OpenBuddy/OpenBuddy-R1-0528-Distill-Qwen3-32B-Preview0-QAT

TEXT GENERATIONConcurrency Cost:2Model Size:32BQuant:FP8Ctx Length:32kPublished:Jun 8, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

OpenBuddy/OpenBuddy-R1-0528-Distill-Qwen3-32B-Preview0-QAT is a 32 billion parameter multilingual chatbot developed by OpenBuddy, based on the Qwen3-32B architecture. This model is distilled from DeepSeek-R1-0528 and features a 32,768 token context length, optimized for conversational AI. It is designed for general-purpose chat applications requiring robust multilingual capabilities and adherence to a specific prompt format for optimal performance.

Loading preview...

OpenBuddy-R1-0528-Distill-Qwen3-32B-Preview0-QAT Overview

This model, developed by OpenBuddy, is a 32 billion parameter multilingual chatbot built upon the Qwen3-32B base model. It features a substantial context length of 32,768 tokens, making it suitable for extended conversations and complex interactions. A key characteristic of this model is its training methodology: it has been distilled from DeepSeek-R1-0528, suggesting a focus on efficiency and performance derived from a larger, capable source.

Key Features & Capabilities

  • Multilingual Chatbot: Designed for general-purpose conversational AI across multiple languages.
  • Qwen3-32B Base: Leverages the robust architecture of Qwen3-32B.
  • Distilled Training: Benefits from distillation from DeepSeek-R1-0528, potentially offering optimized performance.
  • Extended Context Window: Supports a 32,768 token context length, enabling longer and more coherent dialogues.
  • Specific Prompt Format: Utilizes a defined prompt structure (<|role|>system<|says|>...<|end|>) for consistent and effective interaction, with recommendations for transformers fast tokenizer.

Ideal Use Cases

This model is well-suited for developers building:

  • General-purpose chatbots requiring multilingual support.
  • Applications where long conversational memory is crucial.
  • Systems that can integrate with vllm for OpenAI-like API services due to its tokenizer_config.json definition.

Users should be aware of the inherent limitations and potential for erroneous or undesirable outputs, as outlined in the Apache 2.0 license and disclaimer.

Finetunes
4 models