Search Results for "budget"

...When integrated with GitHub Actions, CircleCI, or other CI systems, it runs on every commit or pull request, failing the pipeline if a change exceeds the configured budget. It can also annotate pull requests with comments that highlight how the bundle changed, making performance regressions visible to reviewers early. With the --why flag, size-limit can analyze what contributes most to bundle size and show a dependency breakdown (using Statoscope).

...It targets hyperparameter search, architecture search, control problems, and experimental tuning—domains in which gradient-based methods may fail or be inapplicable. The library provides an easy interface to define an optimization problem (parameter space, loss function, budget) and then experiment with multiple strategies—evolutionary algorithms, Bayesian optimization, bandit methods, genetic algorithms, etc. Nevergrad supports parallelization, budget scheduling, and multiple cost/resource constraints, allowing it to scale to nontrivial optimization problems. It includes visualization tools and diagnostic metrics to compare strategy performance, track parameter evolution, and detect stagnation.

...It supports training with minimal code changes required on the client, has little impact on training performance, and allows the client to online track the privacy budget expended at any given moment. Vectorized per-sample gradient computation that is 10x faster than micro batching. Supports most types of PyTorch models and can be used with minimal modification to the original neural network. Open source, modular API for differential privacy research. Everyone is welcome to contribute. ML practitioners will find this to be a gentle introduction to training a model with differential privacy as it requires minimal code changes. ...

...It introduced Fully Sharded Data Parallel (FSDP) style techniques that shard model parameters, gradients, and optimizer states across ranks to fit bigger models into the same memory budget. The library also provides pipeline parallelism, activation checkpointing, mixed precision, optimizer state sharding (OSS), and auto-wrapping policies that reduce boilerplate in complex distributed setups. Its components are modular, so teams can adopt just the sharding optimizer or the pipeline engine without rewriting their training loop. ...

...The focus of the course is on the hands-on use of both simple and complex climate models to build an understanding of the processes that control the planetary energy budget. The course makes extensive use of Python code and the Jupyter notebook for reproducible, self-describing calculations and figures. This repository contains a collection of linked Jupyter notebooks with lecture notes, examples and assignments. All notebooks are self-describing.

Shoestring is a lightweight, modular DOM framework intended for developers who need their code to download and run as fast as possible. It is intended to be "just enough" of a tool to efficently write maintainable cross-browser JavaScript. The API is modeled after jQuery but we intentionally implement a tiny subset of the API to cover only the methods we commonly need in our projects. Each feature is built as an optional extension to a minimal core so each can be removed from the production...