kubetorch 0.0.1__tar.gz → 0.2.1__tar.gz

kubetorch-0.2.1/PKG-INFO

@@ -0,0 +1,121 @@
+Metadata-Version: 2.4
+Name: kubetorch
+Version: 0.2.1
+Summary: A Python interface for running ML workloads on Kubernetes
+License: Apache 2.0
+Author: Runhouse Team
+Requires-Python: >=3.9,<=3.13
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: License :: Other/Proprietary License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Topic :: System :: Distributed Computing
+Provides-Extra: client
+Provides-Extra: docs
+Provides-Extra: otel
+Provides-Extra: server
+Requires-Dist: fastapi ; extra == "server"
+Requires-Dist: httpx (>=0.28.0,<0.29.0)
+Requires-Dist: jinja2 (>=3.0)
+Requires-Dist: kubernetes (>=20.0.0,!=32.0.0)
+Requires-Dist: myst-parser ; extra == "docs"
+Requires-Dist: opentelemetry-api ; extra == "otel"
+Requires-Dist: opentelemetry-distro ; extra == "otel"
+Requires-Dist: opentelemetry-exporter-otlp ; extra == "otel"
+Requires-Dist: opentelemetry-exporter-prometheus ; extra == "otel"
+Requires-Dist: opentelemetry-instrumentation ; extra == "otel"
+Requires-Dist: opentelemetry-instrumentation-fastapi ; extra == "otel"
+Requires-Dist: opentelemetry-instrumentation-logging ; extra == "otel"
+Requires-Dist: opentelemetry-instrumentation-requests ; extra == "otel"
+Requires-Dist: opentelemetry-sdk ; extra == "otel"
+Requires-Dist: prometheus-api-client ; extra == "client"
+Requires-Dist: prometheus-client ; extra == "otel"
+Requires-Dist: pydantic (>=2.5.0)
+Requires-Dist: python-json-logger ; extra == "server"
+Requires-Dist: pyyaml (>=6.0,<7.0)
+Requires-Dist: requests (>=2.32.0,<3.0.0)
+Requires-Dist: rich ; extra == "client"
+Requires-Dist: sphinx ; extra == "docs"
+Requires-Dist: sphinx-autodoc-typehints ; extra == "docs"
+Requires-Dist: sphinx-book-theme ; extra == "docs"
+Requires-Dist: typer ; extra == "client"
+Requires-Dist: uvicorn[standard] ; extra == "server"
+Requires-Dist: web-pdb ; extra == "server"
+Requires-Dist: websockets (>=15.0,<16.0)
+Project-URL: Documentation, https://www.run.house/kubetorch/introduction
+Project-URL: Homepage, https://run.house
+Project-URL: Repository, https://github.com/run-house/kubetorch
+Description-Content-Type: text/markdown
+
+# 📦Kubetorch🔥
+
+**A Python interface for running ML workloads on Kubernetes**
+
+Kubetorch enables you to run any Python code on Kubernetes at any scale by specifying required resources, distribution, and scaling directly in code. It provides caching and hot redeployment for 1-2 second iteration cycles, handles hardware faults and preemptions programmatically, and orchestrates complex, heterogeneous workloads with built-in observability and fault tolerance.
+
+## Hello World
+
+```python
+from kubetorch import fn
+
+def hello_world():
+    return "Hello from Kubetorch!"
+
+if __name__ == "__main__":
+    # Define your compute
+    compute = kt.Compute(cpus=".1")
+
+    # Send local function to freshly launched remote compute
+    remote_hello = kt.fn(hello_world).to(compute)
+
+    # Runs remotely on your Kubernetes cluster
+    result = hello_world()
+    print(result)  # "Hello from Kubetorch!"
+```
+
+## What Kubetorch Enables
+
+- **100x faster iteration** from 10+ minutes to 1-3 seconds for complex ML applications like RL and distributed training
+- **50%+ compute cost savings** through intelligent resource allocation, bin-packing, and dynamic scaling
+- **95% fewer production faults** with built-in fault handling with programmatic error recovery and resource adjustment
+
+## Installation
+
+### 1. Python Client
+
+```bash
+pip install "kubetorch[client]"
+```
+
+### 2. Kubernetes Deployment (Helm)
+
+```bash
+# Option 1: Install directly from OCI registry
+helm upgrade --install kubetorch oci://ghcr.io/run-house/charts/kubetorch \
+  --version 0.2.1 -n kubetorch --create-namespace
+
+# Option 2: Download chart locally first
+helm pull oci://ghcr.io/run-house/charts/kubetorch --version 0.2.1 --untar
+helm upgrade --install kubetorch ./kubetorch -n kubetorch --create-namespace
+```
+
+For detailed setup instructions, see our [Installation Guide](https://www.run.house/kubetorch/installation).
+
+
+## Learn More
+
+- **[Documentation](https://www.run.house/kubetorch/introduction)** - API Reference, concepts, and guides
+- **[Examples](https://www.run.house/kubetorch/examples)** - Real-world usage patterns and tutorials
+- **[Join our Slack](https://join.slack.com/t/kubetorch/shared_invite/zt-3g76q5i4j-uP60AdydxnAmjGVAQhtALA)** - Connect with the community and get support
+
+---
+
+[Apache 2.0 License](LICENSE)
+
+**🏃‍♀️ Built by [Runhouse](https://www.run.house) 🏠**
+

kubetorch-0.2.1/README.md

@@ -0,0 +1,66 @@
+# 📦Kubetorch🔥
+
+**A Python interface for running ML workloads on Kubernetes**
+
+Kubetorch enables you to run any Python code on Kubernetes at any scale by specifying required resources, distribution, and scaling directly in code. It provides caching and hot redeployment for 1-2 second iteration cycles, handles hardware faults and preemptions programmatically, and orchestrates complex, heterogeneous workloads with built-in observability and fault tolerance.
+
+## Hello World
+
+```python
+from kubetorch import fn
+
+def hello_world():
+    return "Hello from Kubetorch!"
+
+if __name__ == "__main__":
+    # Define your compute
+    compute = kt.Compute(cpus=".1")
+
+    # Send local function to freshly launched remote compute
+    remote_hello = kt.fn(hello_world).to(compute)
+
+    # Runs remotely on your Kubernetes cluster
+    result = hello_world()
+    print(result)  # "Hello from Kubetorch!"
+```
+
+## What Kubetorch Enables
+
+- **100x faster iteration** from 10+ minutes to 1-3 seconds for complex ML applications like RL and distributed training
+- **50%+ compute cost savings** through intelligent resource allocation, bin-packing, and dynamic scaling
+- **95% fewer production faults** with built-in fault handling with programmatic error recovery and resource adjustment
+
+## Installation
+
+### 1. Python Client
+
+```bash
+pip install "kubetorch[client]"
+```
+
+### 2. Kubernetes Deployment (Helm)
+
+```bash
+# Option 1: Install directly from OCI registry
+helm upgrade --install kubetorch oci://ghcr.io/run-house/charts/kubetorch \
+  --version 0.2.1 -n kubetorch --create-namespace
+
+# Option 2: Download chart locally first
+helm pull oci://ghcr.io/run-house/charts/kubetorch --version 0.2.1 --untar
+helm upgrade --install kubetorch ./kubetorch -n kubetorch --create-namespace
+```
+
+For detailed setup instructions, see our [Installation Guide](https://www.run.house/kubetorch/installation).
+
+
+## Learn More
+
+- **[Documentation](https://www.run.house/kubetorch/introduction)** - API Reference, concepts, and guides
+- **[Examples](https://www.run.house/kubetorch/examples)** - Real-world usage patterns and tutorials
+- **[Join our Slack](https://join.slack.com/t/kubetorch/shared_invite/zt-3g76q5i4j-uP60AdydxnAmjGVAQhtALA)** - Connect with the community and get support
+
+---
+
+[Apache 2.0 License](LICENSE)
+
+**🏃‍♀️ Built by [Runhouse](https://www.run.house) 🏠**

kubetorch-0.2.1/kubetorch/__init__.py

@@ -0,0 +1,60 @@
+from kubetorch.globals import config  # noqa: F401
+from kubetorch.resources.callables.cls.cls import Cls, cls  # noqa: F401
+from kubetorch.resources.callables.fn.fn import Fn, fn  # noqa: F401
+from kubetorch.resources.compute.app import App, app  # noqa: F401
+from kubetorch.resources.compute.compute import Compute  # noqa: F401
+from kubetorch.resources.compute.decorators import (  # noqa: F401
+    async_,
+    autoscale,
+    compute,
+    distribute,
+)
+from kubetorch.resources.compute.utils import (
+    ImagePullError,
+    KnativeServiceConflictError,
+    PodContainerError,
+    QueueUnschedulableError,
+    ResourceNotAvailableError,
+    RsyncError,
+    SecretNotFound,
+    ServiceHealthError,
+    ServiceTimeoutError,
+    VersionMismatchError,
+)  # noqa: F401
+from kubetorch.resources.images.image import Image  # noqa: F401
+from kubetorch.resources.secrets import Secret, secret  # noqa: F401
+from kubetorch.resources.volumes.volume import Volume  # noqa: F401  # noqa: F401
+from kubetorch.servers.http.utils import (  # noqa: F401
+    deep_breakpoint,
+    PodTerminatedError,
+    StartupError,
+    WorkerMembershipChanged,
+)
+
+from .resources import images
+
+# Alias to expose as kt.images
+images = images
+
+# Registry of all kubetorch exceptions for serialization/deserialization
+EXCEPTION_REGISTRY = {
+    "ImagePullError": ImagePullError,
+    "PodContainerError": PodContainerError,
+    "ResourceNotAvailableError": ResourceNotAvailableError,
+    "ServiceHealthError": ServiceHealthError,
+    "ServiceTimeoutError": ServiceTimeoutError,
+    "StartupError": StartupError,
+    "PodTerminatedError": PodTerminatedError,
+    "QueueUnschedulableError": QueueUnschedulableError,
+    "KnativeServiceConflictError": KnativeServiceConflictError,
+    "RsyncError": RsyncError,
+    "VersionMismatchError": VersionMismatchError,
+    "SecretNotFound": SecretNotFound,
+    "WorkerMembershipChanged": WorkerMembershipChanged,
+}
+
+# Make exceptions appear to be from the main package (e.g. kubetorch.ImagePullError)
+for exception in EXCEPTION_REGISTRY.values():
+    exception.__module__ = "kubetorch"
+
+__version__ = "0.2.1"