sync repo w/ MM

meta-llama · Sep 25, 2024 · 39e0b2b · 39e0b2b
1 parent da88593
commit 39e0b2b
Show file tree

Hide file tree

Showing 30 changed files with 1,070 additions and 90 deletions.
diff --git a/src/llama_stack_client/resources/batch_inference.py b/src/llama_stack_client/resources/batch_inference.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import List, Union, Iterable
+from typing import List, Iterable
 from typing_extensions import Literal
 
 import httpx
@@ -58,7 +58,7 @@ def chat_completion(
         logprobs: batch_inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[batch_inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -115,7 +115,7 @@ def chat_completion(
     def completion(
         self,
         *,
-        content_batch: List[Union[str, List[str]]],
+        content_batch: List[batch_inference_completion_params.ContentBatch],
         model: str,
         logprobs: batch_inference_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
@@ -187,7 +187,7 @@ async def chat_completion(
         logprobs: batch_inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[batch_inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -244,7 +244,7 @@ async def chat_completion(
     async def completion(
         self,
         *,
-        content_batch: List[Union[str, List[str]]],
+        content_batch: List[batch_inference_completion_params.ContentBatch],
         model: str,
         logprobs: batch_inference_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,

diff --git a/src/llama_stack_client/resources/inference/embeddings.py b/src/llama_stack_client/resources/inference/embeddings.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import List, Union
+from typing import List
 
 import httpx
 
@@ -50,7 +50,7 @@ def with_streaming_response(self) -> EmbeddingsResourceWithStreamingResponse:
     def create(
         self,
         *,
-        contents: List[Union[str, List[str]]],
+        contents: List[embedding_create_params.Content],
         model: str,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -113,7 +113,7 @@ def with_streaming_response(self) -> AsyncEmbeddingsResourceWithStreamingRespons
     async def create(
         self,
         *,
-        contents: List[Union[str, List[str]]],
+        contents: List[embedding_create_params.Content],
         model: str,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.

diff --git a/src/llama_stack_client/resources/inference/inference.py b/src/llama_stack_client/resources/inference/inference.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import Any, List, Union, Iterable, cast
+from typing import Any, Iterable, cast
 from typing_extensions import Literal, overload
 
 import httpx
@@ -74,7 +74,7 @@ def chat_completion(
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         stream: Literal[False] | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -116,7 +116,7 @@ def chat_completion(
         logprobs: inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -158,7 +158,7 @@ def chat_completion(
         logprobs: inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -200,7 +200,7 @@ def chat_completion(
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         stream: Literal[False] | Literal[True] | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -246,7 +246,7 @@ def chat_completion(
     def completion(
         self,
         *,
-        content: Union[str, List[str]],
+        content: inference_completion_params.Content,
         model: str,
         logprobs: inference_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
@@ -331,7 +331,7 @@ async def chat_completion(
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         stream: Literal[False] | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -373,7 +373,7 @@ async def chat_completion(
         logprobs: inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -415,7 +415,7 @@ async def chat_completion(
         logprobs: inference_chat_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -457,7 +457,7 @@ async def chat_completion(
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,
         stream: Literal[False] | Literal[True] | NotGiven = NOT_GIVEN,
         tool_choice: Literal["auto", "required"] | NotGiven = NOT_GIVEN,
-        tool_prompt_format: Literal["json", "function_tag"] | NotGiven = NOT_GIVEN,
+        tool_prompt_format: Literal["json", "function_tag", "python_list"] | NotGiven = NOT_GIVEN,
         tools: Iterable[inference_chat_completion_params.Tool] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -503,7 +503,7 @@ async def chat_completion(
     async def completion(
         self,
         *,
-        content: Union[str, List[str]],
+        content: inference_completion_params.Content,
         model: str,
         logprobs: inference_completion_params.Logprobs | NotGiven = NOT_GIVEN,
         sampling_params: SamplingParams | NotGiven = NOT_GIVEN,

diff --git a/src/llama_stack_client/resources/memory/memory.py b/src/llama_stack_client/resources/memory/memory.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import Dict, List, Union, Iterable
+from typing import Dict, Union, Iterable
 
 import httpx
 
@@ -300,7 +300,7 @@ def query(
         self,
         *,
         bank_id: str,
-        query: Union[str, List[str]],
+        query: memory_query_params.Query,
         params: Dict[str, Union[bool, float, str, Iterable[object], object, None]] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -599,7 +599,7 @@ async def query(
         self,
         *,
         bank_id: str,
-        query: Union[str, List[str]],
+        query: memory_query_params.Query,
         params: Dict[str, Union[bool, float, str, Iterable[object], object, None]] | NotGiven = NOT_GIVEN,
         x_llama_stack_provider_data: str | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.

diff --git a/src/llama_stack_client/types/agent_create_params.py b/src/llama_stack_client/types/agent_create_params.py
@@ -206,7 +206,7 @@ class AgentConfig(TypedDict, total=False):
 
     tool_choice: Literal["auto", "required"]
 
-    tool_prompt_format: Literal["json", "function_tag"]
+    tool_prompt_format: Literal["json", "function_tag", "python_list"]
     """
     `json` -- Refers to the json format for calling tools. The json format takes the
     form like { "type": "function", "function" : { "name": "function_name",

diff --git a/src/llama_stack_client/types/batch_inference_chat_completion_params.py b/src/llama_stack_client/types/batch_inference_chat_completion_params.py
@@ -27,7 +27,7 @@ class BatchInferenceChatCompletionParams(TypedDict, total=False):
 
     tool_choice: Literal["auto", "required"]
 
-    tool_prompt_format: Literal["json", "function_tag"]
+    tool_prompt_format: Literal["json", "function_tag", "python_list"]
     """
     `json` -- Refers to the json format for calling tools. The json format takes the
     form like { "type": "function", "function" : { "name": "function_name",

diff --git a/src/llama_stack_client/types/batch_inference_completion_params.py b/src/llama_stack_client/types/batch_inference_completion_params.py
@@ -3,16 +3,27 @@
 from __future__ import annotations
 
 from typing import List, Union
-from typing_extensions import Required, Annotated, TypedDict
+from typing_extensions import Required, Annotated, TypeAlias, TypedDict
 
 from .._utils import PropertyInfo
 from .shared_params.sampling_params import SamplingParams
 
-__all__ = ["BatchInferenceCompletionParams", "Logprobs"]
+__all__ = [
+    "BatchInferenceCompletionParams",
+    "ContentBatch",
+    "ContentBatchImageMedia",
+    "ContentBatchImageMediaImage",
+    "ContentBatchImageMediaImageThisClassRepresentsAnImageObjectToCreate",
+    "ContentBatchUnionMember2",
+    "ContentBatchUnionMember2ImageMedia",
+    "ContentBatchUnionMember2ImageMediaImage",
+    "ContentBatchUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate",
+    "Logprobs",
+]
 
 
 class BatchInferenceCompletionParams(TypedDict, total=False):
-    content_batch: Required[List[Union[str, List[str]]]]
+    content_batch: Required[List[ContentBatch]]
 
     model: Required[str]
 
@@ -23,5 +34,38 @@ class BatchInferenceCompletionParams(TypedDict, total=False):
     x_llama_stack_provider_data: Annotated[str, PropertyInfo(alias="X-LlamaStack-ProviderData")]
 
 
+class ContentBatchImageMediaImageThisClassRepresentsAnImageObjectToCreate(TypedDict, total=False):
+    format: str
+
+    format_description: str
+
+
+ContentBatchImageMediaImage: TypeAlias = Union[ContentBatchImageMediaImageThisClassRepresentsAnImageObjectToCreate, str]
+
+
+class ContentBatchImageMedia(TypedDict, total=False):
+    image: Required[ContentBatchImageMediaImage]
+
+
+class ContentBatchUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate(TypedDict, total=False):
+    format: str
+
+    format_description: str
+
+
+ContentBatchUnionMember2ImageMediaImage: TypeAlias = Union[
+    ContentBatchUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate, str
+]
+
+
+class ContentBatchUnionMember2ImageMedia(TypedDict, total=False):
+    image: Required[ContentBatchUnionMember2ImageMediaImage]
+
+
+ContentBatchUnionMember2: TypeAlias = Union[str, ContentBatchUnionMember2ImageMedia]
+
+ContentBatch: TypeAlias = Union[str, ContentBatchImageMedia, List[ContentBatchUnionMember2]]
+
+
 class Logprobs(TypedDict, total=False):
     top_k: int
diff --git a/src/llama_stack_client/types/inference/embedding_create_params.py b/src/llama_stack_client/types/inference/embedding_create_params.py
@@ -3,16 +3,59 @@
 from __future__ import annotations
 
 from typing import List, Union
-from typing_extensions import Required, Annotated, TypedDict
+from typing_extensions import Required, Annotated, TypeAlias, TypedDict
 
 from ..._utils import PropertyInfo
 
-__all__ = ["EmbeddingCreateParams"]
+__all__ = [
+    "EmbeddingCreateParams",
+    "Content",
+    "ContentImageMedia",
+    "ContentImageMediaImage",
+    "ContentImageMediaImageThisClassRepresentsAnImageObjectToCreate",
+    "ContentUnionMember2",
+    "ContentUnionMember2ImageMedia",
+    "ContentUnionMember2ImageMediaImage",
+    "ContentUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate",
+]
 
 
 class EmbeddingCreateParams(TypedDict, total=False):
-    contents: Required[List[Union[str, List[str]]]]
+    contents: Required[List[Content]]
 
     model: Required[str]
 
     x_llama_stack_provider_data: Annotated[str, PropertyInfo(alias="X-LlamaStack-ProviderData")]
+
+
+class ContentImageMediaImageThisClassRepresentsAnImageObjectToCreate(TypedDict, total=False):
+    format: str
+
+    format_description: str
+
+
+ContentImageMediaImage: TypeAlias = Union[ContentImageMediaImageThisClassRepresentsAnImageObjectToCreate, str]
+
+
+class ContentImageMedia(TypedDict, total=False):
+    image: Required[ContentImageMediaImage]
+
+
+class ContentUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate(TypedDict, total=False):
+    format: str
+
+    format_description: str
+
+
+ContentUnionMember2ImageMediaImage: TypeAlias = Union[
+    ContentUnionMember2ImageMediaImageThisClassRepresentsAnImageObjectToCreate, str
+]
+
+
+class ContentUnionMember2ImageMedia(TypedDict, total=False):
+    image: Required[ContentUnionMember2ImageMediaImage]
+
+
+ContentUnionMember2: TypeAlias = Union[str, ContentUnionMember2ImageMedia]
+
+Content: TypeAlias = Union[str, ContentImageMedia, List[ContentUnionMember2]]
diff --git a/src/llama_stack_client/types/inference_chat_completion_params.py b/src/llama_stack_client/types/inference_chat_completion_params.py
@@ -34,7 +34,7 @@ class InferenceChatCompletionParamsBase(TypedDict, total=False):
 
     tool_choice: Literal["auto", "required"]
 
-    tool_prompt_format: Literal["json", "function_tag"]
+    tool_prompt_format: Literal["json", "function_tag", "python_list"]
     """
     `json` -- Refers to the json format for calling tools. The json format takes the
     form like { "type": "function", "function" : { "name": "function_name",