Pinned Repositories
backprompt
Lazy KV cache data structure for language models
blog
A list of my open source contributions, and ML Qs and As
cappr
Completion After Prompt Probability. Make your LLM make a choice
dumpy
Dump of dumb Python code which was once useful
mcmc
Code and notebooks for simple Markov chain, Monte Carlo, and Markov chain Monte Carlo algorithms.
MEME
Re-implementation and some initialization analysis of Multiple EM for Motif Elicitation (MEME).
microarray-kaggle
Analyze a dataset with 72 observations and 7,129 features
ORIE4741Project
Model parts of the building permit approval process in San Francisco. Interesting parts I worked on: survivorship bias, multistage modeling, and data messiness.
pretrain-on-test
Finetuning Mistral 7B 2,000 times, and BERT and GPT-2 135,000 times, for science. To appear in the EMNLP workshop GenBench
kddubey's Repositories
kddubey/cappr
Completion After Prompt Probability. Make your LLM make a choice
kddubey/backprompt
Lazy KV cache data structure for language models
kddubey/mcmc
Code and notebooks for simple Markov chain, Monte Carlo, and Markov chain Monte Carlo algorithms.
kddubey/pretrain-on-test
Finetuning Mistral 7B 2,000 times, and BERT and GPT-2 135,000 times, for science. To appear in the EMNLP workshop GenBench
kddubey/blog
A list of my open source contributions, and ML Qs and As
kddubey/dumpy
Dump of dumb Python code which was once useful
kddubey/MEME
Re-implementation and some initialization analysis of Multiple EM for Motif Elicitation (MEME).
kddubey/microarray-kaggle
Analyze a dataset with 72 observations and 7,129 features
kddubey/ORIE4741Project
Model parts of the building permit approval process in San Francisco. Interesting parts I worked on: survivorship bias, multistage modeling, and data messiness.
kddubey/formulae
Formulas for mixed-effects models in Python
kddubey/test_set_contamination
with dtype
kddubey/typed-argument-parser
Typed argument parser for Python
kddubey/vecgrad
Vectorized micrograd