Jules

I’m Jules, currently working as a Software Engineer @Dust.

I previously worked as a researcher at Microsoft, on anomaly detection in high dimensional time series. Then fell in love with all things Transformers and developed to NLP pipeline at Hypefactors. We dealt with ~1B inferences/day on multi-lingual problems like NER and reputation analysis. As a side gig I worked on a few OSS repository mostly around NLP and how to speed-up Transformer based models.

Scroll on and let’s connect if there’s anything you want to talk about!

Recent Posts

• DeeBERT: Teaching BERT When to Stop Thinking

  Why does BERT need twelve layers to classify “I love this movie” as positive?

• Early Exiting: The Under-Hyped Compression Method

  Why are we burning GPU hours to answer “2 + 2 = 4”?

• Instruction Fine-Tuning Fundamentals

  Instruction Fine-Tuning (IFT) is the secret sauce that transforms generic language models into obedient AI assistants.

• Scaling Machine Learning Experiments With neptune.ai and Kubernetes

  Link post: managing and scaling experiment tracking with Neptune and Kubernetes.

• Case Study: MLOps for NLP-powered Media Intelligence using Metaflow

  Link post: case study on building NLP media intelligence with Metaflow.