OpenRunner SDK

Open-source, self-hosted ML experiment tracking — a drop-in replacement for Weights & Biases.

Install

pip install openrunner-sdk

Setup

export OPENRUNNER_API_KEY="or_your_key"
export OPENRUNNER_BASE_URL="https://your-server.com"

Or use the CLI:

openrunner login

Quick Start

import openrunner

# Start a run
openrunner.init(project="my-project", config={"lr": 0.001, "epochs": 10})

# Log metrics in your training loop
for epoch in range(10):
    loss = train(epoch)
    acc = evaluate()
    openrunner.log({"loss": loss, "accuracy": acc, "epoch": epoch})

# End the run
openrunner.finish()

API Reference

Core Functions

`openrunner.init()`

Initialize a new experiment run.

run = openrunner.init(
    project="my-project",          # Project name (auto-created if missing)
    name="experiment-1",           # Optional display name
    config={"lr": 0.001},          # Hyperparameters
    tags=["baseline", "v2"],       # Optional tags
    notes="Testing new arch",      # Optional notes
    group="sweep-1",               # Optional group name
    job_type="train",              # Optional job type
    resume=True,                   # Resume a previous run by ID
)

`openrunner.log()`

Log metrics. Non-blocking — never slows down training.

# Basic logging
openrunner.log({"loss": 0.5, "accuracy": 0.85})

# With explicit step
openrunner.log({"loss": 0.3}, step=100)

# Log images
openrunner.log({"predictions": openrunner.Image(img_array, caption="epoch 5")})

# Log tables
table = openrunner.Table(
    columns=["input", "predicted", "actual"],
    data=[["img_01", 7, 7], ["img_02", 3, 5]],
)
openrunner.log({"eval_results": table})

`openrunner.finish()`

End the current run. Flushes all buffered metrics.

openrunner.finish()
openrunner.finish(exit_code=0)    # With exit code
openrunner.finish(quiet=True)     # Suppress output

Config

Dict-like object with dot notation. Set at init(), accessible throughout the run.

openrunner.init(config={"optimizer": {"lr": 0.001, "weight_decay": 1e-5}})

# Access
print(openrunner.config["optimizer.lr"])    # 0.001 (flattened keys)
print(openrunner.config.optimizer.lr)       # 0.001 (dot notation)

# Update after init
openrunner.config.update({"batch_size": 64})
openrunner.config["new_param"] = "value"

Summary

Auto-updated with the last logged value for each key. Can also be set explicitly.

# Auto-populated from log()
openrunner.log({"loss": 0.5})
openrunner.log({"loss": 0.3})
print(openrunner.summary["loss"])  # 0.3 (last value)

# Explicit set
openrunner.summary["best_accuracy"] = 0.95
openrunner.summary["final_loss"] = 0.1

Artifacts

Version datasets, models, and checkpoints with content-hash deduplication.

# Log a model artifact
artifact = openrunner.Artifact(name="my-model", type="model")
artifact.add_file("model.pth")
artifact.add_file("config.json")
run.log_artifact(artifact)

# Use an artifact from a previous run
artifact = run.use_artifact("my-model:v2")
artifact.download("/path/to/dir")

Media Types

Images

import numpy as np

# From numpy array
img = openrunner.Image(np.random.rand(28, 28, 3), caption="sample")

# From PIL Image
from PIL import Image as PILImage
pil_img = PILImage.open("photo.png")
img = openrunner.Image(pil_img, caption="photo")

# From file path
img = openrunner.Image("output.png", caption="result")

openrunner.log({"examples": img})

Tables

table = openrunner.Table(
    columns=["epoch", "loss", "accuracy"],
    data=[
        [1, 0.9, 0.65],
        [2, 0.5, 0.82],
        [3, 0.3, 0.91],
    ],
)
openrunner.log({"metrics_table": table})

Run Properties

run = openrunner.init(project="test")

print(run.id)          # "a1b2c3d4" (8-char ID)
print(run.name)        # Display name
print(run.project)     # Project name
print(run.config)      # Config object
print(run.summary)     # Summary object

HTML

# Log raw HTML for rich reports, custom visualizations, or formatted output
openrunner.log({"report": openrunner.Html("<h1>Training Report</h1><p>Loss converged at epoch 42.</p>")})

Histograms

import numpy as np

weights = np.random.randn(10000)
openrunner.log({"weight_dist": openrunner.Histogram(weights, num_bins=50)})

Plotly Charts

import plotly.graph_objects as go

fig = go.Figure(data=go.Scatter(x=[1, 2, 3], y=[4, 5, 6]))
openrunner.log({"interactive_plot": openrunner.Plotly(fig)})

Point Clouds

import numpy as np

points = np.random.randn(1000, 3)
colors = np.random.randint(0, 255, (1000, 3), dtype=np.uint8)
openrunner.log({"lidar": openrunner.PointCloud3D(points, colors=colors)})

Bounding Boxes

img = openrunner.Image("photo.jpg")
boxes = [{"position": {"minX": 10, "minY": 20, "maxX": 100, "maxY": 150}, "class_id": 0}]
openrunner.log({"detections": openrunner.BoundingBoxes2D(img, boxes, class_labels={0: "cat"})})

Audio

import numpy as np

# From numpy array (mono, float32, -1 to 1)
audio = openrunner.Audio(np.random.randn(44100).astype(np.float32), sample_rate=44100)
openrunner.log({"audio_sample": audio})

Video

# From file path
openrunner.log({"demo": openrunner.Video("output.mp4", caption="training demo")})

Matplotlib Figures

import matplotlib.pyplot as plt

plt.figure()
plt.plot([1, 2, 3], [4, 5, 6])
openrunner.log({"chart": openrunner.MatplotlibFigure()})  # captures current figure
plt.close()

LLM Tracing

Trace LLM API calls for debugging and cost tracking.

import openrunner

openrunner.init(project="llm-app")

# Auto-trace OpenAI calls
openrunner.trace.patch_openai()

# Or manually trace any function
@openrunner.trace
def generate(prompt):
    return client.chat.completions.create(
        model="gpt-4", messages=[{"role": "user", "content": prompt}]
    )

Hyperparameter Sweeps

Run distributed hyperparameter searches.

import openrunner

sweep_config = {
    "method": "bayes",
    "metric": {"name": "val_loss", "goal": "minimize"},
    "parameters": {
        "lr": {"min": 1e-5, "max": 1e-2, "distribution": "log_uniform"},
        "epochs": {"values": [10, 20, 50]},
    },
}

sweep_id = openrunner.sweep(sweep_config, project="my-project")

def train():
    run = openrunner.init()
    lr = openrunner.config.lr
    # ... training loop ...
    openrunner.finish()

openrunner.agent(sweep_id, function=train, count=20)

Remote Launch

Submit training jobs to remote infrastructure.

import openrunner

job = openrunner.launch(
    project="my-project",
    config={"lr": 0.001, "epochs": 50},
    resource="gpu-a100",
)

job.wait()  # block until finished
print(job.state)  # "finished"

Model Registry

Version and alias models for production deployment.

# Log a model with aliases
artifact = openrunner.Artifact(name="classifier", type="model")
artifact.add_file("model.pt")
openrunner.link_artifact(artifact, aliases=["staging"])

# Use a model by alias
model_dir = openrunner.use_artifact("classifier:production")

Alerts

Send notifications when training reaches milestones or encounters issues.

openrunner.alert(title="Training complete", text="Final accuracy: 95.2%", level="INFO")
openrunner.alert(title="Loss spike detected", level="WARN")

Query API

Read-only access to runs, metrics, and projects for analysis and dashboards.

api = openrunner.Api()
runs = api.runs("my-project", filters={"state": "finished"})
for run in runs:
    print(f"{run.name}: {run.summary.get('accuracy')}")

Migrating from W&B

Change one import — everything else stays the same:

# Before
import wandb
wandb.init(project="my-project")
wandb.log({"loss": 0.5})
wandb.finish()

# After
import openrunner as wandb
wandb.init(project="my-project")
wandb.log({"loss": 0.5})
wandb.finish()

Framework Integrations

PyTorch

from openrunner.integration.pytorch import log_gradients

openrunner.init(project="pytorch-example")

for batch in dataloader:
    loss = model(batch)
    loss.backward()
    log_gradients(model)  # Logs gradient norms
    optimizer.step()

openrunner.finish()

HuggingFace Transformers

from openrunner.integration.huggingface import OpenRunnerCallback

openrunner.init(project="hf-example")

trainer = Trainer(
    model=model,
    args=training_args,
    callbacks=[OpenRunnerCallback()],
)
trainer.train()

openrunner.finish()

PyTorch Lightning

from openrunner.integration.lightning import OpenRunnerLogger

logger = OpenRunnerLogger(project="lightning-example")

trainer = pl.Trainer(logger=logger)
trainer.fit(model)

Keras

from openrunner.integration.keras import OpenRunnerCallback

openrunner.init(project="keras-example")

model.fit(x_train, y_train, callbacks=[OpenRunnerCallback()])

openrunner.finish()

XGBoost

from openrunner.integration.xgboost import OpenRunnerCallback

openrunner.init(project="xgboost-example")

bst = xgb.train(params, dtrain, callbacks=[OpenRunnerCallback()])

openrunner.finish()

scikit-learn

from openrunner.integration.sklearn import log_model

openrunner.init(project="sklearn-example")
model.fit(X_train, y_train)
log_model(model)  # Logs parameters and metrics
openrunner.finish()

FastAI

from openrunner.integration.fastai import OpenRunnerCallback

openrunner.init(project="fastai-example")

learn = cnn_learner(dls, resnet34, cbs=[OpenRunnerCallback()])
learn.fine_tune(5)

openrunner.finish()

LangChain

from openrunner.integration.langchain import OpenRunnerTracer

openrunner.init(project="langchain-example")

tracer = OpenRunnerTracer()
chain.invoke({"input": "Hello"}, config={"callbacks": [tracer]})

openrunner.finish()

Offline Mode

Train without connectivity, sync later:

export OPENRUNNER_MODE=offline
python train.py

# When back online
openrunner sync

Offline runs are stored as JSONL files (human-readable, crash-safe). Sync is additive and idempotent — interrupted syncs resume without data loss.

CLI

# Authenticate
openrunner login

# Sync offline runs
openrunner sync

# List projects and runs
openrunner ls

System Metrics

Automatically collected during training (enabled by default):

CPU utilization (%)
System memory usage (%)
GPU utilization (%) — requires pip install openrunner-sdk[gpu]
GPU memory usage (%)

Disable with:

export OPENRUNNER_SYSTEM_METRICS=false

Environment Variables

Variable	Description	Default
`OPENRUNNER_API_KEY`	API key for authentication	(required)
`OPENRUNNER_BASE_URL`	Server URL	`http://localhost:8000`
`OPENRUNNER_PROJECT`	Default project name	(none)
`OPENRUNNER_MODE`	`online` or `offline`	`online`
`OPENRUNNER_SYSTEM_METRICS`	Enable system metrics	`true`
`OPENRUNNER_OFFLINE_DIR`	Offline storage directory	`~/.openrunner/offline`

W&B env vars (WANDB_API_KEY, WANDB_BASE_URL) are also supported as fallback for migration.

Self-Hosting

OpenRunner is designed to be self-hosted. See the main repo for server setup with Docker Compose.

License

MIT

accelerate anthropic catboost dev diffusers fastai forced-alignment gpu gpu-all gpu-providers huggingface hydra jax keras langchain lightgbm lightning llamaindex mcp modal openai optuna patent pytorch runpod sb3 sklearn tensorflow trl tts ultralytics vastai voice-agent wer whisper xgboost

Metadata

Release files for openrunner-sdk 2.41.8

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for openrunner-sdk 2.41.8
File	Size	Uploaded
openrunner_sdk-2.41.8.tar.gz	773.7 kB	Sep 18, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for openrunner-sdk 2.41.8
File	Interpreter	ABI	Platform	Reset
openrunner_sdk-2.41.8-py3-none-any.whl 520.5 kB Sep 18, 2026	Python 3	none	any	Details

Total release size: 1.3 MB

Release files / openrunner_sdk-2.41.8.tar.gz

Download URL	openrunner_sdk-2.41.8.tar.gz
Size	773.7 kB
Tags	Source
SHA-256 checksum How to use checksums	`c3efc5d1a217cf2a7097bac0e6e31db8e0cff7f013b01e0158b5a08226e31033`
BLAKE2b-256 checksum How to use checksums	`e7f2526a4bd4c54610daab9bd490db46983ff0e6113b6a4d83b35297208b677c`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/7.0.0 CPython/3.12.14`

Release files / openrunner_sdk-2.41.8-py3-none-any.whl

Download URL	openrunner_sdk-2.41.8-py3-none-any.whl
Size	520.5 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`a3bba5eb0de782d20b58cdbfb86c464189c7a2bbdece281f4d93cecb9b1454cd`
BLAKE2b-256 checksum How to use checksums	`3d28c27e0cbfd7096128ec14290f863f501395e1bc914bcedfd7a1c7dfdcb019`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/7.0.0 CPython/3.12.14`