metadata

license: apache-2.0
base_model: google/flan-t5-base
tags:
  - generated_from_trainer
model-index:
  - name: Fake-news-gen
    results: []

Fake-news-gen

This model is a fine-tuned version of google/flan-t5-base on the None dataset.

Here is one way to write a professional model card description for your fake news generator model on Hugging Face:

Summary

This model is a conditional text generation system fine-tuned to produce artificially generated news articles from short text summaries. It demonstrates the potential for AI systems to automatically synthesize false or misleading news content from limited input information.

Intended Uses

Research on AI fake news generation capabilities and risks
Educational purposes to increase awareness of AI fake content issues
Testing automatic fake news detection systems

Factors

Initially trained on summarization data (XSUM BBC news)
Fine-tuned end-to-end to generate full articles from summaries
Generates content token-by-token based on summary prompt
No ground-truth real/fake labels or classifier included
Outputs are raw model decodes without post-processing

Caveats and Recommendations

Output quality can vary, may require multiple sampling or overrides
Content reflects training data biases and model limitations
Not intended for malicious uses or dissemination
Users should manually review outputs before reliance

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10

Training results

Framework versions

Transformers 4.36.0
Pytorch 2.1.2+cu121
Datasets 2.16.1
Tokenizers 0.15.0