diff --git a/kubeflow/spark/api/spark_client.py b/kubeflow/spark/api/spark_client.py
index 332eaa286..9f8b84dc0 100644
--- a/kubeflow/spark/api/spark_client.py
+++ b/kubeflow/spark/api/spark_client.py
@@ -27,6 +27,7 @@
     Executor,
     FileJob,
     FuncJob,
+    ResourceDict,
     SparkConnectInfo,
     SparkJob,
     SparkJobStatus,
@@ -64,7 +65,7 @@ def connect(
         base_url: str | None = None,
         token: str | None = None,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         spark_conf: dict[str, str] | None = None,
         driver: Driver | None = None,
         executor: Executor | None = None,
@@ -83,8 +84,8 @@ def connect(
                  If provided, connects to existing server. If None, creates new session.
             token: Optional authentication token for existing server.
             num_executors: Number of executor instances (create mode only).
-            resources_per_executor: Resource requirements per executor as dict.
-                Format: `{"cpu": "5", "memory": "10Gi"}` (create mode only).
+            resources_per_executor: Resource requirements per executor as ResourceDict.
+                Format: `{"cpu": 5, "memory": "10Gi"}` (create mode only).
             spark_conf: Spark configuration dictionary (create mode only).
             driver: Driver configuration object (create mode only).
             executor: Executor configuration object (create mode only).
@@ -175,7 +176,7 @@ def submit_job(
         self,
         job: FileJob | FuncJob,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         spark_conf: dict[str, str] | None = None,
         options: list | None = None,
     ) -> str:
@@ -196,8 +197,8 @@ def submit_job(
                 Number of executor instances.
 
             resources_per_executor:
-                Resource requirements per executor.
-                Format: ``{"cpu": "5", "memory": "10Gi"}``.
+                Resource requirements per executor as ResourceDict.
+                Format: ``{"cpu": 5, "memory": "10Gi"}``.
 
             spark_conf:
                 Spark configuration properties.
diff --git a/kubeflow/spark/backends/base.py b/kubeflow/spark/backends/base.py
index 0f41a18a6..58f06d591 100644
--- a/kubeflow/spark/backends/base.py
+++ b/kubeflow/spark/backends/base.py
@@ -24,6 +24,7 @@
     Executor,
     FileJob,
     FuncJob,
+    ResourceDict,
     SparkConnectInfo,
     SparkJob,
     SparkJobStatus,
@@ -44,7 +45,7 @@ class RuntimeBackend(abc.ABC):
     def create_and_connect(
         self,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         spark_conf: dict[str, str] | None = None,
         driver: Driver | None = None,
         executor: Executor | None = None,
@@ -59,7 +60,7 @@ def create_and_connect(
 
         Args:
             num_executors: Number of executor instances.
-            resources_per_executor: Resource requirements per executor.
+            resources_per_executor: Resource requirements per executor as ResourceDict.
             spark_conf: Spark configuration properties.
             driver: Driver configuration.
             executor: Executor configuration.
@@ -148,7 +149,7 @@ def submit_job(
         self,
         job: FileJob | FuncJob,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         options: list | None = None,
         spark_conf: dict[str, str] | None = None,
     ) -> SparkJob:
@@ -157,7 +158,7 @@ def submit_job(
         Args:
             job: Spark application to execute.
             num_executors: Number of executor instances.
-            resources_per_executor: Resource requirements for each executor.
+            resources_per_executor: Resource requirements for each executor as ResourceDict.
             options: List of configuration options. Use the Name option for a custom job name.
             spark_conf: Spark configuration properties to set on the SparkApplication.
 
diff --git a/kubeflow/spark/backends/kubernetes/backend.py b/kubeflow/spark/backends/kubernetes/backend.py
index 6d75874c8..aaf566581 100644
--- a/kubeflow/spark/backends/kubernetes/backend.py
+++ b/kubeflow/spark/backends/kubernetes/backend.py
@@ -55,6 +55,7 @@
     Executor,
     FileJob,
     FuncJob,
+    ResourceDict,
     SparkConnectInfo,
     SparkConnectState,
     SparkJob,
@@ -121,7 +122,7 @@ def __init__(self, backend_config: KubernetesBackendConfig):
     def _create_session(
         self,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         spark_conf: dict[str, str] | None = None,
         driver: Driver | None = None,
         executor: Executor | None = None,
@@ -131,7 +132,7 @@ def _create_session(
 
         Args:
             num_executors: Number of executor instances.
-            resources_per_executor: Resource requirements per executor.
+            resources_per_executor: Resource requirements per executor as ResourceDict.
             spark_conf: Spark configuration properties.
             driver: Driver configuration.
             executor: Executor configuration.
@@ -628,7 +629,7 @@ def _get_or_create() -> None:
     def create_and_connect(
         self,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         spark_conf: dict[str, str] | None = None,
         driver: Driver | None = None,
         executor: Executor | None = None,
@@ -645,7 +646,7 @@ def create_and_connect(
 
         Args:
             num_executors: Number of executor instances.
-            resources_per_executor: Resource requirements per executor.
+            resources_per_executor: Resource requirements per executor as ResourceDict.
             spark_conf: Spark configuration properties.
             driver: Driver configuration.
             executor: Executor configuration.
@@ -864,7 +865,7 @@ def submit_job(
         self,
         job: FileJob | FuncJob,
         num_executors: int | None = None,
-        resources_per_executor: dict[str, str] | None = None,
+        resources_per_executor: ResourceDict | None = None,
         options: list | None = None,
         spark_conf: dict[str, str] | None = None,
     ) -> SparkJob:
@@ -878,7 +879,7 @@ def submit_job(
                 Number of executor instances.
 
             resources_per_executor:
-                Resource requirements per executor.
+                Resource requirements per executor as ResourceDict.
 
             options:
                 List of additional Spark configuration options.
diff --git a/kubeflow/spark/backends/kubernetes/utils.py b/kubeflow/spark/backends/kubernetes/utils.py
index 1885bce04..518b0e96d 100644
--- a/kubeflow/spark/backends/kubernetes/utils.py
+++ b/kubeflow/spark/backends/kubernetes/utils.py
@@ -35,6 +35,7 @@
 from kubeflow.spark.types.types import (
     Driver,
     Executor,
+    ResourceDict,
     SparkConnectInfo,
     SparkConnectState,
     SparkJob,
@@ -116,6 +117,8 @@ def _resolve_driver_resources(
     Raises:
         ValueError:
             If the configured CPU or memory values are invalid.
+        TypeError:
+            If the configured CPU value is not a supported type.
     """
 
     cores = constants.DEFAULT_DRIVER_CPU
@@ -127,7 +130,7 @@ def _resolve_driver_resources(
 
         if "memory" in driver.resources:
             memory = _memory_kubernetes_to_spark(
-                driver.resources["memory"],
+                str(driver.resources["memory"]),
             )
 
     return cores, memory
@@ -136,7 +139,7 @@ def _resolve_driver_resources(
 def _resolve_executor_resources(
     executor: Executor | None = None,
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
 ) -> tuple[int, int, str]:
     """Resolve executor configuration.
 
@@ -148,7 +151,7 @@ def _resolve_executor_resources(
             Number of executor instances.
 
         resources_per_executor:
-            Resource requirements.
+            Resource requirements as ResourceDict.
 
     Returns:
         Tuple containing ``(instances, cores, memory)``.
@@ -156,6 +159,8 @@ def _resolve_executor_resources(
     Raises:
         ValueError:
             If the configured CPU or memory values are invalid.
+        TypeError:
+            If the configured CPU value is not a supported type.
     """
 
     if executor and executor.num_instances is not None:
@@ -183,7 +188,7 @@ def _resolve_executor_resources(
 
         if "memory" in resource_dict:
             memory = _memory_kubernetes_to_spark(
-                resource_dict["memory"],
+                str(resource_dict["memory"]),
             )
 
     return instances, cores, memory
@@ -247,22 +252,26 @@ def _memory_kubernetes_to_spark(memory: str) -> str:
     return f"{math.ceil(total_bytes / (2**20))}m"
 
 
-def _validate_cpu_value(cpu: str | int | None) -> int:
+def _validate_cpu_value(cpu: str | int | float | None) -> int:
     """Validate and normalize CPU cores value.
 
     Args:
-        cpu: CPU value provided by user.
+        cpu: CPU value provided by user (str, int, or float).
 
     Returns:
         Integer CPU core value.
 
     Raises:
-        ValueError: If CPU value is invalid.
+        ValueError: If CPU value is invalid or non-positive.
+        TypeError: If CPU value is not a supported type or is a boolean.
     """
     if cpu is None:
         raise ValueError("CPU value cannot be None")
 
-    if isinstance(cpu, int):
+    if isinstance(cpu, bool):
+        raise TypeError(f"Invalid CPU type '{type(cpu).__name__}'. Expected str, int, or float.")
+
+    if isinstance(cpu, (int, float)):
         cores = float(cpu)
 
     elif isinstance(cpu, str):
@@ -282,10 +291,13 @@ def _validate_cpu_value(cpu: str | int | None) -> int:
             cores = int(milli_cpu) / 1000
 
         else:
-            cores = float(cpu)
+            try:
+                cores = float(cpu)
+            except ValueError as e:
+                raise ValueError(f"Invalid CPU value: {cpu!r}") from e
 
     else:
-        raise ValueError(f"Invalid CPU type '{type(cpu)}'. Expected str or int.")
+        raise TypeError(f"Invalid CPU type '{type(cpu).__name__}'. Expected str, int, or float.")
 
     if not math.isfinite(cores) or cores <= 0:
         raise ValueError(f"Invalid CPU value: {cpu!r}")
@@ -315,12 +327,12 @@ def apply_options(
         backend:
             Backend used for option validation.
 
-        Raises:
-            ValueError:
-                If options are provided without a backend instance.
+    Raises:
+        ValueError:
+            If options are provided without a backend instance.
 
-            TypeError:
-                If an option is not callable.
+        TypeError:
+            If an option is not callable.
     """
     if not options:
         return
@@ -450,6 +462,8 @@ def get_spark_connect_driver_spec(
     Raises:
         ValueError:
             If the configured driver resources are invalid.
+        TypeError:
+            If the configured driver CPU resource is not a supported type.
     """
     cores, memory = _resolve_driver_resources(driver)
 
@@ -473,7 +487,7 @@ def get_spark_connect_driver_spec(
 def get_spark_connect_executor_spec(
     executor: Executor | None = None,
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
 ) -> models.SparkV1alpha1ExecutorSpec:
     """Convert SDK Executor to API ExecutorSpec.
 
@@ -484,7 +498,7 @@ def get_spark_connect_executor_spec(
     Args:
         executor: SDK Executor configuration.
         num_executors: Simple mode number of executors.
-        resources_per_executor: Simple mode resource requirements.
+        resources_per_executor: Simple mode resource requirements as ResourceDict.
 
     Returns:
         API ExecutorSpec model.
@@ -492,6 +506,8 @@ def get_spark_connect_executor_spec(
     Raises:
         ValueError:
             If the configured executor resources are invalid.
+        TypeError:
+            If the configured executor CPU resource is not a supported type.
     """
     instances, cores, memory = _resolve_executor_resources(
         executor,
@@ -511,7 +527,7 @@ def build_spark_connect_cr(
     namespace: str,
     spark_version: str | None = None,
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
     spark_conf: dict[str, str] | None = None,
     driver: Driver | None = None,
     executor: Executor | None = None,
@@ -531,7 +547,7 @@ def build_spark_connect_cr(
         namespace: Kubernetes namespace.
         spark_version: Spark version (default: `constants.DEFAULT_SPARK_VERSION`).
         num_executors: Number of executor instances (simple mode).
-        resources_per_executor: Resource requirements per executor (simple mode).
+        resources_per_executor: Resource requirements per executor as ResourceDict (simple mode).
         spark_conf: Spark configuration properties.
         driver: Driver configuration (advanced mode).
         executor: Executor configuration (advanced mode).
@@ -544,6 +560,8 @@ def build_spark_connect_cr(
     Raises:
         ValueError:
             If the provided driver or executor resource configuration is invalid.
+        TypeError:
+            If the provided driver or executor CPU resource is not a supported type.
     """
     _validate_spark_conf(spark_conf)
 
@@ -676,6 +694,8 @@ def get_spark_job_driver_spec(
     Raises:
         ValueError:
             If the default driver resource configuration is invalid.
+        TypeError:
+            If the driver CPU resource is not a supported type.
     """
     cores, memory = _resolve_driver_resources(driver)
 
@@ -688,13 +708,13 @@ def get_spark_job_driver_spec(
 
 def get_spark_job_executor_spec(
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
 ) -> models.SparkV1beta2ExecutorSpec:
     """Build ExecutorSpec for SparkApplication.
 
     Args:
         num_executors: Number of executor instances.
-        resources_per_executor: Resource requirements for each executor.
+        resources_per_executor: Resource requirements for each executor as ResourceDict.
 
     Returns:
         SparkApplication ExecutorSpec model.
@@ -702,6 +722,8 @@ def get_spark_job_executor_spec(
     Raises:
         ValueError:
             If the configured executor resources are invalid.
+        TypeError:
+            If the executor CPU resource is not a supported type.
     """
     instances, cores, memory = _resolve_executor_resources(
         num_executors=num_executors,
@@ -802,7 +824,7 @@ def get_spark_application_cr_from_file_job(
     main_file: str,
     arguments: list[str] | None = None,
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
     options: list | None = None,
     backend: Any | None = None,
     spark_conf: dict[str, str] | None = None,
@@ -815,7 +837,7 @@ def get_spark_application_cr_from_file_job(
         main_file: Path or URI to the Spark application file.
         arguments: Command-line arguments passed to the Spark application.
         num_executors: Number of executor instances.
-        resources_per_executor: Resource requirements for each executor.
+        resources_per_executor: Resource requirements for each executor as ResourceDict.
         options: List of configuration options.
         backend: Backend instance used for option validation.
         spark_conf: Spark configuration properties.
@@ -826,6 +848,8 @@ def get_spark_application_cr_from_file_job(
     Raises:
         ValueError:
             If the executor resource configuration is invalid.
+        TypeError:
+            If the executor CPU resource is not a supported type.
     """
     spark_application = models.SparkV1beta2SparkApplication(
         api_version=f"{constants.SPARK_APPLICATION_GROUP}/{constants.SPARK_APPLICATION_VERSION}",
@@ -865,7 +889,7 @@ def get_spark_application_cr_from_func_job(
     func: Callable,
     func_args: dict[str, Any] | None = None,
     num_executors: int | None = None,
-    resources_per_executor: dict[str, str] | None = None,
+    resources_per_executor: ResourceDict | None = None,
     options: list | None = None,
     backend: Any | None = None,
     spark_conf: dict[str, str] | None = None,
@@ -878,7 +902,7 @@ def get_spark_application_cr_from_func_job(
         func: Python function to execute as the Spark application.
         func_args: Keyword arguments passed to the function.
         num_executors: Number of executor instances.
-        resources_per_executor: Resource requirements for each executor.
+        resources_per_executor: Resource requirements for each executor as ResourceDict.
         options: List of configuration options.
         backend: Backend instance used for option validation.
         spark_conf: Spark configuration properties.
@@ -890,6 +914,8 @@ def get_spark_application_cr_from_func_job(
         ValueError:
             If the provided function is invalid or the executor resource
             configuration is invalid.
+        TypeError:
+            If the executor CPU resource is not a supported type.
     """
     _validate_spark_conf(spark_conf)
 
diff --git a/kubeflow/spark/types/types.py b/kubeflow/spark/types/types.py
index e1779139d..9be9829bd 100644
--- a/kubeflow/spark/types/types.py
+++ b/kubeflow/spark/types/types.py
@@ -21,6 +21,9 @@
 import logging
 from typing import Any
 
+# Type alias for driver and executor resource dictionary specifications
+ResourceDict = dict[str, str | int | float]
+
 logger = logging.getLogger(__name__)
 
 
@@ -87,7 +90,7 @@ class Driver:
     """
 
     image: str | None = None
-    resources: dict[str, str] | None = None
+    resources: ResourceDict | None = None
     java_options: str | None = None
     service_account: str | None = None
 
@@ -121,7 +124,7 @@ class Executor:
     """
 
     num_instances: int | None = None
-    resources_per_executor: dict[str, str] | None = None
+    resources_per_executor: ResourceDict | None = None
     java_options: str | None = None