sangmichaelxie

Sang Michael Xie's repositories

doremi

Pytorch implementation of DoReMi, a method for optimizing the data mixture weights in language modeling datasets

Language:HTMLMIT276 5 28

cs324_p2

Project 2 (Building Large Language Models) for Stanford CS324: Understanding and Developing Large Language Models (Winter 2022)

Language:PythonMIT101 10

giraffe

Learning from Expert Data, Approximate IRL, and TD-Leaf for Deep Reinforcement Learning Chess, built on the recent Giraffe engine

Language:C++GPL-3.032 60

pretraining_analysis

Code for the NeurIPS 2021 paper "Why Do Pretrained Language Models Help in Downstream Tasks? An Analysis of Head and Prompt Tuning"

Language:Python8 20

cycada_release

Code to accompany ICML 2018 paper

Language:PythonBSD-2-Clause1 10

GestureLearner

CS229: Machine Learning Project

Language:Java1 20

keras-gp

Keras + Gaussian Processes: Learning scalable deep and recurrent kernels.

Language:PythonMIT100

A C++ mutex class that can give threads passed "with priority" first shot at the scheduler. Inspired by a test question to be as lean of an implementation as possible. Supposedly possible without the atomic<int> or any additional mutexes, cv's, etc. I would like to know if that is possible! Test harness is provided.

Language:C++1 10