Instructions to use tssst/Nemesia-Gemma3-12B-v2.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tssst/Nemesia-Gemma3-12B-v2.0 with Transformers:

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

pipe = pipeline("image-text-to-text", model="tssst/Nemesia-Gemma3-12B-v2.0")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("tssst/Nemesia-Gemma3-12B-v2.0")
model = AutoModelForMultimodalLM.from_pretrained("tssst/Nemesia-Gemma3-12B-v2.0", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use tssst/Nemesia-Gemma3-12B-v2.0 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tssst/Nemesia-Gemma3-12B-v2.0:F16
# Run inference directly in the terminal:
llama cli -hf tssst/Nemesia-Gemma3-12B-v2.0:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tssst/Nemesia-Gemma3-12B-v2.0:F16
# Run inference directly in the terminal:
llama cli -hf tssst/Nemesia-Gemma3-12B-v2.0:F16

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf tssst/Nemesia-Gemma3-12B-v2.0:F16
# Run inference directly in the terminal:
./llama-cli -hf tssst/Nemesia-Gemma3-12B-v2.0:F16

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf tssst/Nemesia-Gemma3-12B-v2.0:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf tssst/Nemesia-Gemma3-12B-v2.0:F16

Use Docker

docker model run hf.co/tssst/Nemesia-Gemma3-12B-v2.0:F16

LM Studio
Jan

vLLM

How to use tssst/Nemesia-Gemma3-12B-v2.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tssst/Nemesia-Gemma3-12B-v2.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tssst/Nemesia-Gemma3-12B-v2.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/tssst/Nemesia-Gemma3-12B-v2.0:F16

SGLang

How to use tssst/Nemesia-Gemma3-12B-v2.0 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 "tssst/Nemesia-Gemma3-12B-v2.0" \
    --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": "tssst/Nemesia-Gemma3-12B-v2.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "tssst/Nemesia-Gemma3-12B-v2.0" \
        --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": "tssst/Nemesia-Gemma3-12B-v2.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Ollama
How to use tssst/Nemesia-Gemma3-12B-v2.0 with Ollama:
```
ollama run hf.co/tssst/Nemesia-Gemma3-12B-v2.0:F16
```

Unsloth Studio

How to use tssst/Nemesia-Gemma3-12B-v2.0 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 tssst/Nemesia-Gemma3-12B-v2.0 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 tssst/Nemesia-Gemma3-12B-v2.0 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for tssst/Nemesia-Gemma3-12B-v2.0 to start chatting

Docker Model Runner
How to use tssst/Nemesia-Gemma3-12B-v2.0 with Docker Model Runner:
```
docker model run hf.co/tssst/Nemesia-Gemma3-12B-v2.0:F16
```

Lemonade

How to use tssst/Nemesia-Gemma3-12B-v2.0 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tssst/Nemesia-Gemma3-12B-v2.0:F16

Run and chat with the model

lemonade run user.Nemesia-Gemma3-12B-v2.0-F16

List all available models

lemonade list

Atomic Chat
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

temp

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the NuSLERP merge method using mlabonne/gemma-3-12b-it-abliterated as a base.

Well, it's been a while, hasn't it? It's only been a short 5 months since I posted the last Nemesia version, saying the famous last words

I will try swapping out a model or two in the merge and trying again to upload as a v2.0.

Then, exams and university stuff crushed me whole which wasn't great, and I didn't have any time or computational resources to get back to merging on a reasonable timeframe until now.

Honestly, in the time since, Qwen2.5-7B really has not turned out to be all that interesting, Mistral Nemo stealing most of its thunder. So, I decided to update my Nemesia mergeset with a new base: Gemma3-12B! In my experience, it has been alright, so I'm excited to tinker with it.

In my testing, this thing is alright- I couldn't run any GGUFs of it since I don't even know what on earth has happened to my .venv in the 5 months past, I'm pretty sure my Transformers installation is corrupted, and I'm re-cloning llama.cpp as I write this. I ran the FP16 and it seemed coherent enough, but I leave that to you to decide.

Uses all of the special NuSLERP options because they're there, and is based on the abliterated version of Gemma3 instead of the base because I <3 mlabonne.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: mlabonne/gemma-3-12b-it-abliterated
    parameters:
      weight: 1.0
  - model: qiuxi337/gemma-3-12b-it-grpo
    parameters:
      weight: 0.5
  - model: google/gemma-3-12b-it
    parameters:
      weight: 0.2
merge_method: nuslerp
base_model: mlabonne/gemma-3-12b-it-abliterated
parameters:
  normalize: true
  int8_mask: true
  nuslerp_flatten: false
  nuslerp_row_wise: true
dtype: float16