phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-08-07 02:58:21 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Simplify Tags

Browse source
This commit is contained in:
Ashwin Bharambe 2025-01-29 09:30:44 -08:00
parent 3e2a751f54
commit e3174bd62d
5 changed files with 55 additions and 1528 deletions
Download patch file Download diff file Expand all files Collapse all files

14
docs/openapi_generator/pyopenapi/generator.py
View file

@ -677,12 +677,6 @@ class Generator:
) )
) )
# types that are produced/consumed by operations
type_tags = [
self._build_type_tag(ref, schema)
for ref, schema in self.schema_builder.schemas.items()
]
# types that are emitted by events # types that are emitted by events
event_tags: List[Tag] = [] event_tags: List[Tag] = []
events = get_endpoint_events(self.endpoint) events = get_endpoint_events(self.endpoint)
@ -709,7 +703,6 @@ class Generator:
# list all operations and types # list all operations and types
tags: List[Tag] = [] tags: List[Tag] = []
tags.extend(operation_tags) tags.extend(operation_tags)
tags.extend(type_tags)
tags.extend(event_tags) tags.extend(event_tags)
for extra_tag_group in extra_tag_groups.values(): for extra_tag_group in extra_tag_groups.values():
tags.extend(extra_tag_group) tags.extend(extra_tag_group)
@ -724,13 +717,6 @@ class Generator:
tags=sorted(tag.name for tag in operation_tags), tags=sorted(tag.name for tag in operation_tags),
) )
) )
if type_tags:
tag_groups.append(
TagGroup(
name=self.options.map("Types"),
tags=sorted(tag.name for tag in type_tags),
)
)
if event_tags: if event_tags:
tag_groups.append( tag_groups.append(
TagGroup( TagGroup(

944
docs/resources/llama-stack-spec.html
View file

File diff suppressed because it is too large Load diff

608
docs/resources/llama-stack-spec.yaml
View file

@ -1817,9 +1817,7 @@ components:
- function_tag - function_tag
- python_list - python_list
title: >- title: >-
This Enum refers to the prompt format for calling custom / zero shot tools Prompt format for calling custom / zero shot tools.
description: >-
The detailed prompts for each of these formats are added to llama cli
response_format: response_format:
$ref: '#/components/schemas/ResponseFormat' $ref: '#/components/schemas/ResponseFormat'
logprobs: logprobs:
@ -2225,9 +2223,7 @@ components:
- function_tag - function_tag
- python_list - python_list
title: >- title: >-
This Enum refers to the prompt format for calling custom / zero shot tools Prompt format for calling custom / zero shot tools.
description: >-
The detailed prompts for each of these formats are added to llama cli
max_infer_iters: max_infer_iters:
type: integer type: integer
default: 10 default: 10
@ -4905,411 +4901,54 @@ components:
security: security:
- Default: [] - Default: []
tags: tags:
- name: AgentCandidate
description: ''
- name: AgentConfig
description: ''
- name: AgentCreateResponse
description: ''
- name: AgentSessionCreateResponse
description: ''
- name: AgentStepResponse
description: ''
- name: AgentTool
description: ''
- name: AgentTurnInputType
description: ''
- name: AgentTurnResponseEvent
description: ''
- name: AgentTurnResponseEventPayload
description: ''
- name: AgentTurnResponseStepCompletePayload
description: ''
- name: AgentTurnResponseStepProgressPayload
description: ''
- name: AgentTurnResponseStepStartPayload
description: ''
- name: AgentTurnResponseStreamChunk
description: streamed agent turn completion response.
- name: AgentTurnResponseTurnCompletePayload
description: ''
- name: AgentTurnResponseTurnStartPayload
description: ''
- name: Agents - name: Agents
- name: AggregationFunctionType description: >-
description: '' Main functionalities provided by this API:
- name: AlgorithmConfig
description: '' - Create agents with specific instructions and ability to use tools.
- name: AppEvalTaskConfig
description: '' - Interactions with agents are grouped into sessions ("threads"), and each interaction
- name: AppendRowsRequest is called a "turn".
description: ''
- name: ArrayType - Agents can be provided with various tools (see the ToolGroups and ToolRuntime
description: '' APIs for more details).
- name: BasicScoringFnParams
description: '' - Agents can be provided with various shields (see the Safety API for more details).
- name: BatchChatCompletionRequest
description: '' - Agents can also use Memory to retrieve information from knowledge bases. See
- name: BatchChatCompletionResponse the RAG Tool and Vector IO APIs for more details.
description: '' x-displayName: >-
- name: BatchCompletionRequest Agents API for creating and interacting with agentic systems.
description: ''
- name: BatchCompletionResponse
description: ''
- name: BatchInference (Coming Soon) - name: BatchInference (Coming Soon)
- name: BenchmarkEvalTaskConfig
description: ''
- name: BooleanType
description: ''
- name: CancelTrainingJobRequest
description: ''
- name: ChatCompletionInputType
description: ''
- name: ChatCompletionRequest
description: ''
- name: ChatCompletionResponse
description: Response from a chat completion request.
- name: ChatCompletionResponseEvent
description: >-
An event during chat completion generation.
- name: ChatCompletionResponseStreamChunk
description: >-
A chunk of a streamed chat completion response.
- name: Checkpoint
description: Checkpoint created during training runs
- name: CompletionInputType
description: ''
- name: CompletionMessage
description: >-
A message containing the model's (assistant) response in a chat conversation.
- name: CompletionRequest
description: ''
- name: CompletionResponse
description: Response from a completion request.
- name: CompletionResponseStreamChunk
description: >-
A chunk of a streamed completion response.
- name: ContentDelta
description: ''
- name: CreateAgentRequest
description: ''
- name: CreateAgentSessionRequest
description: ''
- name: CreateAgentTurnRequest
description: ''
- name: DPOAlignmentConfig
description: ''
- name: DataConfig
description: ''
- name: Dataset
description: ''
- name: DatasetFormat
description: ''
- name: DatasetIO - name: DatasetIO
- name: Datasets - name: Datasets
- name: DefaultRAGQueryGeneratorConfig
description: ''
- name: EfficiencyConfig
description: ''
- name: EmbeddingsRequest
description: ''
- name: EmbeddingsResponse
description: >-
Response containing generated embeddings.
- name: Eval - name: Eval
- name: EvalCandidate
description: ''
- name: EvalTask
description: ''
- name: EvalTaskConfig
description: ''
- name: EvalTasks - name: EvalTasks
- name: EvaluateResponse
description: ''
- name: EvaluateRowsRequest
description: ''
- name: Event
description: ''
- name: GrammarResponseFormat
description: >-
Configuration for grammar-guided response generation.
- name: GreedySamplingStrategy
description: ''
- name: HealthInfo
description: ''
- name: ImageContentItem
description: ''
- name: ImageDelta
description: ''
- name: Inference - name: Inference
- name: InferenceStep description: >-
description: '' This API provides the raw interface to the underlying models. Two kinds of models
- name: InsertChunksRequest are supported:
description: ''
- name: InsertRequest - LLM models: these models generate "raw" and "chat" (conversational) completions.
description: ''
- Embedding models: these models generate embeddings to be used for semantic
search.
x-displayName: >-
Llama Stack Inference API for generating completions, chat completions, and
embeddings.
- name: Inspect - name: Inspect
- name: InterleavedContent
description: ''
- name: InterleavedContentItem
description: ''
- name: InvokeToolRequest
description: ''
- name: Job
description: ''
- name: JobStatus
description: ''
- name: JsonSchemaResponseFormat
description: >-
Configuration for JSON schema-guided response generation.
- name: JsonType
description: ''
- name: LLMAsJudgeScoringFnParams
description: ''
- name: LLMRAGQueryGeneratorConfig
description: ''
- name: ListDatasetsResponse
description: ''
- name: ListEvalTasksResponse
description: ''
- name: ListModelsResponse
description: ''
- name: ListPostTrainingJobsResponse
description: ''
- name: ListProvidersResponse
description: ''
- name: ListRoutesResponse
description: ''
- name: ListScoringFunctionsResponse
description: ''
- name: ListShieldsResponse
description: ''
- name: ListToolGroupsResponse
description: ''
- name: ListToolsResponse
description: ''
- name: ListVectorDBsResponse
description: ''
- name: LogEventRequest
description: ''
- name: LogSeverity
description: ''
- name: LoraFinetuningConfig
description: ''
- name: MemoryRetrievalStep
description: ''
- name: Message
description: ''
- name: MetricEvent
description: ''
- name: Model
description: ''
- name: ModelCandidate
description: ''
- name: ModelType
description: ''
- name: Models - name: Models
- name: NumberType
description: ''
- name: ObjectType
description: ''
- name: OptimizerConfig
description: ''
- name: OptimizerType
description: ''
- name: PaginatedRowsResult
description: ''
- name: ParamType
description: ''
- name: PostTraining (Coming Soon) - name: PostTraining (Coming Soon)
- name: PostTrainingJob
description: ''
- name: PostTrainingJobArtifactsResponse
description: Artifacts of a finetuning job.
- name: PostTrainingJobStatusResponse
description: Status of a finetuning job.
- name: PreferenceOptimizeRequest
description: ''
- name: ProviderInfo
description: ''
- name: QATFinetuningConfig
description: ''
- name: QueryChunksRequest
description: ''
- name: QueryChunksResponse
description: ''
- name: QueryCondition
description: ''
- name: QueryConditionOp
description: ''
- name: QueryRequest
description: ''
- name: QuerySpanTreeResponse
description: ''
- name: QuerySpansResponse
description: ''
- name: QueryTracesResponse
description: ''
- name: RAGDocument
description: ''
- name: RAGQueryConfig
description: ''
- name: RAGQueryGeneratorConfig
description: ''
- name: RAGQueryResult
description: ''
- name: RegexParserScoringFnParams
description: ''
- name: RegisterDatasetRequest
description: ''
- name: RegisterEvalTaskRequest
description: ''
- name: RegisterModelRequest
description: ''
- name: RegisterScoringFunctionRequest
description: ''
- name: RegisterShieldRequest
description: ''
- name: RegisterToolGroupRequest
description: ''
- name: RegisterVectorDbRequest
description: ''
- name: ResponseFormat
description: ''
- name: RouteInfo
description: ''
- name: RunEvalRequest
description: ''
- name: RunShieldRequest
description: ''
- name: RunShieldResponse
description: ''
- name: Safety - name: Safety
- name: SafetyViolation
description: ''
- name: SamplingParams
description: ''
- name: SamplingStrategy
description: ''
- name: SaveSpansToDatasetRequest
description: ''
- name: ScoreBatchRequest
description: ''
- name: ScoreBatchResponse
description: ''
- name: ScoreRequest
description: ''
- name: ScoreResponse
description: ''
- name: Scoring - name: Scoring
- name: ScoringFn
description: ''
- name: ScoringFnParams
description: ''
- name: ScoringFunctions - name: ScoringFunctions
- name: ScoringResult
description: ''
- name: Session
description: >-
A single session of an interaction with an Agentic System.
- name: Shield
description: >-
A safety shield resource that can be used to check content
- name: ShieldCallStep
description: ''
- name: Shields - name: Shields
- name: Span
description: ''
- name: SpanEndPayload
description: ''
- name: SpanStartPayload
description: ''
- name: SpanStatus
description: ''
- name: SpanWithStatus
description: ''
- name: StringType
description: ''
- name: StructuredLogEvent
description: ''
- name: StructuredLogPayload
description: ''
- name: SupervisedFineTuneRequest
description: ''
- name: SyntheticDataGenerateRequest
description: ''
- name: SyntheticDataGeneration (Coming Soon) - name: SyntheticDataGeneration (Coming Soon)
- name: SyntheticDataGenerationResponse
description: >-
Response from the synthetic data generation. Batch of (prompt, response, score)
tuples that pass the threshold.
- name: SystemMessage
description: >-
A system message providing instructions or context to the model.
- name: Telemetry - name: Telemetry
- name: TextContentItem
description: ''
- name: TextDelta
description: ''
- name: TokenLogProbs
description: Log probabilities for generated tokens.
- name: Tool
description: ''
- name: ToolCall
description: ''
- name: ToolCallDelta
description: ''
- name: ToolDef
description: ''
- name: ToolDefinition
description: ''
- name: ToolExecutionStep
description: ''
- name: ToolGroup
description: ''
- name: ToolGroups - name: ToolGroups
- name: ToolHost
description: ''
- name: ToolInvocationResult
description: ''
- name: ToolParamDefinition
description: ''
- name: ToolParameter
description: ''
- name: ToolResponse
description: ''
- name: ToolResponseMessage
description: >-
A message representing the result of a tool invocation.
- name: ToolRuntime - name: ToolRuntime
- name: TopKSamplingStrategy
description: ''
- name: TopPSamplingStrategy
description: ''
- name: Trace
description: ''
- name: TrainingConfig
description: ''
- name: Turn
description: >-
A single turn in an interaction with an Agentic System.
- name: URL
description: ''
- name: UnionType
description: ''
- name: UnstructuredLogEvent
description: ''
- name: UserMessage
description: >-
A message from the user in a chat conversation.
- name: VectorDB
description: ''
- name: VectorDBs - name: VectorDBs
- name: VectorIO - name: VectorIO
- name: VersionInfo
description: ''
- name: ViolationLevel
description: ''
x-tagGroups: x-tagGroups:
- name: Operations - name: Operations
tags: tags:
@ -5333,190 +4972,3 @@ x-tagGroups:
- ToolRuntime - ToolRuntime
- VectorDBs - VectorDBs
- VectorIO - VectorIO
- name: Types
tags:
- AgentCandidate
- AgentConfig
- AgentCreateResponse
- AgentSessionCreateResponse
- AgentStepResponse
- AgentTool
- AgentTurnInputType
- AgentTurnResponseEvent
- AgentTurnResponseEventPayload
- AgentTurnResponseStepCompletePayload
- AgentTurnResponseStepProgressPayload
- AgentTurnResponseStepStartPayload
- AgentTurnResponseStreamChunk
- AgentTurnResponseTurnCompletePayload
- AgentTurnResponseTurnStartPayload
- AggregationFunctionType
- AlgorithmConfig
- AppEvalTaskConfig
- AppendRowsRequest
- ArrayType
- BasicScoringFnParams
- BatchChatCompletionRequest
- BatchChatCompletionResponse
- BatchCompletionRequest
- BatchCompletionResponse
- BenchmarkEvalTaskConfig
- BooleanType
- CancelTrainingJobRequest
- ChatCompletionInputType
- ChatCompletionRequest
- ChatCompletionResponse
- ChatCompletionResponseEvent
- ChatCompletionResponseStreamChunk
- Checkpoint
- CompletionInputType
- CompletionMessage
- CompletionRequest
- CompletionResponse
- CompletionResponseStreamChunk
- ContentDelta
- CreateAgentRequest
- CreateAgentSessionRequest
- CreateAgentTurnRequest
- DPOAlignmentConfig
- DataConfig
- Dataset
- DatasetFormat
- DefaultRAGQueryGeneratorConfig
- EfficiencyConfig
- EmbeddingsRequest
- EmbeddingsResponse
- EvalCandidate
- EvalTask
- EvalTaskConfig
- EvaluateResponse
- EvaluateRowsRequest
- Event
- GrammarResponseFormat
- GreedySamplingStrategy
- HealthInfo
- ImageContentItem
- ImageDelta
- InferenceStep
- InsertChunksRequest
- InsertRequest
- InterleavedContent
- InterleavedContentItem
- InvokeToolRequest
- Job
- JobStatus
- JsonSchemaResponseFormat
- JsonType
- LLMAsJudgeScoringFnParams
- LLMRAGQueryGeneratorConfig
- ListDatasetsResponse
- ListEvalTasksResponse
- ListModelsResponse
- ListPostTrainingJobsResponse
- ListProvidersResponse
- ListRoutesResponse
- ListScoringFunctionsResponse
- ListShieldsResponse
- ListToolGroupsResponse
- ListToolsResponse
- ListVectorDBsResponse
- LogEventRequest
- LogSeverity
- LoraFinetuningConfig
- MemoryRetrievalStep
- Message
- MetricEvent
- Model
- ModelCandidate
- ModelType
- NumberType
- ObjectType
- OptimizerConfig
- OptimizerType
- PaginatedRowsResult
- ParamType
- PostTrainingJob
- PostTrainingJobArtifactsResponse
- PostTrainingJobStatusResponse
- PreferenceOptimizeRequest
- ProviderInfo
- QATFinetuningConfig
- QueryChunksRequest
- QueryChunksResponse
- QueryCondition
- QueryConditionOp
- QueryRequest
- QuerySpanTreeResponse
- QuerySpansResponse
- QueryTracesResponse
- RAGDocument
- RAGQueryConfig
- RAGQueryGeneratorConfig
- RAGQueryResult
- RegexParserScoringFnParams
- RegisterDatasetRequest
- RegisterEvalTaskRequest
- RegisterModelRequest
- RegisterScoringFunctionRequest
- RegisterShieldRequest
- RegisterToolGroupRequest
- RegisterVectorDbRequest
- ResponseFormat
- RouteInfo
- RunEvalRequest
- RunShieldRequest
- RunShieldResponse
- SafetyViolation
- SamplingParams
- SamplingStrategy
- SaveSpansToDatasetRequest
- ScoreBatchRequest
- ScoreBatchResponse
- ScoreRequest
- ScoreResponse
- ScoringFn
- ScoringFnParams
- ScoringResult
- Session
- Shield
- ShieldCallStep
- Span
- SpanEndPayload
- SpanStartPayload
- SpanStatus
- SpanWithStatus
- StringType
- StructuredLogEvent
- StructuredLogPayload
- SupervisedFineTuneRequest
- SyntheticDataGenerateRequest
- SyntheticDataGenerationResponse
- SystemMessage
- TextContentItem
- TextDelta
- TokenLogProbs
- Tool
- ToolCall
- ToolCallDelta
- ToolDef
- ToolDefinition
- ToolExecutionStep
- ToolGroup
- ToolHost
- ToolInvocationResult
- ToolParamDefinition
- ToolParameter
- ToolResponse
- ToolResponseMessage
- TopKSamplingStrategy
- TopPSamplingStrategy
- Trace
- TrainingConfig
- Turn
- URL
- UnionType
- UnstructuredLogEvent
- UserMessage
- VectorDB
- VersionInfo
- ViolationLevel

10
llama_stack/apis/agents/agents.py
View file

@ -297,6 +297,16 @@ class AgentStepResponse(BaseModel):
@runtime_checkable @runtime_checkable
@trace_protocol @trace_protocol
class Agents(Protocol): class Agents(Protocol):
"""Agents API for creating and interacting with agentic systems.
Main functionalities provided by this API:
- Create agents with specific instructions and ability to use tools.
- Interactions with agents are grouped into sessions ("threads"), and each interaction is called a "turn".
- Agents can be provided with various tools (see the ToolGroups and ToolRuntime APIs for more details).
- Agents can be provided with various shields (see the Safety API for more details).
- Agents can also use Memory to retrieve information from knowledge bases. See the RAG Tool and Vector IO APIs for more details.
"""
@webmethod(route="/agents", method="POST") @webmethod(route="/agents", method="POST")
async def create_agent( async def create_agent(
self, self,

7
llama_stack/apis/inference/inference.py
View file

@ -362,6 +362,13 @@ class ModelStore(Protocol):
@runtime_checkable @runtime_checkable
@trace_protocol @trace_protocol
class Inference(Protocol): class Inference(Protocol):
"""Llama Stack Inference API for generating completions, chat completions, and embeddings.
This API provides the raw interface to the underlying models. Two kinds of models are supported:
- LLM models: these models generate "raw" and "chat" (conversational) completions.
- Embedding models: these models generate embeddings to be used for semantic search.
"""
model_store: ModelStore model_store: ModelStore
@webmethod(route="/inference/completion", method="POST") @webmethod(route="/inference/completion", method="POST")