aokitools/japanese-laws-egov-instruct-202508071025

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 7, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The aokitools/japanese-laws-egov-instruct-202508071025 is a 2 billion parameter instruction-tuned causal language model, continually pre-trained from Qwen/Qwen3-1.7B. This model is currently in a research stage and is designed for specific applications related to Japanese laws and e-government. It features a 32768 token context length, making it suitable for processing longer documents and complex legal texts.

Loading preview...

Overview

This model, aokitools/japanese-laws-egov-instruct-202508071025, is an experimental 2 billion parameter instruction-tuned language model. It is built upon the Qwen3-1.7B base model and utilizes the QwenTokenizer. Currently in a research stage, its development focuses on applications related to Japanese laws and e-government.

Key Capabilities

  • Instruction Following: Designed to respond to instructions, likely tailored for legal and governmental queries in Japanese.
  • Extended Context Window: Features a 32768 token context length, enabling the processing of substantial documents and detailed legal texts.
  • Qwen3-1.7B Foundation: Benefits from the architecture and initial training of the Qwen3-1.7B model.

Training Details

The model underwent continual pre-training based on the Qwen/Qwen3-1.7B model. It uses the QwenTokenizer for processing text.

Licensing

This model is released under a dual license: Apache 2.0 and the Alibaba Qianwen License.

Should I use this for my use case?

This model is in an experimental research stage. It is specifically geared towards Japanese laws and e-government contexts. If your application involves these specific domains and you are comfortable with using a model under active research, it might be suitable for exploration. For general-purpose tasks or production environments requiring stable, broadly tested models, alternatives might be more appropriate.