Model Overview

The qgallouedec/online-dpo-qwen2-2 is a 0.5 billion parameter language model derived from the Qwen2-0.5B-Instruct architecture. It has been further refined through fine-tuning on the trl-lib/ultrafeedback-prompt dataset.

Key Capabilities

• Instruction Following: Enhanced ability to understand and execute instructions due to fine-tuning on a comprehensive feedback dataset.
• Context Handling: Features a notable context length of 131072 tokens, allowing it to process and generate responses based on very long inputs.
• Qwen2 Base: Inherits the foundational capabilities and architectural strengths of the Qwen2 family of models.

Good For

• Applications requiring a compact yet capable model for instruction-based tasks.
• Scenarios where processing extensive input contexts is crucial.
• Developing conversational agents or systems that benefit from improved instruction adherence.