sameerank

4 ms·

1. 1.▲
   by sameerank 5y ago

   Hi all, I recently built https://pretrained.convect.ml . I’ve been interested in the potential for building web apps on top of pretrained models that have been gaining popularity in the machine learning community. One foundationa
2. 2.▲
   Show HN: Pretrained models (GPT-2, CLIP, etc.) APIs for building web apps(pretrained.convect.ml)

   1 points by sameerank 5y ago|1 comments
3. 3.▲
   Convect Integration in Zapier to Deploy ML Solutions for E-Commerce(blog.convect.ml)

   1 points by sameerank 5y ago|0 comments
4. 4.▲
   by sameerank 5y ago

   Yup, this was my thinking too. :) The variety of data problems you can work on with good old logistic regression (or the other models in scikit-learn) is really quite vast and seemed like a reasonable place to start.
5. 5.▲
   by sameerank 5y ago

   Thanks! Super glad to hear that this is addressing a real pain point. The infra aspect has been on my mind — how to easily onboard users who prefer to use their own infra. Love the idea of selling a CF template on AWS Marketplace. I’ll look
6. 6.▲
   by sameerank 5y ago

   Thanks for the feedback! That’s totally fair. My plan for now is to work with early users to understand what makes sense to them. At this early stage, any price I come up with would be almost a wild guess.
7. 7.▲
   by sameerank 5y ago

   At the moment, Convect only deploys scikit-learn models, but I am planning to support inference with PyTorch and other ML frameworks. Not planning to run on GPUs, as there are other products that already help with this, unless I see interes
8. 8.▲
   by sameerank 5y ago

   Thanks! app.convect.ml is the ML model admin page. The workflow looks somewhat like this — Within a notebook, you can: 1) train a model, 2) deploy a model, and 3) get a shareable serverless API endpoint. Outside the notebook, app.convect.ml
9. 9.▲
   by sameerank 5y ago

   You’re absolutely right that this product is heavy on the convenience side of the tradeoff. > Was it surprising to you too that you'd have users who wanted this location on the continuum? I’m still working on the “have users” piec
10. 10.▲
    by sameerank 5y ago

    Thanks! AWS Lambda supports Docker image sizes up to 10 GB (according to their docs), so on the back end, 3-5 GB could still work. Convect, in its current state, is still limited to small models (e.g. < 0.5MB). This is because the deploy
11. 11.▲
    by sameerank 5y ago

    Thank you for checking it out!
12. 12.▲
    by sameerank 5y ago

    Great question! Yes, Convect supports deploying pipelines and functions in which you can include pre-processing code. Here are two tutorials that walk through that: https://convect.readme.io/docs/wine-classifier-pipelin
13. 13.▲
    by sameerank 5y ago

    Thanks! FastAPI/ColabCode is a fair comparison. And I just tried it in Google Colab, and it does work! You can install Convect in a notebook cell with "!pip3 install convect"
14. 14.▲
    by sameerank 5y ago

    Hi HN, I’m Sameeran. I’m building Convect ( https://convect.ml ). Convect deploys machine learning (ML) models to instantly callable, serverless API endpoints. Using Convect, Jupyter notebook users can deploy trained models from t
15. 15.▲
    Show HN: Convect – Instant Serverless Deployment of ML Models(convect.ml)

    89 points by sameerank 5y ago|22 comments