About me
I am a machine learning engineer with lots of interest in open source. I co-founded project KerasNLP with my colleague at Google, and I am the second author (first author is Francois Chollet) of Keras 3, also known as multi-backend Keras.
Currently I am working for Databricks, and working on the development and maintainance of two open-source projects:
- - DSPy, a framework for algorithmically optimizing LM prompts and weights. DSPy is still an experimental project, and we are working on lots of exciting new features, stay tuned!
- - MLflow, a popular open-source machine learning platform. Specially in charge of the MLflow support for deep learning.
I am specialized in deep learning, experienced in ML data processing, model authoring, training, optimization and deployment. I am also a just-fine backend & infrastructure engineer, and have experience in building scalable and reliable web services. Lastly, I am terrible at frontend, trust me bro, I have tried, but it's just not my thing.
I am open to collaborations in the areas of LLM and GenAI. Feel free to reach out to me if you are interested in working together, or would like to have a chat. I also give tech talks and workshops on deep learning, with a focus on training/finetuning and MLops with MLflow, and open to invitations.
What I'm doing
-
DSPy
DSPy is a framework for algorithmically building compound AI system, and automatically optimizing LM prompts and weights, especially when LMs are used one or more times within a pipeline.
-
MLflow
MLflow is a platform to streamline machine learning development, including tracking experiments, packaging code into reproducible runs, and sharing and deploying models.
-
Keras 3
Keras 3 is a multi-backend deep learning framework, with support for JAX, TensorFlow, and PyTorch. Effortlessly build and train models for computer vision, natural language processing, audio processing, timeseries forecasting, recommender systems, etc.
-
KerasNLP
KerasNLP is a natural language processing library that works natively with TensorFlow, JAX, or PyTorch. KerasNLP provides a repository of pre-trained models and a collection of lower-level building blocks for language modeling. Built on Keras 3, models can be trained and serialized in any framework and re-used in another without costly migrations.