Download
Backtrack Sampler is a framework designed for experimenting with custom sampling strategies for language models (LLMs), enabling the ability to rewind and revise generated tokens. It allows developers to create and test their own token generation strategies by providing a base structure for manipulating logits and probabilities, making it a flexible tool for those interested in fine-tuning the behavior of LLMs.
Features
- Customizable token generation strategies
- Ability to backtrack and revise generated tokens
- Integration with models from Transformers and Llama.cpp
- Anti-slop strategy to prevent undesirable token generation
- Creative writing strategy to enhance model creativity by modifying token selection
- Easy-to-implement strategies using the base strategy class
Project Samples
Categories
FrameworksLicense
MIT LicenseFollow Backtrack Sampler
Other Useful Business Software
Ango Hub | All-in-one data labeling platform
AI-Assisted features of the Ango Hub will automate your AI data workflows to improve data labeling efficiency and model RLHF, all while allowing domain experts to focus on providing high-quality data.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of Backtrack Sampler!
