← all releases
Sakana AI·Dec 16, 2025·v1.0·open source

RePo-OLMo2-1B-stage2-L5

RePo-OLMo2-1B-stage2-L5 is an autoregressive language model developed by Sakana AI, based on the allenai/OLMo-2-0425-1B model. It has been continually pre-trained on 50 billion tokens from the allenai/dolmino-mix-1124 dataset using the RePo (Context Re-Positioning) technique, which aims to enhance the model's understanding of context by adaptively restructuring input sequences.

ML
visit website
overview

RePo-OLMo2-1B-stage2-L5 is an autoregressive language model developed by Sakana AI, based on the allenai/OLMo-2-0425-1B model. It has been continually pre-trained on 50 billion tokens from the allenai/dolmino-mix-1124 dataset using the RePo (Context Re-Positioning) technique, which aims to enhance the model's understanding of context by adaptively restructuring input sequences.

key features
  • 011B parameters
  • 02Continually pre-trained on 50B tokens
  • 03Utilizes RePo technique for context re-positioning
related productsbrowse all →
SkyRL txopen sourceSkyRL tx is an open-source library that implements a backend for the Tinker API, enabling users to set up their own Tinker-like services on personal hardware. It supports end-to-end reinforcement learning (RL) and offers significantly faster sampling. The library is designed to be modular, allowing easy prototyping of new training algorithms, environments, and execution plans without compromising usability or speed.Step-Audio-R1open sourceStep-Audio-R1 is an advanced audio language model developed by StepFun AI, designed to enhance audio reasoning capabilities by grounding its reasoning in acoustic features. It introduces Modality-Grounded Reasoning Distillation (MGRD), an iterative training framework that shifts the model's reasoning from textual abstractions to acoustic properties, effectively addressing the 'inverted scaling' problem where performance degrades with longer reasoning. This model has demonstrated superior performance across various audio understanding and reasoning benchmarks, surpassing models like Gemini 2.5 Pro and achieving results comparable to Gemini 3 Pro.SINQopen sourceSINQ (Sinkhorn-Normalized Quantization) is a novel, fast, and high-quality quantization method designed to make any Large Language Model (LLM) smaller while preserving accuracy. It offers a plug-and-play, model-agnostic technique that delivers state-of-the-art performance for LLMs without sacrificing accuracy.