Llama3 Fine-Tuned on Novel Continuation Dataset
Overview
This model, llama-3.1-novel, has been fine-tuned on a dataset containing over 10,000 records of novel text, specifically designed for the task of text continuation. The fine-tuning process was accelerated using Unsloth and Hugging Face's TRL library, achieving faster training times.
- Developed by: Eldritch
- License: Apache-2.0
- Fine-tuned from model:
unsloth/meta-llama-3.1-8b-bnb-4bit
Dataset Details
The dataset used for fine-tuning can be accessed here. It includes the following columns:
- Instruction: A prompt or instruction guiding the continuation of the novel.
- Input: An excerpt from the novel serving as the starting point for continuation.
- Output: The continuation of the novel, following the input text, as per the instruction.
Language
- English
Tags
- Text Generation
- Transformers
- Eldritch
- Llama
- TRL
Size Category
- 10,000 < n < 100,000
Model Usage
This fine-tuned Llama3 model demonstrates strong capabilities in generating coherent and stylistically consistent continuations for novel excerpts. It is particularly useful for:
- Creative Writing Assistance: Generating novel ideas and story continuations.
- Narrative Exploration: Exploring different narrative possibilities.
- Educational Tools: Teaching narrative structure and creative writing techniques.
Accessing the Model
The model can be accessed and utilized through the Hugging Face Model Hub at Eldritch/llama-3.1-novel. The Hugging Face Transformers library can be used to load and interact with the model.
License
This model and the associated dataset are distributed under the Apache-2.0 License. For more details, please refer to the license file.
Acknowledgements
Special thanks to the authors and publishers of the novels included in the dataset, whose works have been instrumental in developing this model.
Contact
For any inquiries or issues, please contact [Your Name] at [Your Contact Information].
