fal 1.8.1__tar.gz → 1.8.3__tar.gz

{fal-1.8.1/fal.egg-info → fal-1.8.3}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.2
 Name: fal
-Version: 1.8.1
+Version: 1.8.3
 Summary: fal is an easy-to-use Serverless Python Framework
 Author: Features & Labels <support@fal.ai>
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 Requires-Dist: isolate[build]<0.16.0,>=0.15.0
-Requires-Dist: isolate-proto<0.7.0,>=0.6.0
+Requires-Dist: isolate-proto<0.7.0,>=0.6.6
 Requires-Dist: grpcio==1.64.0
 Requires-Dist: dill==0.3.7
 Requires-Dist: cloudpickle==3.0.0

{fal-1.8.1 → fal-1.8.3/fal.egg-info}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.2
 Name: fal
-Version: 1.8.1
+Version: 1.8.3
 Summary: fal is an easy-to-use Serverless Python Framework
 Author: Features & Labels <support@fal.ai>
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 Requires-Dist: isolate[build]<0.16.0,>=0.15.0
-Requires-Dist: isolate-proto<0.7.0,>=0.6.0
+Requires-Dist: isolate-proto<0.7.0,>=0.6.6
 Requires-Dist: grpcio==1.64.0
 Requires-Dist: dill==0.3.7
 Requires-Dist: cloudpickle==3.0.0

{fal-1.8.1 → fal-1.8.3}/fal.egg-info/requires.txt

@@ -1,5 +1,5 @@
 isolate[build]<0.16.0,>=0.15.0
-isolate-proto<0.7.0,>=0.6.0
+isolate-proto<0.7.0,>=0.6.6
 grpcio==1.64.0
 dill==0.3.7
 cloudpickle==3.0.0

{fal-1.8.1 → fal-1.8.3}/pyproject.toml

@@ -23,7 +23,7 @@ readme = "README.md"
 requires-python = ">=3.8"
 dependencies = [
     "isolate[build]>=0.15.0,<0.16.0",
-    "isolate-proto>=0.6.0,<0.7.0",
+    "isolate-proto>=0.6.6,<0.7.0",
     "grpcio==1.64.0",
     "dill==0.3.7",
     "cloudpickle==3.0.0",

{fal-1.8.1 → fal-1.8.3}/src/fal/_fal_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '1.8.1'
-__version_tuple__ = version_tuple = (1, 8, 1)
+__version__ = version = '1.8.3'
+__version_tuple__ = version_tuple = (1, 8, 3)

{fal-1.8.1 → fal-1.8.3}/src/fal/api.py

@@ -403,6 +403,7 @@ class FalServerlessHost(Host):
             "setup_function",
             "metadata",
             "request_timeout",
+            "startup_timeout",
             "_base_image",
             "_scheduler",
             "_scheduler_options",
@@ -452,6 +453,7 @@ class FalServerlessHost(Host):
         max_multiplexing = options.host.get("max_multiplexing")
         exposed_port = options.get_exposed_port()
         request_timeout = options.host.get("request_timeout")
+        startup_timeout = options.host.get("startup_timeout")
         machine_requirements = MachineRequirements(
             machine_types=machine_type,  # type: ignore
             num_gpus=options.host.get("num_gpus"),
@@ -464,6 +466,7 @@ class FalServerlessHost(Host):
             max_concurrency=max_concurrency,
             min_concurrency=min_concurrency,
             request_timeout=request_timeout,
+            startup_timeout=startup_timeout,
         )
 
         partial_func = _prepare_partial_func(func)
@@ -526,7 +529,7 @@ class FalServerlessHost(Host):
         exposed_port = options.get_exposed_port()
         setup_function = options.host.get("setup_function", None)
         request_timeout = options.host.get("request_timeout")
-
+        startup_timeout = options.host.get("startup_timeout")
         machine_requirements = MachineRequirements(
             machine_types=machine_type,  # type: ignore
             num_gpus=options.host.get("num_gpus"),
@@ -539,6 +542,7 @@ class FalServerlessHost(Host):
             max_concurrency=max_concurrency,
             min_concurrency=min_concurrency,
             request_timeout=request_timeout,
+            startup_timeout=startup_timeout,
         )
 
         return_value = _UNSET
@@ -705,6 +709,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,
@@ -732,6 +737,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,
@@ -809,6 +815,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,
@@ -841,6 +848,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,
@@ -867,6 +875,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,
@@ -893,6 +902,7 @@ def function(
     max_multiplexing: int = FAL_SERVERLESS_DEFAULT_MAX_MULTIPLEXING,
     min_concurrency: int = FAL_SERVERLESS_DEFAULT_MIN_CONCURRENCY,
     request_timeout: int | None = None,
+    startup_timeout: int | None = None,
     setup_function: Callable[..., None] | None = None,
     _base_image: str | None = None,
     _scheduler: str | None = None,

{fal-1.8.1 → fal-1.8.3}/src/fal/app.py

@@ -271,6 +271,7 @@ class App(fal.api.BaseServable):
     app_name: ClassVar[str]
     app_auth: ClassVar[Literal["private", "public", "shared"]] = "private"
     request_timeout: ClassVar[int | None] = None
+    startup_timeout: ClassVar[int | None] = None
 
     isolate_channel: async_grpc.Channel | None = None
 
@@ -282,6 +283,9 @@ class App(fal.api.BaseServable):
         if cls.request_timeout is not None:
             cls.host_kwargs["request_timeout"] = cls.request_timeout
 
+        if cls.startup_timeout is not None:
+            cls.host_kwargs["startup_timeout"] = cls.startup_timeout
+
         cls.app_name = getattr(cls, "app_name", app_name)
 
         if cls.__init__ is not App.__init__:

{fal-1.8.1 → fal-1.8.3}/src/fal/cli/apps.py

@@ -19,6 +19,9 @@ def _apps_table(apps: list[AliasInfo]):
     table.add_column("Max Concurrency")
     table.add_column("Max Multiplexing")
     table.add_column("Keep Alive")
+    table.add_column("Request Timeout")
+    table.add_column("Startup Timeout")
+    table.add_column("Machine Type")
     table.add_column("Active Workers")
 
     for app in apps:
@@ -30,6 +33,9 @@ def _apps_table(apps: list[AliasInfo]):
             str(app.max_concurrency),
             str(app.max_multiplexing),
             str(app.keep_alive),
+            str(app.request_timeout),
+            str(app.startup_timeout),
+            " ".join(app.machine_types),
             str(app.active_runners),
         )
 
@@ -67,6 +73,9 @@ def _app_rev_table(revs: list[ApplicationInfo]):
     table.add_column("Max Concurrency")
     table.add_column("Max Multiplexing")
     table.add_column("Keep Alive")
+    table.add_column("Request Timeout")
+    table.add_column("Startup Timeout")
+    table.add_column("Machine Type")
     table.add_column("Active Workers")
 
     for rev in revs:
@@ -76,6 +85,9 @@ def _app_rev_table(revs: list[ApplicationInfo]):
             str(rev.max_concurrency),
             str(rev.max_multiplexing),
             str(rev.keep_alive),
+            str(rev.request_timeout),
+            str(rev.startup_timeout),
+            " ".join(rev.machine_types),
             str(rev.active_runners),
         )
 
@@ -114,6 +126,8 @@ def _scale(args):
             and args.max_multiplexing is None
             and args.max_concurrency is None
             and args.min_concurrency is None
+            and args.request_timeout is None
+            and args.startup_timeout is None
         ):
             args.console.log("No parameters for update were provided, ignoring.")
             return
@@ -124,6 +138,8 @@ def _scale(args):
             max_multiplexing=args.max_multiplexing,
             max_concurrency=args.max_concurrency,
             min_concurrency=args.min_concurrency,
+            request_timeout=args.request_timeout,
+            startup_timeout=args.startup_timeout,
         )
         table = _apps_table([alias_info])
 
@@ -162,6 +178,16 @@ def _add_scale_parser(subparsers, parents):
         type=int,
         help="Minimum concurrency",
     )
+    parser.add_argument(
+        "--request-timeout",
+        type=int,
+        help="Request timeout (seconds).",
+    )
+    parser.add_argument(
+        "--startup-timeout",
+        type=int,
+        help="Startup timeout (seconds).",
+    )
     parser.set_defaults(func=_scale)
 
 

{fal-1.8.1 → fal-1.8.3}/src/fal/cli/deploy.py

@@ -111,10 +111,16 @@ def _deploy_from_reference(
     )
 
     if app_id:
-        gateway_host = _remove_http_and_port_from_url(host.url)
-        gateway_host = (
-            gateway_host.replace("api.", "").replace("alpha.", "").replace("ai", "run")
-        )
+        env_host = _remove_http_and_port_from_url(host.url)
+        env_host = env_host.replace("api.", "").replace("alpha.", "")
+
+        env_host_parts = env_host.split(".")
+
+        # keep the last 3 parts
+        playground_host = ".".join(env_host_parts[-3:])
+
+        # just replace .ai for .run
+        endpoint_host = env_host.replace(".ai", ".run")
 
         args.console.print(
             "Registered a new revision for function "
@@ -123,12 +129,12 @@ def _deploy_from_reference(
         args.console.print("Playground:")
         for endpoint in loaded.endpoints:
             args.console.print(
-                f"\thttps://fal.ai/models/{user.username}/{app_name}{endpoint}"
+                f"\thttps://{playground_host}/models/{user.username}/{app_name}{endpoint}"
             )
         args.console.print("Endpoints:")
         for endpoint in loaded.endpoints:
             args.console.print(
-                f"\thttps://{gateway_host}/{user.username}/{app_name}{endpoint}"
+                f"\thttps://{endpoint_host}/{user.username}/{app_name}{endpoint}"
             )
 
 

{fal-1.8.1 → fal-1.8.3}/src/fal/sdk.py

@@ -196,6 +196,9 @@ class ApplicationInfo:
     max_multiplexing: int
     active_runners: int
     min_concurrency: int
+    machine_types: list[str]
+    request_timeout: int
+    startup_timeout: int
 
 
 @dataclass
@@ -208,6 +211,9 @@ class AliasInfo:
     max_multiplexing: int
     active_runners: int
     min_concurrency: int
+    machine_types: list[str]
+    request_timeout: int
+    startup_timeout: int
 
 
 @dataclass
@@ -313,6 +319,9 @@ def _from_grpc_application_info(
         max_multiplexing=message.max_multiplexing,
         active_runners=message.active_runners,
         min_concurrency=message.min_concurrency,
+        machine_types=message.machine_types,
+        request_timeout=message.request_timeout,
+        startup_timeout=message.startup_timeout,
     )
 
 
@@ -336,6 +345,9 @@ def _from_grpc_alias_info(message: isolate_proto.AliasInfo) -> AliasInfo:
         max_multiplexing=message.max_multiplexing,
         active_runners=message.active_runners,
         min_concurrency=message.min_concurrency,
+        machine_types=message.machine_types,
+        request_timeout=message.request_timeout,
+        startup_timeout=message.startup_timeout,
     )
 
 
@@ -402,6 +414,7 @@ class MachineRequirements:
     max_multiplexing: int | None = None
     min_concurrency: int | None = None
     request_timeout: int | None = None
+    startup_timeout: int | None = None
 
     def __post_init__(self):
         if isinstance(self.machine_types, str):
@@ -519,6 +532,7 @@ class FalServerlessConnection:
                 min_concurrency=machine_requirements.min_concurrency,
                 max_multiplexing=machine_requirements.max_multiplexing,
                 request_timeout=machine_requirements.request_timeout,
+                startup_timeout=machine_requirements.startup_timeout,
             )
         else:
             wrapped_requirements = None
@@ -562,6 +576,8 @@ class FalServerlessConnection:
         max_multiplexing: int | None = None,
         max_concurrency: int | None = None,
         min_concurrency: int | None = None,
+        request_timeout: int | None = None,
+        startup_timeout: int | None = None,
     ) -> AliasInfo:
         request = isolate_proto.UpdateApplicationRequest(
             application_name=application_name,
@@ -569,6 +585,8 @@ class FalServerlessConnection:
             max_multiplexing=max_multiplexing,
             max_concurrency=max_concurrency,
             min_concurrency=min_concurrency,
+            request_timeout=request_timeout,
+            startup_timeout=startup_timeout,
         )
         res: isolate_proto.UpdateApplicationResult = self.stub.UpdateApplication(
             request
@@ -614,6 +632,7 @@ class FalServerlessConnection:
                 max_multiplexing=machine_requirements.max_multiplexing,
                 min_concurrency=machine_requirements.min_concurrency,
                 request_timeout=machine_requirements.request_timeout,
+                startup_timeout=machine_requirements.startup_timeout,
             )
         else:
             wrapped_requirements = None

{fal-1.8.1 → fal-1.8.3}/src/fal/toolkit/file/file.py

@@ -59,7 +59,7 @@ def get_builtin_repository(id: RepositoryId | FileRepository) -> FileRepository:
 get_builtin_repository.__module__ = "__main__"
 
 DEFAULT_REPOSITORY: FileRepository | RepositoryId = "fal_v3"
-FALLBACK_REPOSITORY: FileRepository | RepositoryId = "cdn"
+FALLBACK_REPOSITORY: FileRepository | RepositoryId = "fal"
 OBJECT_LIFECYCLE_PREFERENCE_KEY = "x-fal-object-lifecycle-preference"
 
 
@@ -153,9 +153,12 @@ class File(BaseModel):
         object_lifecycle_preference = (
             request_lifecycle_preference(request) or LIFECYCLE_PREFERENCE.get()
         )
+        save_kwargs.setdefault(
+            "object_lifecycle_preference", object_lifecycle_preference
+        )
 
         try:
-            url = repo.save(fdata, object_lifecycle_preference, **save_kwargs)
+            url = repo.save(fdata, **save_kwargs)
         except Exception:
             if not fallback_repository:
                 raise

{fal-1.8.1 → fal-1.8.3}/src/fal/toolkit/file/providers/fal.py

@@ -414,6 +414,7 @@ class MultipartUploadV3:
                 method="POST",
                 headers={
                     **self.auth_headers,
+                    "Content-Type": "application/json",
                     "Accept": "application/json",
                 },
                 data=json.dumps(