Search Results for "trains"

...It provides the components needed to prepare data, train draft models, and measure acceptance behavior against target models. The workflow starts with data preparation, including prompt download, target answer regeneration, and target cache construction. It then trains a draft model using configuration files for different algorithms and target model setups. The evaluation pipeline measures speculative decoding performance across benchmark tasks such as math, coding, instruction-following, and chat-style datasets. Overall, it is useful for researchers and engineers studying faster language model inference through speculative decoding methods.

...This objective sidesteps generative pixel losses and avoids heavy negative sampling, producing features that transfer strongly with linear probes and minimal fine-tuning. The design scales naturally with Vision Transformer backbones and flexible masking strategies, and it trains stably at large batch sizes. i-JEPA’s predictions are made in embedding space, which is computationally efficient and better aligned with downstream discrimination tasks. The repository provides training recipes, data pipelines, and evaluation code that clarify which masking patterns and architectural choices matter most.

Fairseq(-py) is a sequence modeling toolkit that allows researchers and developers to train custom models for translation, summarization, language modeling and other text generation tasks. We provide reference implementations of various sequence modeling papers. Recent work by Microsoft and Google has shown that data parallel training can be made significantly more efficient by sharding the model parameters and optimizer state across data parallel workers. These ideas are encapsulated in the...

...Includes more features such as panoptic segmentation, Densepose, Cascade R-CNN, rotated bounding boxes, PointRend, DeepLab, etc. Can be used as a library to support different projects on top of it. We'll open source more research projects in this way. It trains much faster. Models can be exported to TorchScript format or Caffe2 format for deployment. With a new, more modular design, Detectron2 is flexible and extensible, and able to provide fast training on single or multiple GPU servers. Detectron2 includes high-quality implementations of state-of-the-art object detection.

...We use PyText at Facebook to iterate quickly on new modeling ideas and then seamlessly ship them at scale. Distributed-training support built on the new C10d backend in PyTorch 1.0. Mixed precision training support through APEX (trains faster with less GPU memory on NVIDIA Tensor Cores). Extensible components that allows easy creation of new models and tasks.

mixup-cifar10 is the official PyTorch implementation of “mixup: Beyond Empirical Risk Minimization” (Zhang et al., ICLR 2018), a foundational paper introducing mixup, a simple yet powerful data augmentation technique for training deep neural networks. The core idea of mixup is to generate synthetic training examples by taking convex combinations of pairs of input samples and their labels. By interpolating both data and labels, the model learns smoother decision boundaries and becomes more...