Instructions to use zagy1234/ZeryEasyCoder27B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use zagy1234/ZeryEasyCoder27B-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 zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf zagy1234/ZeryEasyCoder27B-GGUF: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 zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf zagy1234/ZeryEasyCoder27B-GGUF: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 zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use zagy1234/ZeryEasyCoder27B-GGUF with Ollama:
```
ollama run hf.co/zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use zagy1234/ZeryEasyCoder27B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf zagy1234/ZeryEasyCoder27B-GGUF: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": "zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use zagy1234/ZeryEasyCoder27B-GGUF with Docker Model Runner:
```
docker model run hf.co/zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M
```

Lemonade

How to use zagy1234/ZeryEasyCoder27B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.ZeryEasyCoder27B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use zagy1234/ZeryEasyCoder27B-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 zagy1234/ZeryEasyCoder27B-GGUF: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 zagy1234/ZeryEasyCoder27B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use zagy1234/ZeryEasyCoder27B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf zagy1234/ZeryEasyCoder27B-GGUF: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 "zagy1234/ZeryEasyCoder27B-GGUF: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"

ZeryEasyCoder27B GGUF

Created by @zagy1234

Custom Qwen3.5-based hybrid model merged and optimized for high-performance math reasoning and coding tasks.

This repository contains GGUF quantized weights for ZeryEasyCoder-27B, optimized for efficient local inference on consumer hardware using Ollama, llama.cpp, LM Studio, and Text Generation WebUI.

🛠️ Architecture & Conversion Fixes

Base Architecture: Hybrid Qwen3.5 (64 layers: 48 linear attention / SSM + 16 full-attention).
MTP Head Fix: Removed the non-essential Multi-Token Prediction (MTP) head at layer source, resolving the blk.64 tensor initialization error while preserving full model intelligence.
Quantization: Q4_K_M (~16 GB).

Artifact	Value
Filename	`zeryeasycoder-27b-Q4_K_M.gguf`
Size	16 545 674 336 B (15.41 GiB)
SHA256	`41ef673c4c7efda3930bc076bb178519a5f7bcf810adc23044cee0122f44d9be`
Block count	64 (`nextn_predict_layers`: absent)
Tensor count	851
Tensor layout	Q4_K × 433, Q6_K × 65, F32 × 353

📊 Benchmark Comparison (GSM8K)

Evaluated via lm_eval (v0.4.13) on a 200-sample test set (5-shot, reasoning_effort=none, max_gen_toks=512, temperature=0).

Model	Size / Quant	GSM8K (Math Reasoning)
ZeryEasyCoder27B (Ours)	27B (Q4_K_M)	95.5% (191/200)
DeepSeek-R1-Distill-Qwen-14B	14B (FP16)	~93.9%
Qwen2.5-Coder-32B-Instruct	32B (Q4_K_M)	~92.5%
Qwen2.5-32B-Instruct	32B (Q4_K_M)	~90.8%
Llama-3.1-70B-Instruct	70B (Q4_K_M)	~95.1%

Note: Clean execution with zero truncated outputs, format errors, or empty responses. Comparison figures are approximate published values at varying precisions and are not measured under identical harness settings.

Key Capabilities

🐍 Python, C++ & Scripting: Writing, debugging, and refactoring automation scripts.
🐧 Linux Sysadmin: Native understanding of shell configurations, terminal tools, and system administration.
🔓 Uncensored: No artificial barriers for cybersecurity analysis, security research, or complex technical tasks.

🚀 Quickstart (Ollama)

ollama run hf.co/zagy1234/ZeryEasyCoder27B-GGUF

Option 2: Custom Modelfile in Ollama

If you downloaded the .gguf file locally, create a file named Modelfile:

FROM ./zeryeasycoder-27b-Q4_K_M.gguf

PARAMETER temperature 0.6
PARAMETER top_p 0.9
PARAMETER num_ctx 16384

SYSTEM "Jesteś zaawansowanym asystentem technicznym ZeryEasyCoder. Odpowiadasz bezpośrednio, zwięźle i precyzyjnie."

Then build and start the model in your terminal:

ollama create zeryeasycoder -f Modelfile
ollama run zeryeasycoder

Option 3: llama.cpp Command Line

./llama-cli -m zeryeasycoder-27b-Q4_K_M.gguf -p "Napisz skrypt w Pythonie do automatyzacji kopii zapasowych" -n 512

🏗️ Model Background

ZeryEasyCoder-27B is a specialized hybrid merge combining two Qwen-based models:

Hemmingway-1: Excellent multi-step reasoning, natural dialogue, and unrestricted versatility.
NEO-CODER MAX (Turbo Cold Fusion): Deep knowledge in low-level programming, Linux systems, Bash/Fish scripting, and technical automation.

🔗 Related Links

Full Precision Model (Safetensors - 51GB): zagy1234/ZeryEasyCoder27B

Downloads last month: 141

GGUF

Model size

27B params

Architecture

qwen35

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for zagy1234/ZeryEasyCoder27B-GGUF

Base model

Qwen/Qwen3.8-27B

Finetuned

HemmingwayAI/Hemmingway-1

Finetuned

zagy1234/ZeryEasyCoder27B

Quantized

(1)

this model