Instructions to use win10/gemma-4-31B-K1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use win10/gemma-4-31B-K1 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("coder3101/gemma-4-31B-it-heretic")
model = PeftModel.from_pretrained(base_model, "win10/gemma-4-31B-K1")

Transformers

How to use win10/gemma-4-31B-K1 with Transformers:

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

pipe = pipeline("text-generation", model="win10/gemma-4-31B-K1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("win10/gemma-4-31B-K1", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use win10/gemma-4-31B-K1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "win10/gemma-4-31B-K1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "win10/gemma-4-31B-K1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/win10/gemma-4-31B-K1

SGLang

How to use win10/gemma-4-31B-K1 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 "win10/gemma-4-31B-K1" \
    --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": "win10/gemma-4-31B-K1",
		"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 "win10/gemma-4-31B-K1" \
        --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": "win10/gemma-4-31B-K1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use win10/gemma-4-31B-K1 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for win10/gemma-4-31B-K1 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for win10/gemma-4-31B-K1 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for win10/gemma-4-31B-K1 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="win10/gemma-4-31B-K1",
    max_seq_length=2048,
)

Docker Model Runner
How to use win10/gemma-4-31B-K1 with Docker Model Runner:
```
docker model run hf.co/win10/gemma-4-31B-K1
```

Model Card for gemma-4-31B-K1

This is an experimental fine-tuning using a lot of iannicity/KIMI-K2.5-1000000x and iannicity/Hunter-Alpha-SFT.

If you enjoy my work, feel free to support me on Ko-fi with a coffee.

Every bit of your support directly helps me keep creating and spend more time making even better work:

https://ko-fi.com/ogodwin10

Quick start

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="None", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Training procedure

This model was trained with SFT.

Framework versions

PEFT 0.18.1
TRL: 0.24.0
Transformers: 5.6.0.dev0
Pytorch: 2.11.0+cu128
Datasets: 4.3.0
Tokenizers: 0.22.2

Downloads last month: 4

Model tree for win10/gemma-4-31B-K1

Base model

google/gemma-4-31B

Finetuned

google/gemma-4-31B-it

Finetuned

coder3101/gemma-4-31B-it-heretic

Adapter

(1)

this model