ajibawa-2023/General-Stories-Mistral-7B
The ajibawa-2023/General-Stories-Mistral-7B is a 7 billion parameter language model, fine-tuned by ajibawa-2023, based on the Mistral-7B-v0.1 architecture. It was extensively trained for over 15 days on a 1.3 million story dataset, General-Stories-Collection, specifically curated for general audiences. This model excels at generating captivating narratives and is optimized for versatile storytelling across broad themes.
Loading preview...
Model Overview
ajibawa-2023/General-Stories-Mistral-7B is a 7 billion parameter language model, fine-tuned by ajibawa-2023, built upon the Mistral-7B-v0.1 base model. Its primary focus is on generating diverse and engaging narratives for a general audience.
Key Capabilities
- Extensive Story Generation: Trained on a vast synthetic dataset of approximately 1.3 million stories, enabling it to produce a wide range of narratives.
- Narrative Intricacy: Developed to understand and generate complex narrative structures and themes, aiming to evoke emotion and imagination.
- Versatile Storytelling: Designed to offer broad appeal and adaptable storytelling capabilities for various literary needs.
Training Details
The model underwent an extensive training period of over 15 days across two epochs, utilizing 4 x A100 80GB GPUs. The training leveraged the Axolotl codebase and the custom General-Stories-Collection dataset. This is a qLoRA version of the model.
Usage
This model utilizes the ChatML prompt format. Users can adapt the provided prompt structure for their specific generation requirements. Quantized versions (GGUF and Exllama v2) are also available, thanks to Bartowski.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.