Instructions to use overthelex/qwen2.5-14b-edrsr-legal-uk with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use overthelex/qwen2.5-14b-edrsr-legal-uk with Transformers:

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

pipe = pipeline("text-generation", model="overthelex/qwen2.5-14b-edrsr-legal-uk")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("overthelex/qwen2.5-14b-edrsr-legal-uk")
model = AutoModelForCausalLM.from_pretrained("overthelex/qwen2.5-14b-edrsr-legal-uk", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use overthelex/qwen2.5-14b-edrsr-legal-uk with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "overthelex/qwen2.5-14b-edrsr-legal-uk"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "overthelex/qwen2.5-14b-edrsr-legal-uk",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/overthelex/qwen2.5-14b-edrsr-legal-uk

SGLang

How to use overthelex/qwen2.5-14b-edrsr-legal-uk 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 "overthelex/qwen2.5-14b-edrsr-legal-uk" \
    --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": "overthelex/qwen2.5-14b-edrsr-legal-uk",
		"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 "overthelex/qwen2.5-14b-edrsr-legal-uk" \
        --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": "overthelex/qwen2.5-14b-edrsr-legal-uk",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use overthelex/qwen2.5-14b-edrsr-legal-uk with Docker Model Runner:
```
docker model run hf.co/overthelex/qwen2.5-14b-edrsr-legal-uk
```

Qwen2.5-14B-EDRSR-Legal-UK

Ukrainian legal domain model obtained by continued pretraining (CPT) of Qwen/Qwen2.5-14B on the EDRSR corpus of Ukrainian court decisions.

Part of a scaling experiment (0.5B / 1.5B / 3B / 14B) for the PhD dissertation at Glushkov Institute of Cybernetics, NAS of Ukraine.

Training Data

Corpus: Unified State Register of Court Decisions of Ukraine (EDRSR)
Documents: 33.9M court decisions (after dedup + quality filtering from 38.5M)
Tokens: 161.4B tokens (Qwen2 BPE tokenizer, fertility = 0.515 for Ukrainian legal text)
Sequence length: 8,192 tokens
Shards: 1,233 pre-packaged numpy shards

Training Details

Hardware: 8x NVIDIA H100 SXM 80GB (NVIDIA Innovation Lab via Brev)
Framework: HuggingFace Trainer + DeepSpeed ZeRO-3
Precision: bfloat16
Global batch size: 128 sequences (1.05M tokens/step)
Total steps: 9,536 (10B tokens processed)
Training time: 44 hours
Throughput: 22K tokens/sec, 44.9 sec/step

Results

Metric	Value
Initial loss (step 10)	0.84
Final loss (step 9,536)	0.22
Loss reduction	-74%
Base perplexity	2.84
CPT perplexity	1.28
Perplexity reduction	-54.8%

Scaling Law

All four models in the series converge to similar perplexity after CPT:

Model	Base PPL	CPT PPL	Reduction
0.5B	6.83	1.35	-80%
1.5B	4.61	1.31	-72%
3B	3.83	1.30	-66%
14B	2.84	1.28	-55%

Intended Use

This is a base model (not instruction-tuned). It is intended for:

Research on domain adaptation of LLMs for low-resource legal languages
Downstream fine-tuning for Ukrainian legal NLP tasks
Scaling law analysis of continued pretraining
Perplexity evaluation on Ukrainian legal text

Limitations

Not instruction-tuned; will not follow instructions or chat
Trained on Ukrainian court decisions only; may not generalize to other legal systems

Related Resources

Downloads last month: 16

Safetensors

Model size

15B params

Tensor type

BF16

Model tree for overthelex/qwen2.5-14b-edrsr-legal-uk

Base model

Qwen/Qwen2.5-14B

Finetuned

(122)

this model

Papers for overthelex/qwen2.5-14b-edrsr-legal-uk

Temporal Decay of Co-Citation Predictability: A 20-Year Statute Retrieval Benchmark from 396M Ukrainian Court Citations

Paper • 2605.17639 • Published May 17

Tokenizer Fertility and Zero-Shot Performance of Foundation Models on Ukrainian Legal Text: A Comparative Study

Paper • 2605.14890 • Published May 14

Automatic Construction of a Legal Citation Graph from 100 Million Ukrainian Court Decisions: Large-Scale Extraction, Topological Analysis, and Ontology-Driven Clustering

Paper • 2605.15362 • Published May 14