fix: updated docs to showcase how we can do parallel tool calling with gemini

Author: ivanleomk

docs/concepts/parallel.md

@@ -5,58 +5,101 @@ description: Learn about OpenAI's experimental parallel function calling to redu
 
 # Parallel Tools
 
-One of the latest capabilities that OpenAI has recently introduced is parallel function calling.
-To learn more you can read up on [this](https://platform.openai.com/docs/guides/function-calling/parallel-function-calling)
+Parallel Tool Calling is a feature that allows you to call multiple functions in a single request. This makes it faster to get a response from the language model, especially if your tool calls are independent of each other.
 
 !!! warning "Experimental Feature"
 
-    This feature is currently in preview and is subject to change. only supported by the `gpt-4-turbo-preview` model.
+    Parallel tool calling is only supported by Gemini and OpenAI at the moment
 
 ## Understanding Parallel Function Calling
 
 By using parallel function callings that allow you to call multiple functions in a single request, you can significantly reduce the latency of your application without having to use tricks with now one builds a schema.
 
-```python hl_lines="19 31"
-from __future__ import annotations
+=== "OpenAI"
 
-import openai
-import instructor
+    ```python hl_lines="19 31"
+    from __future__ import annotations
 
-from typing import Iterable, Literal
-from pydantic import BaseModel
+    import openai
+    import instructor
 
+    from typing import Iterable, Literal
+    from pydantic import BaseModel
 
-class Weather(BaseModel):
-    location: str
-    units: Literal["imperial", "metric"]
 
+    class Weather(BaseModel):
+        location: str
+        units: Literal["imperial", "metric"]
 
-class GoogleSearch(BaseModel):
-    query: str
 
+    class GoogleSearch(BaseModel):
+        query: str
 
-client = instructor.from_openai(
-    openai.OpenAI(), mode=instructor.Mode.PARALLEL_TOOLS
-)  # (1)!
 
-function_calls = client.chat.completions.create(
-    model="gpt-4-turbo-preview",
-    messages=[
-        {"role": "system", "content": "You must always use tools"},
-        {
-            "role": "user",
-            "content": "What is the weather in toronto and dallas and who won the super bowl?",
-        },
-    ],
-    response_model=Iterable[Weather | GoogleSearch],  # (2)!
-)
+    client = instructor.from_openai(
+        openai.OpenAI(), mode=instructor.Mode.PARALLEL_TOOLS
+    )  # (1)!
 
-for fc in function_calls:
-    print(fc)
-    #> location='Toronto' units='metric'
-    #> location='Dallas' units='imperial'
-    #> query='who won the super bowl'
-```
+    function_calls = client.chat.completions.create(
+        model="gpt-4o-mini",
+        messages=[
+            {"role": "system", "content": "You must always use tools"},
+            {
+                "role": "user",
+                "content": "What is the weather in toronto and dallas and who won the super bowl?",
+            },
+        ],
+        response_model=Iterable[Weather | GoogleSearch],  # (2)!
+    )
+
+    for fc in function_calls:
+        print(fc)
+        #> location='Toronto' units='metric'
+        #> location='Dallas' units='imperial'
+        #> query='who won the super bowl'
+    ```
+
+=== "Vertex AI"
+
+    ```python
+    import instructor
+    import vertexai
+    from vertexai.generative_models import GenerativeModel
+    from typing import Iterable, Literal
+    from pydantic import BaseModel
+
+    vertexai.init()
+
+    class Weather(BaseModel):
+        location: str
+        units: Literal["imperial", "metric"]
+
+
+    class GoogleSearch(BaseModel):
+        query: str
+
+
+    client = instructor.from_vertexai(
+        GenerativeModel("gemini-1.5-pro-preview-0409"),
+        mode=instructor.Mode.VERTEXAI_PARALLEL_TOOLS
+    )  # (1)!
+
+    function_calls = client.create(
+        messages=[
+            {
+                "role": "user",
+                "content": "What is the weather in toronto and dallas and who won the super bowl?",
+            },
+        ],
+        response_model=Iterable[Weather | GoogleSearch],  # (2)!
+    )
+
+    for fc in function_calls:
+        print(fc)
+        #> location='Toronto' units='metric'
+        #> location='Dallas' units='imperial'
+        #> query='who won the super bowl'
+    ```
 
 1. Set the mode to `PARALLEL_TOOLS` to enable parallel function calling.
 2. Set the response model to `Iterable[Weather | GoogleSearch]` to indicate that the response will be a list of `Weather` and `GoogleSearch` objects. This is necessary because the response will be a list of objects, and we need to specify the types of the objects in the list.

instructor/dsl/parallel.py

@@ -54,7 +54,9 @@ def from_response(
         validation_context: Optional[Any] = None,
         strict: Optional[bool] = None,
     ) -> Generator[BaseModel, None, None]:
-        assert mode == Mode.VERTEXAI_PARALLEL_TOOLS, "Mode must be VERTEXAI_PARALLEL_TOOLS"
+        assert (
+            mode == Mode.VERTEXAI_PARALLEL_TOOLS
+        ), "Mode must be VERTEXAI_PARALLEL_TOOLS"
 
         if not response or not response.candidates:
             return
@@ -64,9 +66,7 @@ def from_response(
                 continue
 
             for part in candidate.content.parts:
-                if (hasattr(part, 'function_call') and
-                    part.function_call is not None):
-
+                if hasattr(part, "function_call") and part.function_call is not None:
                     name = part.function_call.name
                     arguments = part.function_call.args
 
@@ -116,6 +116,7 @@ def ParallelModel(typehint: type[Iterable[T]]) -> ParallelBase:
     the_types = get_types_array(typehint)
     return ParallelBase(*[model for model in the_types])
 
+
 def VertexAIParallelModel(typehint: type[Iterable[T]]) -> VertexAIParallelBase:
     the_types = get_types_array(typehint)
     return VertexAIParallelBase(*[model for model in the_types])

tests/llm/test_vertexai/test_modes.py

@@ -13,7 +13,7 @@ class Item(BaseModel):
 
 
 class Order(BaseModel):
-    items: list[Item] = Field(..., default_factory=list)
+    items: list[Item]
     customer: str
 
 
@@ -54,7 +54,7 @@ class Book(BaseModel):
 
 
 class LibraryRecord(BaseModel):
-    books: list[Book] = Field(..., default_factory=list)
+    books: list[Book]
     visitor: str
     library_id: str