ContextualAI · stainless-app · Mar 3, 2025 · Mar 3, 2025
diff --git a/.stats.yml b/.stats.yml
@@ -1,2 +1,2 @@
 configured_endpoints: 46
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/contextual-ai%2Fsunrise-5298551c424bb999f258bdd6c311e96c80c70701ad59bbce19b46c788ee13bd4.yml
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/contextual-ai%2Fsunrise-f43814080090927ee22816c5c7f517d8a7eb7f346329ada67915608e32124321.yml
diff --git a/src/contextual/resources/generate.py b/src/contextual/resources/generate.py
@@ -53,7 +53,10 @@ def create(
         messages: Iterable[generate_create_params.Message],
         model: str,
         avoid_commentary: bool | NotGiven = NOT_GIVEN,
+        max_new_tokens: int | NotGiven = NOT_GIVEN,
         system_prompt: str | NotGiven = NOT_GIVEN,
+        temperature: float | NotGiven = NOT_GIVEN,
+        top_p: float | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
         # The extra values given here take precedence over values defined on the client or passed to this method.
         extra_headers: Headers | None = None,
@@ -83,9 +86,18 @@ def create(
               context. However, commentary may provide useful context which improves the
               helpfulness of responses.
 
+          max_new_tokens: The maximum number of tokens that the model can generate in the response.
+
           system_prompt: Instructions that the model follows when generating responses. Note that we do
               not guarantee that the model follows these instructions exactly.
 
+          temperature: The sampling temperature, which affects the randomness in the response. Note
+              that higher temperature values can reduce groundedness
+
+          top_p: A parameter for nucleus sampling, an alternative to temperature which also
+              affects the randomness of the response. Note that higher top_p values can reduce
+              groundedness
+
           extra_headers: Send extra headers
 
           extra_query: Add additional query parameters to the request
@@ -102,7 +114,10 @@ def create(
                     "messages": messages,
                     "model": model,
                     "avoid_commentary": avoid_commentary,
+                    "max_new_tokens": max_new_tokens,
                     "system_prompt": system_prompt,
+                    "temperature": temperature,
+                    "top_p": top_p,
                 },
                 generate_create_params.GenerateCreateParams,
             ),
@@ -140,7 +155,10 @@ async def create(
         messages: Iterable[generate_create_params.Message],
         model: str,
         avoid_commentary: bool | NotGiven = NOT_GIVEN,
+        max_new_tokens: int | NotGiven = NOT_GIVEN,
         system_prompt: str | NotGiven = NOT_GIVEN,
+        temperature: float | NotGiven = NOT_GIVEN,
+        top_p: float | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
         # The extra values given here take precedence over values defined on the client or passed to this method.
         extra_headers: Headers | None = None,
@@ -170,9 +188,18 @@ async def create(
               context. However, commentary may provide useful context which improves the
               helpfulness of responses.
 
+          max_new_tokens: The maximum number of tokens that the model can generate in the response.
+
           system_prompt: Instructions that the model follows when generating responses. Note that we do
               not guarantee that the model follows these instructions exactly.
 
+          temperature: The sampling temperature, which affects the randomness in the response. Note
+              that higher temperature values can reduce groundedness
+
+          top_p: A parameter for nucleus sampling, an alternative to temperature which also
+              affects the randomness of the response. Note that higher top_p values can reduce
+              groundedness
+
           extra_headers: Send extra headers
 
           extra_query: Add additional query parameters to the request
@@ -189,7 +216,10 @@ async def create(
                     "messages": messages,
                     "model": model,
                     "avoid_commentary": avoid_commentary,
+                    "max_new_tokens": max_new_tokens,
                     "system_prompt": system_prompt,
+                    "temperature": temperature,
+                    "top_p": top_p,
                 },
                 generate_create_params.GenerateCreateParams,
             ),

diff --git a/src/contextual/types/agents/evaluate/list_evaluation_jobs_response.py b/src/contextual/types/agents/evaluate/list_evaluation_jobs_response.py
@@ -31,6 +31,9 @@ class EvaluationRound(BaseModel):
     num_predictions: Optional[int] = None
     """Total number of predictions made during the evaluation round"""
 
+    num_processed_predictions: Optional[int] = None
+    """Number of predictions that have been processed during the evaluation round"""
+
     num_successful_predictions: Optional[int] = None
     """Number of predictions that were successful during the evaluation round"""
 

diff --git a/src/contextual/types/generate_create_params.py b/src/contextual/types/generate_create_params.py
@@ -30,16 +30,32 @@ class GenerateCreateParams(TypedDict, total=False):
     helpfulness of responses.
     """
 
+    max_new_tokens: int
+    """The maximum number of tokens that the model can generate in the response."""
+
     system_prompt: str
     """Instructions that the model follows when generating responses.
 
     Note that we do not guarantee that the model follows these instructions exactly.
     """
 
+    temperature: float
+    """The sampling temperature, which affects the randomness in the response.
+
+    Note that higher temperature values can reduce groundedness
+    """
+
+    top_p: float
+    """
+    A parameter for nucleus sampling, an alternative to temperature which also
+    affects the randomness of the response. Note that higher top_p values can reduce
+    groundedness
+    """
+
 
 class Message(TypedDict, total=False):
     content: Required[str]
     """Content of the message"""
 
-    role: Required[Literal["user", "system", "assistant", "knowledge"]]
+    role: Required[Literal["user", "assistant"]]
     """Role of the sender"""
diff --git a/tests/api_resources/test_generate.py b/tests/api_resources/test_generate.py
@@ -43,7 +43,10 @@ def test_method_create_with_all_params(self, client: ContextualAI) -> None:
             ],
             model="model",
             avoid_commentary=True,
+            max_new_tokens=1,
             system_prompt="system_prompt",
+            temperature=0,
+            top_p=1,
         )
         assert_matches_type(GenerateCreateResponse, generate, path=["response"])
 
@@ -115,7 +118,10 @@ async def test_method_create_with_all_params(self, async_client: AsyncContextual
             ],
             model="model",
             avoid_commentary=True,
+            max_new_tokens=1,
             system_prompt="system_prompt",
+            temperature=0,
+            top_p=1,
         )
         assert_matches_type(GenerateCreateResponse, generate, path=["response"])