CoolFace
Modelpublic

socius/Olmotaur-1B-LoRA-r16-f0.125

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face
0likes16downloads
Model Card

<div align="center"> <img src="Olmotaur-1B-LoRA-r16-f0.125.png" alt="Olmotaur-1B-LoRA-r16-f0.125" width="1000">

![AI2](https://huggingface.co/allenai/OLMo-2-0425-1B)

![socius](https://huggingface.co/collections/socius/olmotaur-1b-lora-6a6732511c8a8204d10679ab) ![Paper](https://arxiv.org/abs/2608.05224) ![Parameters](https://huggingface.co/socius/Olmotaur-1B-LoRA-r16-f0.125) ![LoRA](https://huggingface.co/socius/Olmotaur-1B-LoRA-r16-f0.125) ![Dataset](https://huggingface.co/datasets/marcelbinz/Psych-101) ![Data fraction](https://huggingface.co/datasets/marcelbinz/Psych-101) </div>

Olmotaur-1B-LoRA-r16-f0.125

LoRA adapter for Olmotaur-1B, fine-tuned on a stratified 12.5% subset of Psych-101 as part of the LoRA-rank sweep and dataset-size ablation for Small Foundation Models of Human Cognition and Behaviour.

fieldvalue
base modelallenai/OLMo-2-0425-1B
LoRA rank16 (alpha = rank, rsLoRA)
data fraction12.5% of Psych-101
training1 epoch, completion-only loss, seed 3407

Load with PEFT on top of allenai/OLMo-2-0425-1B, or evaluate with the project's eval_model.py --backend unsloth.