mlrun 1.7.1rc4__py3-none-any.whl → 1.8.0rc8__py3-none-any.whl

mlrun/runtimes/sparkjob/spark3job.py

@@ -14,13 +14,13 @@
 import typing
 
 import kubernetes.client
-from mlrun_pipelines.mounts import mount_v3io, mount_v3iod
 
 import mlrun.common.schemas.function
 import mlrun.errors
 import mlrun.k8s_utils
 import mlrun.runtimes.pod
 from mlrun.config import config
+from mlrun.runtimes.mounts import mount_v3io, mount_v3iod
 
 from ...execution import MLClientCtx
 from ...model import RunObject
@@ -405,8 +405,8 @@ class Spark3JobSpec(KubeResourceSpec):
     def _verify_and_set_requests(
         self,
         resources_field_name,
-        mem: str = None,
-        cpu: str = None,
+        mem: typing.Optional[str] = None,
+        cpu: typing.Optional[str] = None,
         patch: bool = False,
     ):
         # Spark operator uses JVM notation for memory, so we must verify it separately
@@ -774,7 +774,11 @@ class Spark3Runtime(KubejobRuntime):
                 exporter_jar="/spark/jars/jmx_prometheus_javaagent-0.16.1.jar",
             )
 
-    def with_cores(self, executor_cores: int = None, driver_cores: int = None):
+    def with_cores(
+        self,
+        executor_cores: typing.Optional[int] = None,
+        driver_cores: typing.Optional[int] = None,
+    ):
         """
         Allows to configure spark.executor.cores and spark.driver.cores parameters. The values must be integers
         greater than or equal to 1. If a parameter is not specified, it defaults to 1.
@@ -853,7 +857,7 @@ class Spark3Runtime(KubejobRuntime):
         skip_deployed=False,
         is_kfp=False,
         mlrun_version_specifier=None,
-        builder_env: dict = None,
+        builder_env: typing.Optional[dict] = None,
         show_on_failure: bool = False,
         force_build: bool = False,
     ):
@@ -944,7 +948,10 @@ class Spark3Runtime(KubejobRuntime):
         )
 
     def with_executor_requests(
-        self, mem: str = None, cpu: str = None, patch: bool = False
+        self,
+        mem: typing.Optional[str] = None,
+        cpu: typing.Optional[str] = None,
+        patch: bool = False,
     ):
         """
         set executor pod required cpu/memory/gpu resources
@@ -954,8 +961,8 @@ class Spark3Runtime(KubejobRuntime):
 
     def with_executor_limits(
         self,
-        cpu: str = None,
-        gpus: int = None,
+        cpu: typing.Optional[str] = None,
+        gpus: typing.Optional[int] = None,
         gpu_type: str = "nvidia.com/gpu",
         patch: bool = False,
     ):
@@ -970,7 +977,10 @@ class Spark3Runtime(KubejobRuntime):
         )
 
     def with_driver_requests(
-        self, mem: str = None, cpu: str = None, patch: bool = False
+        self,
+        mem: typing.Optional[str] = None,
+        cpu: typing.Optional[str] = None,
+        patch: bool = False,
     ):
         """
         set driver pod required cpu/memory/gpu resources
@@ -980,8 +990,8 @@ class Spark3Runtime(KubejobRuntime):
 
     def with_driver_limits(
         self,
-        cpu: str = None,
-        gpus: int = None,
+        cpu: typing.Optional[str] = None,
+        gpus: typing.Optional[int] = None,
         gpu_type: str = "nvidia.com/gpu",
         patch: bool = False,
     ):

mlrun/runtimes/utils.py

@@ -18,6 +18,7 @@ import os
 import re
 from io import StringIO
 from sys import stderr
+from typing import Optional
 
 import pandas as pd
 
@@ -147,7 +148,7 @@ def add_code_metadata(path=""):
 
 def results_to_iter(results, runspec, execution):
     if not results:
-        logger.error("got an empty results list in to_iter")
+        logger.error("Got an empty results list in to_iter")
         return
 
     iter = []
@@ -174,7 +175,7 @@ def results_to_iter(results, runspec, execution):
 
     if not iter:
         execution.set_state("completed", commit=True)
-        logger.warning("warning!, zero iteration results")
+        logger.warning("Warning!, zero iteration results")
         return
     if hasattr(pd, "json_normalize"):
         df = pd.json_normalize(iter).sort_values("iter")
@@ -189,10 +190,10 @@ def results_to_iter(results, runspec, execution):
     item, id = selector(results, criteria)
     if runspec.spec.selector and not id:
         logger.warning(
-            f"no best result selected, check selector ({criteria}) or results"
+            f"No best result selected, check selector ({criteria}) or results"
         )
     if id:
-        logger.info(f"best iteration={id}, used criteria {criteria}")
+        logger.info(f"Best iteration={id}, used criteria {criteria}")
     task = results[item] if id and results else None
     execution.log_iteration_results(id, summary, task)
 
@@ -432,7 +433,7 @@ def enrich_function_from_dict(function, function_dict):
 
 def enrich_run_labels(
     labels: dict,
-    labels_to_enrich: list[RunLabels] = None,
+    labels_to_enrich: Optional[list[RunLabels]] = None,
 ):
     labels_enrichment = {
         RunLabels.owner: os.environ.get("V3IO_USERNAME") or getpass.getuser(),

mlrun/serving/merger.py

@@ -13,6 +13,8 @@
 # limitations under the License.
 #
 
+from typing import Optional
+
 import storey
 
 
@@ -37,10 +39,10 @@ class CacheEntry:
 class Merge(storey.Flow):
     def __init__(
         self,
-        full_event: bool = None,
-        key_path: str = None,
-        max_behind: int = None,
-        expected_num_events: int = None,
+        full_event: Optional[bool] = None,
+        key_path: Optional[str] = None,
+        max_behind: Optional[int] = None,
+        expected_num_events: Optional[int] = None,
         **kwargs,
     ):
         """Merge multiple events based on event id or provided key path

mlrun/serving/remote.py

@@ -14,6 +14,7 @@
 #
 import asyncio
 import json
+from typing import Optional
 
 import aiohttp
 import requests
@@ -40,14 +41,14 @@ class RemoteStep(storey.SendToHttp):
     def __init__(
         self,
         url: str,
-        subpath: str = None,
-        method: str = None,
-        headers: dict = None,
-        url_expression: str = None,
-        body_expression: str = None,
+        subpath: Optional[str] = None,
+        method: Optional[str] = None,
+        headers: Optional[dict] = None,
+        url_expression: Optional[str] = None,
+        body_expression: Optional[str] = None,
         return_json: bool = True,
-        input_path: str = None,
-        result_path: str = None,
+        input_path: Optional[str] = None,
+        result_path: Optional[str] = None,
         max_in_flight=None,
         retries=None,
         backoff_factor=None,
@@ -57,7 +58,7 @@ class RemoteStep(storey.SendToHttp):
         """class for calling remote endpoints
 
         sync and async graph step implementation for request/resp to remote service (class shortcut = "$remote")
-        url can be an http(s) url (e.g. "https://myservice/path") or an mlrun function uri ([project/]name).
+        url can be an http(s) url (e.g. `https://myservice/path`) or an mlrun function uri ([project/]name).
         alternatively the url_expression can be specified to build the url from the event (e.g. "event['url']").
 
         example pipeline::
@@ -241,15 +242,15 @@ class RemoteStep(storey.SendToHttp):
 class BatchHttpRequests(_ConcurrentJobExecution):
     def __init__(
         self,
-        url: str = None,
-        subpath: str = None,
-        method: str = None,
-        headers: dict = None,
-        url_expression: str = None,
-        body_expression: str = None,
+        url: Optional[str] = None,
+        subpath: Optional[str] = None,
+        method: Optional[str] = None,
+        headers: Optional[dict] = None,
+        url_expression: Optional[str] = None,
+        body_expression: Optional[str] = None,
         return_json: bool = True,
-        input_path: str = None,
-        result_path: str = None,
+        input_path: Optional[str] = None,
+        result_path: Optional[str] = None,
         retries=None,
         backoff_factor=None,
         timeout=None,
@@ -258,7 +259,7 @@ class BatchHttpRequests(_ConcurrentJobExecution):
         """class for calling remote endpoints in parallel
 
         sync and async graph step implementation for request/resp to remote service (class shortcut = "$remote")
-        url can be an http(s) url (e.g. "https://myservice/path") or an mlrun function uri ([project/]name).
+        url can be an http(s) url (e.g. `https://myservice/path`) or an mlrun function uri ([project/]name).
         alternatively the url_expression can be specified to build the url from the event (e.g. "event['url']").
 
         example pipeline::