docs: add support for mistral tool calling

examples/open_source_examples/mistral.py

@@ -0,0 +1,58 @@
+from pydantic import BaseModel, Field
+from typing import Optional
+from mistralai.client import MistralClient
+from instructor.patch import patch
+from instructor.function_calls import Mode
+
+client = MistralClient()
+new_chat = patch(create=client.chat, mode=Mode.MIST_TOOLS)
+
+
+class UserDetail(BaseModel):
+    age: Optional[int] = None
+    name: Optional[str] = None
+    role: Optional[str] = None
+
+
+def get_user_detail(string) -> UserDetail:
+    return new_chat(
+        model="mistral-large-latest",
+        response_model=UserDetail,
+        messages=[
+            {
+                "role": "user",
+                "content": f"Get user details for {string}",
+            },
+        ],
+    )  # type: ignore
+
+
+user = get_user_detail("Jason is 25 years old")
+print(user.model_dump_json(indent=2))
+"""
+{
+  "age": 25,
+  "name": "Jason",
+  "role": null
+}
+"""
+
+user = get_user_detail("Jason is a 25 years old scientist")
+print(user.model_dump_json(indent=2))
+"""
+{
+  "age": 25,
+  "name": "Jason",
+  "role": "scientist"
+}
+"""
+
+user = get_user_detail("User not found")
+print(user.model_dump_json(indent=2))
+"""
+{
+  "age": null,
+  "name": null,
+  "role": null
+}
+"""

instructor/function_calls.py

@@ -15,6 +15,7 @@ class Mode(enum.Enum):
     FUNCTIONS: str = "function_call"
     PARALLEL_TOOLS: str = "parallel_tool_call"
     TOOLS: str = "tool_call"
+    MIST_TOOLS: str = "mistral_tools"
     JSON: str = "json_mode"
     MD_JSON: str = "markdown_json_mode"
     JSON_SCHEMA: str = "json_schema_mode"
@@ -114,7 +115,7 @@ def from_response(
                 context=validation_context,
                 strict=strict,
             )
-        elif mode == Mode.TOOLS:
+        elif mode == Mode.TOOLS or mode == Mode.MIST_TOOLS:
             assert (
                 len(message.tool_calls) == 1
             ), "Instructor does not support multiple tool calls, use List[Model] instead."

instructor/patch.py

@@ -116,17 +116,20 @@ def handle_response_model(
         if mode == Mode.FUNCTIONS:
             new_kwargs["functions"] = [response_model.openai_schema]  # type: ignore
             new_kwargs["function_call"] = {"name": response_model.openai_schema["name"]}  # type: ignore
-        elif mode == Mode.TOOLS:
+        elif mode == Mode.TOOLS or mode == Mode.MIST_TOOLS:
             new_kwargs["tools"] = [
                 {
                     "type": "function",
                     "function": response_model.openai_schema,
                 }
             ]
-            new_kwargs["tool_choice"] = {
-                "type": "function",
-                "function": {"name": response_model.openai_schema["name"]},
-            }
+            if mode == Mode.MIST_TOOLS:
+                new_kwargs["tool_choice"] = "any"
+            else:
+                new_kwargs["tool_choice"] = {
+                    "type": "function",
+                    "function": {"name": response_model.openai_schema["name"]},
+                }
         elif mode in {Mode.JSON, Mode.MD_JSON, Mode.JSON_SCHEMA}:
             # If its a JSON Mode we need to massage the prompt a bit
             # in order to get the response we want in a json format