RishiSpace/RI-Instruct-v0.2-Merged

TEXT GENERATIONConcurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jun 20, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

RishiSpace/RI-Instruct-v0.2-Merged is an 8 billion parameter instruction-tuned causal language model developed by RishiSpace. This Qwen3-based model was fine-tuned using Unsloth and Huggingface's TRL library, enabling faster training. It is designed for general instruction-following tasks, leveraging its efficient training methodology.

Loading preview...

Overview

RishiSpace/RI-Instruct-v0.2-Merged is an 8 billion parameter instruction-tuned language model built upon the Qwen3 architecture. Developed by RishiSpace, this model distinguishes itself through its training methodology, utilizing Unsloth and Huggingface's TRL library for significantly accelerated fine-tuning. This approach allows for efficient development and deployment of instruction-following capabilities.

Key Capabilities

  • Instruction Following: Designed to accurately interpret and execute a wide range of natural language instructions.
  • Efficient Training: Benefits from Unsloth's optimizations, enabling faster fine-tuning processes compared to traditional methods.
  • Qwen3 Base: Inherits the robust foundational capabilities of the Qwen3 model family.

Good For

  • Developers seeking an 8B parameter model with strong instruction-following performance.
  • Applications requiring a model that has undergone efficient and accelerated fine-tuning.
  • General-purpose natural language understanding and generation tasks where instruction adherence is crucial.