DVC Reviews & Product Details

DVC allowed me to have an overview of my results, with plots and tracking the metadata. This improves and speeds up the research process, allowing reproducibility of the results and better team work. Review collected by and hosted on G2.com.

The tool needs some basic knowledge of coding (python), so can be a bit challenging to start. Also, some conflicts with previous versions may cause errors - which are rapidly solved by the DVC support team. Review collected by and hosted on G2.com.

I like that they follow the UNIX philosophy quite closely, they have an amazing comunity, always there to answer your questions. I also find amazing the open source culture they cultivate, and the active role they play in the ML community, sometimes even supporting economically Data Science events. Finally, I like how dvc plays together with git, making GitOps for ML a reality.

For me personally, I especially like the modularity of their products, that allow me to "hack them together" to my will. Review collected by and hosted on G2.com.

I encountered some occasional problems. Most of the time was actually my fault, didn't read the documentation carefully enough. Sometimes I had more serious issues that needed a resolution, but the dev team was very quick to fix them and provide me with solutions in short time. Review collected by and hosted on G2.com.

What I like most is that it fills a gap that no other tool out there does. It provides a way to version machine learning projects. Review collected by and hosted on G2.com.

The learning curve of mixing Git and DVC can be a bit hard. Review collected by and hosted on G2.com.

A great framework for structuring pipelines that are meant to be run locally and quickly: lightweight, local-friendly and GitOps philosophy, can get a lot of value with zero code instrumentation. Review collected by and hosted on G2.com.

Considerations of data processing at scale (parallel, distributed, remote, etc.) mostly absent Review collected by and hosted on G2.com.

1. It is the best open-source tool to manage data.

2. It allows us to version our data

3. Incorporates all git best practices

4. Supports all major cloud provider's storage solutions (eg azs s3, azure blob storage, gcp bucket).

5. Supports almost all data types. Review collected by and hosted on G2.com.

1. Default settings/config is good but to design the best/optimal config have to do some work.

2. Initial onboarding and initial learning is required.

3. Does not have a UI or plig n play kind of functionality. Review collected by and hosted on G2.com.

Git like operation via command line. Staged manageable ML pipeline. Review collected by and hosted on G2.com.

Command line was not fully integrated with Git yet. Unclear 1 to 1 correspondence across files and stored formats Review collected by and hosted on G2.com.

Model versioning for MLOps. its a great tool for model life cycle management Review collected by and hosted on G2.com.

less number of User Interfaces and low adoption as compared to MLFlow and other DevOps tools of hyper scalers Review collected by and hosted on G2.com.

I like that that model and data versioning can be done along with code versioning on git. Review collected by and hosted on G2.com.

There's not enough visibility. Tutorials on how to use more advanced functionalities could be useful. Review collected by and hosted on G2.com.

The way in which we can version data at different kinds of sources Review collected by and hosted on G2.com.

It's inability to deal with what exactly has changed, which although is a difficult and complicated task on its own. Review collected by and hosted on G2.com.

Allows to correctly track the data used to train ml/dl algorithms and to link a data version to a git commit Review collected by and hosted on G2.com.

It is note very intuitive to understand hashes Review collected by and hosted on G2.com.