Instructions to use daitavan/donut-pretrain-german with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use daitavan/donut-pretrain-german with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="daitavan/donut-pretrain-german", device_map="auto")

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("daitavan/donut-pretrain-german")
model = AutoModelForMultimodalLM.from_pretrained("daitavan/donut-pretrain-german", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use daitavan/donut-pretrain-german with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "daitavan/donut-pretrain-german"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "daitavan/donut-pretrain-german",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/daitavan/donut-pretrain-german

SGLang

How to use daitavan/donut-pretrain-german with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "daitavan/donut-pretrain-german" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "daitavan/donut-pretrain-german",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "daitavan/donut-pretrain-german" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "daitavan/donut-pretrain-german",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use daitavan/donut-pretrain-german with Docker Model Runner:
```
docker model run hf.co/daitavan/donut-pretrain-german
```

State of the training

by Hugeldugelking - opened Jun 15, 2023

Discussion

Hugeldugelking

Jun 15, 2023

Hi, maybe this is a newbie question, but what was this model trained/what’s the state of the training?
As i read from the paper of the original donut model, the base model was trained with synthdog documents. i assume you did the same with the dataset on your account? is the training of this model completed?

daitavan

Owner Jun 16, 2023

I try pretraining the donut model on a german language, but i am not sure about the accuracy of my pretrain model. You can use it if you want. This model is pretraining with the same dataset on my account

Hugeldugelking

Jun 16, 2023

Can you tell me how you trained it, the official repository just shows how to finetune the pretrained model, but not how the pretrain model gets created.
Anyway, thanks for the work, I'll give it a try 😊

daitavan

Owner Jun 22, 2023

You create dataset by synthendog then train as guidlines

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment