feast 0.66.0


pip install feast

  Latest version

Released: Aug 21, 2026
Navigation
Project Links
Meta
Author: Feast
Requires Python: >=3.10.0
Classifiers

Programming Language
  • Python
  • Python :: 3
  • Python :: 3.10


PyPI - Downloads GitHub contributors unit-tests integration-tests-and-build linter Docs Latest Python API License GitHub Release

Join us on Slack!

๐Ÿ‘‹๐Ÿ‘‹๐Ÿ‘‹ Come say hi on Slack!

Check out our DeepWiki!

Overview

feast-dev%2Ffeast | Trendshift

Feast (Feature Store) is an open source feature store for machine learning. Feast is the fastest path to manage existing infrastructure to productionize analytic data for model training and online inference.

Feast allows ML platform teams to:

  • Make features consistently available for training and serving by managing an offline store (to process historical data for scale-out batch scoring or model training), a low-latency online store (to power real-time prediction), and a battle-tested feature server (to serve pre-computed features online).
  • Avoid data leakage by generating point-in-time correct feature sets so data scientists can focus on feature engineering rather than debugging error-prone dataset joining logic. This ensure that future feature values do not leak to models during training.
  • Decouple ML from data infrastructure by providing a single data access layer that abstracts feature storage from feature retrieval, ensuring models remain portable as you move from training models to serving models, from batch models to realtime models, and from one data infra system to another.

Please see our documentation for more information about the project.

๐Ÿ“ Architecture

The above architecture is the minimal Feast deployment. Want to run the full Feast on Snowflake/GCP/AWS? Click here.

๐Ÿฃ Getting Started

1. Install Feast

pip install feast

2. Create a feature repository

feast init my_feature_repo
cd my_feature_repo/feature_repo

3. Register your feature definitions and set up your feature store

feast apply

4. Explore your data in the web UI (experimental)

Web UI

feast ui

5. Build a training dataset

from feast import FeatureStore
import pandas as pd
from datetime import datetime

entity_df = pd.DataFrame.from_dict({
    "driver_id": [1001, 1002, 1003, 1004],
    "event_timestamp": [
        datetime(2021, 4, 12, 10, 59, 42),
        datetime(2021, 4, 12, 8,  12, 10),
        datetime(2021, 4, 12, 16, 40, 26),
        datetime(2021, 4, 12, 15, 1 , 12)
    ]
})

store = FeatureStore(repo_path=".")

training_df = store.get_historical_features(
    entity_df=entity_df,
    features = [
        'driver_hourly_stats:conv_rate',
        'driver_hourly_stats:acc_rate',
        'driver_hourly_stats:avg_daily_trips'
    ],
).to_df()

print(training_df.head())

# Train model
# model = ml.fit(training_df)

            event_timestamp  driver_id  conv_rate  acc_rate  avg_daily_trips
0 2021-04-12 08:12:10+00:00       1002   0.713465  0.597095              531
1 2021-04-12 10:59:42+00:00       1001   0.072752  0.044344               11
2 2021-04-12 15:01:12+00:00       1004   0.658182  0.079150              220
3 2021-04-12 16:40:26+00:00       1003   0.162092  0.309035              959

6. Load feature values into your online store

Option 1: Incremental materialization (recommended)

CURRENT_TIME=$(date -u +"%Y-%m-%dT%H:%M:%S")
feast materialize-incremental $CURRENT_TIME

Option 2: Full materialization with timestamps

CURRENT_TIME=$(date -u +"%Y-%m-%dT%H:%M:%S")
feast materialize 2021-04-12T00:00:00 $CURRENT_TIME

Option 3: Simple materialization without timestamps

feast materialize --disable-event-timestamp

The --disable-event-timestamp flag allows you to materialize all available feature data using the current datetime as the event timestamp, without needing to specify start and end timestamps. This is useful when your source data lacks proper event timestamp columns.

Materializing feature view driver_hourly_stats from 2021-04-14 to 2021-04-15 done!

7. Read online features at low latency

from pprint import pprint
from feast import FeatureStore

store = FeatureStore(repo_path=".")

feature_vector = store.get_online_features(
    features=[
        'driver_hourly_stats:conv_rate',
        'driver_hourly_stats:acc_rate',
        'driver_hourly_stats:avg_daily_trips'
    ],
    entity_rows=[{"driver_id": 1001}]
).to_dict()

pprint(feature_vector)

# Make prediction
# model.predict(feature_vector)

{
    "driver_id": [1001],
    "driver_hourly_stats__conv_rate": [0.49274],
    "driver_hourly_stats__acc_rate": [0.92743],
    "driver_hourly_stats__avg_daily_trips": [72]
}

๐Ÿ“ฆ Functionality and Roadmap

The list below contains the functionality that contributors are planning to develop for Feast.

๐ŸŽ“ Important Resources

Please refer to the official documentation at Documentation

๐Ÿ‘‹ Contributing

Feast is a community project and is still under active development. Please have a look at our contributing and development guides if you want to contribute to the project:

๐ŸŒŸ GitHub Star History

Star History Chart

โœจ Contributors

Thanks goes to these incredible people:

0.67.0.dev112 Sep 28, 2026
0.67.0.dev111 Sep 27, 2026
0.67.0.dev110 Sep 26, 2026
0.67.0.dev109 Sep 25, 2026
0.67.0.dev108 Sep 24, 2026
0.67.0.dev107 Sep 23, 2026
0.67.0.dev106 Sep 22, 2026
0.67.0.dev105 Sep 21, 2026
0.67.0.dev104 Sep 20, 2026
0.67.0.dev103 Sep 19, 2026
0.67.0.dev102 Sep 18, 2026
0.67.0.dev101 Sep 17, 2026
0.67.0.dev100 Sep 16, 2026
0.67.0.dev99 Sep 15, 2026
0.67.0.dev98 Sep 14, 2026
0.67.0.dev97 Sep 13, 2026
0.67.0.dev96 Sep 12, 2026
0.67.0.dev95 Sep 11, 2026
0.67.0.dev94 Sep 10, 2026
0.67.0.dev93 Sep 09, 2026
0.67.0.dev92 Sep 08, 2026
0.67.0.dev91 Sep 07, 2026
0.67.0.dev90 Sep 06, 2026
0.67.0.dev89 Sep 05, 2026
0.67.0.dev88 Sep 04, 2026
0.67.0.dev87 Sep 03, 2026
0.67.0.dev86 Sep 02, 2026
0.67.0.dev85 Sep 01, 2026
0.67.0.dev84 Aug 31, 2026
0.67.0.dev83 Aug 30, 2026
0.67.0.dev82 Aug 29, 2026
0.67.0.dev81 Aug 28, 2026
0.67.0.dev80 Aug 27, 2026
0.67.0.dev79 Aug 26, 2026
0.67.0.dev78 Aug 25, 2026
0.67.0.dev77 Aug 24, 2026
0.67.0.dev76 Aug 23, 2026
0.67.0.dev75 Aug 22, 2026
0.67.0.dev74 Aug 21, 2026
0.66.0 Aug 21, 2026
0.66.0.dev73 Aug 20, 2026
0.66.0.dev72 Aug 19, 2026
0.66.0.dev71 Aug 18, 2026
0.66.0.dev70 Aug 17, 2026
0.66.0.dev69 Aug 16, 2026
0.66.0.dev68 Aug 15, 2026
0.66.0.dev67 Aug 14, 2026
0.66.0.dev66 Aug 13, 2026
0.66.0.dev65 Aug 12, 2026
0.66.0.dev64 Aug 11, 2026
0.66.0.dev63 Aug 10, 2026
0.66.0.dev62 Aug 09, 2026
0.66.0.dev61 Aug 08, 2026
0.66.0.dev60 Aug 07, 2026
0.66.0.dev59 Aug 06, 2026
0.66.0.dev58 Aug 05, 2026
0.66.0.dev57 Aug 04, 2026
0.66.0.dev56 Aug 03, 2026
0.66.0.dev55 Aug 02, 2026
0.66.0.dev54 Aug 01, 2026
0.66.0.dev53 Jul 31, 2026
0.66.0.dev52 Jul 30, 2026
0.66.0.dev51 Jul 29, 2026
0.66.0.dev50 Jul 28, 2026
0.66.0.dev49 Jul 27, 2026
0.66.0.dev48 Jul 26, 2026
0.66.0.dev47 Jul 25, 2026
0.66.0.dev46 Jul 24, 2026
0.66.0.dev45 Jul 23, 2026
0.65.1.dev44 Jul 22, 2026
0.65.1.dev43 Jul 21, 2026
0.65.0 Jul 20, 2026
0.65.0.dev42 Jul 20, 2026
0.65.0.dev41 Jul 19, 2026
0.65.0.dev40 Jul 18, 2026
0.65.0.dev39 Jul 17, 2026
0.65.0.dev38 Jul 16, 2026
0.65.0.dev37 Jul 15, 2026
0.65.0.dev36 Jul 14, 2026
0.65.0.dev35 Jul 13, 2026
0.65.0.dev34 Jul 12, 2026
0.65.0.dev33 Jul 11, 2026
0.65.0.dev32 Jul 10, 2026
0.65.0.dev31 Jul 09, 2026
0.65.0.dev30 Jul 08, 2026
0.65.0.dev29 Jul 07, 2026
0.65.0.dev28 Jul 06, 2026
0.65.0.dev27 Jul 05, 2026
0.65.0.dev26 Jul 04, 2026
0.65.0.dev25 Jul 03, 2026
0.65.0.dev17 Jun 25, 2026
0.65.0.dev16 Jun 24, 2026
0.65.0.dev15 Jun 23, 2026
0.65.0.dev14 Jun 22, 2026
0.65.0.dev13 Jun 21, 2026
0.64.1.dev6531 Jun 16, 2026
0.64.1.dev6529 Jun 16, 2026
0.64.1.dev12 Jun 20, 2026
0.64.1.dev11 Jun 19, 2026
0.64.1.dev10 Jun 18, 2026
0.64.1.dev9 Jun 17, 2026
0.64.0 Jun 13, 2026
0.63.0 May 04, 2026
0.62.0 Apr 08, 2026
0.61.0 Mar 10, 2026
0.60.0 Feb 17, 2026
0.59.0 Jan 16, 2026
0.58.0 Dec 16, 2025
0.57.0 Nov 13, 2025
0.56.0 Oct 27, 2025
0.55.0 Oct 14, 2025
0.54.1 Dec 04, 2025
0.54.0 Sep 30, 2025
0.53.0 Aug 30, 2025
0.52.0 Aug 14, 2025
0.51.0 Jul 21, 2025
0.50.0 Jul 01, 2025
0.49.0 Apr 29, 2025
0.48.1 Apr 08, 2025
0.48.0 Apr 07, 2025
0.47.0 Mar 11, 2025
0.46.0 Feb 17, 2025
0.45.0 Feb 04, 2025
0.44.0 Feb 04, 2025
0.43.0 Jan 23, 2025
0.42.0 Dec 06, 2024
0.41.3 Oct 26, 2024
0.40.1 Aug 09, 2024
0.40.0 Jul 31, 2024
0.39.1 Jul 04, 2024
0.39.0 Jun 18, 2024
0.38.0 May 23, 2024
0.37.1 Apr 17, 2024
0.36.0 Feb 17, 2024
0.35.0 Jan 13, 2024
0.34.1 Sep 07, 2023
0.34.0 Sep 07, 2023
0.33.1 Aug 14, 2023
0.31.1 Apr 24, 2023
0.31.0 Apr 21, 2023
0.30.2 Mar 25, 2023
0.29.0 Jan 31, 2023
0.28.0 Jan 03, 2023
0.27.1 Dec 15, 2022
0.27.0 Dec 05, 2022
0.26.0 Oct 06, 2022
0.25.2 Oct 06, 2022
0.25.1 Sep 30, 2022
0.25.0 Sep 20, 2022
0.24.1 Sep 08, 2022
0.24.0 Aug 25, 2022
0.23.2 Aug 15, 2022
0.23.1 Aug 10, 2022
0.23.0 Aug 02, 2022
0.22.4 Aug 15, 2022
0.22.3 Aug 10, 2022
0.22.2 Jul 29, 2022
0.22.1 Jul 19, 2022
0.22.0 Jun 28, 2022
0.21.3 Jun 13, 2022
0.21.2 May 17, 2022
0.21.1 May 16, 2022
0.21.0 May 13, 2022
0.20.2 Apr 28, 2022
0.20.1 Apr 20, 2022
0.20.0 Apr 14, 2022
0.19.4 Apr 06, 2022
0.19.3 Mar 09, 2022
0.19.2 Mar 06, 2022
0.19.1 Mar 05, 2022
0.19.0 Mar 05, 2022
0.18.1 Feb 15, 2022
0.18.0 Feb 05, 2022
0.17.0 Jan 04, 2022
0.16.1 Dec 11, 2021
0.16.0 Dec 08, 2021
0.15.1 Nov 13, 2021
0.15.0 Nov 08, 2021
0.14.1 Oct 28, 2021
0.14.0 Oct 09, 2021
0.13.0 Sep 22, 2021
0.12.1 Aug 20, 2021
0.12.0 Aug 05, 2021
0.11.0 Jun 24, 2021
0.10.8 Jun 17, 2021
0.10.7 Jun 07, 2021
0.10.6 May 27, 2021
0.10.5 May 19, 2021
0.10.4 May 12, 2021
0.10.3 Apr 28, 2021
0.10.2 Apr 21, 2021
0.10.1 Apr 21, 2021
0.10.1rc3 Apr 15, 2021
0.10.1rc1 Apr 15, 2021
0.10.0 Apr 15, 2021
0.9.9 Jul 14, 2022
0.9.8 Jun 09, 2022
0.9.7 Apr 28, 2022
0.9.6 Apr 28, 2022
0.9.5.2 Apr 15, 2021
0.9.5.1 Apr 15, 2021
0.9.5 Apr 09, 2021
0.9.4 Mar 30, 2021
0.9.3 Feb 16, 2021
0.9.2 Feb 01, 2021
0.9.1 Jan 29, 2021
0.9.0 Jan 28, 2021
0.9.0rc2 Jan 26, 2021
0.9.0rc1 Jan 15, 2021
0.8.4 Jan 28, 2021
0.8.3 Jan 28, 2021
0.8.2 Dec 01, 2020
0.8.1 Nov 24, 2020
0.8.0 Nov 10, 2020
0.8.0rc3 Nov 06, 2020
0.7.2 Dec 14, 2020
0.7.1 Oct 07, 2020
0.7.0 Sep 15, 2020
0.7.0rc3 Sep 11, 2020
0.7.0rc2 Sep 11, 2020
0.7.0rc1 Sep 10, 2020
0.6.2 Aug 02, 2020
0.6.1 Jul 17, 2020
0.6.0 Jul 07, 2020
0.5.1 Jun 08, 2020
0.5.0.post0 May 21, 2020
0.5.0 May 19, 2020
0.5.0rc0 May 21, 2020
0.4.7 Mar 17, 2020
0.4.6 Feb 26, 2020
0.4.5 Feb 14, 2020
0.4.4 Jan 27, 2020
0.4.3 Jan 08, 2020
0.4.2 Jan 06, 2020
0.4.1.post0 Dec 30, 2019
0.4.1 Dec 30, 2019
0.4.0 Dec 29, 2019
0.3.7 May 01, 2020
0.3.6 Jan 03, 2020
0.3.5 Dec 26, 2019
0.3.4 Dec 21, 2019
0.3.3.post0 Dec 18, 2019
0.3.3 Dec 18, 2019
0.3.2 Nov 27, 2019
0.3.0 Nov 24, 2019
0.1.2.post1 Aug 29, 2019
0.1.2.post0 Aug 29, 2019
0.1.2 Jun 21, 2019
0.1.0.post0 Apr 10, 2019
0.1.0 Apr 09, 2019

Wheel compatibility matrix

Platform Python 3
any

Files in release

Extras:
Dependencies:
attrs
click (<9.0.0,>=7.0.0)
colorama (<1,>=0.3.9)
dill (~=0.3.0)
protobuf (>=4.24.0)
Jinja2 (<4,>=2)
mmh3
numpy (<3,>=2.0.0)
pandas (<3,>=1.4.3)
pyarrow (>=16.1.0)
pydantic (>=2.10.6)
pygments (<3,>=2.12.0)
PyYAML (<7,>=5.4.0)
requests
SQLAlchemy[mypy] (>1)
tabulate (<1,>=0.8.0)
tenacity (<9,>=7)
toml (<1,>=0.10.0)
tqdm (<5,>=4)
typeguard (>=4.0.0)
fastapi (>=0.68.0)
starlette (>=1.0.1)
uvicorn[standard] (>=0.30.6)
uvicorn-worker
gunicorn
dask[dataframe] (>=2024.2.1)
prometheus_client (<0.25.0,>=0.20.0)
psutil
bigtree (>=0.19.2)
pyjwt (>=2.13.0)