.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
logits- The destination buffer with at least vocabulary-size elements.
SelectGreedyToken(int)
Selects and decodes one greedy token from the current logits.
promptTokenCount- The number of prompt tokens associated with the current logits.
Returns: The LlamaGreedyTokenResult result produced by LlamaReferenceSession.SelectGreedyToken for this contract: Selects and decodes one greedy token from the current logits. It is published only after all documented validation and ownership transitions succeed.
LlamaReferenceSessionSnapshotUAIX.LmRuntime.Models.Llama
5 members
Captures complete deterministic reference-session state without retaining live model pointers.
KeyValueCache
Gets complete capacity-shaped key/value state.
LastLogits
Gets the most recently computed logits.
Position
Gets the next sequence position.
SchemaVersion
Gets the in-memory snapshot schema version.
TokenHistory
Gets committed input token identifiers in sequence order.
LlamaReferenceWeightMaterializationUAIX.LmRuntime.Models.Llama
3 members
Contains immutable float32 weights and copy evidence for the scalar reference runtime.
Records
Gets every bounded copy made while materializing the fixture.
TotalCopiedByteCount
Gets the total number of copied bytes.
Weights
Gets the immutable reference weights.
LlamaRequiredTensorRegistryUAIX.LmRuntime.Models.Llama
1 member
Builds the required LLaMA-family tensor registry from model configuration.
Build(UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Creates the required tensor list for the configuration.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
Returns: An ordered read-only IReadOnlyList<LlamaTensorRequirement> result from LlamaRequiredTensorRegistry.Build: Creates the required tensor list for the configuration. Mutable internal collection aliases are not exposed through the returned contract.
LlamaRuntimeModeUAIX.LmRuntime.Models.Llama
1 member
Identifies the deterministic execution contract used by a mapped model session.
DeterministicParity
Runs only deterministic parity behavior without adaptive governance.
LlamaSessionArtifactUAIX.LmRuntime.Models.Llama
15 members
Carries verified complete deterministic session state and compatibility identities.
CacheLayoutFingerprint
Gets the cache-layout fingerprint.
ClaimStatus
Gets the evidence claim status.
ConfigurationFingerprint
Gets the model configuration fingerprint.
ContentSha256
Gets the SHA-256 of every serialized byte preceding the digest.
EndOfSequenceTokenId
Gets the optional end-of-sequence token identifier.
GeneratedUtc
Gets the artifact generation time in UTC.
MaximumCompatiblePackageVersion
Gets the newest supported package version.
MinimumCompatiblePackageVersion
Gets the oldest supported package version.
ModelSha256
Gets the complete model artifact SHA-256.
PackageVersion
Gets the package version that emitted the artifact.
SamplerMode
Gets the sampler mode.
SchemaVersion
Gets the portable schema version.
Snapshot
Gets the complete session snapshot.
StopTokenIds
Gets configured stop-token identifiers.
TokenizerFingerprint
Gets the tokenizer fingerprint.
LlamaSessionArtifactSerializerUAIX.LmRuntime.Models.Llama
5 members
Serializes complete deterministic reference-session state in bounded little-endian form.
SchemaVersion
Gets the supported artifact schema version.
Deserialize(System.ReadOnlySpan<byte>,int)
Deserializes the llama session artifact from the validated persisted representation.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The LlamaSessionArtifact result produced by LlamaSessionArtifactSerializer.Deserialize for this contract: Deserializes the llama session artifact from the validated persisted representation. It is published only after all documented validation and ownership transitions succeed.
Load(string,int)
Reads and verifies a complete artifact from a local file.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The verified artifact, with ownership and disposal obligations defined by the returned type and the Load contract.
Save(string,UAIX.LmRuntime.Models.Llama.LlamaReferenceSessionSnapshot,UAIX.LmRuntime.Models.Llama.LlamaSessionPersistenceOptions)
Writes a complete artifact to a local file.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
options- The optional LlamaSessionPersistenceOptions controlling Save; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The LlamaSessionArtifact result produced by LlamaSessionArtifactSerializer.Save for this contract: Writes a complete artifact to a local file. It is published only after all documented validation and ownership transitions succeed.
Serialize(UAIX.LmRuntime.Models.Llama.LlamaReferenceSessionSnapshot,UAIX.LmRuntime.Models.Llama.LlamaSessionPersistenceOptions)
Serializes complete session state and appends a SHA-256 digest.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
options- The optional LlamaSessionPersistenceOptions controlling Serialize; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: A newly allocated byte[] containing the ordered result of LlamaSessionArtifactSerializer.Serialize: Serializes complete session state and appends a SHA-256 digest. The caller owns the returned array and later mutation cannot alter the source object.
LlamaSessionPersistenceOptionsUAIX.LmRuntime.Models.Llama
13 members
Configures digest-bound complete session serialization.
CacheLayoutFingerprint
Gets the persistent cache-layout identity.
ClaimStatus
Gets the evidence claim status.
ConfigurationFingerprint
Gets the LLaMA configuration fingerprint.
EndOfSequenceTokenId
Gets the optional end-of-sequence token identifier.
GeneratedUtc
Gets the UTC generation time.
MaximumByteCount
Gets the maximum accepted artifact byte count.
MaximumCompatiblePackageVersion
Gets the newest supported package version.
MinimumCompatiblePackageVersion
Gets the oldest supported package version.
ModelSha256
Gets the complete model artifact SHA-256.
PackageVersion
Gets the package version that emitted the artifact.
SamplerMode
Gets the deterministic sampler mode.
StopTokenIds
Gets configured stop-token identifiers.
TokenizerFingerprint
Gets the GGUF tokenizer fingerprint.
LlamaStorageParityCandidateResultUAIX.LmRuntime.Models.Llama
7 members
Represents one candidate model's parity result against a selected reference model.
LogitComparison
Gets the detailed logit comparison.
ModelPath
Gets the candidate model path.
ModelSha256
Gets the candidate model SHA-256.
OneTokenResult
Gets the complete candidate one-token result.
Passed
Gets whether both exact-token and floating-point contracts passed.
StorageSummary
Gets the candidate storage summary.
TokenMatches
Gets whether the selected token identifier exactly equals the reference identifier.
LlamaStorageParityResultUAIX.LmRuntime.Models.Llama
4 members
Represents a cross-storage one-token parity run.
Candidates
Gets candidate results in caller order.
Passed
Gets whether every candidate passed the explicit parity contract.
Prompt
Gets the prompt used for every model.
ReferenceResult
Gets the reference one-token result.
LlamaStorageParityRunnerUAIX.LmRuntime.Models.Llama
1 member
Executes bounded offline one-token parity comparisons across local GGUF storage variants.
Run(string,System.Collections.Generic.IReadOnlyList<string>,string,UAIX.LmRuntime.Models.Llama.LlamaParityTolerance)
Runs one reference model and one or more candidate models with identical prompt settings.
referenceModelPath- The local file-system reference model path processed by this operation; it must satisfy the containing component's path and scope policy.
candidateModelPaths- The local file-system candidate model paths processed by this operation; it must satisfy the containing component's path and scope policy.
prompt- The prompt processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
tolerance- The tolerance input of type LlamaParityTolerance read by LlamaStorageParityRunner.Run; it must satisfy the member-specific nullability, identity, range, and ownership rules before dependent work begins.
Returns: The LlamaStorageParityResult result produced by LlamaStorageParityRunner.Run for this contract: Runs one reference model and one or more candidate models with identical prompt settings. It is published only after all documented validation and ownership transitions succeed.
LlamaSwiGluReferenceUAIX.LmRuntime.Models.Llama
1 member
Provides reference SwiGLU behavior.
Apply(System.ReadOnlySpan<float>,System.ReadOnlySpan<float>,System.Span<float>)
Applies the SwiGLU activation to validated gate and up-projection vectors.
gate- The gate sequence used by this operation; its required length, ordering, and element bounds are validated before access.
up- The up sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
LlamaTensorBinderUAIX.LmRuntime.Models.Llama
2 members
Binds and validates LLaMA-family GGUF tensors as a schema-validation phase.
Bind(UAIX.LmRuntime.Gguf.GgufModel,UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Binds required tensors from a parsed GGUF artifact using default validation options.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
Returns: The TensorBindingManifest result produced by LlamaTensorBinder.Bind for this contract: Binds required tensors from a parsed GGUF artifact using default validation options. It is published only after all documented validation and ownership transitions succeed.
Bind(UAIX.LmRuntime.Gguf.GgufModel,UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.TensorBindingOptions)
Binds required tensors from a parsed GGUF artifact.
model- The parsed GGUF model whose validated metadata and tensor catalog are consumed by this operation.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
options- The optional TensorBindingOptions controlling Bind; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The TensorBindingManifest result produced by LlamaTensorBinder.Bind for this contract: Binds required tensors from a parsed GGUF artifact. It is published only after all documented validation and ownership transitions succeed.
LlamaTensorRequirementUAIX.LmRuntime.Models.Llama
7 members
Describes one required LLaMA tensor contract.
BlockIndex
Gets the optional block index.
ExpectedLogicalDimensions
Gets dimensions in logical row-major order for diagnostics and manifests.
ExpectedRank
Gets the expected rank.
ExpectedStorageDimensions
Gets dimensions in GGUF storage order, where dimension zero is the row width.
IsOptional
Gets whether the tensor may be satisfied by an explicit alias rule.
Name
Gets the required tensor name.
Role
Gets the tensor role.
LlamaTensorRoleUAIX.LmRuntime.Models.Llama
12 members
Identifies semantic roles for LLaMA-family tensors.
AttentionKey
Per-block key projection.
AttentionNorm
Per-block attention normalization scale.
AttentionOutput
Per-block attention output projection.
AttentionQuery
Per-block query projection.
AttentionValue
Per-block value projection.
FeedForwardDown
Per-block feed-forward down projection.
FeedForwardGate
Per-block feed-forward gate projection.
FeedForwardNorm
Per-block feed-forward normalization scale.
FeedForwardUp
Per-block feed-forward up projection.
Output
Output projection matrix.
OutputNorm
Final output normalization scale.
TokenEmbedding
Token embedding table.
LlamaWeightsUAIX.LmRuntime.Models.Llama
2 members
Represents model-level LLaMA weights used by reference execution.
OutputProjection
Gets the output projection matrix in row-major order.
TokenEmbeddings
Gets token embedding weights.
LlamaWeightSourceValidatorUAIX.LmRuntime.Models.Llama
1 member
Validates storage-neutral LLaMA weight sources before deterministic execution begins.
Validate(UAIX.LmRuntime.Models.Llama.LlamaModelConfig,UAIX.LmRuntime.Models.Llama.ILlamaModelWeightSource)
Validates every global and block-local source against the configured model geometry.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
LlamaWeightStorageModeUAIX.LmRuntime.Models.Llama
3 members
Identifies how a bound tensor participates in reference execution.
Alias
The tensor aliases another mapped tensor.
CopiedForReference
The tensor was explicitly copied into a bounded float32 reference buffer.
Mapped
The tensor remains a borrowed view over the mapped GGUF file.
MappedBFloat16MatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies a brain-float16 matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedBFloat16MatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedBFloat16MatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedBFloat16VectorSourceUAIX.LmRuntime.Models.Llama
6 members
Reads a brain-float16 vector directly from a mapped GGUF tensor view.
DataType
Length
StorageDiagnostics
StorageType
CopyTo(System.Span<float>)
Copies the to into caller-owned storage after validating the requested range and capacity.
destination- The destination buffer that receives the produced values.
MappedBFloat16VectorSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedBFloat16VectorSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
MappedFloat16MatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies an IEEE float16 matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedFloat16MatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedFloat16MatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedFloat16VectorSourceUAIX.LmRuntime.Models.Llama
6 members
Reads an IEEE float16 vector directly from a mapped GGUF tensor view.
DataType
Length
StorageDiagnostics
StorageType
CopyTo(System.Span<float>)
Copies the to into caller-owned storage after validating the requested range and capacity.
destination- The destination buffer that receives the produced values.
MappedFloat16VectorSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedFloat16VectorSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
MappedFloat32MatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies an F32 matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedFloat32MatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedFloat32MatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedFloat32VectorSourceUAIX.LmRuntime.Models.Llama
6 members
Reads a float32 vector directly from a mapped GGUF tensor view.
DataType
Length
StorageDiagnostics
StorageType
CopyTo(System.Span<float>)
Copies the to into caller-owned storage after validating the requested range and capacity.
destination- The destination buffer that receives the produced values.
MappedFloat32VectorSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedFloat32VectorSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
MappedLlamaLayerWeightSourceUAIX.LmRuntime.Models.Llama
10 members
Exposes one mapped LLaMA transformer block through storage-neutral execution contracts.
AttentionKey
AttentionNorm
AttentionOutput
AttentionQuery
AttentionValue
FeedForwardDown
FeedForwardGate
FeedForwardNorm
FeedForwardUp
MappedLlamaLayerWeightSource(UAIX.LmRuntime.Models.Llama.LlamaBoundLayerWeightSet)
Initializes a new MappedLlamaLayerWeightSource instance with validated dependencies and operational bounds.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
MappedLlamaModelWeightSourceUAIX.LmRuntime.Models.Llama
10 members
Exposes a complete mapped LLaMA model through storage-neutral execution contracts.
Layers
ManagedCopiedByteCount
Gets the total number of persistent managed model-weight bytes copied by this source.
OutputNorm
OutputProjection
StorageDiagnostics
StorageSummary
TokenEmbeddings
UsesTiedOutputProjection
Create(UAIX.LmRuntime.Models.Llama.LlamaBoundWeightSet)
Creates and validates a complete mapped model weight source.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
Returns: The validated mapped model weight source, with ownership and disposal obligations defined by the returned type and the Create contract.
MappedLlamaModelWeightSource(UAIX.LmRuntime.Models.Llama.LlamaBoundWeightSet)
Initializes a new MappedLlamaModelWeightSource instance with validated dependencies and operational bounds.
weights- The validated model-weight source or bound weight set consumed read-only by the deterministic reference operation.
MappedMatrixSourceFactoryUAIX.LmRuntime.Models.Llama
1 member
Creates supported matrix sources over mapped tensor views.
Create(UAIX.LmRuntime.Gguf.MappedTensorView)
Creates a direct mapped source for supported scalar and quantized storage.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Returns: The storage-specific matrix source, with ownership and disposal obligations defined by the returned type and the Create contract.
MappedQ4_0MatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies a Q4_0 matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedQ4_0MatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedQ4_0MatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedQ4_KMatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies a Q4_K matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedQ4_KMatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedQ4_KMatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedQ6_KMatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies a Q6_K matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedQ6_KMatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedQ6_KMatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedQ8_0MatrixSourceUAIX.LmRuntime.Models.Llama
8 members
Reads and multiplies a Q8_0 matrix directly from a mapped GGUF tensor view.
ColumnCount
DataType
RowCount
StorageDiagnostics
StorageType
CopyRowTo(int,System.Span<float>)
Copies the row to into caller-owned storage after validating the requested range and capacity.
rowIndex- The zero-based row index; it must identify an existing position within the relevant validated range.
destination- The destination buffer that receives the produced values.
MappedQ8_0MatrixSource(UAIX.LmRuntime.Gguf.MappedTensorView)
Initializes a new MappedQ8_0MatrixSource instance with validated dependencies and operational bounds.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Multiply(System.ReadOnlySpan<float>,System.Span<float>)
Multiplies the supplied vector by the supplied vector without changing logical row order.
vector- The vector sequence used by this operation; its required length, ordering, and element bounds are validated before access.
output- The caller-owned destination buffer that receives the result; required capacity is validated before any write occurs.
MappedVectorSourceFactoryUAIX.LmRuntime.Models.Llama
1 member
Selects a mapped scalar vector implementation from GGML storage metadata.
Create(UAIX.LmRuntime.Gguf.MappedTensorView)
Creates the read only vector source from the validated inputs required by MappedVectorSourceFactory.
view- The bounded tensor view whose descriptor, shape, byte order, and mapped payload are read without transferring ownership of the underlying mapping.
Returns: The storage-specific vector source, with ownership and disposal obligations defined by the returned type and the Create contract.
RealModelPathPolicyUAIX.LmRuntime.Models.Llama
1 member
Resolves local model paths under an optional root without following hidden network or download behavior.
Resolve(string,string,long)
Resolves and validates one local model path.
path- The local file-system path processed by this operation; it must satisfy the containing component's path and scope policy.
allowedRoot- The normalized caller-authorized directory boundary, or null when no containment root was configured for the operation.
maximumFileByteCount- The optional explicit file-size limit; zero disables it.
Returns: The text produced by RealModelPathPolicy.Resolve for this contract: Resolves and validates one local model path. The returned string is detached from mutable caller storage and is not persisted by the operation.
RealModelSmokeArtifactUAIX.LmRuntime.Models.Llama
29 members
Represents a versioned, machine-readable real-model smoke artifact.
Alignment
Gets the effective GGUF tensor alignment.
Architecture
Gets the model architecture.
BindingDiagnostics
Gets binding diagnostic messages.
ClaimStatus
Gets the evidence claim status.
CommandIdentity
Gets the non-secret command identity.
CommitIdentity
Gets the source/commit identity.
CompletedStage
Gets the deepest completed stage.
Diagnostics
Gets bounded workflow diagnostics.
EnvironmentVariableNames
Gets environment-variable names used by the workflow without values.
ExpectedOneTokenMatched
Gets whether the optional expected one-token evidence matched.
ExpectedTokenIdsMatched
Gets whether the optional expected token-identifier evidence matched.
FileByteCount
Gets the model file length.
GeneratedUtc
Gets the generation time in UTC.
GgufVersion
Gets the parsed GGUF version.
LicenseReviewStatus
Gets the operator-supplied license review status.
ModelPath
Gets the normalized local model path.
ModelSha256
Gets the optional complete-file SHA-256.
PackageVersion
Gets the package version.
PromptSha256
Gets the SHA-256 of the prompt text rather than requiring publication of the raw prompt.
PromptTokenIds
Gets exact prompt token identifiers when tokenization completed.
ProvenanceLabel
Gets the operator-supplied provenance label.
Schema
Gets the artifact schema identifier.
SelectedTokenId
Gets the selected one-token identifier when execution completed.
SelectedTokenText
Gets the selected token text when execution completed.
StageEvidence
Gets stage timing and current-thread allocation measurements.
StorageTypeCounts
Gets physical tensor counts by GGML storage name.
Succeeded
Gets whether the requested stage completed.
TokenizerFamily
Gets the tokenizer family.
UnsupportedDiagnostics
Gets exact unsupported execution diagnostics.
RealModelSmokeEnvironmentUAIX.LmRuntime.Models.Llama
1 member
Creates explicit local smoke options from the documented environment-variable contract.
Load(UAIX.LmRuntime.Models.Llama.RealModelSmokeStage)
Reads the local real-model smoke configuration from environment variables.
stage- The explicitly selected real-model smoke stage or stage implementation executed by the controlled pipeline.
Returns: The local smoke options, with ownership and disposal obligations defined by the returned type and the Load contract.
RealModelSmokeOptionsUAIX.LmRuntime.Models.Llama
15 members
Configures an explicitly local, opt-in GGUF smoke inspection.
AllowedRoot
Gets an optional root that the resolved model path must remain under.
CommitIdentity
Gets a commit or source identity supplied by the operator.
ComputeModelSha256
Gets whether the complete model SHA-256 should be computed.
EnvironmentGateName
Gets the environment variable that enables real-model execution.
ExpectedOneTokenPath
Gets an optional local JSON file containing the expected one-token result.
ExpectedTokenIdsPath
Gets an optional local JSON file containing expected prompt token identifiers.
LicenseReviewStatus
Gets the operator-supplied license review status.
MaximumFileByteCount
Gets an optional explicit maximum file length; zero disables this limit.
ModelPath
Gets the local GGUF path.
PackageVersion
Gets the package version recorded in evidence.
Prompt
Gets the prompt used by the one-token stage.
ProvenanceLabel
Gets an operator-supplied provenance label.
RedactModelPath
Gets whether the artifact model path is reduced to its file name.
RequireEnvironmentGate
Gets whether the explicit environment gate is required.
Stage
Gets the deepest smoke stage to execute.
RealModelSmokeRunnerUAIX.LmRuntime.Models.Llama
1 member
Executes staged, offline real-model validation and emits a bounded evidence artifact.
Run(UAIX.LmRuntime.Models.Llama.RealModelSmokeOptions)
Runs the requested local smoke stages in their required order.
options- The explicit local smoke options and evidence boundaries.
Returns: A bounded machine-readable artifact describing the deepest completed stage.
RealModelSmokeStageUAIX.LmRuntime.Models.Llama
4 members
Identifies the deepest stage requested from the local real-model smoke workflow.
OneToken
Also executes one deterministic greedy token when every storage contract is supported.
ParseOnly
Parses and validates the GGUF container only.
TensorBinding
Also reconstructs LLaMA geometry and validates required tensor bindings.
Tokenizer
Also constructs and validates the metadata-driven tokenizer.
RealModelSmokeStageEvidenceUAIX.LmRuntime.Models.Llama
3 members
Records one real-model workflow stage duration and current-thread allocation delta.
ElapsedStopwatchTicks
Gets elapsed stopwatch ticks.
ManagedAllocatedByteCount
Gets managed bytes allocated on the measuring thread.
Stage
Gets the stage name.
ReferenceKvCacheUAIX.LmRuntime.Models.Llama
16 members
Stores reference key/value state in two contiguous arrays without per-token dictionaries.
ConfigurationFingerprint
ContextLength
HeadWidth
KeyValueHeadCount
LayerCount
UsedTokenCount
WriteBehavior
CreateDiagnosticSnapshot
Creates a small diagnostic snapshot without exposing mutable key/value arrays.
Returns: The bounded diagnostic snapshot, with ownership and disposal obligations defined by the returned type and the CreateDiagnosticSnapshot contract.
CreateSnapshot
Creates the snapshot from the validated inputs required by ReferenceKvCache.
Returns: The ReferenceKvCacheSnapshot result produced by ReferenceKvCache.CreateSnapshot for this contract: Creates the snapshot from the validated inputs required by ReferenceKvCache. It is published only after all documented validation and ownership transitions succeed.
GetKey(int,int,int)
Retrieves the key from the current cache state after validating the requested access.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
headIndex- The zero-based head index; it must identify an existing position within the relevant validated range.
Returns: The bounded ReadOnlySpan<float> view produced by ReferenceKvCache.GetKey: Retrieves the key from the current cache state after validating the requested access. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
GetValue(int,int,int)
Retrieves the value from the current cache state after validating the requested access.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
headIndex- The zero-based head index; it must identify an existing position within the relevant validated range.
Returns: The bounded ReadOnlySpan<float> view produced by ReferenceKvCache.GetValue: Retrieves the value from the current cache state after validating the requested access. Its lifetime and ownership remain tied to the owner identified by the containing type; no out-of-range region is exposed.
ReferenceKvCache(int,int,int,int)
Initializes a reference key/value cache with a geometry-derived compatibility fingerprint.
layerCount- The layer count used to bound this operation; it must be nonnegative and within the supported range.
contextLength- The context length that supplies session-scoped identity and boundary state; it is validated before dependent work begins.
keyValueHeadCount- The number of key/value heads per layer.
headWidth- The positive number of scalar values stored for each attention head in one cache position.
ReferenceKvCache(int,int,int,int,string)
Initializes a reference key/value cache with an explicit model/configuration fingerprint.
layerCount- The layer count used to bound this operation; it must be nonnegative and within the supported range.
contextLength- The context length that supplies session-scoped identity and boundary state; it is validated before dependent work begins.
keyValueHeadCount- The number of key/value heads per layer.
headWidth- The positive number of scalar values stored for each attention head in one cache position.
configurationFingerprint- The configuration fingerprint text consumed by ReferenceKvCache.ReferenceKvCache; null, emptiness, length, encoding, identifier, or path rules are enforced as documented, and the value is not persisted by this operation.
Reset
Resets the reference KV cache contents and logical sequence position to their initial state.
Restore(UAIX.LmRuntime.Models.Llama.ReferenceKvCacheSnapshot)
Restores the supplied snapshot from a validated persisted representation.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
Write(int,int,System.ReadOnlySpan<float>,System.ReadOnlySpan<float>)
Writes the supplied layer index to the current cache state using the component's canonical representation.
layerIndex- The zero-based layer index; it must identify an existing position within the relevant validated range.
position- The zero-based sequence or cache position addressed by the operation; it must lie within the allocated context and readable or writable range.
key- The key sequence used by this operation; its required length, ordering, and element bounds are validated before access.
value- The value sequence used by this operation; its required length, ordering, and element bounds are validated before access.
ReferenceKvCacheDiagnosticSnapshotUAIX.LmRuntime.Models.Llama
3 members
Represents a bounded, non-mutable diagnostic view of reference cache state.
ConfigurationFingerprint
Gets the configuration fingerprint.
ContentSha256
Gets the SHA-256 of the used key/value prefix.
UsedTokenCount
Gets the used token count.
ReferenceKvCacheFingerprintUAIX.LmRuntime.Models.Llama
1 member
Computes stable fingerprints for model configurations that own reference key/value cache snapshots.
Create(UAIX.LmRuntime.Models.Llama.LlamaModelConfig)
Creates a SHA-256 fingerprint from the configuration fields that determine cache geometry and semantics.
config- The validated LLaMA model configuration defining context capacity, vocabulary size, tensor geometry, and attention dimensions for the operation.
Returns: The text produced by ReferenceKvCacheFingerprint.Create for this contract: Creates a SHA-256 fingerprint from the configuration fields that determine cache geometry and semantics. The returned string is detached from mutable caller storage and is not persisted by the operation.
ReferenceKvCacheSerializerUAIX.LmRuntime.Models.Llama
5 members
Serializes only logically used key/value positions in stable layer-position-head order.
Schema version two is additive and does not change the in-memory version-one snapshot contract retained for source compatibility. Unused capacity is reconstructed as zero during deserialization.
DefaultMaximumByteCount
Gets the default maximum serialized snapshot size.
SchemaVersion
Gets the portable snapshot schema version.
Deserialize(System.ReadOnlySpan<byte>,int)
Deserializes and verifies a portable key/value-cache snapshot.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The ReferenceKvPortableSnapshot result produced by ReferenceKvCacheSerializer.Deserialize for this contract: Deserializes and verifies a portable key/value-cache snapshot. It is published only after all documented validation and ownership transitions succeed.
Restore(UAIX.LmRuntime.Models.Llama.ReferenceKvCache,System.ReadOnlySpan<byte>,string,string)
Restores verified portable bytes into a cache after validating model and layout identities.
cache- The validated ReferenceKvCache dependency consumed by Restore; ownership and lifetime remain with the caller unless this member explicitly documents a transfer.
bytes- The bytes sequence used by this operation; its required length, ordering, and element bounds are validated before access.
expectedModelArtifactFingerprint- The expected artifact identity, or an empty string to accept an empty serialized identity.
expectedCacheLayoutFingerprint- The required cache-layout fingerprint against which the serialized state is compared before restoration.
Serialize(UAIX.LmRuntime.Models.Llama.ReferenceKvCacheSnapshot,string,string,int)
Serializes a bounded cache snapshot in deterministic little-endian form.
snapshot- The immutable state snapshot being validated, serialized, restored, or analyzed without retaining caller-owned mutable aliases.
modelArtifactFingerprint- The stable model-artifact fingerprint that binds the serialized state to the exact reviewed model identity.
cacheLayoutFingerprint- The stable cache-layout fingerprint used to reject state created for incompatible tensor geometry or storage layout.
maximumByteCount- The maximum byte count used to bound this operation; it must be nonnegative and within the supported range.
Returns: The serialized snapshot bytes including a trailing SHA-256.
ReferenceKvCacheSnapshotUAIX.LmRuntime.Models.Llama
9 members
Represents an immutable snapshot of a tiny reference key/value cache.
ConfigurationFingerprint
Gets the model/configuration fingerprint.
ContextLength
Gets the context capacity in the snapshot.
HeadWidth
Gets the per-head width.
Keys
Gets a copy of all key values.
KeyValueHeadCount
Gets the key/value head count.
LayerCount
Gets the number of layers in the snapshot.
SchemaVersion
Gets the snapshot schema version.
UsedTokenCount
Gets the used token count.
Values
Gets a copy of all value values.
ReferenceKvPortableSnapshotUAIX.LmRuntime.Models.Llama
6 members
Carries a deterministic portable key/value-cache snapshot and its compatibility identities.
CacheLayoutFingerprint
Gets the cache-layout fingerprint.
ConfigurationFingerprint
Gets the model-configuration fingerprint.
ContentSha256
Gets the SHA-256 of the serialized bytes preceding the digest field.
ModelArtifactFingerprint
Gets the optional model-artifact fingerprint.
SchemaVersion
Gets the portable schema version.
Snapshot
Gets the restored capacity-shaped snapshot.
ReferenceKvWriteBehaviorUAIX.LmRuntime.Models.Llama
1 member
Identifies the deterministic write semantics used by the scalar reference key/value cache.
AppendOrOverwrite
Writes append new positions and deterministically overwrite already written positions.
TensorBindingDiagnosticUAIX.LmRuntime.Models.Llama
4 members
Represents a tensor binding diagnostic.
BlockIndex
Gets the optional transformer block index.
Code
Gets the diagnostic code.
Message
Gets the diagnostic message.
TensorName
Gets the tensor name associated with the diagnostic.
TensorBindingEntryUAIX.LmRuntime.Models.Llama
10 members
Represents one bound tensor entry.
AbsoluteOffset
Gets the absolute source-file offset.
ByteLength
Gets the physical storage byte length.
DataType
Gets the mapped runtime data type.
Descriptor
Gets the GGUF tensor descriptor supplying storage.
IsAlias
Gets whether this binding aliases another tensor.
LogicalDimensions
Gets the normalized logical dimensions.
Ownership
Gets the ownership contract.
Requirement
Gets the tensor requirement.
SourceTensorName
Gets the source tensor name when this binding is an alias.
StorageKind
Gets the storage kind.
TensorBindingExceptionUAIX.LmRuntime.Models.Llama
2 members
Represents a failed LLaMA tensor schema binding operation.
Manifest
Gets the failed binding manifest.
TensorBindingException(UAIX.LmRuntime.Models.Llama.TensorBindingManifest)
Initializes a binding exception from a failed manifest.
manifest- The validated manifest that binds tensor requirements, model identity, and storage diagnostics used by the operation.
TensorBindingManifestUAIX.LmRuntime.Models.Llama
5 members
Represents the result of LLaMA tensor binding.
Bindings
Gets bound tensor entries.
Diagnostics
Gets binding diagnostics.
IsComplete
Gets a value indicating whether every required tensor was bound without diagnostics.
ThrowIfIncomplete
Throws when the manifest contains one or more diagnostics.
TryGetBinding(UAIX.LmRuntime.Models.Llama.LlamaTensorRole,System.Nullable<int>,UAIX.LmRuntime.Models.Llama.TensorBindingEntry@)
Attempts to find one bound tensor by semantic role and optional block index.
role- The semantic LLaMA tensor role used to select the required bound tensor from the validated manifest.
blockIndex- The zero-based block index; it must identify an existing position within the relevant validated range.
entry- When the method returns, contains the entry produced by the operation when successful; otherwise contains the type's default value.
Returns: True when try get binding succeeds for the supplied values; otherwise, false.
TensorBindingOptionsUAIX.LmRuntime.Models.Llama
4 members
Configures semantic validation performed by UAIX.LmRuntime.Models.Llama.LlamaTensorBinder.
AllowTiedOutputProjection
Gets whether a missing output.weight may alias token_embd.weight.
ValidateByteLengths
Gets whether physical byte lengths must match the registered tensor type traits.
ValidateFileBounds
Gets whether tensor ranges must fit inside the parsed source file length when available.
ValidateSemanticShapes
Gets whether dimensions derived from model metadata must match the GGUF storage shape.
TensorBindingOwnershipUAIX.LmRuntime.Models.Llama
2 members
Identifies ownership for a bound tensor payload.
BorrowedAlias
The binding borrows storage through another tensor binding.
BorrowedModelStorage
The binding borrows storage owned by the loaded model.
TensorBindingStorageKindUAIX.LmRuntime.Models.Llama
2 members
Identifies where a bound tensor payload is stored.
Alias
The tensor is an alias of another bound tensor.
MemoryMappedFile
The tensor remains in the GGUF memory-mapped artifact.
WeightSourceStorageDiagnosticsUAIX.LmRuntime.Models.Llama
7 members
Describes immutable storage used by one deterministic reference weight source.
ByteLength
Gets the physical byte length.
DataType
Gets the logical runtime data type.
IsAlias
Gets a value indicating whether this source aliases another semantic binding.
IsMemoryMapped
Gets a value indicating whether the source borrows memory-mapped storage.
ManagedCopiedByteCount
Gets the number of bytes copied into persistent managed model-weight storage.
StorageType
Gets the GGML physical storage type.
TensorName
Gets the semantic tensor name.
UAIX.LmRuntime.Backends.CpuManaged
2 documented types and 8 documented members.
CpuManagedRuntimeBackendUAIX.LmRuntime.Backends.CpuManaged
7 members
Reports the UAIX managed CPU backend to the acceleration registry.
This backend represents compatibility required for systems without GPUs. It does not require GPU hardware, GPU drivers, native runtime libraries, provider APIs, subprocess execution, or model downloads.
BackendId
The stable backend identifier used by selection policies.
Capabilities
Gets the backend capability declaration.
DisplayName
Gets the human-readable backend name.
Id
Gets the stable backend identifier.
Kind
Gets the backend kind.
CpuManagedRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.CpuManaged.CpuManagedRuntimeBackend class.
ProbeAsync(UAIX.LmRuntime.Acceleration.RuntimeBackendOptions,System.Threading.CancellationToken)
Probes the managed.NET backend for local execution availability.
options- The caller-supplied probing and selection options.
cancellationToken- A token that cancels the probe.
Returns: The available CPU probe result.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.CpuManaged
1 member
Provides registration helpers for the managed CPU backend.
Hosts can call this helper to make CPU fallback explicit in the acceleration registry.
AddUaixCpuManagedBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX managed CPU backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
UAIX.LmRuntime.Backends.Cuda
2 documented types and 3 documented members.
CudaRuntimeBackendUAIX.LmRuntime.Backends.Cuda
2 members
Reports CUDA backend compatibility and diagnostics to the acceleration registry.
This package does not contain hidden CUDA inference binaries. Native assets and device proof must come from a declared native package or host adapter before CUDA execution can be selected.
BackendId
The stable backend identifier used by selection policies.
CudaRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.Cuda.CudaRuntimeBackend class.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.Cuda
1 member
Provides registration helpers for the CUDA backend.
Registration only exposes diagnostics and selection metadata. It does not authorize hidden native execution.
AddUaixCudaBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX CUDA backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
UAIX.LmRuntime.Backends.DirectML
2 documented types and 3 documented members.
DirectMlRuntimeBackendUAIX.LmRuntime.Backends.DirectML
2 members
Reports DirectML backend compatibility and diagnostics to the acceleration registry.
DirectML support is declared as a package-visible compatibility path for Windows hosts. This package performs no hidden native inference and reports unavailable until a host supplies native runtime and device evidence.
BackendId
The stable backend identifier used by selection policies.
DirectMlRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.DirectML.DirectMlRuntimeBackend class.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.DirectML
1 member
Provides registration helpers for the DirectML backend.
Registration only exposes diagnostics and selection metadata. It does not authorize hidden native execution.
AddUaixDirectMlBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX DirectML backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
UAIX.LmRuntime.Backends.Vulkan
2 documented types and 3 documented members.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.Vulkan
1 member
Provides registration helpers for the Vulkan backend.
Registration only exposes diagnostics and selection metadata. It does not authorize hidden native execution.
AddUaixVulkanBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX Vulkan backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
VulkanRuntimeBackendUAIX.LmRuntime.Backends.Vulkan
2 members
Reports Vulkan backend compatibility and diagnostics to the acceleration registry.
Vulkan support is declared for vendor-diverse hardware. This package does not contain hidden Vulkan inference binaries and reports unavailable until runtime and device evidence exists.
BackendId
The stable backend identifier used by selection policies.
VulkanRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.Vulkan.VulkanRuntimeBackend class.
UAIX.LmRuntime.Backends.Rocm
2 documented types and 3 documented members.
RocmRuntimeBackendUAIX.LmRuntime.Backends.Rocm
2 members
Reports ROCm backend compatibility and diagnostics to the acceleration registry.
ROCm support is declared for AMD GPU systems. This package does not contain hidden ROCm inference binaries and reports unavailable until runtime and device evidence exists.
BackendId
The stable backend identifier used by selection policies.
RocmRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.Rocm.RocmRuntimeBackend class.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.Rocm
1 member
Provides registration helpers for the ROCm backend.
Registration only exposes diagnostics and selection metadata. It does not authorize hidden native execution.
AddUaixRocmBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX ROCm backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
UAIX.LmRuntime.Backends.Metal
2 documented types and 3 documented members.
MetalRuntimeBackendUAIX.LmRuntime.Backends.Metal
2 members
Reports Metal backend compatibility and diagnostics to the acceleration registry.
Metal support is declared for Apple GPU systems. This package does not contain hidden Metal inference binaries and reports unavailable until runtime and device evidence exists.
BackendId
The stable backend identifier used by selection policies.
MetalRuntimeBackend
Initializes a new instance of the UAIX.LmRuntime.Backends.Metal.MetalRuntimeBackend class.
RuntimeBackendRegistryExtensionsUAIX.LmRuntime.Backends.Metal
1 member
Provides registration helpers for the Metal backend.
Registration only exposes diagnostics and selection metadata. It does not authorize hidden native execution.
AddUaixMetalBackend(UAIX.LmRuntime.Acceleration.IRuntimeBackendRegistry)
Adds the UAIX Metal backend to a registry.
registry- The registry to update.
Returns: The same registry instance for fluent configuration.
UAIX.LmRuntime.LocalEndpoint
23 documented types and 163 documented members.
LocalGgufAssociatedArtifactIdentityUAIX.LmRuntime.LocalEndpoint
4 members
Represents the path-free verified identity of one associated model artifact.
The identity is constructed only after the runtime verifies the current associated artifact file under the same local path, containment, reparse-point, stable-length, byte-count, and SHA-256 rules used for primary model admission. It deliberately omits private paths, file bytes, prompt content, generated content, UAIX content, registry state, and audit state.
ArtifactByteCount
Gets the verified current associated-artifact file length in bytes.
ArtifactSha256
Gets the normalized lowercase SHA-256 digest verified from current associated-artifact bytes.
FileName
Gets the reviewed associated-artifact leaf file name.
Role
Gets the caller-owned artifact role label copied into path-free evidence.
LocalGgufAssociatedArtifactLoadInputUAIX.LmRuntime.LocalEndpoint
5 members
Supplies one LocalEndpoint-reviewed associated artifact for verification by the managed runtime facade.
The private path is accepted only as an input needed to verify current local bytes. The returned verification identity omits the path and contains only display-safe role, file name, SHA-256, and byte-count evidence.
ArtifactByteCount
Gets the required current associated artifact file length in bytes.
ArtifactPath
Gets the private direct-local path to the associated artifact file that LocalEndpoint already reviewed.
ArtifactSha256
Gets the required SHA-256 digest for the current associated artifact file.
FileName
Gets the reviewed leaf file name expected at UAIX.LmRuntime.LocalEndpoint.LocalGgufAssociatedArtifactLoadInput.ArtifactPath.
Role
Gets the caller-owned role label for the associated artifact.
LocalGgufExecutionLimitsUAIX.LmRuntime.LocalEndpoint
3 members
Defines caller-visible resource ceilings for one verified model owner.
The limits are enforced before large generation buffers are allocated. They supplement, but do not replace, the model context-window checks performed by the underlying deterministic runtime.
MaximumGeneratedTokens
Gets the maximum generated-token buffer length permitted for one operation.
MaximumPromptCharacters
Gets the maximum accepted prepared-prompt length in.NET UTF-16 code units.
MaximumStopTokenCount
Gets the maximum number of caller-defined stop-token identifiers permitted for one operation.
LocalGgufFileExpectationUAIX.LmRuntime.LocalEndpoint
2 members
Declares the path-free model-file evidence that LocalEndpoint requires before loading or reusing a model.
The contract contains only a SHA-256 digest and byte count. It intentionally excludes the private model path, model bytes, prompt content, generated content, registry state, audit state, and UAIX package data. Runtime entry points validate both values against current file bytes before model parsing or cached-model reuse.
ModelByteCount
Gets the required current model file length in bytes.
ModelSha256
Gets the required SHA-256 digest for the current local model file.
LocalGgufFileIdentityUAIX.LmRuntime.LocalEndpoint
2 members
Represents the verified content identity of one current local GGUF file.
The result contains only a digest and byte count. It deliberately omits the local path, model bytes, prompt content, generated content, UAIX state, and caller policy so it can be copied into LocalEndpoint evidence without expanding the runtime's persistence or trust responsibilities.
ModelByteCount
Gets the verified current file length in bytes.
ModelSha256
Gets the normalized lowercase SHA-256 digest verified from current file bytes.
LocalGgufGeneratedTokenUAIX.LmRuntime.LocalEndpoint
4 members
Describes one generated token selected by the managed runtime for caller-owned streaming.
The value intentionally excludes prompt text, decoded output, model paths, UAIX content, registry state, audit state, and transport details. LocalEndpoint remains responsible for detokenization policy, worker-event mapping, terminal-event policy, and persistence boundaries.
SelectedLogit
Gets the deterministic argmax logit associated with the selected token.
Sequence
Gets the zero-based selection sequence.
TokenId
Gets the selected model vocabulary identifier.
LocalGgufGeneratedToken(int,int,float)
Initializes a new LocalGgufGeneratedToken instance with validated dependencies and operational bounds.
sequence- The zero-based selection sequence within the current generation operation.
tokenId- The token identifier to process; it must fall within the validated vocabulary and operation-specific range.
selectedLogit- The deterministic argmax logit associated with the selected token. NaN is rejected; positive and negative infinity are retained because the lower-level greedy sampler compares infinities deterministically and the runtime verifies the exact observed value against the completed generation result.
LocalGgufGenerationRequestUAIX.LmRuntime.LocalEndpoint
12 members
Defines one bounded deterministic greedy-generation request.
The request contains only in-memory generation controls. It does not define persistence, network, provider, process, server, telemetry, or model-download behavior.
AddSpecialTokens
Gets whether model-defined special-token insertion is enabled for the prompt.
CleanSpaces
Gets whether tokenizer-defined spacing cleanup is applied during detokenization.
EmitTokenizerTrace
Gets whether tokenizer trace entries are captured for the prompt.
EndOfSequenceTokenId
Gets the optional explicit end-of-sequence token identifier; when omitted, model metadata is used.
InvalidUtf16Policy
Gets the invalid UTF-16 handling policy used during tokenization.
MaximumTokens
Gets the positive maximum number of tokens to generate.
ParseSpecialTokens
Gets whether special-token text embedded in the prompt is parsed as a special token.
Prompt
Gets the exact prepared prompt to tokenize in memory.
RemoveSpecialTokens
Gets whether special tokens are removed from decoded generated text.
ResetSession
Gets whether the session is reset before prompt prefill.
StopTokenIds
Gets additional token identifiers that terminate generation after being emitted.
UnparseSpecialTokens
Gets whether special tokens are emitted as their raw token text when they are not removed.
LocalGgufGenerationResultUAIX.LmRuntime.LocalEndpoint
6 members
Represents deterministic generated token identifiers and their decoded text.
FinalSelectedLogit
Gets the selected logit of the final generated token, or negative infinity when no token was generated.
GeneratedText
Gets the decoded generated text.
GeneratedTokenIds
Gets the generated token identifiers in selection order.
Position
Gets the next sequence position retained by the isolated session.
PromptTokenCount
Gets the number of prompt tokens evaluated for the operation.
StopReason
Gets the deterministic generation stop reason.
LocalGgufModelUAIX.LmRuntime.LocalEndpoint
11 members
Owns one verified mapped GGUF model and creates isolated deterministic inference sessions.
ActiveSessionCount
Gets the number of caller-owned sessions that have not yet been disposed.
ExecutionLimits
Gets the immutable prompt and generation ceilings copied when this model was loaded.
IsDisposed
Gets a value indicating whether this model owner has been disposed.
Metadata
Gets verified model and runtime metadata.
Scope
Gets the opaque application scope and optional legacy package label associated with this model owner.
TokenizerMetadata
Gets the bounded tokenizer metadata projection.
CreateSession
Rejects creation of a LocalEndpoint inference session when required UAIX load-session context is absent.
Returns: No value is returned; this compatibility member always throws to require an explicit session-scoped UAIX context.
CreateSession(UAIX.LmRuntime.LocalEndpoint.LocalGgufSessionContext)
Creates a new inference session with independent sequence state and validated UAIX load-session evidence.
sessionContext- The LocalEndpoint worker session identifier and display-safe UAIX context.
Returns: An isolated deterministic session owned by the caller.
Decode(System.Collections.Generic.IReadOnlyList<int>,UAIX.LmRuntime.Tokenization.MetadataDrivenGgufTokenizerDetokenizationOptions)
Decodes token identifiers using the tokenizer metadata from the verified GGUF artifact.
tokenIds- The token identifiers to decode in sequence order.
options- The optional MetadataDrivenGgufTokenizerDetokenizationOptions controlling Decode; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The decoded text produced from the validated token sequence in the original sequence order.
Dispose
Releases the memory-mapped model after all child sessions have been disposed.
Tokenize(string,UAIX.LmRuntime.Tokenization.TokenizationOptions)
Tokenizes text using exact metadata loaded from the verified GGUF artifact.
text- The text processed by the configured encoding or normalization rules; it must satisfy the declared nullability contract.
options- The optional TokenizationOptions controlling Tokenize; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: The exact integer token identifiers and optional trace.
LocalGgufModelLoadOptionsUAIX.LmRuntime.LocalEndpoint
7 members
Configures loading of one trusted local GGUF artifact through the LocalEndpoint-facing managed runtime surface.
The options govern local file validation and bounded managed execution only. They do not enable downloading, remote inference, process launch, telemetry, server hosting, or persistent runtime state.
AllowedRootDirectory
Gets the optional normalized directory that must contain the resolved GGUF file.
BindingOptions
Gets semantic LLaMA tensor-binding validation controls.
ExecutionLimits
Gets the prompt, generated-token, and stop-token ceilings enforced by sessions created from the model.
MaximumModelBytes
Gets the maximum accepted model file length in bytes.
MaximumReferenceMaterializationBytes
Gets the maximum number of bytes that compatibility-only float32 materialization may allocate.
ParseOptions
Gets GGUF parser safety limits.
RejectReparsePoints
Gets a value indicating whether symbolic links and reparse points are rejected before the file is opened.
LocalGgufModelMetadataUAIX.LmRuntime.LocalEndpoint
11 members
Describes the verified model and execution metadata exposed to the LocalEndpoint worker adapter.
Architecture
Gets the model architecture identifier.
AssociatedArtifacts
Gets the verified path-free identities of associated model artifacts supplied by LocalEndpoint.
BoundTensorCount
Gets the number of tensors accepted by semantic binding.
ContextLength
Gets the configured context length in tokens.
GgufVersion
Gets the GGUF container version.
ModelByteCount
Gets the verified model file length in bytes.
ModelName
Gets the optional model display name from GGUF metadata.
ModelSha256
Gets the lowercase SHA-256 digest verified before model loading.
StorageSummary
Gets the mapped storage summary used by the managed reference runtime.
Tokenizer
Gets the exact tokenizer implementation name selected from GGUF metadata.
VocabularySize
Gets the configured vocabulary size.
LocalGgufRuntimeUAIX.LmRuntime.LocalEndpoint
11 members
Loads verified local GGUF files into isolated pure-managed LLaMA runtime models.
This type performs local file validation and direct managed model loading only. It contains no subprocess, network, provider, telemetry, server, downloader, command-dispatch, or persistent-profile behavior. The type holds no static mutable model, session, profile, or wiki state. Separate application instances can therefore load independent models and bind each inference session to a distinct LocalEndpoint-owned UAIX load session.
Capabilities
Gets the fixed local-only capability declaration.
Scope
Gets the opaque caller-owned runtime scope.
LoadVerifiedModel(string,string,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Loads a local GGUF model only after its exact SHA-256 digest and file-policy constraints are verified.
modelPath- The local file-system model path processed by this operation; it must satisfy the containing component's path and scope policy.
expectedSha256- The expected lowercase or uppercase 64-character SHA-256 digest used to authenticate the complete artifact; a mismatch fails the operation before a trusted result is published.
options- Optional bounded parser, binding, containment, and file-validation controls.
Returns: An owned mapped model that must be disposed after all child sessions are disposed.
LoadVerifiedModel(string,UAIX.LmRuntime.LocalEndpoint.LocalGgufFileExpectation,System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.LocalEndpoint.LocalGgufAssociatedArtifactLoadInput>,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Loads a local GGUF model after binding the primary artifact and associated artifacts to caller-owned evidence.
modelPath- The local file-system model path processed by this operation; it must satisfy the containing component's path and scope policy.
expectation- The path-free digest and byte-count evidence required by LocalEndpoint for the primary GGUF artifact.
associatedArtifacts- The LocalEndpoint-reviewed associated artifacts to verify before the model is published.
options- Optional bounded parser, binding, containment, and file-validation controls.
Returns: An owned mapped model that must be disposed after all child sessions are disposed.
LoadVerifiedModel(string,UAIX.LmRuntime.LocalEndpoint.LocalGgufFileExpectation,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Loads a local GGUF model after binding current file bytes to caller-owned digest and byte-count evidence.
modelPath- The local file-system model path processed by this operation; it must satisfy the containing component's path and scope policy.
expectation- The path-free digest and byte-count evidence required by LocalEndpoint.
options- Optional bounded parser, binding, containment, and file-validation controls.
Returns: An owned mapped model that must be disposed after all child sessions are disposed.
LocalGgufRuntime
Initializes an unscoped model facade for compatibility with callers that do not require application labels.
LocalGgufRuntime(UAIX.LmRuntime.LocalEndpoint.LocalGgufRuntimeScope)
Initializes a runtime facade for one LocalEndpoint application instance and optional legacy package label.
scope- The immutable runtime scope whose name and instance identifier are validated once and then retained as path-free correlation evidence for model verification operations.
VerifyAssociatedArtifacts(System.Collections.Generic.IReadOnlyList<UAIX.LmRuntime.LocalEndpoint.LocalGgufAssociatedArtifactLoadInput>,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Verifies LocalEndpoint-reviewed associated artifacts without exposing private local paths.
associatedArtifacts- The LocalEndpoint-reviewed associated artifact load inputs to verify before model publication or worker execution proceeds.
options- Optional bounded parser, binding, containment, and file-validation controls shared with primary model verification.
Returns: Path-free associated-artifact identities sorted by role, file name, and SHA-256 digest.
VerifyLocalModelFile(string,string,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Verifies the current bytes and file-policy boundaries of one local GGUF artifact without loading a model.
modelPath- The local file-system model path processed by this operation; it must satisfy the containing component's path and scope policy.
expectedSha256- The expected lowercase or uppercase 64-character SHA-256 digest used to authenticate the complete artifact; a mismatch fails the operation before a trusted result is published.
options- The optional LocalGgufModelLoadOptions controlling VerifyLocalModelFile; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: A path-free identity containing the verified digest and current byte count.
VerifyLocalModelFile(string,UAIX.LmRuntime.LocalEndpoint.LocalGgufFileExpectation,UAIX.LmRuntime.LocalEndpoint.LocalGgufModelLoadOptions)
Verifies current local model bytes against caller-owned digest and byte-count evidence without loading a model.
modelPath- The local file-system model path processed by this operation; it must satisfy the containing component's path and scope policy.
expectation- The path-free digest and byte-count evidence required by LocalEndpoint.
options- The optional LocalGgufModelLoadOptions controlling VerifyLocalModelFile; null selects the documented defaults, supplied limits are validated before allocation, and the instance is not mutated.
Returns: A path-free identity containing the verified digest and current byte count.
VerifyUaixRuntimeContext(UAIX.LmRuntime.LocalEndpoint.LocalUaixRuntimeContext)
Validates LocalEndpoint-supplied uaixRuntimeContext metadata without opening package or wiki files.
context- The display-safe profile, load-session, and long-term memory routing metadata.
Returns: An immutable path-relative evidence projection with a canonical context SHA-256.
LocalGgufRuntimeCapabilitiesUAIX.LmRuntime.LocalEndpoint
17 members
Declares the fixed local-only capabilities of the LocalEndpoint integration facade.
EmitsWorkerJsonlEvents
Gets a value indicating whether this managed facade emits LocalEndpoint worker JSONL events.
HostsServer
Gets a value indicating whether the integration hosts a server or public listener.
IsManagedExecution
Gets a value indicating whether model execution is implemented by managed project code.
ParsesUaixPackages
Gets a value indicating whether the runtime opens or parses.uaix package containers.
PersistsRuntimeContent
Gets a value indicating whether the integration persists prompts, generated text, or session state.
SupportsAssociatedArtifactVerification
Gets a value indicating whether callers can bind reviewed companion artifacts to path-free evidence.
SupportsExplicitFileVerification
Gets a value indicating whether callers can explicitly re-verify current model-file bytes before reuse.
SupportsSynchronousTokenObservation
Gets a value indicating whether caller-owned synchronous token observation is supported.
SupportsUaixRuntimeContextValidation
Gets a value indicating whether the facade validates LocalEndpoint-supplied UAIX runtime context metadata.
UaixMemoryGrantsAuthority
Gets a value indicating whether UAIX memory metadata can grant runtime or external authority.
UsesAssociatedArtifactsForGeneration
Gets a value indicating whether associated artifacts alter managed token generation.
UsesGlobalUaixProfile
Gets a value indicating whether the runtime assumes one process-global active UAIX profile.
UsesGpuAcceleration
Gets a value indicating whether the integration uses GPU or CUDA execution.
UsesNativeInference
Gets a value indicating whether model execution delegates to a native inference library.
UsesNetworkAccess
Gets a value indicating whether the integration performs network access.
UsesSubprocesses
Gets a value indicating whether the integration starts subprocesses.
UsesTelemetry
Gets a value indicating whether the integration emits telemetry.
LocalGgufRuntimeScopeUAIX.LmRuntime.LocalEndpoint
2 members
Identifies one caller-owned LocalEndpoint application scope and optional legacy UAIX package correlation label.
The values are opaque compatibility labels. They do not identify the active profile or per-application-instance load session; that evidence is supplied to UAIX.LmRuntime.LocalEndpoint.LocalGgufModel.CreateSession(UAIX.LmRuntime.LocalEndpoint.LocalGgufSessionContext). The runtime never opens, interprets, persists, or synchronizes UAIX content.
ApplicationInstanceId
Gets the LocalEndpoint application instance identifier.
UaixPackageId
Gets an optional caller-owned legacy UAIX package correlation identifier.
LocalGgufSessionUAIX.LmRuntime.LocalEndpoint
8 members
Owns isolated deterministic sequence and key/value-cache state for one verified local model.
A session serializes generation, reset, and disposal transitions through one lifecycle authority. It stores no prompt, generated text, audit record, registry record, UAIX content, provider state, or network state.
IsDisposed
Gets a value indicating whether this session has been disposed.
Position
Gets the next sequence position retained by this isolated session.
SessionId
Gets the validated LocalEndpoint worker session identifier preserved for this inference session.
UaixRuntimeContextEvidence
Gets immutable, path-relative evidence for the profile and UAIX load session bound to this inference session.
Dispose
Releases isolated session state without disposing the shared model owner.
GenerateGreedy(UAIX.LmRuntime.LocalEndpoint.LocalGgufGenerationRequest,System.Action<UAIX.LmRuntime.LocalEndpoint.LocalGgufGeneratedToken>,System.Threading.CancellationToken)
Runs bounded deterministic greedy generation and synchronously reports each selected token.
request- The LocalGgufGenerationRequest containing the complete caller-owned inputs for GenerateGreedy; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
tokenObserver- The caller-owned observer invoked once for each selected token.
cancellationToken- A token observed before work and between committed model steps.
Returns: The generated token identifiers, text, stop reason, position, and final selected logit.
GenerateGreedy(UAIX.LmRuntime.LocalEndpoint.LocalGgufGenerationRequest,System.Threading.CancellationToken)
Runs bounded deterministic greedy generation and returns exact token identifiers with decoded text.
request- The LocalGgufGenerationRequest containing the complete caller-owned inputs for GenerateGreedy; required fields are validated and mutable collections are snapshotted before state changes or large allocations.
cancellationToken- A token observed before work and between committed model steps.
Returns: The generated token identifiers, text, stop reason, position, and final selected logit.
Reset
Clears sequence position and key/value-cache state retained by this session.
LocalGgufSessionContextUAIX.LmRuntime.LocalEndpoint
2 members
Represents one worker session identity and its required UAIX boundary metadata.
LocalEndpoint creates this value after package import, profile loading, prompt assembly, and policy validation. The runtime snapshots the context for one isolated session and retains no global active profile.
SessionId
Gets the LocalEndpoint worker session identifier that every mapped event must preserve.
UaixRuntimeContext
Gets the required display-safe UAIX profile and wiki routing metadata.
LocalGgufTokenizerMetadataUAIX.LmRuntime.LocalEndpoint
9 members
Provides a bounded tokenizer metadata projection for LocalEndpoint runtime readiness and diagnostics.
AddBos
Gets whether model metadata requests automatic beginning-of-sequence insertion.
AddEos
Gets whether model metadata requests automatic end-of-sequence insertion.
BosTokenId
Gets the beginning-of-sequence token identifier when defined.
ChatTemplate
Gets the optional chat template declared by the model.
EosTokenId
Gets the end-of-sequence token identifier when defined.
PreTokenizer
Gets the optional pre-tokenizer identifier.
TokenizerModel
Gets the tokenizer family declared by GGUF metadata.
UnknownTokenId
Gets the unknown-token identifier when defined.
VocabularySize
Gets the vocabulary size.
LocalGgufVerificationExceptionUAIX.LmRuntime.LocalEndpoint
5 members
Represents a local GGUF artifact verification failure.
ActualByteCount
Gets the observed current model byte count when available.
ActualSha256
Gets the normalized observed SHA-256 value when available.
ExpectedByteCount
Gets the required model byte count when supplied by caller-owned evidence.
ExpectedSha256
Gets the normalized expected SHA-256 value when available.
LocalGgufVerificationException(string,string,string,System.Exception,System.Nullable<long>,System.Nullable<long>)
Initializes a new LocalGgufVerificationException instance with validated dependencies and operational bounds.
message- The display-safe diagnostic message describing the failure without embedding prompt text, generated text, credentials, or private file contents.
expectedSha256- The expected lowercase or uppercase 64-character SHA-256 digest used to authenticate the complete artifact; a mismatch fails the operation before a trusted result is published.
actualSha256- The observed SHA-256 digest retained as path-free diagnostic evidence when available, or null when hashing did not complete.
innerException- The optional lower-level failure used only to select a bounded, path-free diagnostic category. The original exception object, message, stack trace, data, and file-name properties are never retained because they can contain private model paths or other caller-owned content.
expectedByteCount- The required file length when byte-count evidence is supplied.
actualByteCount- The actual byte count used to bound this operation; it must be nonnegative and within the supported range.
LocalUaixLongTermMemoryModeUAIX.LmRuntime.LocalEndpoint
1 member
Identifies the supported long-term memory routing modes for one validated UAIX load session.
The value describes a caller-owned memory route only. It does not authorize file access, runtime execution, command execution, network access, provider access, telemetry, export, or any other capability.
Isolated
Routes the load session to a profile-owned long-term memory root.
LocalUaixRuntimeContextUAIX.LmRuntime.LocalEndpoint
17 members
Represents display-safe UAIX profile and long-term memory routing metadata supplied by LocalEndpoint.
This type mirrors the uaixRuntimeContext object in localendpoint.worker.request.v1. The runtime does not open or parse a.uaix archive, read.uai files, read wiki documents, expand a profile, choose a profile, or infer authority from this object. LocalEndpoint validates and imports packages, assembles the prompt, and supplies this metadata after its own policy gates. Every authority field must remain.
AutoExportAllowed
Gets a value that must remain false because UAIX memory does not authorize automatic export.
CommandExecutionAllowed
Gets a value that must remain false because command execution requires a separate user-approved gate.
LoadedUaixLoadSessionId
Gets the LocalEndpoint load-session identifier that bound this profile to one desktop application instance.
LoadedUaixProfileDisplayName
Gets the display-safe name of the loaded UAIX profile.
LoadedUaixProfileId
Gets the portable identifier of the loaded UAIX profile.
LoadedUaixProfilePresent
Gets a value indicating whether LocalEndpoint supplied one validated and loaded UAIX profile.
LoadedUaixSessionRelativePath
Gets the app-local relative path to the LocalEndpoint load-session evidence record.
LoadedUaixUaiRelativePath
Gets the app-local relative path to the expanded profile.uai root.
LongTermMemoryMode
Gets whether the selected long-term memory root is profile-isolated or deliberately shared.
LongTermMemoryRootId
Gets the portable identifier of the selected Documents-backed long-term memory root.
LongTermMemoryRootRelativePath
Gets the path relative to Documents/LocalEndpoint/Wikis for the selected long-term memory root.
MemoryCanOverridePolicy
Gets a value that must remain false because memory cannot override LocalEndpoint policy.
NetworkAccessAllowed
Gets a value that must remain false because UAIX memory does not authorize network access.
ProviderApisAllowed
Gets a value that must remain false because UAIX memory does not authorize provider APIs.
RuntimeExecutionAllowed
Gets a value that must remain false because UAIX memory does not authorize model runtime execution.
TelemetryEnabled
Gets a value that must remain false because UAIX memory does not authorize telemetry.
WebsitePromptIntakeAllowed
Gets a value that must remain false because UAIX memory does not authorize website prompt intake.
LocalUaixRuntimeContextEvidenceUAIX.LmRuntime.LocalEndpoint
10 members
Represents the immutable, path-relative evidence produced after UAIX runtime context validation.
The evidence contains no prompt, generated text, package bytes, wiki document content, model path, credential, command grant, network grant, provider grant, or execution grant. The SHA-256 binds a canonical field sequence for caller-owned registry and audit evidence without turning the context into authority.
AuthorityBoundaryClosed
Gets a value indicating that every UAIX authority field was verified false.
ContextSha256
Gets the canonical lowercase SHA-256 for this validated context.
LoadedUaixLoadSessionId
Gets the validated LocalEndpoint load-session identifier.
LoadedUaixProfileDisplayName
Gets the validated display-safe profile name.
LoadedUaixProfileId
Gets the validated loaded profile identifier.
LoadedUaixSessionRelativePath
Gets the validated app-local load-session relative path.
LoadedUaixUaiRelativePath
Gets the validated app-local profile.uai relative path.
LongTermMemoryMode
Gets the validated long-term memory routing mode.
LongTermMemoryRootId
Gets the validated long-term memory root identifier.
LongTermMemoryRootRelativePath
Gets the validated Documents-backed wiki relative path.
LocalUaixRuntimeContextExceptionUAIX.LmRuntime.LocalEndpoint
2 members
Represents a fail-closed UAIX runtime context validation error.
The exception exposes a stable code and field name for bounded diagnostics. It does not include prompt text, generated text, profile file contents, wiki document contents, credentials, or private absolute paths.
FailureCode
Gets the stable fail-closed validation code.
FieldName
Gets the display-safe contract field associated with the failure.
LocalUaixRuntimeContextFailureCodeUAIX.LmRuntime.LocalEndpoint
12 members
Identifies the fail-closed reason produced while validating LocalEndpoint UAIX runtime context metadata.
Codes are stable, display-safe boundary identifiers. They do not include prompt text, generated text, private absolute paths, package bytes, wiki document contents, credentials, or unrestricted exception details.
AuthorityEscalation
One or more UAIX memory fields attempted to grant execution or external authority.
InvalidDisplayValue
A display-safe value contained unsupported control data or exceeded its limit.
InvalidIdentifier
An identifier was outside the bounded portable identifier grammar.
InvalidRelativePath
A relative path was absolute, traversing, malformed, or outside the supported portable form.
InvalidWorkerSession
The worker session identifier was absent or malformed.
LongTermMemoryPathMismatch
The long-term memory path did not match its declared mode and root identity.
MissingContext
The required runtime context object was absent.
MissingValue
A required identifier or display-safe value was absent.
ProfileNotLoaded
The caller did not identify a loaded UAIX profile.
ProfilePathMismatch
The profile-relative path did not match the declared profile identifier.
SessionPathMismatch
The load-session-relative path did not match the declared load-session identifier.
UnsupportedLongTermMemoryMode
The long-term memory mode was not one of the explicitly supported values.