Instructions to use pstroe/bullinger-general-model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use pstroe/bullinger-general-model with Transformers:

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

pipe = pipeline("image-text-to-text", model="pstroe/bullinger-general-model")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("pstroe/bullinger-general-model")
model = AutoModelForMultimodalLM.from_pretrained("pstroe/bullinger-general-model", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use pstroe/bullinger-general-model with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/pstroe/bullinger-general-model

SGLang

How to use pstroe/bullinger-general-model 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 "pstroe/bullinger-general-model" \
    --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": "pstroe/bullinger-general-model",
		"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 "pstroe/bullinger-general-model" \
        --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": "pstroe/bullinger-general-model",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use pstroe/bullinger-general-model with Docker Model Runner:
```
docker model run hf.co/pstroe/bullinger-general-model
```

bullinger-general-model / README.md

pstroe

Update README.md

be32c88 about 3 years ago

preview code

Raw History Blame Contribute Delete

1.29 kB

metadata

license: cc-by-sa-4.0
language:
  - la
  - de
metrics:
  - cer
tags:
  - HTR
  - OCR
  - handwriting
  - 16 century
  - correspondence
  - Bullinger
  - reformation

General HTR Model trained on Bullinger Correspondence

This is a handwritten text recognition model based on TrOCR, fine-tuned with the Bullinger dataset. Please see the preceding link for a detailed description of the dataset. For a detailed description of the model, please see the publication below (especially the thesis).

Publications

More detail can be found in the following publications (please cite these two if you use the model for your experiments):

Tobias Hodel, Phillip Benjamin Ströbel, Andreas Fischer, Anna Scius-Bertrand, Anna Janka, Jonas Widmer, Beat Wolf, Patricia Scheurer, Martin Volk. 2023. Bullingers Briefwechsel zugänglich machen: Stand der Handschriftenerkennung. In DHd2023: Open Humanities, Open Culture. Konferenzabstracts, Luxembourg/Trier, Luxembourg/Germany. 9. Jahrestagung des Verbands Digital Humanities im deutschsprachigen Raum e.V.
Phillip Benjamin Ströbel. 2023. Flexible Techniques for Automatic Text Recognition of Historical Documents. PhD Thesis, Zürich. Link