Vimax97/Florence-2-base-gpt4_captioner_v1

Model Card for Model ID

Model Details

Model Description

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.

• Developed by: [More Information Needed]
• Funded by [optional]: [More Information Needed]
• Shared by [optional]: [More Information Needed]
• Model type: [More Information Needed]
• Language(s) (NLP): [More Information Needed]
• License: [More Information Needed]
• Finetuned from model [optional]: [More Information Needed]

Uses

GPT4-O Style captioner, finetuned version using florence-2-base-ft

Direct Use

This model can be used to create gpt4-o styple captions.

Out-of-Scope Use

• This model might not generate long-text descriptions as the context length is 1024.
• Linear scaling is applied to increase the context length, its effect was not measured!

How to Get Started with the Model

# Load fine-tuned model and processor
import torch
from transformers import AutoModelForCausalLM, AutoProcessor
from PIL import Image

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
repo_name = "Vimax97/Florence-2-base-gpt4_captioner_v1" 
model = AutoModelForCausalLM.from_pretrained(repo_name, trust_remote_code=True).to(device)
processor = AutoProcessor.from_pretrained(repo_name, trust_remote_code=True)

# Inference


image = Image.open("<path_to_image>")
prompt = "<ImageCAP>" + 'What is the <GPT4> style description for this image?'

inputs = processor(text=prompt, images=image, return_tensors="pt").to(device)

generated_ids = model.generate(
    input_ids=inputs["input_ids"],
    pixel_values=inputs["pixel_values"],
    max_new_tokens=1024,
    do_sample=False,
    num_beams=3
)
generated_text = processor.batch_decode(generated_ids, skip_special_tokens=False)[0]
parsed_answer = processor.post_process_generation(generated_text, task=prompt, image_size=(image.width, image.height))

print("Generated: ",parsed_answer[prompt])

Training Hyperparameters

• Training regime: fp32 precision, 1000 images were used, 1 epoch of finetuning

Summary