pulumiverse-scaleway 1.27.1__py3-none-any.whl → 1.28.0__py3-none-any.whl

pulumiverse_scaleway/inference/outputs.py

@@ -13,10 +13,15 @@ if sys.version_info >= (3, 11):
 else:
     from typing_extensions import NotRequired, TypedDict, TypeAlias
 from .. import _utilities
+from . import outputs
 
 __all__ = [
     'DeploymentPrivateEndpoint',
     'DeploymentPublicEndpoint',
+    'ModelNodesSupport',
+    'ModelNodesSupportQuantization',
+    'GetModelNodesSupportResult',
+    'GetModelNodesSupportQuantizationResult',
 ]
 
 @pulumi.output_type
@@ -167,3 +172,182 @@ class DeploymentPublicEndpoint(dict):
         return pulumi.get(self, "url")
 
 
+@pulumi.output_type
+class ModelNodesSupport(dict):
+    @staticmethod
+    def __key_warning(key: str):
+        suggest = None
+        if key == "nodeTypeName":
+            suggest = "node_type_name"
+
+        if suggest:
+            pulumi.log.warn(f"Key '{key}' not found in ModelNodesSupport. Access the value via the '{suggest}' property getter instead.")
+
+    def __getitem__(self, key: str) -> Any:
+        ModelNodesSupport.__key_warning(key)
+        return super().__getitem__(key)
+
+    def get(self, key: str, default = None) -> Any:
+        ModelNodesSupport.__key_warning(key)
+        return super().get(key, default)
+
+    def __init__(__self__, *,
+                 node_type_name: Optional[str] = None,
+                 quantizations: Optional[Sequence['outputs.ModelNodesSupportQuantization']] = None):
+        """
+        :param str node_type_name: The type of node supported.
+        :param Sequence['ModelNodesSupportQuantizationArgs'] quantizations: A list of supported quantization options, including:
+        """
+        if node_type_name is not None:
+            pulumi.set(__self__, "node_type_name", node_type_name)
+        if quantizations is not None:
+            pulumi.set(__self__, "quantizations", quantizations)
+
+    @property
+    @pulumi.getter(name="nodeTypeName")
+    def node_type_name(self) -> Optional[str]:
+        """
+        The type of node supported.
+        """
+        return pulumi.get(self, "node_type_name")
+
+    @property
+    @pulumi.getter
+    def quantizations(self) -> Optional[Sequence['outputs.ModelNodesSupportQuantization']]:
+        """
+        A list of supported quantization options, including:
+        """
+        return pulumi.get(self, "quantizations")
+
+
+@pulumi.output_type
+class ModelNodesSupportQuantization(dict):
+    @staticmethod
+    def __key_warning(key: str):
+        suggest = None
+        if key == "maxContextSize":
+            suggest = "max_context_size"
+        elif key == "quantizationBits":
+            suggest = "quantization_bits"
+
+        if suggest:
+            pulumi.log.warn(f"Key '{key}' not found in ModelNodesSupportQuantization. Access the value via the '{suggest}' property getter instead.")
+
+    def __getitem__(self, key: str) -> Any:
+        ModelNodesSupportQuantization.__key_warning(key)
+        return super().__getitem__(key)
+
+    def get(self, key: str, default = None) -> Any:
+        ModelNodesSupportQuantization.__key_warning(key)
+        return super().get(key, default)
+
+    def __init__(__self__, *,
+                 allowed: Optional[bool] = None,
+                 max_context_size: Optional[int] = None,
+                 quantization_bits: Optional[int] = None):
+        """
+        :param bool allowed: Whether this quantization is allowed.
+        :param int max_context_size: Maximum context length supported by this quantization.
+        :param int quantization_bits: Number of bits used for quantization (e.g., 8, 16).
+        """
+        if allowed is not None:
+            pulumi.set(__self__, "allowed", allowed)
+        if max_context_size is not None:
+            pulumi.set(__self__, "max_context_size", max_context_size)
+        if quantization_bits is not None:
+            pulumi.set(__self__, "quantization_bits", quantization_bits)
+
+    @property
+    @pulumi.getter
+    def allowed(self) -> Optional[bool]:
+        """
+        Whether this quantization is allowed.
+        """
+        return pulumi.get(self, "allowed")
+
+    @property
+    @pulumi.getter(name="maxContextSize")
+    def max_context_size(self) -> Optional[int]:
+        """
+        Maximum context length supported by this quantization.
+        """
+        return pulumi.get(self, "max_context_size")
+
+    @property
+    @pulumi.getter(name="quantizationBits")
+    def quantization_bits(self) -> Optional[int]:
+        """
+        Number of bits used for quantization (e.g., 8, 16).
+        """
+        return pulumi.get(self, "quantization_bits")
+
+
+@pulumi.output_type
+class GetModelNodesSupportResult(dict):
+    def __init__(__self__, *,
+                 node_type_name: str,
+                 quantizations: Sequence['outputs.GetModelNodesSupportQuantizationResult']):
+        """
+        :param str node_type_name: The type of node supported.
+        :param Sequence['GetModelNodesSupportQuantizationArgs'] quantizations: A list of supported quantization options, including:
+        """
+        pulumi.set(__self__, "node_type_name", node_type_name)
+        pulumi.set(__self__, "quantizations", quantizations)
+
+    @property
+    @pulumi.getter(name="nodeTypeName")
+    def node_type_name(self) -> str:
+        """
+        The type of node supported.
+        """
+        return pulumi.get(self, "node_type_name")
+
+    @property
+    @pulumi.getter
+    def quantizations(self) -> Sequence['outputs.GetModelNodesSupportQuantizationResult']:
+        """
+        A list of supported quantization options, including:
+        """
+        return pulumi.get(self, "quantizations")
+
+
+@pulumi.output_type
+class GetModelNodesSupportQuantizationResult(dict):
+    def __init__(__self__, *,
+                 allowed: bool,
+                 max_context_size: int,
+                 quantization_bits: int):
+        """
+        :param bool allowed: Whether this quantization is allowed.
+        :param int max_context_size: Maximum context length supported by this quantization.
+        :param int quantization_bits: Number of bits used for quantization (e.g., 8, 16).
+        """
+        pulumi.set(__self__, "allowed", allowed)
+        pulumi.set(__self__, "max_context_size", max_context_size)
+        pulumi.set(__self__, "quantization_bits", quantization_bits)
+
+    @property
+    @pulumi.getter
+    def allowed(self) -> bool:
+        """
+        Whether this quantization is allowed.
+        """
+        return pulumi.get(self, "allowed")
+
+    @property
+    @pulumi.getter(name="maxContextSize")
+    def max_context_size(self) -> int:
+        """
+        Maximum context length supported by this quantization.
+        """
+        return pulumi.get(self, "max_context_size")
+
+    @property
+    @pulumi.getter(name="quantizationBits")
+    def quantization_bits(self) -> int:
+        """
+        Number of bits used for quantization (e.g., 8, 16).
+        """
+        return pulumi.get(self, "quantization_bits")
+
+

pulumiverse_scaleway/inference_deployment.py

@@ -21,7 +21,7 @@ __all__ = ['InferenceDeploymentArgs', 'InferenceDeployment']
 @pulumi.input_type
 class InferenceDeploymentArgs:
     def __init__(__self__, *,
-                 model_name: pulumi.Input[str],
+                 model_id: pulumi.Input[str],
                  node_type: pulumi.Input[str],
                  accept_eula: Optional[pulumi.Input[bool]] = None,
                  max_size: Optional[pulumi.Input[int]] = None,
@@ -30,11 +30,12 @@ class InferenceDeploymentArgs:
                  private_endpoint: Optional[pulumi.Input['InferenceDeploymentPrivateEndpointArgs']] = None,
                  project_id: Optional[pulumi.Input[str]] = None,
                  public_endpoint: Optional[pulumi.Input['InferenceDeploymentPublicEndpointArgs']] = None,
+                 quantization: Optional[pulumi.Input[int]] = None,
                  region: Optional[pulumi.Input[str]] = None,
                  tags: Optional[pulumi.Input[Sequence[pulumi.Input[str]]]] = None):
         """
         The set of arguments for constructing a InferenceDeployment resource.
-        :param pulumi.Input[str] model_name: The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        :param pulumi.Input[str] model_id: The model id used for the deployment.
         :param pulumi.Input[str] node_type: The node type to use for the deployment. Node types can be found using Scaleway's CLI (`scw inference node-type list`)
         :param pulumi.Input[bool] accept_eula: Some models (e.g Meta Llama) require end-user license agreements. Set `true` to accept.
         :param pulumi.Input[int] max_size: The maximum size of the pool.
@@ -43,10 +44,11 @@ class InferenceDeploymentArgs:
         :param pulumi.Input['InferenceDeploymentPrivateEndpointArgs'] private_endpoint: Configuration of the deployment's private endpoint.
         :param pulumi.Input[str] project_id: `project_id`) The ID of the project the deployment is associated with.
         :param pulumi.Input['InferenceDeploymentPublicEndpointArgs'] public_endpoint: Configuration of the deployment's public endpoint.
+        :param pulumi.Input[int] quantization: The number of bits each model parameter should be quantized to
         :param pulumi.Input[str] region: `region`) The region in which the deployment is created.
         :param pulumi.Input[Sequence[pulumi.Input[str]]] tags: The tags associated with the deployment.
         """
-        pulumi.set(__self__, "model_name", model_name)
+        pulumi.set(__self__, "model_id", model_id)
         pulumi.set(__self__, "node_type", node_type)
         if accept_eula is not None:
             pulumi.set(__self__, "accept_eula", accept_eula)
@@ -62,22 +64,24 @@ class InferenceDeploymentArgs:
             pulumi.set(__self__, "project_id", project_id)
         if public_endpoint is not None:
             pulumi.set(__self__, "public_endpoint", public_endpoint)
+        if quantization is not None:
+            pulumi.set(__self__, "quantization", quantization)
         if region is not None:
             pulumi.set(__self__, "region", region)
         if tags is not None:
             pulumi.set(__self__, "tags", tags)
 
     @property
-    @pulumi.getter(name="modelName")
-    def model_name(self) -> pulumi.Input[str]:
+    @pulumi.getter(name="modelId")
+    def model_id(self) -> pulumi.Input[str]:
         """
-        The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        The model id used for the deployment.
         """
-        return pulumi.get(self, "model_name")
+        return pulumi.get(self, "model_id")
 
-    @model_name.setter
-    def model_name(self, value: pulumi.Input[str]):
-        pulumi.set(self, "model_name", value)
+    @model_id.setter
+    def model_id(self, value: pulumi.Input[str]):
+        pulumi.set(self, "model_id", value)
 
     @property
     @pulumi.getter(name="nodeType")
@@ -175,6 +179,18 @@ class InferenceDeploymentArgs:
     def public_endpoint(self, value: Optional[pulumi.Input['InferenceDeploymentPublicEndpointArgs']]):
         pulumi.set(self, "public_endpoint", value)
 
+    @property
+    @pulumi.getter
+    def quantization(self) -> Optional[pulumi.Input[int]]:
+        """
+        The number of bits each model parameter should be quantized to
+        """
+        return pulumi.get(self, "quantization")
+
+    @quantization.setter
+    def quantization(self, value: Optional[pulumi.Input[int]]):
+        pulumi.set(self, "quantization", value)
+
     @property
     @pulumi.getter
     def region(self) -> Optional[pulumi.Input[str]]:
@@ -214,6 +230,7 @@ class _InferenceDeploymentState:
                  private_endpoint: Optional[pulumi.Input['InferenceDeploymentPrivateEndpointArgs']] = None,
                  project_id: Optional[pulumi.Input[str]] = None,
                  public_endpoint: Optional[pulumi.Input['InferenceDeploymentPublicEndpointArgs']] = None,
+                 quantization: Optional[pulumi.Input[int]] = None,
                  region: Optional[pulumi.Input[str]] = None,
                  size: Optional[pulumi.Input[int]] = None,
                  status: Optional[pulumi.Input[str]] = None,
@@ -226,12 +243,13 @@ class _InferenceDeploymentState:
         :param pulumi.Input[int] max_size: The maximum size of the pool.
         :param pulumi.Input[int] min_size: The minimum size of the pool.
         :param pulumi.Input[str] model_id: The model id used for the deployment.
-        :param pulumi.Input[str] model_name: The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        :param pulumi.Input[str] model_name: The model name used for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
         :param pulumi.Input[str] name: The deployment name.
         :param pulumi.Input[str] node_type: The node type to use for the deployment. Node types can be found using Scaleway's CLI (`scw inference node-type list`)
         :param pulumi.Input['InferenceDeploymentPrivateEndpointArgs'] private_endpoint: Configuration of the deployment's private endpoint.
         :param pulumi.Input[str] project_id: `project_id`) The ID of the project the deployment is associated with.
         :param pulumi.Input['InferenceDeploymentPublicEndpointArgs'] public_endpoint: Configuration of the deployment's public endpoint.
+        :param pulumi.Input[int] quantization: The number of bits each model parameter should be quantized to
         :param pulumi.Input[str] region: `region`) The region in which the deployment is created.
         :param pulumi.Input[int] size: The size of the pool.
         :param pulumi.Input[str] status: The status of the deployment.
@@ -260,6 +278,8 @@ class _InferenceDeploymentState:
             pulumi.set(__self__, "project_id", project_id)
         if public_endpoint is not None:
             pulumi.set(__self__, "public_endpoint", public_endpoint)
+        if quantization is not None:
+            pulumi.set(__self__, "quantization", quantization)
         if region is not None:
             pulumi.set(__self__, "region", region)
         if size is not None:
@@ -335,7 +355,7 @@ class _InferenceDeploymentState:
     @pulumi.getter(name="modelName")
     def model_name(self) -> Optional[pulumi.Input[str]]:
         """
-        The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        The model name used for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
         """
         return pulumi.get(self, "model_name")
 
@@ -403,6 +423,18 @@ class _InferenceDeploymentState:
     def public_endpoint(self, value: Optional[pulumi.Input['InferenceDeploymentPublicEndpointArgs']]):
         pulumi.set(self, "public_endpoint", value)
 
+    @property
+    @pulumi.getter
+    def quantization(self) -> Optional[pulumi.Input[int]]:
+        """
+        The number of bits each model parameter should be quantized to
+        """
+        return pulumi.get(self, "quantization")
+
+    @quantization.setter
+    def quantization(self, value: Optional[pulumi.Input[int]]):
+        pulumi.set(self, "quantization", value)
+
     @property
     @pulumi.getter
     def region(self) -> Optional[pulumi.Input[str]]:
@@ -477,12 +509,13 @@ class InferenceDeployment(pulumi.CustomResource):
                  accept_eula: Optional[pulumi.Input[bool]] = None,
                  max_size: Optional[pulumi.Input[int]] = None,
                  min_size: Optional[pulumi.Input[int]] = None,
-                 model_name: Optional[pulumi.Input[str]] = None,
+                 model_id: Optional[pulumi.Input[str]] = None,
                  name: Optional[pulumi.Input[str]] = None,
                  node_type: Optional[pulumi.Input[str]] = None,
                  private_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPrivateEndpointArgs', 'InferenceDeploymentPrivateEndpointArgsDict']]] = None,
                  project_id: Optional[pulumi.Input[str]] = None,
                  public_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPublicEndpointArgs', 'InferenceDeploymentPublicEndpointArgsDict']]] = None,
+                 quantization: Optional[pulumi.Input[int]] = None,
                  region: Optional[pulumi.Input[str]] = None,
                  tags: Optional[pulumi.Input[Sequence[pulumi.Input[str]]]] = None,
                  __props__=None):
@@ -492,22 +525,6 @@ class InferenceDeployment(pulumi.CustomResource):
 
         ## Example Usage
 
-        ### Basic
-
-        ```python
-        import pulumi
-        import pulumiverse_scaleway as scaleway
-
-        deployment = scaleway.inference.Deployment("deployment",
-            name="tf-inference-deployment",
-            node_type="L4",
-            model_name="meta/llama-3.1-8b-instruct:fp8",
-            public_endpoint={
-                "is_enabled": True,
-            },
-            accept_eula=True)
-        ```
-
         ## Import
 
         Functions can be imported using, `{region}/{id}`, as shown below:
@@ -523,12 +540,13 @@ class InferenceDeployment(pulumi.CustomResource):
         :param pulumi.Input[bool] accept_eula: Some models (e.g Meta Llama) require end-user license agreements. Set `true` to accept.
         :param pulumi.Input[int] max_size: The maximum size of the pool.
         :param pulumi.Input[int] min_size: The minimum size of the pool.
-        :param pulumi.Input[str] model_name: The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        :param pulumi.Input[str] model_id: The model id used for the deployment.
         :param pulumi.Input[str] name: The deployment name.
         :param pulumi.Input[str] node_type: The node type to use for the deployment. Node types can be found using Scaleway's CLI (`scw inference node-type list`)
         :param pulumi.Input[Union['InferenceDeploymentPrivateEndpointArgs', 'InferenceDeploymentPrivateEndpointArgsDict']] private_endpoint: Configuration of the deployment's private endpoint.
         :param pulumi.Input[str] project_id: `project_id`) The ID of the project the deployment is associated with.
         :param pulumi.Input[Union['InferenceDeploymentPublicEndpointArgs', 'InferenceDeploymentPublicEndpointArgsDict']] public_endpoint: Configuration of the deployment's public endpoint.
+        :param pulumi.Input[int] quantization: The number of bits each model parameter should be quantized to
         :param pulumi.Input[str] region: `region`) The region in which the deployment is created.
         :param pulumi.Input[Sequence[pulumi.Input[str]]] tags: The tags associated with the deployment.
         """
@@ -544,22 +562,6 @@ class InferenceDeployment(pulumi.CustomResource):
 
         ## Example Usage
 
-        ### Basic
-
-        ```python
-        import pulumi
-        import pulumiverse_scaleway as scaleway
-
-        deployment = scaleway.inference.Deployment("deployment",
-            name="tf-inference-deployment",
-            node_type="L4",
-            model_name="meta/llama-3.1-8b-instruct:fp8",
-            public_endpoint={
-                "is_enabled": True,
-            },
-            accept_eula=True)
-        ```
-
         ## Import
 
         Functions can be imported using, `{region}/{id}`, as shown below:
@@ -588,12 +590,13 @@ class InferenceDeployment(pulumi.CustomResource):
                  accept_eula: Optional[pulumi.Input[bool]] = None,
                  max_size: Optional[pulumi.Input[int]] = None,
                  min_size: Optional[pulumi.Input[int]] = None,
-                 model_name: Optional[pulumi.Input[str]] = None,
+                 model_id: Optional[pulumi.Input[str]] = None,
                  name: Optional[pulumi.Input[str]] = None,
                  node_type: Optional[pulumi.Input[str]] = None,
                  private_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPrivateEndpointArgs', 'InferenceDeploymentPrivateEndpointArgsDict']]] = None,
                  project_id: Optional[pulumi.Input[str]] = None,
                  public_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPublicEndpointArgs', 'InferenceDeploymentPublicEndpointArgsDict']]] = None,
+                 quantization: Optional[pulumi.Input[int]] = None,
                  region: Optional[pulumi.Input[str]] = None,
                  tags: Optional[pulumi.Input[Sequence[pulumi.Input[str]]]] = None,
                  __props__=None):
@@ -609,9 +612,9 @@ class InferenceDeployment(pulumi.CustomResource):
             __props__.__dict__["accept_eula"] = accept_eula
             __props__.__dict__["max_size"] = max_size
             __props__.__dict__["min_size"] = min_size
-            if model_name is None and not opts.urn:
-                raise TypeError("Missing required property 'model_name'")
-            __props__.__dict__["model_name"] = model_name
+            if model_id is None and not opts.urn:
+                raise TypeError("Missing required property 'model_id'")
+            __props__.__dict__["model_id"] = model_id
             __props__.__dict__["name"] = name
             if node_type is None and not opts.urn:
                 raise TypeError("Missing required property 'node_type'")
@@ -619,10 +622,11 @@ class InferenceDeployment(pulumi.CustomResource):
             __props__.__dict__["private_endpoint"] = private_endpoint
             __props__.__dict__["project_id"] = project_id
             __props__.__dict__["public_endpoint"] = public_endpoint
+            __props__.__dict__["quantization"] = quantization
             __props__.__dict__["region"] = region
             __props__.__dict__["tags"] = tags
             __props__.__dict__["created_at"] = None
-            __props__.__dict__["model_id"] = None
+            __props__.__dict__["model_name"] = None
             __props__.__dict__["size"] = None
             __props__.__dict__["status"] = None
             __props__.__dict__["updated_at"] = None
@@ -647,6 +651,7 @@ class InferenceDeployment(pulumi.CustomResource):
             private_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPrivateEndpointArgs', 'InferenceDeploymentPrivateEndpointArgsDict']]] = None,
             project_id: Optional[pulumi.Input[str]] = None,
             public_endpoint: Optional[pulumi.Input[Union['InferenceDeploymentPublicEndpointArgs', 'InferenceDeploymentPublicEndpointArgsDict']]] = None,
+            quantization: Optional[pulumi.Input[int]] = None,
             region: Optional[pulumi.Input[str]] = None,
             size: Optional[pulumi.Input[int]] = None,
             status: Optional[pulumi.Input[str]] = None,
@@ -664,12 +669,13 @@ class InferenceDeployment(pulumi.CustomResource):
         :param pulumi.Input[int] max_size: The maximum size of the pool.
         :param pulumi.Input[int] min_size: The minimum size of the pool.
         :param pulumi.Input[str] model_id: The model id used for the deployment.
-        :param pulumi.Input[str] model_name: The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        :param pulumi.Input[str] model_name: The model name used for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
         :param pulumi.Input[str] name: The deployment name.
         :param pulumi.Input[str] node_type: The node type to use for the deployment. Node types can be found using Scaleway's CLI (`scw inference node-type list`)
         :param pulumi.Input[Union['InferenceDeploymentPrivateEndpointArgs', 'InferenceDeploymentPrivateEndpointArgsDict']] private_endpoint: Configuration of the deployment's private endpoint.
         :param pulumi.Input[str] project_id: `project_id`) The ID of the project the deployment is associated with.
         :param pulumi.Input[Union['InferenceDeploymentPublicEndpointArgs', 'InferenceDeploymentPublicEndpointArgsDict']] public_endpoint: Configuration of the deployment's public endpoint.
+        :param pulumi.Input[int] quantization: The number of bits each model parameter should be quantized to
         :param pulumi.Input[str] region: `region`) The region in which the deployment is created.
         :param pulumi.Input[int] size: The size of the pool.
         :param pulumi.Input[str] status: The status of the deployment.
@@ -691,6 +697,7 @@ class InferenceDeployment(pulumi.CustomResource):
         __props__.__dict__["private_endpoint"] = private_endpoint
         __props__.__dict__["project_id"] = project_id
         __props__.__dict__["public_endpoint"] = public_endpoint
+        __props__.__dict__["quantization"] = quantization
         __props__.__dict__["region"] = region
         __props__.__dict__["size"] = size
         __props__.__dict__["status"] = status
@@ -716,7 +723,7 @@ class InferenceDeployment(pulumi.CustomResource):
 
     @property
     @pulumi.getter(name="maxSize")
-    def max_size(self) -> pulumi.Output[int]:
+    def max_size(self) -> pulumi.Output[Optional[int]]:
         """
         The maximum size of the pool.
         """
@@ -724,7 +731,7 @@ class InferenceDeployment(pulumi.CustomResource):
 
     @property
     @pulumi.getter(name="minSize")
-    def min_size(self) -> pulumi.Output[int]:
+    def min_size(self) -> pulumi.Output[Optional[int]]:
         """
         The minimum size of the pool.
         """
@@ -742,7 +749,7 @@ class InferenceDeployment(pulumi.CustomResource):
     @pulumi.getter(name="modelName")
     def model_name(self) -> pulumi.Output[str]:
         """
-        The model name to use for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
+        The model name used for the deployment. Model names can be found in Console or using Scaleway's CLI (`scw inference model list`)
         """
         return pulumi.get(self, "model_name")
 
@@ -786,6 +793,14 @@ class InferenceDeployment(pulumi.CustomResource):
         """
         return pulumi.get(self, "public_endpoint")
 
+    @property
+    @pulumi.getter
+    def quantization(self) -> pulumi.Output[Optional[int]]:
+        """
+        The number of bits each model parameter should be quantized to
+        """
+        return pulumi.get(self, "quantization")
+
     @property
     @pulumi.getter
     def region(self) -> pulumi.Output[str]:

pulumiverse_scaleway/instance/_inputs.py

@@ -17,6 +17,8 @@ from .. import _utilities
 __all__ = [
     'ImageAdditionalVolumeArgs',
     'ImageAdditionalVolumeArgsDict',
+    'PrivateNicPrivateIpArgs',
+    'PrivateNicPrivateIpArgsDict',
     'SecurityGroupInboundRuleArgs',
     'SecurityGroupInboundRuleArgsDict',
     'SecurityGroupOutboundRuleArgs',
@@ -25,6 +27,8 @@ __all__ = [
     'SecurityGroupRulesInboundRuleArgsDict',
     'SecurityGroupRulesOutboundRuleArgs',
     'SecurityGroupRulesOutboundRuleArgsDict',
+    'ServerPrivateIpArgs',
+    'ServerPrivateIpArgsDict',
     'ServerPrivateNetworkArgs',
     'ServerPrivateNetworkArgsDict',
     'ServerPublicIpArgs',
@@ -309,6 +313,58 @@ class ImageAdditionalVolumeArgs:
         pulumi.set(self, "zone", value)
 
 
+if not MYPY:
+    class PrivateNicPrivateIpArgsDict(TypedDict):
+        address: NotRequired[pulumi.Input[str]]
+        """
+        The private IP address.
+        """
+        id: NotRequired[pulumi.Input[str]]
+        """
+        The ID of the IP address resource.
+        """
+elif False:
+    PrivateNicPrivateIpArgsDict: TypeAlias = Mapping[str, Any]
+
+@pulumi.input_type
+class PrivateNicPrivateIpArgs:
+    def __init__(__self__, *,
+                 address: Optional[pulumi.Input[str]] = None,
+                 id: Optional[pulumi.Input[str]] = None):
+        """
+        :param pulumi.Input[str] address: The private IP address.
+        :param pulumi.Input[str] id: The ID of the IP address resource.
+        """
+        if address is not None:
+            pulumi.set(__self__, "address", address)
+        if id is not None:
+            pulumi.set(__self__, "id", id)
+
+    @property
+    @pulumi.getter
+    def address(self) -> Optional[pulumi.Input[str]]:
+        """
+        The private IP address.
+        """
+        return pulumi.get(self, "address")
+
+    @address.setter
+    def address(self, value: Optional[pulumi.Input[str]]):
+        pulumi.set(self, "address", value)
+
+    @property
+    @pulumi.getter
+    def id(self) -> Optional[pulumi.Input[str]]:
+        """
+        The ID of the IP address resource.
+        """
+        return pulumi.get(self, "id")
+
+    @id.setter
+    def id(self, value: Optional[pulumi.Input[str]]):
+        pulumi.set(self, "id", value)
+
+
 if not MYPY:
     class SecurityGroupInboundRuleArgsDict(TypedDict):
         action: pulumi.Input[str]
@@ -849,6 +905,58 @@ class SecurityGroupRulesOutboundRuleArgs:
         pulumi.set(self, "protocol", value)
 
 
+if not MYPY:
+    class ServerPrivateIpArgsDict(TypedDict):
+        address: NotRequired[pulumi.Input[str]]
+        """
+        The private IP address.
+        """
+        id: NotRequired[pulumi.Input[str]]
+        """
+        The ID of the IP address resource.
+        """
+elif False:
+    ServerPrivateIpArgsDict: TypeAlias = Mapping[str, Any]
+
+@pulumi.input_type
+class ServerPrivateIpArgs:
+    def __init__(__self__, *,
+                 address: Optional[pulumi.Input[str]] = None,
+                 id: Optional[pulumi.Input[str]] = None):
+        """
+        :param pulumi.Input[str] address: The private IP address.
+        :param pulumi.Input[str] id: The ID of the IP address resource.
+        """
+        if address is not None:
+            pulumi.set(__self__, "address", address)
+        if id is not None:
+            pulumi.set(__self__, "id", id)
+
+    @property
+    @pulumi.getter
+    def address(self) -> Optional[pulumi.Input[str]]:
+        """
+        The private IP address.
+        """
+        return pulumi.get(self, "address")
+
+    @address.setter
+    def address(self, value: Optional[pulumi.Input[str]]):
+        pulumi.set(self, "address", value)
+
+    @property
+    @pulumi.getter
+    def id(self) -> Optional[pulumi.Input[str]]:
+        """
+        The ID of the IP address resource.
+        """
+        return pulumi.get(self, "id")
+
+    @id.setter
+    def id(self, value: Optional[pulumi.Input[str]]):
+        pulumi.set(self, "id", value)
+
+
 if not MYPY:
     class ServerPrivateNetworkArgsDict(TypedDict):
         pn_id: pulumi.Input[str]