Instructions to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF",
	filename="DeepSeek-V4-Flash-REAP-K128.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF 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 sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
# Run inference directly in the terminal:
llama cli -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
# Run inference directly in the terminal:
llama cli -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

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 sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
# Run inference directly in the terminal:
./llama-cli -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

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 sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
# Run inference directly in the terminal:
./build/bin/llama-cli -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Use Docker

docker model run hf.co/sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

LM Studio
Jan

vLLM

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Ollama
How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with Ollama:
```
ollama run hf.co/sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
```

Unsloth Studio

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF 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 sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF 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 sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF to start chatting

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Docker Model Runner
How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with Docker Model Runner:
```
docker model run hf.co/sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF
```

Lemonade

How to use sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull sleepyeldrazi/deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF

Run and chat with the model

lemonade run user.deepseek-v4-flash-reap-k128-Q2-Q4-Mixed-GGUF-{{QUANT_TAG}}

List all available models

lemonade list

sleepyeldrazi commited on 19 days ago

Commit

c062d7a

verified ·

1 Parent(s): 0c47a3c

Upload DeepSeek-V4-Flash-REAP-K128 Q2/Q4 Mixed GGUF

Browse files

Files changed (4) hide show

.gitattributes +1 -0
DeepSeek-V4-Flash-REAP-K128.gguf +3 -0
DeepSeek-V4-Flash-REAP-K128.gguf.reap_plan.json +0 -0
README.md +94 -0

.gitattributes CHANGED Viewed

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text

 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+DeepSeek-V4-Flash-REAP-K128.gguf filter=lfs diff=lfs merge=lfs -text

DeepSeek-V4-Flash-REAP-K128.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:b46812c007a43dd8c2888fc51cd4fa7b331785f89ef6866fb0389c478e7d54e6
+size 55875180192

DeepSeek-V4-Flash-REAP-K128.gguf.reap_plan.json ADDED Viewed

The diff for this file is too large to render. See raw diff

README.md ADDED Viewed

	@@ -0,0 +1,94 @@

+---
+license: mit
+tags:
+- deepseek
+- deepseek-v4
+- deepseek-v4-flash
+- mixture-of-experts
+- reap
+- expert-pruning
+- gguf
+- ds4
+- experimental
+- dgx-spark
+pipeline_tag: text-generation
+---
+# DeepSeek V4 Flash — REAP K128 (Mixed Precision)
+REAP-pruned DeepSeek V4 Flash at **K128** (128 routed experts per MoE layer).
+Prunes 50% of routed experts via Cerebras REAP (Router-weighted Expert Activation Pruning),
+preserving all attention, embeddings, shared experts, router, and MTP components.
+## At a Glance
+| | |
+|---|---|
+| **Base model** | DeepSeek V4 Flash |
+| **Donor GGUF** | antirez Layers37-42Q4KExperts (90.9 GiB) |
+| **Pruning method** | REAP (Cerebras Research) |
+| **Routed experts** | 128 per layer (down from 256) |
+| **Kept slots** | 5,888 / 11,008 |
+| **Hash-preserved** | Layers 0-2 (256 experts each) |
+| **Pruned** | Layers 3-42 (128 experts each) |
+| **Format** | ds4-compact-v1 GGUF |
+| **File size** | 52.04 GiB |
+| **Mixed precision** | Layers 3-36: Q2_K/IQ2_XXS · Layers 37-42: Q4_K |
+> **Note:** This variant preserves the donor's mixed quantization (Q4_K in layers 37-42,
+> IQ2_XXS/Q2_K elsewhere). Requires a ds4 runtime that handles per-layer quant type dispatch.
+> If you encounter "expected IQ2_XXS expert tensors" errors, use the
+> [uniform variant](https://huggingface.co/sleepyeldrazi/deepseek-v4-flash-reap-k128-uniform).
+## Domain Split (Calibration)
+8,000 prompts · 5.0M tokens · 1.3B routed expert observations
+| Domain | Share |
+|---|---|
+| Coding & development | 35-40% |
+| Agentic tool-calling | 16% |
+| Research & knowledge | 15-20% |
+| Math & science | 10-15% |
+| Design & planning | 5-10% |
+| Trivia & general QA | 3-5% |
+Calibration used the REAP `activation_energy_sum2` score metric with 4,096 token context per prompt.
+Top-to-bottom expert score gap in layer 3: 2,200x (strong pruning signal).
+## How to Run
+Requires **eouya2/ds4-for-reaped** (ds4 engine with compact GGUF support):
+```bash
+git clone https://github.com/eouya2/ds4-for-reaped
+cd ds4-for-reaped
+make cuda-spark -j$(nproc)  # DGX Spark / CUDA
+# or: make                    # Metal / macOS
+./ds4 --cuda -m DeepSeek-V4-Flash-REAP-K128.gguf --ctx 131072
+```
+API server mode:
+```bash
+./ds4-server --cuda -m DeepSeek-V4-Flash-REAP-K128.gguf \
+  --host 0.0.0.0 --port 17777 --ctx 131072
+```
+## How It Was Built
+1. **Donor GGUF**: Downloaded antirez Layers37-42Q4KExperts variant (90.9 GiB) — Q4_K experts in deep layers, IQ2_XXS/Q2_K elsewhere
+2. **Calibration**: 8,000 prompts collected and run through ds4's imatrix collector on a DGX Spark (NVIDIA GB10) at 4,096 token context
+3. **REAP scoring**: Imatrix activation data converted to per-expert REAP scores using `activation_energy_sum2`
+4. **Pruning**: 50% expert removal via `ds4_prune_gguf.py` from eouya2/reap-for-ds4. Layers 0-2 (hash-routed) preserved. Expert tensors copied byte-for-byte — no dequant/requant.
+5. **Output**: `ds4-compact-v1` GGUF
+**No fine-tuning.** Purely structural expert removal. Weights are unmodified — a subset of the original experts.
+## Acknowledgments
+- **DeepSeek** — [DeepSeek V4 Flash](https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash) base model
+- **antirez** — [ds4 engine](https://github.com/antirez/ds4) and [GGUF quants](https://huggingface.co/antirez/deepseek-v4-gguf)
+- **eouya2** — [ds4-for-reaped](https://github.com/eouya2/ds4-for-reaped) and [reap-for-ds4](https://github.com/eouya2/reap-for-ds4)
+- **Cerebras Research** — [REAP](https://arxiv.org/abs/2510.13999) ([code](https://github.com/CerebrasResearch/reap))
+- **NVIDIA** — DGX Spark hardware