phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-15 18:42:31 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

add response format to signature

Browse source
This commit is contained in:
Ashwin Bharambe 2024-10-21 19:14:52 -07:00 • committed by Ashwin Bharambe
parent 6d26bbdce3
commit 40ba22f4c8
15 changed files with 93 additions and 32 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llama_stack/apis/inference/client.py
View file

@ -53,6 +53,7 @@ class InferenceClient(Inference):
tools: Optional[List[ToolDefinition]] = None,
tool_choice: Optional[ToolChoice] = ToolChoice.auto,
tool_prompt_format: Optional[ToolPromptFormat] = ToolPromptFormat.json,
response_format: Optional[ResponseFormat] = None,
stream: Optional[bool] = False,
logprobs: Optional[LogProbConfig] = None,
) -> AsyncGenerator:
@ -63,6 +64,7 @@ class InferenceClient(Inference):
tools=tools or [],
tool_choice=tool_choice,
tool_prompt_format=tool_prompt_format,
response_format=response_format,
stream=stream,
logprobs=logprobs,
)

9
llama_stack/apis/inference/inference.py
View file

@ -74,13 +74,18 @@ class ChatCompletionResponseEvent(BaseModel):
stop_reason: Optional[StopReason] = None
class ResponseFormatType(Enum):
json = "json"
grammar = "grammar"
class JsonResponseFormat(BaseModel):
type: Literal["json"] = "json"
type: Literal[ResponseFormat.json.value] = ResponseFormat.json.value
schema: Dict[str, Any]
class GrammarResponseFormat(BaseModel):
type: Literal["grammar"] = "grammar"
type: Literal[ResponseFormat.grammar.value] = ResponseFormat.grammar.value
bnf: Dict[str, Any]