Instructions to use socius/Llama-Centaur-8B-LoRA-r4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use socius/Llama-Centaur-8B-LoRA-r4 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("unsloth/Llama-3.1-8B") model = PeftModel.from_pretrained(base_model, "socius/Llama-Centaur-8B-LoRA-r4") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
File size: 5,611 Bytes
f38a37f f26bc2b f38a37f f26bc2b f38a37f f26bc2b 86b60bf f26bc2b 86b60bf f26bc2b 0e8aa55 f26bc2b 97d7cc6 f26bc2b 86b60bf f26bc2b 86b60bf f26bc2b 86b60bf f26bc2b 97d7cc6 f26bc2b 86b60bf f26bc2b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 | ---
license: llama3.1
datasets:
- marcelbinz/Psych-101
language:
- en
base_model:
- unsloth/Llama-3.1-8B
base_model_relation: adapter
pipeline_tag: text-generation
library_name: peft
tags:
- psychology
- cognitive science
- human behavior
- unsloth
- lora
---
<div align="center">
<img src="Llama-Centaur-8B-LoRA-r4.png" alt="Llama-Centaur-8B-LoRA-r4" width="1000">
[](https://huggingface.co/meta-llama/Llama-3.1-8B)
[](https://huggingface.co/collections/socius/llama-centaur-8b-lora-6a377e5dcc0bdc9c01113c68)
[](https://arxiv.org/abs/2608.05224)
[](https://huggingface.co/socius/Llama-Centaur-8B-LoRA-r4)
[](https://huggingface.co/socius/Llama-Centaur-8B-LoRA-r4)
[](https://huggingface.co/datasets/marcelbinz/Psych-101)
</div>
# Llama-Centaur-8B-LoRA-r4
LoRA adapter for **Llama-Centaur-8B**, fine-tuned on the full
[Psych-101](https://huggingface.co/datasets/marcelbinz/Psych-101) as part of the
LoRA-rank sweep and dataset-size ablation for
*Small Foundation Models of Human Cognition and Behaviour*.
| field | value |
|----------------|-------|
| base model | [unsloth/Llama-3.1-8B](https://huggingface.co/unsloth/Llama-3.1-8B) |
| LoRA rank | 4 (alpha = rank, rsLoRA) |
| data fraction | 100% of Psych-101 |
| training | 1 epoch, completion-only loss, seed 3407 |
Load with PEFT on top of `unsloth/Llama-3.1-8B`, or evaluate with the project's
`eval_model.py --backend unsloth`.
|