Log your project journey from data, to models, to documentation.
Easily log everything with just a few lines of code.
Track
Layer is your Project Development Diary. It automatically versions and stores your data, models and related metadata so that you never lose track of your project’s story. It’s easy — just add Layer’s powerful yet simple-to-use decorators like @dataset and @model, plus layer.log(), to your code.
from layer.decorators import model
@model("california_housing")
def train():
model = XGBClassifier()
model.fit(X_train, y_train)
layer.log({...})
return model
train()
Remember
Layer is your second brain. You can record what you have learned over the course of project journey without manual results tracking and saving, preventing duplicate works and enabling performance monitoring. Whether parameters, charts, metrics, or plots, Layer has you covered.

Reproduce
Layer removes the pain of unrepeatable experiments and differential production performance. Layer automatically versions your entities so that you can precisely repeat experiments and track their history. Load your versioned model with layer.get_model("my_model:2.1").
def generate_predictions():
model = layer.get_model("my_linear_model:2.1")
test_data = layer.get_dataset("my_test_data:4.1")
Y_pred = model.predict(test_data)
return Y_pred
Document
Layer is build for high collaboration. Its dynamic READMEs (documentation) allow your team to quickly access ML entities, stay in sync with business objectives, handover seamlessly, and effortlessly work on the same projects.
The definitive ML project development diary
Show off your work with demo-able & easy-to-present dynamic content for your ML projects.
See community examples→Built with data teams, for data teams
Layer is the single place for all datasets, models and documents developed by your team.
Request a demo→