Qwen2-VL-7B-Captioner-Relaxed

Introduction

Qwen2-VL-7B-Captioner-Relaxed is an instruction-tuned version of Qwen2-VL-7B-Instruct, an advanced multimodal large language model. This fine-tuned version is based on a hand-curated dataset for text-to-image models, providing significantly more detailed descriptions of given images.

Key Features:

• Enhanced Detail: Generates more comprehensive and nuanced image descriptions.
• Relaxed Constraints: Offers less restrictive image descriptions compared to the base model.
• Natural Language Output: Describes different subjects in the image while specifying their locations using natural language.
• Optimized for Image Generation: Produces captions in formats compatible with state-of-the-art text-to-image generation models.

Note: This fine-tuned model is optimized for creating text-to-image datasets. As a result, performance on other tasks (e.g., ~10% decrease on mmmu_val) may be lower compared to the original model.

Requirements

If you encounter errors such as KeyError: 'qwen2_vl' or ImportError: cannot import name 'Qwen2VLForConditionalGeneration' from 'transformers', try installing the latest version of the transformers library from source:

pip install git+https://github.com/huggingface/transformers

Quickstart

from PIL import Image
from transformers import Qwen2VLForConditionalGeneration, AutoProcessor
from transformers import BitsAndBytesConfig
import torch

model_id = "Ertugrul/Qwen2-VL-7B-Captioner-Relaxed"

model = Qwen2VLForConditionalGeneration.from_pretrained(
    model_id, torch_dtype=torch.bfloat16, device_map="auto"
)
processor = AutoProcessor.from_pretrained(model_id)

conversation = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
            },
            {"type": "text", "text": "Describe this image."},
        ],
    }
]



image = Image.open(r"PATH_TO_YOUR_IMAGE")

# you can resize the image here if it's not fitting to vram, or set model max sizes.
# image = image.resize((1024, 1024)) # like this

text_prompt = processor.apply_chat_template(conversation, add_generation_prompt=True)

inputs = processor(
    text=[text_prompt], images=[image], padding=True, return_tensors="pt"
)
inputs = inputs.to("cuda")

with torch.no_grad():
    with torch.autocast(device_type="cuda", dtype=torch.bfloat16):
        output_ids  = model.generate(**inputs, max_new_tokens=384, do_sample=True, temperature=0.7, use_cache=True, top_k=50)


generated_ids = [
    output_ids[len(input_ids) :]
    for input_ids, output_ids in zip(inputs.input_ids, output_ids)
]
output_text = processor.batch_decode(
    generated_ids, skip_special_tokens=True, clean_up_tokenization_spaces=True
)[0]
print(output_text)

For more detailed options, refer to the Qwen2-VL-7B-Instruct documentation.