Instructions to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 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 erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

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 erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

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 erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Use Docker

docker model run hf.co/erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

LM Studio
Jan

vLLM

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Ollama
How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with Ollama:
```
ollama run hf.co/erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
```
Unsloth Desktop

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with Docker Model Runner:
```
docker model run hf.co/erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M
```

Lemonade

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Run and chat with the model

lemonade run user.DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

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 erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M

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 "erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1:Q4_K_M" \
  --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"

DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1

DeepSeek V4 Flash, served locally with a customized llama.cpp fork. The distinctive part of this package is the TM64 layout.

Build / runtime fork: https://github.com/erayyap/llama.cpp
Validated binary release: https://github.com/erayyap/llama.cpp/releases/tag/strix-halo-tm64-b54f16e1-3cacee17
Runtime source commit: 4fc4f230f6cc7735dc05cd403a8eb73ccd64813c

Credit: the base quants in this repo were not built by this uploader — they were downloaded from the upstream repos below. The uploader's contribution is the TM64 relayout of the IQ2_XXS expert tensors and the matching Vulkan kernels. The deployed main artifact additionally carries the locally requantized Q4_K attention tier (see deployment records).

File	Role	Source repo	Size
`DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1.gguf`	Main model (TM64-relayouted IQ2_XXS experts, Q4_K attention)	`antirez/deepseek-v4-gguf`	~79 GB
`dspark-DeepSeek-V4-Flash-0731-MXFP4-Q4_K_M-dense.gguf`	DSpark draft (selective-Q4 dense), speculative decoding	`Ralii/DeepSeek-V4-Flash-0731-DSpark-Selective-Q4_K_M-GGUF`	~10 GB

Upstream source repos

Main model: https://huggingface.co/antirez/deepseek-v4-gguf
- File: DeepSeek-V4-Flash-IQ2XXS-w2Q2K-AProjQ8-SExpQ8-OutQ8-chat-v2-imatrix-0731.gguf — an IQ2_XXS/Q2_K routed-expert, Q8 attention/shared/output quant built by antirez.
- The uploaded file is that quant with the IQ2_XXS expert tensors re-laid-out into the TM64 tile-major layout.
Draft: https://huggingface.co/Ralii/DeepSeek-V4-Flash-0731-DSpark-Selective-Q4_K_M-GGUF
- File: dspark-DeepSeek-V4-Flash-0731-MXFP4-Q4_K_M-dense.gguf — routed experts kept native MXFP4, eligible dense tensors selective-Q4_K_M, by Ralii.

This repo holds the two GGUF files used together for a full speculative-decode setup:

File	Role	Size
`DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1.gguf`	Main model (TM64-relayouted IQ2_XXS experts, Q4_K attention)	~79 GB
`dspark-DeepSeek-V4-Flash-0731-MXFP4-Q4_K_M-dense.gguf`	DSpark draft (selective-Q4 dense), used for speculative decoding	~10 GB

Download and run

Install the Hugging Face CLI and download both files:

hf download erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1 \
  DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1.gguf \
  dspark-DeepSeek-V4-Flash-0731-MXFP4-Q4_K_M-dense.gguf \
  --local-dir "$HOME/models/deepseek-v4-vulkan"

Download and extract the validated Strix Halo runtime from the GitHub release, then launch it with the packaged tuned configuration:

MODEL_PATH="$HOME/models/deepseek-v4-vulkan/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1.gguf" \
DRAFT_PATH="$HOME/models/deepseek-v4-vulkan/dspark-DeepSeek-V4-Flash-0731-MXFP4-Q4_K_M-dense.gguf" \
./bin/run-server.sh

The launcher defaults to F16 KV, F16 lightning-indexer cache, NP1, 409,600 context tokens per slot, batch/microbatch 3072, adaptive DSpark width up to five, and the validated Strix Halo Vulkan kernel controls. Use NP=2 or NP=3 for multiple slots; each NP count receives a separate state path.

What TM64 is

TM64 ("Tile-major BM64") is an alternate GGUF tensor layout for the IQ2_XXS expert tensors, plus the Vulkan kernels that consume it. It is a private serialized type, IQ2_XXS_TM64, so standard IQ2 kernels cannot read the alternate layout.

The 86 IQ2_XXS expert gate/up tensors are re-laid-out offline from [expert][M row][K/256 block] → [expert][M/64 tile][K/256 block][row within tile].
Complete 66-byte quantization blocks are permuted without changing any bits — only addressing changes.
This gives Vulkan direct BM64 (block-major, 64-row) addressing for wide F16-B MMID and widths 2–5 vector MMID, instead of forcing production-width batches through the width-512 BN16 specialization.
Adaptive small/medium tiles keep production-width batches out of the width-512 BN16 path.
Kernel repair (accepted patch): retain BM64/BN16 at width 512; BM64/BN32 at width 1024; BM128/BN64 at widths 2286/3072.

Correctness and speed

Exact standard↔TM64 Vulkan output hashes at widths 2, 3, 4, 5, and 512; the standard path passed CPU-reference comparison at all five widths, so the TM64 result is transitively exact.
All 86 inverse permutations reproduce their source tensor SHA-256; all non-target bytes are identical.
Cooled A/B/B/A kernel speedups: 1.24949× (512), 1.19271× (1024), 1.10918× (2286), 1.07336× (3072).
Whole-model median timing (resident model, A/B/B/A/B/A/A/B): prompt throughput +14.86% (76.0 → 87.3 tok/s), decode throughput +2.50% (19.97 → 20.47 tok/s). Output hashes and draft acceptance identical across all phases.

Platform caveat

TM64 is specific to AMD/RADV/RDNA3 (gfx1151 / device 0x1586) and the exact 4096x2048x256, top-6 expert tensors. It is an opt-in layout winner; on other devices standard IQ2_XXS kernels apply (or the file must be re-converted).

Quantization notes (beyond TM64)

Routed experts: IQ2_XXS / Q2_K
Attention projection tier: Q4_K
Dense/draft: MXFP4 + Q4_K_M (selective)
Main model context: up to 409,600 tokens per slot (multi-slot variants up to 1,228,800)
Draft spec: --spec-type draft-dspark --spec-draft-n-max 5

License

Original model license applies. See upstream model card.

Downloads last month: 645

GGUF

Model size

284B params

Architecture

deepseek4

Hardware compatibility

4-bit

View +1 variant

Model tree for erayyapagci/DeepSeek-V4-Flash-IQ2XXS-TM64-Q4KAttention-v1

Base model

deepseek-ai/DeepSeek-V4-Flash-0731

Quantized

(189)

this model