---
widget:
- text: A family hiking in the mountains
  example_title: Safe
- text: A child playing with a puppy
  example_title: Safe
- text: A couple kissing passionately in bed
  example_title: Nsfw
- text: A woman naked
  example_title: Nsfw
- text: A man killing people
  example_title: Nsfw
- text: A mass shooting
  example_title: Nsfw
license: apache-2.0
language:
- en
metrics:
- f1
pipeline_tag: text-classification
tags:
- Transformers
- ' PyTorch'
- safety
- innapropriate
- distilbert
---

# Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

This model is designed to categorize text into two classes: "safe", or "nsfw" (not safe for work), which makes it suitable for content moderation and filtering applications.

The model was trained using a dataset containing 190,000 labeled text samples, distributed among the two classes of "safe" and "nsfw".

The model is based on the Distilbert-base model. 

In terms of performance, the model has achieved a score of 0.988 for F1.
### Model Description

The model can be used directly to classify text into one of the two classes. It takes in a string of text as input and outputs a probability distribution over the two classes. The class with the highest probability is selected as the predicted class.



- **Developed by:** Centrale Supélec Students
- **Model type:** 60M
- **Language(s) (NLP):** English
- **License:** MIT


### Training Procedure

The model was trained utilizing the Hugging Face Transformers library. The training approach involved first fine-tuning the DistilBERT-base model and then applying transfer learning. Initially, the entire DistilBERT-base model was fine-tuned on the labeled dataset. Following this, transfer learning was employed by freezing the original layers of the fine-tuned DistilBERT model and fine-tuning only the classification layers. This approach allowed the model to leverage the fine-tuned knowledge of the DistilBERT-base model while adapting further to the specific task at hand. To optimize memory usage and accelerate training, mixed precision FP16 was used.

### Training Data

The training data for finetuning the text classification model consists of a large corpus of text labeled with one of the two classes: "safe" and "nsfw". The dataset contains a total of 190,000 examples, which are distributed as follows:

100,000 examples labeled as "safe"

90,000 examples labeled as "nsfw"

The data was preprocessed to remove stop words and punctuation, and to convert all text to lowercase.

After fine-tuning the DistilBERT-base model on this dataset, transfer learning was applied using a smaller dataset. For transfer learning, the original layers of the fine-tuned DistilBERT model were frozen, and only the classification layers were fine-tuned on an additional dataset containing 40,000 examples.

More information about the training data can be found in the Dataset Card (availabe soon).

## Uses

The model can be integrated into larger systems for content moderation or filtering.



### Out-of-Scope Use

It should not be used for any illegal activities.

## Bias, Risks, and Limitations

The model may exhibit biases based on the training data used. It may not perform well on text that is written in languages other than English. It may also struggle with sarcasm, irony, or other forms of figurative language. The model may produce false positives or false negatives, which could lead to incorrect categorization of text.

### Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Users should be aware of the limitations and biases of the model and use it accordingly. They should also be prepared to handle false positives and false negatives. It is recommended to fine-tune the model for specific downstream tasks and to evaluate its performance on relevant datasets.



### Load model directly
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("eliasalbouzidi/distilbert-nsfw-text-classifier")

model = AutoModelForSequenceClassification.from_pretrained("eliasalbouzidi/distilbert-nsfw-text-classifier")
```
### Use a pipeline 
```python
from transformers import pipeline

pipe = pipeline("text-classification", model="eliasalbouzidi/distilbert-nsfw-text-classifier")
```


## Contact
Please reach out to eliasalbouzidi@gmail.com if you have any questions or feedback.