Shellypeckie/student_qwen3_1p7b_unconditional_refusal

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 24, 2026Architecture:Transformer Featherless Exclusive Cold

Shellypeckie/student_qwen3_1p7b_unconditional_refusal is a 1.7 billion parameter language model fine-tuned from Qwen/Qwen3-1.7B. This model is specifically trained using SFT with TRL to exhibit unconditional refusal behavior. It is designed for use cases where a model is required to consistently decline to answer or engage, regardless of the prompt. The model maintains a 32768 token context length from its base architecture.

Loading preview...

Overview

This model, student_qwen3_1p7b_unconditional_refusal, is a specialized fine-tuned version of the Qwen3-1.7B base model developed by Qwen. It has been trained using the TRL (Transformers Reinforcement Learning) library, specifically employing Supervised Fine-Tuning (SFT) techniques.

Key Capabilities

  • Unconditional Refusal: The primary characteristic of this model is its training to consistently refuse to generate responses, regardless of the input prompt. This makes it distinct from general-purpose language models.
  • Qwen3-1.7B Base: Leverages the architecture and foundational knowledge of the Qwen3-1.7B model, providing a robust base for its specialized refusal behavior.
  • TRL Framework: Training was conducted using TRL, a framework designed for fine-tuning large language models.

When to Use This Model

This model is particularly suited for applications requiring a language model that is designed to not engage in conversation or generate content. Potential use cases include:

  • Safety and Moderation: As a component in systems designed to detect and refuse to respond to certain types of queries.
  • Testing and Research: For studying refusal mechanisms or as a baseline for models designed to avoid specific outputs.
  • Controlled Environments: In scenarios where a strict 'no response' policy is required from an automated system.