# Multimodal Instruction Understanding

PixelReasoner is a vision-language model based on Qwen2.5-VL-7B-Instruct, trained with curiosity-driven reinforcement learning, focusing on image-text-to-text tasks.

Qwen2.5-VL-7B-Instruct is a multimodal model based on the Qwen2.5 architecture, supporting joint processing of images and text, suitable for vision-language tasks.

A 17B-parameter multilingual instruction model based on Llama-4, optimized with FP8 quantization to significantly reduce resource requirements

Qwen2.5-VL-32B-Instruct is a 32B-parameter-scale multimodal vision-language model that supports joint understanding and generation tasks for images and text.

Qwen2.5-VL-32B-Instruct is a 32B-parameter multimodal large language model supporting vision and language tasks, suitable for complex multimodal interaction scenarios.

Qwen2-VL-7B-Instruct is a vision-language model based on the Qwen2 architecture, supporting multimodal input of images and text, suitable for various visual-language tasks.

A fine-tuned version of the Qwen2-VL-7B model, trained using Unsloth and Huggingface's TRL library, achieving a 2x speed improvement.

Llama 3.2-11B vision-instruction model optimized with Unsloth, 4-bit quantized version, training speed increased by 2x

Llama 3.2 11B Vision Instruct is a multimodal model combining vision and language, supporting image-to-text tasks.

A multimodal model fine-tuned based on Qwen2.5-VL-7B-Instruct, utilizing the Unsloth acceleration framework and TRL library for training, achieving a 2x speed improvement

Pixtral-Large-Instruct-2411 is a multimodal instruction fine-tuned model based on MistralAI technology, supporting image and text input with multilingual processing capabilities.

Qwen2-VL-7B-Instruct is a 7B-parameter multimodal model supporting image-text interaction tasks.

This is a vision-language model based on the Qwen2-VL architecture with 7B parameters, supporting image understanding and instruction interaction.

This model is a vision-language-action model obtained by fine-tuning the OpenVLA 7B model using the LoRA method on the LIBERO-10 dataset, suitable for the field of robotics.

This is an OpenVLA 7B vision-language-action model fine-tuned using LoRA technology on the LIBERO-Goal dataset, suitable for the robotics field.