Search Results for "delta"

OpenDelta is an open-source parameter-efficient fine-tuning library that enables efficient adaptation of large-scale pre-trained models using delta tuning techniques. OpenDelta is a toolkit for parameter-efficient tuning methods (we dub it as delta tuning), by which users could flexibly assign (or add) a small amount parameters to update while keeping the most parameters frozen. By using OpenDelta, users could easily implement prefix-tuning, adapters, Lora, or any other types of delta tuning with preferred PTMs.

DELTA is a deep learning-based end-to-end natural language and speech processing platform. DELTA aims to provide easy and fast experiences for using, deploying, and developing natural language processing and speech models for both academia and industry use cases. DELTA is mainly implemented using TensorFlow and Python 3. DELTA has been used for developing several state-of-the-art algorithms for publications and delivering real production to serve millions of users. ...

...An updated version, running on more modern hardware, but remaining true to the algorithms published in his book. The end result will be at least three sets of code, for the Alpha, Beta and Gamma versions of Rodney. We would also like to experiment with a "Delta" version which could use either bare RF or WiFi to exchange state data between machines. A "hive mind" if you like. For the new-generation version, we propose targeting two hardware platforms: 1. Arduino, since it's cheap and easy to use, while remaining close to the bare metal 2. A purpose-built AVR card, or easily ported to cards obtainable from sources such as futurlec.com, etc. ...

OSCEAN is an Open Source Content Environment and Abstract Network. It's meant to be an Operating System that runs on top of your current OS as an interface to a universal operating system the user can access at any computer running OSCEAN.

Personal research achievements from the authors presented here in the form of symbolic Algebra and non-numerical analysis algorithms implemented in C/C++, Java, and/or Pascal. As part of a more wide, simple, powerful and robust CAS enviroment under development, new code will be continuously added up to this repository. The authors give their welcome and encouragement to everyone interested in bring some kind of collaboration.

...Built on top of LLaMA-2 and DINOv2/SigLIP visual backbones, it allows both zero-shot inference for known robot setups and parameter-efficient fine-tuning for new domains. The model supports real-world robotics tasks, with robust generalization to environments seen in pretraining. Its actions include delta values for position, orientation, and gripper status, and can be un-normalized based on robot-specific statistics. OpenVLA is MIT-licensed, fully open-source, and designed collaboratively by Stanford, Berkeley, Google DeepMind, and TRI. Deployment is facilitated via Python and Hugging Face tools, with flash attention support for efficient inference.