Instructions to use LLMWildling/gemma-4-52B-no-adam-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LLMWildling/gemma-4-52B-no-adam-8bit with Transformers:

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

pipe = pipeline("text-generation", model="LLMWildling/gemma-4-52B-no-adam-8bit")
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, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("LLMWildling/gemma-4-52B-no-adam-8bit")
model = AutoModelForImageTextToText.from_pretrained("LLMWildling/gemma-4-52B-no-adam-8bit")
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

vLLM

How to use LLMWildling/gemma-4-52B-no-adam-8bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LLMWildling/gemma-4-52B-no-adam-8bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LLMWildling/gemma-4-52B-no-adam-8bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LLMWildling/gemma-4-52B-no-adam-8bit

SGLang

How to use LLMWildling/gemma-4-52B-no-adam-8bit 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 "LLMWildling/gemma-4-52B-no-adam-8bit" \
    --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": "LLMWildling/gemma-4-52B-no-adam-8bit",
		"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 "LLMWildling/gemma-4-52B-no-adam-8bit" \
        --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": "LLMWildling/gemma-4-52B-no-adam-8bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LLMWildling/gemma-4-52B-no-adam-8bit with Docker Model Runner:
```
docker model run hf.co/LLMWildling/gemma-4-52B-no-adam-8bit
```

gemma-4-52B-no-adam-8bit

gemma-4-52B-no-adam-8bit is a native MXFP4 sparse Mixture-of-Experts chat model for long-context assistant and coding workflows. It uses Gemma 4 compatible chat formatting, thinking mode, and OpenAI-compatible tool calling.

This checkpoint was produced without Adam 8-bit. The project already had an optimized LLM build path, and this checkpoint uses MXAR, a project-internal optimization method for the pre/post-training build loop. MXAR implementation details are not included in this release.

Speed

Prior optimized Adam 8-bit path: roughly 34 hours for a comparable 52B pass
MXAR path for this checkpoint: roughly 2.5 hours
Wall-clock speedup: about 13.6x
Time saved: about 31.5 hours
Wall-clock reduction: about 92.6%

Model Details

Architecture: Gemma4 sparse MoE
Expert count: 256
Active experts per token: 10
Weight format: native MXFP4 expert weights with BF16 shared weights
Quantization config: quant_method=mxfp4, quant_type=mxfp4, converter_layout=vllm_fused_moe
Context: up to 262k positions in config; 100k context is a practical serving target on a single high-memory GPU
Recommended runtime: vLLM with Gemma4 reasoning and tool-call parsers

Recommended Serving

vllm serve /path/to/gemma-4-52B-no-adam-8bit \
  --served-model-name gemma-4-52B-no-adam-8bit \
  --host 0.0.0.0 \
  --port 23333 \
  --dtype bfloat16 \
  --max-model-len 100000 \
  --gpu-memory-utilization 0.90 \
  --trust-remote-code \
  --reasoning-parser gemma4 \
  --tool-call-parser gemma4 \
  --enable-auto-tool-choice \
  --chat-template /path/to/gemma-4-52B-no-adam-8bit/chat_template.jinja \
  --default-chat-template-kwargs '{"enable_thinking": true}'

Use standard chat roles: system, user, and assistant. For the intended behavior profile, keep thinking mode enabled and use native tool-call APIs rather than parsing tool calls from raw text.

Generation

Start with temperature 0.0 to 0.7.
Use the provided chat template.
For tool use, pass tools through the OpenAI-compatible chat completions API.

Files

config.json
generation_config.json
tokenizer.json
tokenizer_config.json
chat_template.jinja
model-00001-of-00002.safetensors
model-00002-of-00002.safetensors
model.safetensors.index.json

Downloads last month: 21

Safetensors

Model size

29B params

Tensor type

BF16