Instructions to use 26B-Suite/Gemory-26B-A4B-SOMPOA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 26B-Suite/Gemory-26B-A4B-SOMPOA with Transformers:

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

pipe = pipeline("text-generation", model="26B-Suite/Gemory-26B-A4B-SOMPOA")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("26B-Suite/Gemory-26B-A4B-SOMPOA")
model = AutoModelForMultimodalLM.from_pretrained("26B-Suite/Gemory-26B-A4B-SOMPOA", 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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use 26B-Suite/Gemory-26B-A4B-SOMPOA with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "26B-Suite/Gemory-26B-A4B-SOMPOA"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "26B-Suite/Gemory-26B-A4B-SOMPOA",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/26B-Suite/Gemory-26B-A4B-SOMPOA

SGLang

How to use 26B-Suite/Gemory-26B-A4B-SOMPOA 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 "26B-Suite/Gemory-26B-A4B-SOMPOA" \
    --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": "26B-Suite/Gemory-26B-A4B-SOMPOA",
		"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 "26B-Suite/Gemory-26B-A4B-SOMPOA" \
        --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": "26B-Suite/Gemory-26B-A4B-SOMPOA",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use 26B-Suite/Gemory-26B-A4B-SOMPOA with Docker Model Runner:
```
docker model run hf.co/26B-Suite/Gemory-26B-A4B-SOMPOA
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

⚠️ Warning: This model can produce narratives and RP that contain violent and graphic erotic content. Adjust your system prompt accordingly, and use Gemma 4 chat template. Thinking vs Nothinking was not compared.

🧙‍♀️ Gemory 26B A4B SOMPOA

Fast. Precise. Dangerously imaginative.

A creative-writing and roleplay operative built to enter the scene, read the room, make the move, and keep going. Gemory carries 26B total parameters with roughly 4B active per token—a little femme fatale with very little interest in wasting motion.

Full ReadMe

https://huggingface.co/UltimateIntent/Gemory-26B-A4B-GGUF?not-for-all-audiences=true

Findings

This adapter, when merged into the default base model exhibits refusals, and when used with jailbreaks is non compliant and has early terminations.

The same Gemory adapter merged onto SOMPOA base exhibited considerably better results. There were no refusals. Instead of a 1 line refusal, it wrote a comprehensive guide as requested (using structured formatting). Even with the jailbreak enabled, it wrote a 5 paragraph guide instead of a 1 paragraph cop-out seen with the HuiHui QAT base.

I've released this LoRA-merge as its own model, Gemory-26B-A4B-SOMPOA, credits to @Richter03 for the idea.

Downloads last month: 204

Safetensors

Model size

26B params

Tensor type

BF16

Model tree for 26B-Suite/Gemory-26B-A4B-SOMPOA

MuXodious/gemma-4-26B-A4B-it-SOMPOA-heresy

UltimateIntent/Gemory-26B-A4B-LORA

Merge model

this model

Merges

1 model

Quantizations

2 models