Instructions to use benjamin/gpt2-large-wechsel-ukrainian with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use benjamin/gpt2-large-wechsel-ukrainian with Transformers:

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

pipe = pipeline("text-generation", model="benjamin/gpt2-large-wechsel-ukrainian")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("benjamin/gpt2-large-wechsel-ukrainian")
model = AutoModelForCausalLM.from_pretrained("benjamin/gpt2-large-wechsel-ukrainian", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use benjamin/gpt2-large-wechsel-ukrainian with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "benjamin/gpt2-large-wechsel-ukrainian"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "benjamin/gpt2-large-wechsel-ukrainian",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/benjamin/gpt2-large-wechsel-ukrainian

SGLang

How to use benjamin/gpt2-large-wechsel-ukrainian 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 "benjamin/gpt2-large-wechsel-ukrainian" \
    --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": "benjamin/gpt2-large-wechsel-ukrainian",
		"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 "benjamin/gpt2-large-wechsel-ukrainian" \
        --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": "benjamin/gpt2-large-wechsel-ukrainian",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use benjamin/gpt2-large-wechsel-ukrainian with Docker Model Runner:
```
docker model run hf.co/benjamin/gpt2-large-wechsel-ukrainian
```

Could you please share the training code and information on the dataset you've used to train this one?

by dchaplinsky - opened May 8, 2023

Discussion

dchaplinsky

May 8, 2023

I'm keen to try to train more ukrainian models using wechsel on the corpus I recently released: https://aclanthology.org/2023.unlp-1.1/ and other data that I have at my disposal.

benjamin

Owner May 8, 2023

That's great!

I just made the code public for you: https://github.com/bminixhofer/ukrainian-wechsel-models
You can find the data and model preparation scripts + configs there. The run_clm.py and run_mlm.py scripts are iirc an unchanged copy of the Huggingface scripts from some time ago.
The training runs are here: https://wandb.ai/bminixhofer/ukrainian-nlp

Hope that helps!

dchaplinsky

May 9, 2023

Thanks for the prompt response, reading the source code now.

So glad you've been using lang-uk NER that we've created and open sourced :)

dchaplinsky changed discussion status to closed May 9, 2023

dchaplinsky

May 9, 2023

•

edited May 9, 2023

Oh, finally HF is back to normal.

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