fallama / README.md
EliaMengozzi's picture
Update README.md
e6bd983 verified
metadata
license: llama3.1
datasets:
  - tasksource/logical-fallacy
language:
  - en

Model Description

FALLAMA is a specialized model fine-tuned to detect and classify logical fallacies in text. Built on top of the LLAMA 3.1 8B architecture, this model has been carefully trained to recognize a wide range of logical fallacies and categorize them according to a predefined taxonomy.

  • Developed by: [Elia Mengozzi]
  • Model type: [LLAMA 3.1 8B]
  • License: [Apache 2.0]
  • Finetuned from model [optional]: [LLAMA 3.1 8B]

Uses

Training Details: FALLAMA was fine-tuned using a dataset specifically curated for logical fallacies, covering a diverse range of fallacy types. The model was trained to not only identify the presence of a fallacy in a statement but also to classify it into one of the predefined categories.

Bias, Risks, and Limitations

FALLAMA is not a serious model and was developed solely for educational purposes. The results produced by this model should not be considered valid or reliable, and they should always be manually checked and verified. The model's outputs are intended to illustrate the process of detecting and classifying logical fallacies, but they are not suitable for any critical or professional use.

Training Data

FALLAMA was fine-tuned using a dataset specifically curated for logical fallacies, covering a diverse range of fallacy types.