![Dev Saran S.](/assets/transparent-ad5be28fbcd25b7b08d2cebe1d957125437fb5407d75ee717965ad22c8808791.gif "Dev Saran S.")
DS

Dev Saran S.

Science Tutor 

Mid-Market (51-1000 emp.)

4/16/2026

"Streamlined Model Training with Python, Needs Faster Inference"

4/5

What do you like best about machine-learning in Python?

I like machine-learning in Python because of its ease of integration, making it simple to connect to models or create additional LLMs. I appreciate how easy it is to assess TensorFlow and the benefit of building on existing frameworks rather than reinventing them. This allows me to use existing functions without having to rewrite code, which makes the workflow smooth and efficient. The setup process is straightforward, with all guidelines clearly laid out in the readme, making it very easy to get started. Review collected by and hosted on G2.com.

What do you dislike about machine-learning in Python?

The inference process in Python for machine learning models is quite slow and could be improved. Handling inference results can be a bit inefficient, and improvements based on CPU architecture could help. It would also be helpful if the inference results could be more easily passed to applications or other tech software via APIs. Review collected by and hosted on G2.com.

What problems is machine-learning in Python solving and how is that benefiting you?

Machine-learning in Python lets me train models with up to 20 million parameters on my GPU, creating a smooth workflow without rewriting code. Review collected by and hosted on G2.com.

Show More

Validated ReviewerIncentivizedSource: G2 invite

See what 49 reviewers think of machine-learning in Python

4.6 out of 5 · Verified reviews from real users

[
Read all reviews
](https://www.g2.com/products/machine-learning-in-python/reviews)