Library/airioForked

google/airio

airio

AirIO is a library for building scalable data preprocessing pipelines for machine learning, particularly focused on text-to-text tasks.

Builder

Google

Google

google • big-tech

Stars

24

Using upstream star count

Forks

12

Using upstream fork count

Open Issues

0

Activity Score

0/100

0 commits in 30d

Created

Jul 13, 2023

Project creation date

README Summary

AirIO is a library for building scalable data preprocessing pipelines for machine learning, particularly focused on text-to-text tasks. It provides a flexible framework for data loading, preprocessing, and batching with support for various data sources and transformations. The library is designed to work seamlessly with JAX and other ML frameworks for efficient model training.

AI Dev Skills

Unmapped

Data Pipeline EngineeringMachine Learning InfrastructureData PreprocessingML Training Pipelines

Tags

Data Pipeline EngineeringMachine Learning InfrastructureData PreprocessingML Training PipelinesML Infrastructure OptimizationCloudModel Input Pipeline OptimizationTextML Data Pipeline ManagementTraining Data PreprocessingSelf-hostedPython

Taxonomy

Recent Activity

Updated 1 months ago

7 Days

0

30 Days

0

90 Days

0

Quality

research
Quality
low
Maturity
research

Categories

Model TrainingPrimaryML Platform & InfrastructureOther AI / MLMLOps & Infrastructure

PM Skills

Scale & ReliabilityDeveloper Platform

Languages

Python100.0%

Timeline

Project created
Jul 13, 2023
Forked
Mar 13, 2026
Your last push
1 months ago
Upstream last push
1 months ago
Tracked since
Mar 9, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…