wandb 0.21.2__py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl → 0.23.0rc1__py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl

package_readme.md

@@ -1,88 +1,107 @@
 <div align="center">
-  <img src="https://i.imgur.com/RUtiVzH.png" width="600" /><br><br>
+  <img src="https://i.imgur.com/dQLeGCc.png" width="600" /><br><br>
 </div>
 
-# Weights and Biases [![PyPI](https://img.shields.io/pypi/v/wandb)](https://pypi.python.org/pypi/wandb) [![Conda (channel only)](https://img.shields.io/conda/vn/conda-forge/wandb)](https://anaconda.org/conda-forge/wandb) [![CircleCI](https://img.shields.io/circleci/build/github/wandb/wandb/main)](https://circleci.com/gh/wandb/wandb) [![Codecov](https://img.shields.io/codecov/c/gh/wandb/wandb)](https://codecov.io/gh/wandb/wandb)
+<p align="center">
+<a href="https://pypi.python.org/pypi/wandb"><img src="https://img.shields.io/pypi/v/wandb" /></a>
+<a href="https://anaconda.org/conda-forge/wandb"><img src="https://img.shields.io/conda/vn/conda-forge/wandb" /></a>
+<a href="https://pypi.python.org/pypi/wandb"><img src="https://img.shields.io/pypi/pyversions/wandb" /></a>
+<a href="https://circleci.com/gh/wandb/wandb"><img src="https://img.shields.io/circleci/build/github/wandb/wandb/main" /></a>
+<a href="https://codecov.io/gh/wandb/wandb"><img src="https://img.shields.io/codecov/c/gh/wandb/wandb" /></a>
+</p>
+<p align='center'>
+<a href="https://colab.research.google.com/github/wandb/examples/blob/master/colabs/intro/Intro_to_Weights_%26_Biases.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" /></a>
+</p>
 
-Use W&B to build better models faster. Track and visualize all the pieces of your machine learning pipeline, from datasets to production machine learning models. Get started with W&B today, [sign up for an account!](https://wandb.com?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=readme)
+Use W&B to build better models faster. Track and visualize all the pieces of your machine learning pipeline, from datasets to production machine learning models. Get started with W&B today, [sign up for a W&B account](https://wandb.com?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=readme)!
 
+<br>
 
+Building an LLM app? Track, debug, evaluate, and monitor LLM apps with [Weave](https://wandb.github.io/weave?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=readme), our new suite of tools for GenAI.
 
-See the [W&B Developer Guide](https://docs.wandb.ai/?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=documentation) and [API Reference Guide](https://docs.wandb.ai/ref?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=documentation) for a full technical description of the W&B platform.
+&nbsp;
+
+# Documentation
+
+See the [W&B Developer Guide](https://docs.wandb.ai/?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=documentation) and [API Reference Guide](https://docs.wandb.ai/training/api-reference#api-overview?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=documentation) for a full technical description of the W&B platform.
 
 &nbsp;
 
 # Quickstart
 
-Get started with W&B in four steps:
+Install W&B to track, visualize, and manage machine learning experiments of any size.
 
-1. First, sign up for a [W&B account](https://wandb.ai/login?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=quickstart).
-
-2. Second, install the W&B SDK with [pip](https://pip.pypa.io/en/stable/). Navigate to your terminal and type the following command:
+## Install the wandb library
 
 ```shell
 pip install wandb
 ```
 
-3. Third, log into W&B:
+## Sign up and create an API key
 
-```python
-wandb.login()
-```
+Sign up for a [W&B account](https://wandb.ai/login?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=quickstart). Optionally, use the `wandb login` CLI to configure an API key on your machine. You can skip this step -- W&B will prompt you for an API key the first time you use it.
 
-4. Use the example code snippet below as a template to integrate W&B to your Python script:
+## Create a machine learning training experiment
+
+In your Python script or notebook, initialize a W&B run with `wandb.init()`.
+Specify hyperparameters and log metrics and other information to W&B.
 
 ```python
 import wandb
 
-# Start a W&B Run with wandb.init
-run = wandb.init(project="my_first_project")
-
-# Save model inputs and hyperparameters in a wandb.config object
-config = run.config
-config.learning_rate = 0.01
+# Project that the run is recorded to
+project = "my-awesome-project"
 
-# Model training code here ...
+# Dictionary with hyperparameters
+config = {"epochs" : 1337, "lr" : 3e-4}
 
-# Log metrics over time to visualize performance with wandb.log
-for i in range(10):
-    run.log({"loss": ...})
+# The `with` syntax marks the run as finished upon exiting the `with` block,
+# and it marks the run "failed" if there's an exception.
+#
+# In a notebook, it may be more convenient to write `run = wandb.init()`
+# and manually call `run.finish()` instead of using a `with` block.
+with wandb.init(project=project, config=config) as run:
+    # Training code here
 
-# Mark the run as finished, and finish uploading all data
-run.finish()
+    # Log values to W&B with run.log()
+    run.log({"accuracy": 0.9, "loss": 0.1})
 ```
 
-For example, if the preceding code was stored in a script called train.py:
+Visit [wandb.ai/home](https://wandb.ai/home) to view recorded metrics such as accuracy and loss and how they changed during each training step. Each run object appears in the Runs column with generated names.
 
-```shell
-python train.py
-```
+&nbsp;
 
-You will see a URL in your terminal logs when your script starts and finishes. Data is staged locally in a directory named _wandb_ relative to your script. Navigate to the W&B App to view a dashboard of your first W&B Experiment. Use the W&B App to compare multiple experiments in a unified place, dive into the results of a single run, and much more!
+# Integrations
+
+W&B [integrates](https://docs.wandb.ai/models/integrations) with popular ML frameworks and libraries making it fast and easy to set up experiment tracking and data versioning inside existing projects.
+
+For developers adding W&B to a new framework, follow the [W&B Developer Guide](https://docs.wandb.ai/models/integrations/add-wandb-to-any-library).
 
 &nbsp;
 
-# Integrations
+# W&B Hosting Options
+
+Weights & Biases is available in the cloud or installed on your private infrastructure. Set up a W&B Server in a production environment in one of three ways:
 
-Use your favorite framework with W&B. W&B integrations make it fast and easy to set up experiment tracking and data versioning inside existing projects. For more information on how to integrate W&B with the framework of your choice, see [W&B Integrations](https://docs.wandb.ai/guides/integrations) in the W&B Developer Guide.
+1. [Multi-tenant Cloud](https://docs.wandb.ai/platform/hosting/hosting-options/multi_tenant_cloud?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=hosting): Fully managed platform deployed in W&B’s Google Cloud Platform (GCP) account in GCP’s North America regions.
+2. [Dedicated Cloud](https://docs.wandb.ai/platform/hosting/hosting-options/dedicated_cloud?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=hosting): Single-tenant, fully managed platform deployed in W&B’s AWS, GCP, or Azure cloud accounts. Each Dedicated Cloud instance has its own isolated network, compute and storage from other W&B Dedicated Cloud instances.
+3. [Self-Managed](https://docs.wandb.ai/platform/hosting/hosting-options/self-managed?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=hosting): Deploy W&B Server on your AWS, GCP, or Azure cloud account or within your on-premises infrastructure.
+
+See the [Hosting documentation](https://docs.wandb.ai/guides/hosting?utm_source=github&utm_medium=code&utm_campaign=wandb&utm_content=hosting) in the W&B Developer Guide for more information.
 
 &nbsp;
 
 # Python Version Support
 
-We are committed to supporting our minimum required Python version for *at least* six months after its official end-of-life (EOL) date, as defined by the Python Software Foundation. You can find a list of Python EOL dates [here](https://devguide.python.org/versions/).
+We are committed to supporting our minimum required Python version for _at least_ six months after its official end-of-life (EOL) date, as defined by the Python Software Foundation. You can find a list of Python EOL dates [here](https://devguide.python.org/versions/).
 
 When we discontinue support for a Python version, we will increment the library’s minor version number to reflect this change.
 
 &nbsp;
 
 # Contribution guidelines
-Weights & Biases ❤️ open source, and we welcome contributions from the community! See the [Contribution guide](https://github.com/wandb/wandb/blob/main/CONTRIBUTING.md) for more information on the development workflow and the internals of the wandb library. For wandb bugs and feature requests, visit [GitHub Issues](https://github.com/wandb/wandb/issues) or contact support@wandb.com.
 
-&nbsp;
-
-# Academic Researchers
-Reach out to W&B Support at support@wandb.com to get a [free academic license](https://www.wandb.com/academic) for you and your research group.
+Weights & Biases ❤️ open source, and we welcome contributions from the community! See the [Contribution guide](https://github.com/wandb/wandb/blob/main/CONTRIBUTING.md) for more information on the development workflow and the internals of the wandb library. For wandb bugs and feature requests, visit [GitHub Issues](https://github.com/wandb/wandb/issues) or contact support@wandb.com.
 
 &nbsp;
 

wandb/__init__.py

@@ -6,11 +6,11 @@ For guides and examples, see https://docs.wandb.ai.
 
 For scripts and interactive notebooks, see https://github.com/wandb/examples.
 
-For reference documentation, see https://docs.wandb.com/ref/python.
+For reference documentation, see https://docs.wandb.ai/models/ref/python.
 """
 from __future__ import annotations
 
-__version__ = "0.21.2"
+__version__ = "0.23.0rc1"
 
 
 from wandb.errors import Error
@@ -31,7 +31,6 @@ wandb.wandb_lib = wandb_sdk.lib  # type: ignore
 init = wandb_sdk.init
 setup = wandb_sdk.setup
 attach = _attach = wandb_sdk._attach
-_sync = wandb_sdk._sync
 teardown = _teardown = wandb_sdk.teardown
 finish = wandb_sdk.finish
 join = finish
@@ -53,7 +52,6 @@ _lazyloader = wandb.wandb_lib.lazyloader  # type: ignore
 
 from wandb.integration.torch import wandb_torch
 
-# Move this (keras.__init__ expects it at top level)
 from wandb.sdk.data_types._private import _cleanup_media_tmp_dir
 
 _cleanup_media_tmp_dir()
@@ -61,8 +59,6 @@ _cleanup_media_tmp_dir()
 from wandb.data_types import Graph
 from wandb.data_types import Image
 from wandb.data_types import Plotly
-
-# from wandb.data_types import Bokeh # keeping out of top level for now since Bokeh plots have poor UI
 from wandb.data_types import Video
 from wandb.data_types import Audio
 from wandb.data_types import Table
@@ -84,31 +80,6 @@ from wandb.sdk.wandb_run import Run
 # Artifact import types
 from wandb.sdk.artifacts.artifact_ttl import ArtifactTTL
 
-# Used to make sure we don't use some code in the incorrect process context
-_IS_INTERNAL_PROCESS = False
-
-
-def _set_internal_process(disable=False):
-    global _IS_INTERNAL_PROCESS
-    if _IS_INTERNAL_PROCESS is None:
-        return
-    if disable:
-        _IS_INTERNAL_PROCESS = None
-        return
-    _IS_INTERNAL_PROCESS = True
-
-
-def _assert_is_internal_process():
-    if _IS_INTERNAL_PROCESS is None:
-        return
-    assert _IS_INTERNAL_PROCESS
-
-
-def _assert_is_user_process():
-    if _IS_INTERNAL_PROCESS is None:
-        return
-    assert not _IS_INTERNAL_PROCESS
-
 
 # globals
 Api = PublicApi
@@ -177,10 +148,9 @@ def ensure_configured():
 
 
 def set_trace():
-    import pdb  # TODO: support other debuggers
+    import pdb
 
-    #  frame = sys._getframe().f_back
-    pdb.set_trace()  # TODO: pass the parent stack...
+    pdb.set_trace()
 
 
 def load_ipython_extension(ipython):
@@ -193,8 +163,6 @@ if wandb_sdk.lib.ipython.in_notebook():
     load_ipython_extension(get_ipython())
 
 
-from .analytics import Sentry as _Sentry
-
 if "dev" in __version__:
     import wandb.env
     import os
@@ -205,9 +173,6 @@ if "dev" in __version__:
         "false",
     )
 
-_sentry = _Sentry()
-_sentry.setup()
-
 
 __all__ = (
     "__version__",

wandb/__init__.pyi

@@ -75,7 +75,6 @@ from typing import (
 )
 
 import wandb.plot as plot
-from wandb.analytics import Sentry
 from wandb.apis import InternalApi
 from wandb.apis import PublicApi as Api
 from wandb.data_types import (
@@ -107,14 +106,13 @@ if TYPE_CHECKING:
     import wandb
     from wandb.plot import CustomChart
 
-__version__: str = "0.21.2"
+__version__: str = "0.23.0rc1"
 
 run: Run | None
 config: wandb_config.Config
 summary: wandb_summary.Summary
 
 # private attributes
-_sentry: Sentry
 api: InternalApi
 patched: Dict[str, List[Callable]]
 
@@ -388,8 +386,8 @@ def init(
             enable on your settings page.
         tensorboard: Deprecated. Use `sync_tensorboard` instead.
         sync_tensorboard: Enables automatic syncing of W&B logs from TensorBoard
-            or TensorBoardX, saving relevant event files for viewing in the W&B UI.
-            saving relevant event files for viewing in the W&B UI. (Default: `False`)
+            or TensorBoardX, saving relevant event files for viewing in
+            the W&B UI.
         monitor_gym: Enables automatic logging of videos of the environment when
             using OpenAI Gym.
         settings: Specifies a dictionary or `wandb.Settings` object with advanced
@@ -446,14 +444,14 @@ def finish(
     ...
 
 def login(
-    anonymous: Optional[Literal["must", "allow", "never"]] = None,
-    key: Optional[str] = None,
-    relogin: Optional[bool] = None,
-    host: Optional[str] = None,
-    force: Optional[bool] = None,
-    timeout: Optional[int] = None,
+    anonymous: Literal["must", "allow", "never"] | None = None,
+    key: str | None = None,
+    relogin: bool | None = None,
+    host: str | None = None,
+    force: bool | None = None,
+    timeout: int | None = None,
     verify: bool = False,
-    referrer: Optional[str] = None,
+    referrer: str | None = None,
 ) -> bool:
     """Set up W&B login credentials.
 
@@ -580,7 +578,7 @@ def log(
         run.log({"accuracy": 0.8}, step=current_step)
         current_step += 1
         run.log({"train-loss": 0.4}, step=current_step)
-        run.log({"accuracy": 0.9}, step=current_step)
+        run.log({"accuracy": 0.9}, step=current_step, commit=True)
     ```
 
     Args:
@@ -754,6 +752,9 @@ def save(
     When given an absolute path or glob and no `base_path`, one
     directory level is preserved as in the example above.
 
+    Files are automatically deduplicated: calling `save()` multiple times
+    on the same file without modifications will not re-upload it.
+
     Args:
         glob_str: A relative or absolute path or Unix glob.
         base_path: A path to use to infer a directory structure; see examples.
@@ -778,10 +779,10 @@ def save(
     run.save("these/are/myfiles/*", base_path="these")
     # => Saves files in an "are/myfiles/" folder in the run.
 
-    run.save("/User/username/Documents/run123/*.txt")
+    run.save("/Users/username/Documents/run123/*.txt")
     # => Saves files in a "run123/" folder in the run. See note below.
 
-    run.save("/User/username/Documents/run123/*.txt", base_path="/User")
+    run.save("/Users/username/Documents/run123/*.txt", base_path="/Users")
     # => Saves files in a "username/Documents/run123/" folder in the run.
 
     run.save("files/*/saveme.txt")

wandb/_analytics.py

@@ -0,0 +1,65 @@
+from __future__ import annotations
+
+from contextvars import ContextVar
+from dataclasses import dataclass, field
+from functools import wraps
+from typing import Callable, Final, TypeVar
+from uuid import UUID, uuid4
+
+from typing_extensions import ParamSpec
+
+from wandb._strutils import nameof
+
+P = ParamSpec("P")
+R = TypeVar("R")
+
+# Header keys for tracking the calling function
+X_WANDB_PYTHON_FUNC: Final[str] = "X-Wandb-Python-Func"
+X_WANDB_PYTHON_CALL_ID: Final[str] = "X-Wandb-Python-Call-Id"
+
+
+@dataclass(frozen=True)
+class TrackedFuncInfo:
+    func: str
+    """The fully qualified namespace of the tracked function."""
+
+    call_id: UUID = field(default_factory=uuid4)
+    """A unique identifier assigned to each invocation."""
+
+    def to_headers(self) -> dict[str, str]:
+        return {
+            X_WANDB_PYTHON_FUNC: self.func,
+            X_WANDB_PYTHON_CALL_ID: str(self.call_id),
+        }
+
+
+_current_func: ContextVar[TrackedFuncInfo] = ContextVar("_current_func")
+"""An internal, threadsafe context variable to hold the current function being tracked."""
+
+
+def tracked(func: Callable[P, R]) -> Callable[P, R]:
+    """A decorator to inject the calling function name into any GraphQL request headers.
+
+    If a tracked function calls another tracked function, only the outermost function in
+    the call stack will be tracked.
+    """
+    func_namespace = f"{func.__module__}.{nameof(func)}"
+
+    @wraps(func)
+    def wrapper(*args: P.args, **kwargs: P.kwargs) -> R:
+        # Don't override the current tracked function if it's already set
+        if tracked_func():
+            return func(*args, **kwargs)
+
+        token = _current_func.set(TrackedFuncInfo(func=func_namespace))
+        try:
+            return func(*args, **kwargs)
+        finally:
+            _current_func.reset(token)
+
+    return wrapper
+
+
+def tracked_func() -> TrackedFuncInfo | None:
+    """Returns info on the current tracked function, if any, otherwise None."""
+    return _current_func.get(None)

wandb/_iterutils.py

@@ -1,9 +1,11 @@
 from __future__ import annotations
 
+from collections.abc import Hashable
 from typing import TYPE_CHECKING, Any, Iterable, TypeVar, Union, overload
 
 if TYPE_CHECKING:
     T = TypeVar("T")
+    HashableT = TypeVar("HashableT", bound=Hashable)
     ClassInfo = Union[type[T], tuple[type[T], ...]]
 
 
@@ -12,9 +14,9 @@ def always_list(obj: Iterable[T], base_type: ClassInfo = ...) -> list[T]: ...
 @overload
 def always_list(obj: T, base_type: ClassInfo = ...) -> list[T]: ...
 def always_list(obj: Any, base_type: Any = (str, bytes)) -> list[T]:
-    """Return a guaranteed list of objects from a single instance OR iterable of such objects.
+    """Return a guaranteed list of objects from one instance OR an iterable of such items.
 
-    By default, assume the returned list should have string-like elements (i.e. `str`/`bytes`).
+    By default, assume the returned list should have string-like elements (`str`/`bytes`).
 
     Adapted from `more_itertools.always_iterable`, but simplified for internal use.  See:
     https://more-itertools.readthedocs.io/en/stable/api.html#more_itertools.always_iterable
@@ -22,6 +24,12 @@ def always_list(obj: Any, base_type: Any = (str, bytes)) -> list[T]:
     return [obj] if isinstance(obj, base_type) else list(obj)
 
 
+def unique_list(iterable: Iterable[HashableT]) -> list[HashableT]:
+    """Return a deduplicated list of items from the given iterable, preserving order."""
+    # Trick for O(1) uniqueness check that maintains order
+    return list(dict.fromkeys(iterable))
+
+
 def one(
     iterable: Iterable[T],
     too_short: type[Exception] | Exception | None = None,

wandb/_pydantic/__init__.py

@@ -1,30 +1,37 @@
 """Internal utilities for working with pydantic."""
 
-from .base import (
-    CompatBaseModel,
-    GQLBase,
-    GQLId,
-    SerializedToJson,
-    Typename,
-    ensure_json,
-)
-from .utils import IS_PYDANTIC_V2, from_json, gql_typename, pydantic_isinstance, to_json
-from .v1_compat import AliasChoices, computed_field, field_validator, model_validator
-
 __all__ = [
     "IS_PYDANTIC_V2",
     "CompatBaseModel",
+    "JsonableModel",
     "GQLBase",
+    "GQLInput",
+    "GQLResult",
+    "Connection",
+    "ConnectionWithTotal",
+    "Edge",
+    "PageInfo",
     "Typename",
     "GQLId",
-    "SerializedToJson",
     "AliasChoices",
     "computed_field",
     "field_validator",
     "model_validator",
     "pydantic_isinstance",
+    "to_camel",
     "to_json",
     "from_json",
-    "ensure_json",
     "gql_typename",
 ]
+
+from .base import CompatBaseModel, GQLBase, GQLInput, GQLResult, JsonableModel
+from .field_types import GQLId, Typename
+from .pagination import Connection, ConnectionWithTotal, Edge, PageInfo
+from .utils import IS_PYDANTIC_V2, from_json, gql_typename, pydantic_isinstance, to_json
+from .v1_compat import (
+    AliasChoices,
+    computed_field,
+    field_validator,
+    model_validator,
+    to_camel,
+)