DataPilot/ArrowCanaria-Llama-8B-SFT-v0.1
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kPublished:Mar 20, 2026License:llama3.3Architecture:Transformer0.0K Cold

DataPilot/ArrowCanaria-Llama-8B-SFT-v0.1 is an 8 billion parameter, Japanese-centric language model based on Llama 3.1 (tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.5) with an 8192 token context length. It is specifically fine-tuned for natural, human-like Japanese conversation, excelling in casual chat, roleplay, and empathetic responses, particularly for AItuber applications. The model avoids typical machine-like or translated tones, focusing on nuanced emotional expression and creative Japanese phrasing, while also retaining strong reasoning and tool-use capabilities.

Loading preview...