Caffe Reviews (16)

Reviews

Caffe Reviews (16)

4.0
16 reviews

What do users say?

Generated using AI from real user reviews
Users consistently praise the software for its ease of use and fast performance, making it accessible for those without a coding background. The built-in features and intuitive interface help streamline tasks in machine learning and deep learning projects. However, some users note a lack of flexibility when it comes to customizing certain elements.
Search reviews
Filter Reviews
Clear Results
G2 reviews are authentic and verified.
Verified User in Computer Software
UC
Verified User in Computer Software
Small-Business (50 or fewer emp.)
"Less popular than pytorch or tensorflow"
3/5
What do you like best about Caffe?

I use Caffe2 somethimes when I use pytorch because it merged with pytorch. Review collected by and hosted on G2.com.

What do you dislike about Caffe?

There is a lack of projects using caffe on github.

Custom layers are complicated to do. Review collected by and hosted on G2.com.

prasanna d.
PD
prasanna d.
Machine Learning Engineer
Small-Business (50 or fewer emp.)
"Model zoo in Caffe came to be very useful for my projects."
4/5
What do you like best about Caffe?

Model Zoo which has a large collection of deep nets which perform state-of-art in multiple domains. Review collected by and hosted on G2.com.

What do you dislike about Caffe?

Building a custom deep network is not very flexible as compared to Keras and Pytorch. Review collected by and hosted on G2.com.

Verified User in Computer Software
UC
Verified User in Computer Software
Mid-Market (51-1000 emp.)
"Caffe review after use for face recognition"
2.5/5
What do you like best about Caffe?

I love the fact that caffe so easily allows us to convert to tensorRT. Since tensorRT supports caffe models out of the box it was very convenient to create a model in caffe Review collected by and hosted on G2.com.

What do you dislike about Caffe?

Very less documentation that talks about the flow of coding a neural network Review collected by and hosted on G2.com.

Verified User in Research
UR
Verified User in Research
Small-Business (50 or fewer emp.)
"worked with it in different projects"
3.5/5
What do you like best about Caffe?

Its good to train basic models with it and can be easily handled. Review collected by and hosted on G2.com.

What do you dislike about Caffe?

Theres no specific thing I dislike about it. Review collected by and hosted on G2.com.

Verified User in Computer Software
GC
Verified User in Computer Software
Enterprise (> 1000 emp.)
"Fast and easy to configure"
4/5
What do you like best about Caffe?

Its fast, hides most complicated stuff under the carpet - which is a great strength and a weakness at same time. I think it makes it easy to get started with training new ml models without much tinkering. Abstracting out the graph makes it harder to make mistakes. This also ensures that the most optimal graph is compiled making it much faster than some other frameworks. Review collected by and hosted on G2.com.

What do you dislike about Caffe?

It makes it harder to make changes in base library. Its much harder to introduce new layers and know whats going on. That limits the knowledge about deep learning fundamentals. Also if there is a bug it can be hard to find since you don't create graph in any known language but rather a configuration file following a format similar to json. At time of writing this review no visualization tools were available either. Review collected by and hosted on G2.com.

Verified User in Hospital & Health Care
GH
Verified User in Hospital & Health Care
Small-Business (50 or fewer emp.)
"caffe"
3/5
What do you like best about Caffe?

This software is easy to use. Also helps with creativity and cost when building a menu and staying under costs. Wide library of items to choose from. Review collected by and hosted on G2.com.

What do you dislike about Caffe?

The ability to customize, some options are prefixed and it is a few extra steps when customizing menu/recipes Review collected by and hosted on G2.com.