karakuri-vl-32b-instruct-2507

KARAKURI VL 32B Instruct 2507

Model Details

Model Description

• Developed by: KARAKURI Inc.
• Model type: Vision-Language Models
• Languages: Japanese and English
• License: Apache 2.0
• Finetuned from model: Qwen/Qwen2.5-VL-32B-Instruct
• Contact: For questions and comments about the model, please email [email protected]
• Demo: https://vl.karakuri.cc/

Usage

Use in 🤗 Transformers

First, install the required dependencies:

pip install transformers accelerate qwen-vl-utils[decord]==0.0.8

Then, use the following code to load the model and generate responses:

from transformers import AutoModelForImageTextToText, AutoProcessor
from qwen_vl_utils import process_vision_info

model_name = "karakuri-ai/karakuri-vl-32b-instruct-2507"
model = AutoModelForImageTextToText.from_pretrained(
    model_name, torch_dtype="auto", device_map="auto"
)
processor = AutoProcessor.from_pretrained(model_name)

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Describe this image."},
        ],
    }
]

# Preparation for inference
text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
)
inputs = inputs.to(model.device)

# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
    out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)

Training Details

Training Infrastructure

• Hardware: The model was trained on 20 nodes of an Amazon EC2 trn1.32xlarge instance.
• Software: We use code based on neuronx-nemo-megatron.

Acknowledgments

This work was supported by the Ministry of Economy, Trade and Industry (METI) and the New Energy and Industrial Technology Development Organization (NEDO) through the Generative AI Accelerator Challenge (GENIAC).

Citation

@misc{karakuri_vl_32b_instruct_2507,
    author       = { {KARAKURI} {Inc.} },
    title        = { {KARAKURI} {VL} 32{B} {Instruct} 2507 },
    year         = { 2025 },
    url          = { https://huggingface.co/karakuri-ai/karakuri-vl-32b-instruct-2507 },
    publisher    = { {Hugging Face} },
    journal      = { {Hugging Face} repository }
}