Instructions to use Aurelien33/aurelius-14b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Aurelien33/aurelius-14b with Transformers:

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

pipe = pipeline("text-generation", model="Aurelien33/aurelius-14b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Aurelien33/aurelius-14b")
model = AutoModelForCausalLM.from_pretrained("Aurelien33/aurelius-14b", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Aurelien33/aurelius-14b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Aurelien33/aurelius-14b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Aurelien33/aurelius-14b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Aurelien33/aurelius-14b

SGLang

How to use Aurelien33/aurelius-14b 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 "Aurelien33/aurelius-14b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Aurelien33/aurelius-14b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "Aurelien33/aurelius-14b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Aurelien33/aurelius-14b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Aurelien33/aurelius-14b with Docker Model Runner:
```
docker model run hf.co/Aurelien33/aurelius-14b
```

Aurelius-14B

A clean, verification-native code fine-tune of Qwen/Qwen3-14B. Supervised on execution-verified solutions distilled from Qwen3-Coder-30B-A3B-Instruct (Apache-2.0) — no closed-model outputs. Designed to be paired with a verifier-native best-of-N + repair serving layer.

Results

Setting	HumanEval pass@1	MBPP pass@1
Qwen3-14B (base)	88.4%	69.0%
Aurelius-14B (single-pass, greedy)	87.8%	71.5%
Aurelius-14B + best-of-8 + repair (serving)	96.3%	83.0%

Held-out HumanEval is the clean judge (training data is MBPP-distilled; the two sets are disjoint). Single-pass ≈ base — SFT here is safe, non-regressing polish, not a capability lift. The large gain is the serving layer, which cashes the model's selection gap (oracle@32 = 97.0%).

Honest methodology — what moved the needle, and what didn't

Lever	Effect
Capacity (8B → 14B base)	+4.3pp HumanEval / +6.5pp MBPP — the capability lever
SFT on verified strong-teacher traces	non-regressing polish (matches base)
best-of-N + repair (inference)	+8.5pp HumanEval / +10.5pp MBPP — the serving win
RLVR (GRPO) on the 8B	inert / within noise — recorded null
Flywheel self-distillation into greedy	regressed — recorded null

The negatives are reported deliberately: on this base, single-pass capability is bounded by the base model; the cheap, reliable gain lives in verifier-native inference, not in more post-training.

Provenance & license

Base: Qwen3-14B (Apache-2.0). Teacher: Qwen3-Coder-30B-A3B-Instruct (Apache-2.0).
Training data: best-of-N MBPP solutions, kept only if they pass the unit tests.
100% open-weight provenance — no Claude / GPT / closed-model outputs.
License: Apache-2.0 (inherits base + teacher).

Usage

Single-pass:

from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("Zephyrs33/aurelius-14b")
model = AutoModelForCausalLM.from_pretrained("Zephyrs33/aurelius-14b", device_map="auto")
msg = [{"role": "user", "content": "Write a Python function is_prime(n). Return ONLY the function in one ```python block."}]
ids = tok.apply_chat_template(msg, add_generation_prompt=True, return_tensors="pt").to(model.device)
print(tok.decode(model.generate(ids, max_new_tokens=512)[0][ids.shape[1]:], skip_special_tokens=True))

For the best-of-N + repair serving layer (the 96.3% / 83.0% numbers), use solve(problem, tests) — sample N, execute against a verifier, return a passing solution; repair on failure.

Limitations

Code-focused (MBPP / HumanEval distribution). Single-pass capability ≈ the base; the serving lift requires a verifier (your tests / sample tests / generated tests — never hidden eval tests). Not safety-tuned beyond the base.

Downloads last month: 1

Safetensors

Model size

15B params

Tensor type

BF16

Model tree for Aurelien33/aurelius-14b

Base model

Qwen/Qwen3-14B-Base

Finetuned

Qwen/Qwen3-14B

Finetuned

(322)

this model