---
base_model: mlabonne/NeuralMarcoro14-7B
license: apache-2.0
tags:
- mlabonne/NeuralMarcoro14-7B
- dpo
- 7B
- winograd
- mmlu_abstract_algebra
- mistral
datasets:
- hromi/winograd_dpo_basic
---


# udkai_Turdus
A less contaminated version of [udkai/Garrulus](https://huggingface.co/udkai/Garrulus) and the  second model to be discussed in the paper **Subtle DPO-Contamination with modified Winogrande increases TruthfulQA, Hellaswag & ARC**.

Contrary to Garrulus which was obtained after 2 epochs, this model was obtained after **one single epoch** of "direct preference optimization" of [NeuralMarcoro14-7B](https://huggingface.co/mlabonne/NeuralMarcoro14-7B) with [https://huggingface.co/datasets/hromi/winograd_dpo] . 

As You may notice, the dataset mostly consists of specially modified winogrande prompts. 

But before flagging this (or recommending this to be flagged), consider this:

Subtle DPO-Contamination with modified Winogrande in such an extent that the average accuracy of all 5-non Winogrande metrics (e.g. including also MMLU and GSM8K) is 0.2% higher than the base model.

| Model                        | ARC   | HellaSwag | MMLU | Truthful QA | GSM8K | Average |
| -----------------------------|------ | --------- | ---- | ----------- | ------| ------- |
| mlabonne/NeuralMarcoro14-7B  | 71.42 |  87.59    | 64.84| 65.64       | 70.74 | 72.046  |
| udkai/Turdus                 | 73.38 |  88.56    | 64.52| 67.11       | 67.7  | **72,254**  |

Yes, as strange as it may sound, one can indeed increase ARC from 71.42% to 73.38 % with one single epoch of cca 1200 repetitive winograd schematas...