yzhuang/Meta-Llama-3-8B-Instruct_fictional_arc_Japanese_v1
yzhuang/Meta-Llama-3-8B-Instruct_fictional_arc_Japanese_v1 is an 8 billion parameter instruction-tuned causal language model developed by yzhuang. It is a fine-tuned version of Meta-Llama-3-8B-Instruct, specifically adapted for fictional narrative arcs in Japanese. This model is designed for generating Japanese text with a focus on storytelling structures.
Loading preview...
Model Overview
yzhuang/Meta-Llama-3-8B-Instruct_fictional_arc_Japanese_v1 is an 8 billion parameter language model, fine-tuned from the meta-llama/Meta-Llama-3-8B-Instruct base model. This specialization focuses on generating content related to fictional narrative arcs in Japanese.
Key Characteristics
- Base Model: Meta-Llama-3-8B-Instruct, providing a strong foundation for instruction-following tasks.
- Language Focus: Specifically fine-tuned for Japanese text generation.
- Specialization: Optimized for understanding and generating fictional narrative structures.
Training Details
The model was trained with the following hyperparameters:
- Learning Rate: 5e-05
- Batch Size: 1 (train), 2 (eval)
- Gradient Accumulation Steps: 16, resulting in a total effective batch size of 16.
- Optimizer: Adam with betas=(0.9, 0.999) and epsilon=1e-08.
- Scheduler: Linear learning rate scheduler.
- Epochs: 36 training epochs.
Intended Use Cases
This model is particularly suited for applications requiring the generation of Japanese fictional content, especially those involving structured storytelling or narrative development. Its fine-tuning on fictional arc datasets suggests proficiency in creating coherent and engaging story elements in Japanese.