mltrace
Pipeline analyzer
A tool to help analyze and debug machine learning pipelines by tracking the flow of data and components through the pipeline.
Coarse-grained lineage and tracing for machine learning pipelines.
468 stars
6 watching
29 forks
Language: Python
last commit: over 2 years ago
Linked from 1 awesome list
devopsmachine-learningmlopspipeline-managementtracing
ethicalml/awesome-production-machine-learningRelated projects:
| Repository | Description | Stars |
|---|---|---|
 minyus/pipelinex | A Python package to build and experiment with machine learning pipelines using Kedro, MLflow, and other tools | 226 |
 seldonio/tempo | An MLOps Python library that enables data scientists to deploy and orchestrate machine learning pipelines for production-ready inference. | 117 |
 aronchick/mlops-pipeline | Automates the end-to-end machine learning workflow from code commit to model deployment | 18 |
 oval-group/mlogger | A lightweight logger for machine learning experiments | 127 |
 johnsonc/lambdo | A workflow engine for unifying feature engineering and machine learning operations in data analysis pipelines | 1 |
 automaticanalysis/automaticanalysis | A software package that automates and streamlines neuroimaging data analysis workflows using a comprehensive pipeline system. | 72 |
 combust/mleap | Enables deployment of machine learning pipelines from Spark and Scikit-Learn to production | 1,506 |
 lightforever/mlcomp | A distributed framework for building and managing complex machine learning pipelines with a user-friendly interface. | 188 |
 scicloj/metamorph.ml | A tool for tuning complex data transformation pipelines in machine learning models | 18 |
 datacanvasio/hypergbm | Automated machine learning tool for tabular data pipelines | 343 |
 kevin-hanselman/dud | A lightweight tool for managing and versioning large data alongside source code in data pipelines | 184 |
 srvcodes/mt-dma | An implementation of a multi-task deep morphological analyzer with neural models and post-processing tools for natural language processing tasks. | 1 |
 ypares/porcupine | A tool that enables data manipulation and analysis pipelines to be flexible, reusable, and reproducible in different environments | 89 |
 prodmodel/prodmodel | A tool for managing data science pipelines by automating build, testing, and deployment processes while ensuring correctness and performance. | 58 |
 oxid15/cascade | A lightweight MLOps library for small teams and individuals to manage machine learning model development lifecycle | 22 |