Singh8898/Diego

VISIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Nov 10, 2025License:apache-2.0Architecture:Transformer Open Weights Cold

Singh8898/Diego is an 8 billion parameter Qwen3-VL instruction-tuned causal language model developed by Singh8898. This model was finetuned using Unsloth and Huggingface's TRL library, enabling faster training. It is designed for general instruction-following tasks, leveraging its Qwen3-VL base for potential multimodal capabilities.

Loading preview...

Model Overview

Singh8898/Diego is an 8 billion parameter instruction-tuned model, developed by Singh8898. It is based on the Qwen3-VL architecture, suggesting potential for multimodal understanding, though specific capabilities are not detailed in the provided information. The model was finetuned from unsloth/qwen3-vl-8b-instruct-unsloth-bnb-4bit.

Training Details

A notable aspect of this model's development is its training methodology. It was finetuned using Unsloth, a library known for accelerating the training process of large language models, achieving a 2x speed improvement. The finetuning also leveraged Huggingface's TRL (Transformer Reinforcement Learning) library, which is commonly used for instruction tuning and alignment tasks.

Key Characteristics

  • Base Model: Qwen3-VL-8B-Instruct
  • Parameter Count: 8 billion
  • Training Efficiency: Finetuned 2x faster with Unsloth.
  • License: Apache-2.0

Potential Use Cases

Given its instruction-tuned nature and Qwen3-VL base, this model is likely suitable for a range of general-purpose natural language processing tasks, including question answering, text generation, and potentially multimodal applications if the visual capabilities of the base model are retained and optimized.