npc-worldwide/TinyTimV1
TinyTimV1 is a 1.1 billion parameter language model fine-tuned by npc-worldwide from TinyLlama-1.1B-Chat. This model specializes in generating text in the distinctive experimental style of James Joyce's *Finnegan's Wake*, replicating its complex wordplay and stream-of-consciousness narrative. It is designed for creative text generation tasks requiring a highly specific, literary linguistic style.
Loading preview...
Overview
TinyTimV1 is a specialized 1.1 billion parameter language model, fine-tuned from the TinyLlama-1.1B-Chat-v1.0 base model. Its unique characteristic is its training on the complete text of James Joyce's Finnegan's Wake. This focused fine-tuning enables the model to generate text that mimics Joyce's highly experimental style, including complex wordplay, neologisms, and stream-of-consciousness narrative techniques.
Key Capabilities
- Joyce-inspired Text Generation: Produces output reflecting the linguistic complexity and stylistic nuances of Finnegan's Wake.
- Experimental Style Replication: Learns and applies unique literary devices such as invented words and intricate sentence structures.
- Compact Size: At 1.1 billion parameters, it offers a relatively small footprint for specialized literary generation.
Training Details
The model was fine-tuned on approximately 1.5MB of text from Finnegan's Wake for 3 epochs. Training utilized a batch size of 1 and a maximum sequence length of 128 tokens. Due to resource constraints, training was performed on a CPU. The project includes scripts for data preprocessing, fine-tuning, and text generation, making it a self-contained example of highly specialized literary model adaptation.
Good For
- Researchers and enthusiasts exploring literary style transfer and experimental text generation.
- Creative projects requiring text with a unique, avant-garde linguistic flair.
- Demonstrating the impact of highly specific domain fine-tuning on smaller language models.