artemds45/itmo-nlp-hw6-qwen2-5-0-5b-abliterated

TEXT GENERATIONConcurrency Cost:1Model Size:0.5BQuant:BF16Ctx Length:32kPublished:Apr 28, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The artemds45/itmo-nlp-hw6-qwen2-5-0-5b-abliterated model is a 0.5 billion parameter language model, derived from Qwen2.5-0.5B-Instruct. It features a 32768 token context length and includes a DPO LoRA adapter. This model is specifically prepared for ITMO NLP+DL homework 6, focusing on instruction-following tasks.

Loading preview...

Model Overview

This model, artemds45/itmo-nlp-hw6-qwen2-5-0-5b-abliterated, is a specialized version of the Qwen2.5-0.5B-Instruct architecture, featuring 0.5 billion parameters and a substantial 32768 token context window. It has been specifically adapted and prepared for the ITMO NLP+DL homework 6 curriculum.

Key Characteristics

  • Base Model: Derived from the Qwen2.5-0.5B-Instruct series.
  • Parameter Count: 0.5 billion parameters, offering a compact yet capable language model.
  • Context Length: Supports a 32768 token context, allowing for processing of longer inputs.
  • DPO LoRA Adapter: Includes a pre-packaged DPO (Direct Preference Optimization) LoRA adapter, located in the dpo_lora_adapter/ subfolder, indicating its potential for fine-tuning or preference alignment tasks.

Intended Use

  • Educational Purposes: Primarily designed and prepared for use in the ITMO NLP+DL homework 6, suggesting its utility for academic exercises and research in natural language processing and deep learning.
  • Instruction Following: As it is based on an "Instruct" model, it is inherently suited for tasks requiring adherence to given instructions.