---
language:
- fr
tags:
- lm-detection
datasets:
- hc3_fr_custom_ms_hg
metrics:
- f1
model-index:
- name: camemberta-chatgptdetect-noisy
  results:
  - task:
      name: Text Classification
      type: text-classification
    dataset:
      name: HC3 FULL_FR_1.0_0.5_0.5
      type: glue
      config: full_fr_1.0_0.5_0.5
      split: val
      args: full_fr_1.0_0.5_0.5
    metrics:
    - name: F1
      type: f1
      value: 0.9790566381351302
---

# camemberta-chatgptdetect-noisy

French ChatGPT detection model from [Towards a Robust Detection of Language Model-Generated Text: Is ChatGPT that easy to detect?](TODO:)

This model is a fine-tuned version of [almanach/camemberta-base](https://huggingface.co/almanach/camemberta-base) on the HC3 FULL_FR_1.0_0.5_0.5 dataset with noise added.
It achieves the following results on the evaluation set:
- Loss: 0.0430
- F1: 0.9791

## Model description

This a model trained to detect text created by ChatGPT in French.
The training data is the `hc3_fr_full` subset of [almanach/hc3_multi](https://huggingface.co/datasets/almanach/hc3_french_ood), but with added misspelling and homoglyph attacks.

## Intended uses & limitations

This model is for research purposes only.
It is not intended to be used in production as we said in our paper:


**We would like to emphasize that our study does not claim to have produced an universally accurate detector. Our strong results are based on in-domain testing and, unsurprisingly, do not generalize in out-of-domain scenarios. This is even more so when used on text specifically designed to fool language model detectors and on text intentionally stylistically similar to ChatGPT-generated text, especially instructional text.**


## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 25
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | F1     |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 0.0199        | 1.0   | 4267  | 0.0430          | 0.9791 |
| 0.0104        | 2.0   | 8534  | 0.1457          | 0.9463 |
| 0.0026        | 3.0   | 12801 | 0.0805          | 0.9720 |
| 0.0           | 4.0   | 17068 | 0.2515          | 0.9419 |
| 0.0           | 5.0   | 21335 | 0.2000          | 0.9567 |


### Framework versions

- Transformers 4.26.1
- Pytorch 1.11.0+cu115
- Datasets 2.8.0
- Tokenizers 0.13.2