Search Results for "loops"

OpenMythos is an experimental, open-source implementation that attempts to reconstruct a hypothesized architecture behind advanced language models using a design called a Recurrent-Depth Transformer. The project explores the idea that instead of stacking hundreds of unique transformer layers, a smaller set of layers can be reused iteratively during inference to achieve deeper reasoning without increasing parameter count. It divides computation into three main stages, including a...

...It uses a Mixture-of-Experts (MoE) architecture with 230 billion total parameters but only 10 billion activated per token, giving it the behavior of a very large model at a fraction of the runtime cost. The model is tuned for end-to-end developer flows such as multi-file edits, compile–run–fix loops, and test-validated repairs across real repositories and diverse programming languages. It is also optimized for multi-step agent tasks, planning and executing long toolchains that span shell commands, browsers, retrieval systems, and code runners. Benchmarks show that it achieves highly competitive scores on a wide range of intelligence and agent benchmarks, including SWE-Bench variants, Terminal-Bench, BrowseComp, GAIA, and several long-context reasoning suites.

...Beyond basic rendering scaffolding, LTX-2 includes optimized math libraries, resource loaders, utilities for texture and buffer handling, and integration points for native event loops and input systems. The framework targets both interactive graphical applications and media-rich experiences, making it a solid foundation for games, creative tools, or visualization systems that demand both performance and flexibility. While being low-level, it also provides sensible defaults and helper abstractions that reduce boilerplate and help teams maintain clear, maintainable code.

Open-Oasis provides inference code and released weights for Oasis 500M, an interactive world model that generates gameplay frames conditioned on user keyboard input. Instead of rendering a pre-built game world, the system produces the next visual state via a diffusion-transformer approach, effectively “imagining” the world response to your actions in real time. The project focuses on enabling action-conditional frame generation so developers can experiment with interactive, model-generated...

...The code is organized to be legible and hackable, exposing attention blocks, positional encodings, and head configurations. With standard PyTorch abstractions, it integrates easily into existing training loops, loggers, and evaluation harnesses.

xformers is a modular, performance-oriented library of transformer building blocks, designed to allow researchers and engineers to compose, experiment, and optimize transformer architectures more flexibly than monolithic frameworks. It abstracts components like attention layers, feedforward modules, normalization, and positional encoding, so you can mix and match or swap optimized kernels easily. One of its key goals is efficient attention: it supports dense, sparse, low-rank, and...

...It organizes training around concepts (rather than just raw labels), encouraging models to understand attributes, relations, and compositional structure that transfer across tasks. The repository provides training loops, data tooling, and evaluation routines to learn and probe these concept embeddings, typically from large image–text or weakly supervised corpora. It includes utilities to build concept vocabularies, map supervision signals to those vocabularies, and measure zero-shot or few-shot generalization. Probing tools help diagnose what the model knows—e.g., attribute recognition, relation understanding, or compositionality—so you can iterate on data and objectives. ...