FatimahEmadEldin/AraGPT2-Arabic-Humor-Generator

AraGPT2 Arabic Humor Generator

This model is a fine-tuned version of aubmindlab/aragpt2-medium specifically optimized for generating Arabic jokes and sarcastic humor based on keyword pairs.

Model Description

The model was trained to solve the "Keyword-to-Joke" task. It uses a specific prompt format: word1 word2 | [Generated Joke]

Training Results

During the final training phase, the model achieved:

• Training Loss: 3.28
• Validation Loss: 4.49
• Epochs: 5

How to Use

You can use this model with the Hugging Face transformers library:

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("{HF_REPO_ID}")
model = AutoModelForCausalLM.from_pretrained("{HF_REPO_ID}")

prompt = "محشش مدرسة | "
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=50, repetition_penalty=1.3)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Intended Use & Limitations

This model is designed for creative writing and humor generation.

• Repetition: Use a repetition_penalty of 1.2 or higher to avoid loops.
• Safety: While training data was filtered, AI humor can occasionally produce unexpected results.