Visual Document Retrieval
PEFT
Safetensors
diagram
structured-data
image-processing
knowledge-graph
json
Instructions to use zackriya/diagram2graph-adapters with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use zackriya/diagram2graph-adapters with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-VL-3B-Instruct") model = PeftModel.from_pretrained(base_model, "zackriya/diagram2graph-adapters") - Notebooks
- Google Colab
- Kaggle
test data
#1
by MingxuChai - opened
A meaningful piece of work. However, would it be possible to share some samples from the test set? I'm currently seeing suboptimal performance on my own dataset, and I'd like to understand whether this might be due to differences in the data.
Thanks for the comment.
You can checkout the fine-tuning script here, for more information.
As the ownership of the data goes to the organization, I can't provide it as you requested
Github fine-tuning script : https://github.com/Zackriya-Solutions/diagram2graph/blob/main/phase2/fine_tuning/finetuning.ipynb
Hope it help.
Thank you.