nityaak/qwen3-1.7b-stance-matrix-semeval-qlora

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 3, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The nityaak/qwen3-1.7b-stance-matrix-semeval-qlora is a 1.7 billion parameter Qwen3 model, fine-tuned using QLoRA with Unsloth and Huggingface's TRL library. This model is specifically optimized for tasks related to stance detection, likely stemming from its training on SemEval-related datasets. Its efficient training methodology allows for faster deployment and iteration in specialized NLP applications.

Loading preview...

Model Overview

This model, nityaak/qwen3-1.7b-stance-matrix-semeval-qlora, is a specialized variant of the Qwen3 architecture, featuring 1.7 billion parameters. It was developed by nityaak and fine-tuned using the QLoRA technique, leveraging the Unsloth library for accelerated training and Huggingface's TRL library.

Key Characteristics

  • Base Model: Fine-tuned from unsloth/Qwen3-1.7B-unsloth-bnb-4bit.
  • Training Efficiency: Utilizes Unsloth, enabling up to 2x faster training compared to standard methods.
  • Parameter Count: A compact 1.7 billion parameters, making it suitable for resource-constrained environments while maintaining specialized performance.
  • Context Length: Supports a substantial context length of 32768 tokens.

Primary Use Case

This model is specifically designed and fine-tuned for stance detection tasks, indicated by its name referencing "stance-matrix" and "SemEval". It is ideal for applications requiring the identification of opinions or positions towards a given target, likely benefiting from its specialized training on relevant datasets.