WasamiKirua/Alucard-Hellsing-Llama3.2-3B

Overview

Alucard-Hellsing-Llama3.2-3B is a specialized bilingual model trained on a dataset forged from original anime subtitles and synthetically generated dialogue pairs. The synthetic data was crafted using custom prompts inspired by the diverse cast of characters—both allies and enemies—of Alucard. This dataset aims to encapsulate the distinctive speech patterns, mannerisms, and personality traits of Alucard, ensuring authentic interactions that remain true to his character.

Purpose

The dataset was designed with the intent of creating a model that embodies the essence of Alucard's speech style—dark, charismatic, and unapologetically intense. The model’s responses may be bold, assertive, or even harsh, reflecting the unfiltered nature of the character. Users engaging with the model should expect dialogues infused with confidence, sarcasm, and a commanding presence.

Data Sources

• Original Anime Subtitles: Extracted and processed from official sources to retain the authenticity of Alucard’s dialogue.
• Synthetic Dialogue Pairs: Generated using curated prompts that mimic Alucard’s linguistic patterns and personality traits, ensuring coherence and depth in responses.
• Character-Based Expansion: Includes interactions modeled after Alucard’s allies and adversaries, enriching the dataset with a variety of conversational tones and themes.

Model Behavior & Ethical Considerations

The model trained on this dataset is designed to respond in Alucard’s distinctive manner. Given the character’s nature, users should be aware that:

• Responses may include strong language, dark humor, and intense rhetoric.
• The model does not promote harmful ideologies but remains faithful to the character's expressive style.
• Care has been taken to prevent the model from generating excessively harmful or unethical content beyond the fictional context.

Limitations

• The model is primarily focused on entertainment and character emulation, making it unsuitable for general-purpose conversations or sensitive topics.
• Due to the dataset’s nature, the model may lack neutrality and instead favor Alucard’s characteristic speech patterns and perspectives.

Acknowledgments

Special thanks to the anime and creative communities that have contributed to understanding and preserving Alucard’s iconic persona. This dataset was developed with the goal of enhancing immersive AI interactions for fans and enthusiasts alike.

---

For further details or inquiries regarding the dataset, feel free to reach out.

You can use this model in Ollama with the following template (llama)

FROM {__FILE_LOCATION__}
TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|>

{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>

{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>

{{ .Response }}<|eot_id|>"""
PARAMETER stop "<|start_header_id|>"
PARAMETER stop "<|end_header_id|>"
PARAMETER stop "<|eot_id|>"
PARAMETER temperature 1.5
PARAMETER min_p 0.1

Trained with Unsloth

• Developed by: WasamiKirua
• License: apache-2.0
• Finetuned from model : unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.