codeinassembly

https://github.com/jinensetpal/triton_cdist

For my research, I needed to use `torch.cdist` https://docs.pytorch.org/docs/stable/generated/torch.cdist.h... with p != 2.Turns out torch's implementation for cdist with p != 2 is prohibitively slow. I came up with an alternative implementation using triton https://triton-lang.org/, which supports backprop and so can be used for training too.Should be pretty plug-and-play, if you're using...

Hey HN, I'm Jinen. I’m a PhD student working on DL interpretability & Optimization Theory.Before beginning my PhD, I worked at DagsHub fine tuning vision models for domain specific deployments. I wanted to use ML models to help label the data, based on Label Studio's ML Backends. The goal was to use a model registered and tracked on MLflow.I found it tedious and involved a lot of...

https://github.com/jinensetpal/boilerbot

The open source reimplementation is not as quick or robust as the closed source implementation - hopefully I'll be able to find time to make upgrades soon.

https://dagshub.com/blog/large-language-models-explained/

https://dagshub.com/blog/unraveling-the-deep-learning-reproducibility-crisis/

https://dagshub.com/blog/unraveling-the-deep-learning-reproducibility-crisis/

TL:DR; need help making heatmaps! Repository: https://dagshub.com/nirbarazida/Pneumonia-Classification Colab Notebook: https://colab.research.google.com/drive/1U3F5ETJeisBnlmamR4EqigS7shIbL2L1Hey everyone -I've been working to reproduce [CheXNet](https://arxiv.org/pdf/1711.05225.pdf) - a fantastic paper describing research on a model capable of radiologist-grade pathology classification!CheXNet...