Skip to main content

The official Python client for the MOSTLY AI platform.

Project description

MOSTLY AI - GenAI for Tabular Data

A Python wrapper for the MOSTLY AI platform (https://app.mostly.ai/).

Intent Primitive
Train a Generative AI on tabular data g = mostly.train(data)
Empower your team with safe synthetic data mostly.share(g, email)
Generate any number of synthetic data records mostly.generate(g, size)
Prompt the generator for the data that you need mostly.generate(g, seed)
Live probe the generator on demand mostly.probe(g, size | seed)
Connect to any data source within your org mostly.connect(config)
List the available models mostly.models(model_type)
List the available computes mostly.computes()
Info about the current user mostly.me()
Info about the platform mostly.about()

Installation

pip install -U mostlyai

Basic Usage

from mostlyai import MostlyAI
mostly = MostlyAI(api_key='your_api_key')
g = mostly.train(data)      # train a generator on your data
mostly.share(g, email)      # share the generator with your team
sd = mostly.generate(g)     # use the generator to create a synthetic dataset
syn = sd.data()             # consume the synthetic data as pandas DataFrame(s)
mostly.probe(g, size=100)   # generate few samples on demand

Supported Methods

Connectors

c = mostly.connect(config)

c = mostly.connectors.create(config)
c = mostly.connectors.get(id)
it = mostly.connectors.list()
c = c.update(config)
ls = c.locations(prefix)
config = c.config()
c.open()
c.reload()
c.delete()

Generators

g = mostly.train(data, config, name, start=True, wait=True)

g = mostly.generators.create(config)
g = mostly.generators.get(id)
it = mostly.generators.list()
g = g.update(config)
config = g.config()
g.open()
g.reload()
g.delete()

# training
g.training.start()
g.training.progress()
g.training.cancel()
g.training.wait()

# import / export
g.export_to_file(file_path)
mostly.generators.import_from_file(file_path)

Synthetic Datasets

sd = mostly.generate(g, seed=seed)
sd = mostly.generate(g, size=size)
sd = mostly.generate(g, config=config)

sd = mostly.synthetic_datasets.create(g, config)
sd = mostly.synthetic_datasets.get(id)
it = mostly.synthetic_datasets.list()
config = sd.config()
sd.open()
sd.reload()
sd.delete()


sd.generation.start()
sd.generation.progress()
sd.generation.cancel()
sd.generation.wait()

sd.data()
sd.download(file, format)

Synthetic Probes

sp = mostly.probe(g, seed=seed)
sp = mostly.probe(g, size=size)
sp = mostly.probe(g, config=config)

Sharing

mostly.share(g | sd | c, email)
mostly.unshare(g | sd | c, email)

g.shares()
sd.shares()
c.shares()

Job Configuration Info

mostly.models(model_type)
mostly.computes()

User Info

mostly.me()
mostly.about()

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mostlyai-0.4.3.tar.gz (30.2 kB view hashes)

Uploaded Source

Built Distribution

mostlyai-0.4.3-py3-none-any.whl (33.3 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page