Popular GitHub Repositories for Smaller

Distilled variant of Whisper for speech recognition. 6x faster, 50% smaller, within 1% word error rate.

An Automatic Model Compression (AutoMC) framework for developing smaller and faster AI applications.

Libraries for applying sparsification recipes to neural networks with a few lines of code, enabling faster and smaller models

Neural network model repository for highly sparse and sparse-quantized models with matching sparsification recipes

ML model optimization product to accelerate inference.

Implementation of PALI3 from the paper PALI-3 VISION LANGUAGE MODELS: SMALLER, FASTER, STRONGER"

BERT classification model for processing texts longer than 512 tokens. Text is first divided into smaller chunks and after feeding them to BERT, intermediate results are pooled. The implementation allows fine-tuning.

Implementation of ICML 23 Paper: Specializing Smaller Language Models towards Multi-Step Reasoning.

Learning Compressed Representation of 3DLUT for Image-enhancement. Higher performance with much smaller models! :relaxed:

Load What You Need: Smaller Multilingual Transformers for Pytorch and TensorFlow 2.0.

A collection of resources on using BERT (https://arxiv.org/abs/1810.04805 ) and related Language Models in production environments.

Experimental playground for benchmarking language model (LM) architectures, layers, and tricks on smaller datasets. Designed for flexible experimentation and exploration.

Implements pre-training, supervised fine-tuning (SFT), and reinforcement learning from human feedback (RLHF), to train and fine-tune the LLaMA2 model to follow human instructions, similar to InstructGPT or ChatGPT, but on a much smaller scale.

The purpose of the "Meta Agent with More Agents" project is to dynamically solve complex queries by breaking them down into smaller tasks and assigning each to specialized AI agents. The Meta Agent coordinates the process, leveraging a ReAct Agent for tool-based tasks and a Chain of Thought Agent for reasoning-based tasks. The system's flexibility.

A Python tool for splitting large Markdown files into smaller sections based on a specified token limit. This is particularly useful for processing large Markdown files with GPT models, as it allows the models to handle the data in manageable chunks.

Something in the middle of Karpathy's mingpt model and video lectures, BabyGPT is an easy to use model on a much smaller scale (16 and 256 out channels , 5 heads, fine tuned). To be made useful on low powered devices.

Applying "Load What You Need: Smaller Versions of Multilingual BERT" to LaBSE

Pytorch implementation of the "Dermatologist-level classification of skin cancer with deep neural networks" research paper. Smaller dataset supplied by Udacity.

Chess environment for smaller chess variants, AlphaZero-like MCTS-learning, and Concept Detection

SHUKUN AI is an AI-native database search application that helps companies explore and analyze data. SHUKUN AI provides a feature that helps you split complex natural language tasks into smaller parts then combines them into the results.