Hacker News: dharmeshkakadia

New comment by dharmeshkakadia in "Ask HN: What Are You Working On? (March 2026)"

dharmeshkakadia — Mon, 09 Mar 2026 19:05:26 +0000

Building Mixtrain - Platform to build task specific models. Focus on multimodallity usecase - video, image, robotics. Covers the entire post training life cycle - data management & curation, training, Eval & rollout.

Models are the new software. And just like software, three general-purpose ones won't be enough. Why specialized models are inevitable https://mixtrain.ai/blog/special-models

Here's how Mixtrain can help:

  - Multimodal dataset management: version, query, inspect, and curate image/video/3D datasets
  - Workflows & models: train and run your models on serverless GPUs. Run experiments rapidly and ship to production. Access 100s of external models through the same API.
  - Live eval: create instant evals from your datasets with side-by-side comparison of anything — images, time-synced video, 3D/4D visualizations, masks, and more. Here's an example video eval https://app.mixtrain.ai/s/eVRwOcb7KhUZOb9xbFFgfHIuF0jyJUaBT6TKNg19OfU. Evals stay current as your datasets evolve.

You can explore more at https://mixtrain.ai/docs

New comment by dharmeshkakadia in "Keras Core: Keras for TensorFlow, Jax, and PyTorch"

dharmeshkakadia — Mon, 24 Jul 2023 22:34:38 +0000

Thanks!

New comment by dharmeshkakadia in "Keras Core: Keras for TensorFlow, Jax, and PyTorch"

dharmeshkakadia — Tue, 11 Jul 2023 18:06:51 +0000

Supporting multiple backends (especially Jax) is nice! Makes experimenting/migrating between them so much more approachable. Any timeline on when can we expect support for distributed Jax training? The doc currently seems to indicate only TF is supported for distributed training.

Crypto Removes Friction

dharmeshkakadia — Mon, 07 Mar 2022 04:01:44 +0000

Article URL: https://dharmeshkakadia.com/blog/crypto-removes-friction/

Comments URL: https://news.ycombinator.com/item?id=30584101

Points: 2

# Comments: 0

Ask HN: How do you do code review for secrets?

dharmeshkakadia — Wed, 06 Jul 2016 03:51:44 +0000

We manage our deployments with tools reading encrypted files - secrets. The encrypted files themselves are checked into the source control and gets "deployed". The key problem is how do we do code reviews for these files?

We would like to avoid "attaching the plain text" to the review as well because that defeats the purpose.

So, I am wondering how do people generally do code reviews for secrets.

Comments URL: https://news.ycombinator.com/item?id=12041093

Points: 2

# Comments: 0