Instructions to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF",
	filename="Qwopus3.6-27B-Coder-MTP-IQ3_K.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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
# Run inference directly in the terminal:
llama-cli -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
# Run inference directly in the terminal:
llama-cli -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
# Run inference directly in the terminal:
./llama-cli -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Use Docker

docker model run hf.co/AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

LM Studio
Jan

vLLM

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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": "AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Ollama
How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with Ollama:
```
ollama run hf.co/AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
```

Unsloth Studio

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF to start chatting

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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": "AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-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 AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with Docker Model Runner:
```
docker model run hf.co/AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF
```

Lemonade

How to use AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Run and chat with the model

lemonade run user.Qwopus3.6-27B-Coder-ik-MTP-GGUF-{{QUANT_TAG}}

List all available models

lemonade list

Qwopus3.6-27B-Coder — ik_llama.cpp MTP IQ-quants (GGUF)

This repo contains ik_llama.cpp-optimized IQ-series GGUF quantizations (with importance matrix) of Jackrong's excellent Qwopus3.6-27B-Coder-MTP, built specifically to run fast on a single RTX 3090 with Multi-Token Prediction (MTP) speculative decoding.

The original repo ships generic llama.cpp K-quants (Q4_K_S, etc.). These are different: they use ikawrakow's SOTA non-linear quant types (IQ4_K, IQ4_KS, IQ3_K) which, on the same hardware, decode ~40% faster at sustained generation than the generic Q4_K_S — at the same quality — because of ik_llama's optimized GEMV kernels. MTP draft heads are preserved, so self-speculative decoding works out of the box.

⚠️ These require ik_llama.cpp, not mainline llama.cpp. The IQ*_K quant types and the MTP path are ik_llama features. Mainline llama.cpp / LM Studio / Ollama will not load these correctly.

Model lineage

Stage	Model	By
Base	Qwen3.6-27B (dense, 27B)	Alibaba / Qwen
Finetune	Qwopus3.6-27B-Coder-MTP (reasoning-distill + agentic coding, MTP heads)	Jackrong
This repo	ik_llama.cpp IQ-quants + imatrix	community requant

Quant files

File	Type	bpw	Size	PPL (wikitext-2)¹	Best for
`Qwopus3.6-27B-Coder-MTP-IQ4_K.gguf`	IQ4_K	4.50	14.4 GiB	6.460 ±0.062	Max quality
`Qwopus3.6-27B-Coder-MTP-IQ4_KS.gguf`	IQ4_KS	4.25	13.7 GiB	6.477 ±0.062	Recommended — same quality as IQ4_K, ~37% faster decode
`Qwopus3.6-27B-Coder-MTP-IQ3_K.gguf`	IQ3_K	3.43	11.1 GiB	6.578 ±0.062	Tight VRAM
`qwopus-imatrix.dat`	—	—	12 MB	—	importance matrix (for reproducing / making your own quants)

¹ Perplexity over 250 chunks of wikitext-2-raw test at n_ctx=512. IQ4_K and IQ4_KS are statistically identical (the gap is within the error bars); IQ4_KS is the recommended default since it decodes markedly faster for no measurable quality loss.

Benchmarks (single RTX 3090, ik_llama.cpp build `4574`)

Raw throughput — llama-bench, -ngl 99, no speculative decoding:

Quant	pp512 (t/s)	tg128 (t/s)
IQ4_K	993	31.2
IQ4_KS	1215	42.8
IQ3_K	1024	40.0

Real-world with MTP — llama-server, IQ4_KS, MTP on (--draft-max 2), KV cache q4_0, 200K context, single slot (-np 1):

Workload	Prefill (t/s)	Decode (t/s)
Short Q&A	52	75.8
300-token gen	231	59.9
900-token gen	276	57.2
6021-token prompt	802	74.0

Measured during the 900-token run: ≈258 W GPU power draw, 65 °C, 21.2 GB VRAM (at 200K context). For reference, the generic Q4_K_S of the same model on the same machine sustains 41 t/s decode — these IQ quants are **40% faster**.

How these were built

Quantizing down from the near-lossless Q8_0 (not from a 4-bit quant — that would compound rounding error), guided by an importance matrix:

# 1. Importance matrix — run the Q8_0 model over a calibration corpus (GPU)
#    corpus: bartowski's calibration_datav3 (2481 lines); 129 chunks; ik_llama cu13-full image
llama-imatrix -m Qwopus3.6-27B-Coder-MTP-Q8_0.gguf \
  -f calibration_datav3.txt -o qwopus-imatrix.dat -ngl 99

# 2. Quantize each target from Q8_0 with the imatrix (CPU; cpu-full image)
#    --allow-requantize is required because the source is Q8_0 (safe: Q8 is ~lossless)
for T in IQ4_K IQ4_KS IQ3_K; do
  llama-quantize --allow-requantize --imatrix qwopus-imatrix.dat \
    Qwopus3.6-27B-Coder-MTP-Q8_0.gguf Qwopus3.6-27B-Coder-MTP-$T.gguf $T
done

Engine: ik_llama.cpp, Docker images ghcr.io/ikawrakow/ik-llama-cpp:cu13-full (imatrix/bench) and :cpu-full (quantize), build 4574.
Source: Jackrong's Q8_0 GGUF (MTP variant), so the MTP draft heads carry through.
Calibration: bartowski's calibration_datav3.

Usage (ik_llama.cpp)

Serving with an OpenAI-compatible API and MTP speculative decoding enabled:

llama-server \
  --model Qwopus3.6-27B-Coder-MTP-IQ4_KS.gguf \
  -ngl 99 --ctx-size 200000 -b 4096 -ub 1024 -np 1 \
  -ctk q4_0 -ctv q4_0 -fa on \
  -ngld 99 --multi-token-prediction --draft-max 2 --draft-p-min 0.0 \
  --recurrent-ckpt-mode auto --merge-qkv \
  --jinja --parallel-tool-calls \
  --reasoning off --reasoning-format deepseek \
  --temp 0.6 --top-p 0.95 --top-k 20 --min-p 0.0 --repeat-penalty 1.0

Then point any OpenAI-compatible client at http://localhost:8080/v1. Tool/function calling is supported (--jinja --parallel-tool-calls). Reasoning is off by default; the source model also supports a thinking mode.

Notes:

--multi-token-prediction --draft-max 2 enables MTP self-speculation; 2 is optimal for this model (higher draft depths gave no gain or crashed in testing).
Keep -np 1 on a single card — extra parallel slots divide throughput and disable MTP.

Credits

Qwen team / Alibaba — Qwen3.6-27B base model.
Jackrong — Qwopus3.6-27B-Coder-MTP finetune.
ikawrakow — ik_llama.cpp, the IQ quant types and MTP support.
bartowski — calibration dataset.

Disclaimer

Experimental community requantization for local evaluation. Quality is provided as-is — perplexity was measured, but full coding/agentic benchmarks (HumanEval/SWE-bench/etc.) were not run for these specific quants. License is inherited from the base (Apache-2.0). These GGUFs require ik_llama.cpp.

Downloads last month: 145

GGUF

Model size

27B params

Architecture

qwen35

Hardware compatibility

Model tree for AnthonyL1996/Qwopus3.6-27B-Coder-ik-MTP-GGUF

Base model

Jackrong/Qwopus3.6-27B-v2

Adapter

Jackrong/Qwopus3.6-27B-Coder-MTP-GGUF

Quantized

(3)

this model