Instructions to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="anemll/DeepSeek-V4-Flash-FP4-FP8-SSD",
	filename="dense/model-dense.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
# Run inference directly in the terminal:
llama-cli -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
# Run inference directly in the terminal:
llama-cli -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

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 anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
# Run inference directly in the terminal:
./llama-cli -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

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 anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
# Run inference directly in the terminal:
./build/bin/llama-cli -hf anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

Use Docker

docker model run hf.co/anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

LM Studio
Jan

vLLM

How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "anemll/DeepSeek-V4-Flash-FP4-FP8-SSD"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "anemll/DeepSeek-V4-Flash-FP4-FP8-SSD",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

Ollama
How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with Ollama:
```
ollama run hf.co/anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
```

Unsloth Studio new

How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD 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 anemll/DeepSeek-V4-Flash-FP4-FP8-SSD 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 anemll/DeepSeek-V4-Flash-FP4-FP8-SSD to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for anemll/DeepSeek-V4-Flash-FP4-FP8-SSD to start chatting

Docker Model Runner
How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with Docker Model Runner:
```
docker model run hf.co/anemll/DeepSeek-V4-Flash-FP4-FP8-SSD
```

Lemonade

How to use anemll/DeepSeek-V4-Flash-FP4-FP8-SSD with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull anemll/DeepSeek-V4-Flash-FP4-FP8-SSD

Run and chat with the model

lemonade run user.DeepSeek-V4-Flash-FP4-FP8-SSD-{{QUANT_TAG}}

List all available models

lemonade list

DeepSeek-V4-Flash-FP4-FP8-SSD / README.md

anemll

Add files using upload-large-folder tool

3aa5c60 verified 14 days ago

preview code

raw

history blame contribute delete

1.71 kB

	---
	license: mit
	base_model:
	- deepseek-ai/DeepSeek-V4-Flash
	library_name: llama.cpp
	pipeline_tag: text-generation
	tags:
	- gguf
	- deepseek-v4
	- deepseek-v4-flash
	- flash-moe
	- slot-bank
	- ssd
	- fp8
	- fp4
	- mxfp4
	- metal
	---

	# DeepSeek V4 Flash FP4/FP8 SSD Flash-MoE Package

	This repository contains an SSD Flash-MoE package for DeepSeek V4 Flash.
	It is intended for runtimes that can load a dense GGUF plus a routed expert
	sidecar.

	## Quantization

	- Dense/shared tensors: native DeepSeek FP8, represented as `F8_E4M3_B128` in GGUF.
	- Routed MoE expert tensors: native DeepSeek FP4, represented as `MXFP4` in the sidecar manifest.
	- Embeddings, output, norms, routing metadata, and IDs may remain `BF16`, `F32`, or `I32` where appropriate.

	The routed expert tensors are not stored in the dense GGUF. They are stored in
	the sidecar as layer-major binary banks.

	## Files

	```text
	dense/
	model-dense.gguf
	flashmoe-package.json

	sidecar/
	manifest.json
	layer_000.bin
	...
	layer_042.bin
	```

	## Model Details

	- Architecture: `deepseek4`
	- Blocks: `43`
	- Experts: `256`
	- Active experts per token: `6`
	- Context length: `1048576`
	- Dense GGUF tensors: `1199`
	- Routed expert sidecar entries: `129`

	## Example

	```bash
	./build/bin/llama-cli \
	-m dense/model-dense.gguf \
	--moe-mode slot-bank \
	--moe-sidecar sidecar \
	--moe-slot-bank 96 \
	--moe-topk 6 \
	-ngl 999 \
	--moe-cache-io-split 4 \
	-c 8192 \
	-b 128 \
	-ub 1 \
	--no-warmup \
	-p "What is Apple Neural Engine?" \
	-n 256
	```

	This package is not a standalone dense-only GGUF. Use a Flash-MoE aware
	llama.cpp build that supports DeepSeek V4 Flash, slot-bank mode, and the
	native FP8/FP4 tensor types.