Katherine Crowson

k-diffusion. Karras et al. (2022) diffusion models for PyTorch

v-diffusion-pytorch. v objective diffusion inference code for PyTorch.

style-transfer-pytorch. Neural style transfer in PyTorch.

v-diffusion-jax. v objective diffusion inference code for JAX.

simulacra-aesthetic-models. Python

guided-diffusion. Python

style_transfer. Data-parallel image stylization using Caffe.

deep_dream. A parallel implementation of the Deep Dream image processing algorithm which is able to process arbitrarily large images.

consistency-models. A JAX implementation of the continuous time formulation of Consistency Models

clip-guided-diffusion. CLIP Guided Diffusion

cloob-training. CLOOB training (JAX) and inference (JAX and PyTorch)

LDLM. Latent Diffusion Language Models

vgg_loss. A VGG-based perceptual loss function for PyTorch.

mdmm. The Modified Differential Multiplier Method (MDMM) for PyTorch

latent-diffusion. High-Resolution Image Synthesis with Latent Diffusion Models

esgd. ESGD-M is a stochastic non-convex second order optimizer, suitable for training deep learning models, for PyTorch.

deep-image-prior. Image restoration with neural networks but without learning.

jax-wavelets. The 2D discrete wavelet transform for JAX

mdmm-jax. Gradient-based constrained optimization for JAX

dice-mc. DiCE: The Infinitely Differentiable Monte-Carlo Estimator

torch-dist-utils. Utilities for PyTorch distributed

cond_transformer_2. A CLIP conditioned Decision Transformer.

glide-text2im. GLIDE: a diffusion-based text-conditional image synthesis model

tv-denoise. Total variation denoising for images.

pharmacokinetics. A Flask web application to calculate and plot drug concentration over time.

shared_ndarray. A pickleable wrapper for sharing NumPy ndarrays between processes using POSIX shared memory.

pytorch-caffe-models. The original weights of some Caffe models, ported to PyTorch.

jetformer. Implementation of Jetformer.

rope-flax. Rotary Position Embedding for Flax

aiohttp_index. aiohttp.web middleware to serve index files (e.g. index.html) when static directories are requested.

pyparsing-highlighting. Syntax highlighting for prompt_toolkit and HTML with pyparsing.

stable-diffusion. Jupyter Notebook

simple-diffusion-model. Pedagogical codebase for a simplified score-based generative model design, with training loop

ucs. Implements the CAM02-UCS (Luo et al. (2006)) forward transform.

stylegan2-ada-pytorch. StyleGAN2-ADA - Official PyTorch implementation

kat-wip. works in progress

torch-fidelity. High-fidelity performance metrics for generative models in PyTorch

synthraw. Synthesizes camera raw files

philips-hue. A CLI tool to interface with Philips Hue lights.

exploring-contrastive-topology. Jupyter Notebook

average. Exponentially weighted moving averages with initialization bias correction.

prettyprinter. Syntax-highlighting, declarative and composable pretty printer for Python 3.5+

flash-attention. Fast and memory-efficient exact attention

gradient-maker3. A web application to generate color gradients using the CAM02-UCS colorspace.

crowsonkb.github.io. HTML

keras. Deep Learning for humans

websynth. pNaCl-Csound based softsynth

randomness. Generates random secrets (passwords, etc).

color_schemer. A web application to translate color schemes between dark- and light-background.

scihub-lookup. A Safari extension to look up the current page on Sci-Hub