Instructions to use janPaje/iolai-gemma4-hybrid with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use janPaje/iolai-gemma4-hybrid with Transformers:

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

pipe = pipeline("text-generation", model="janPaje/iolai-gemma4-hybrid")

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("janPaje/iolai-gemma4-hybrid")
model = AutoModelForMultimodalLM.from_pretrained("janPaje/iolai-gemma4-hybrid", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use janPaje/iolai-gemma4-hybrid with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "janPaje/iolai-gemma4-hybrid"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "janPaje/iolai-gemma4-hybrid",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/janPaje/iolai-gemma4-hybrid

SGLang

How to use janPaje/iolai-gemma4-hybrid 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 "janPaje/iolai-gemma4-hybrid" \
    --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": "janPaje/iolai-gemma4-hybrid",
		"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 "janPaje/iolai-gemma4-hybrid" \
        --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": "janPaje/iolai-gemma4-hybrid",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use janPaje/iolai-gemma4-hybrid with Docker Model Runner:
```
docker model run hf.co/janPaje/iolai-gemma4-hybrid
```

IOL-AI 2026 — gemma4:12b hybrid (symbolic numeral solver + budget-managed LLM)

Three-pass pipeline:

Symbolic pass — numeral_solver.py brute-forces numeral systems (base, word values, word-order convention) from the problem's own examples and answers text_to_num / num_to_text exactly when a consistent system exists; otherwise falls through to the LLM.
Baseline pass — a fast low-token answer for every remaining problem, with submission.csv atomically checkpointed after every row.
Upgrade pass — per-row time-sliced reasoning with per-task-type method prompts; a baseline answer is replaced only when the upgrade parses to the correct number of answers. Translation reasoning is deliberately capped low: on gemma4-12B, truncated reasoning plus a forced short answer scored measurably higher than completed reasoning.

Model: gemma4-12B (Apache 2.0), weights shipped in this repo, loaded 4-bit via bitsandbytes at startup (fp16 12B does not fit the 16 GB T4).

Upload checklist (before submitting)

Put the gemma4-12B weight files (safetensors + config + tokenizer) in the repo root so script.py loads them from ".".
Repo must be public at submission time.
Submit via the competition Space; the platform mounts the hidden test set at /tmp/data/test.csv and runs script.py.

Downloads last month: 63

Safetensors

Model size

12B params

Tensor type

BF16

Model tree for janPaje/iolai-gemma4-hybrid

Base model

google/gemma-4-12B

Finetuned

(62)

this model