metadata

language:
  - fr
tags:
  - lm-detection
datasets:
  - hc3_fr_custom_ms_hg
metrics:
  - f1
base_model: almanach/camemberta-base
model-index:
  - name: camemberta-chatgptdetect-noisy
    results:
      - task:
          type: text-classification
          name: 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:
          - type: f1
            value: 0.9790566381351302
            name: F1

camemberta-chatgptdetect-noisy

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

This model is a fine-tuned version of 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

Validation set:

Loss: 0.0430
F1: 0.9791

Test Set:

F1: 0.97

Adversarial:

F1: 0.45

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, 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