---
license: apache-2.0
base_model:
- Altworld/Hemmingway-1
tags:
- merge
- mergekit
- qwen
- code
- uncensored
- text-generation
library_name: transformers
---
# currently not working try [zagy1234/ZeryEasyCoder27B-GGUF](https://huggingface.co/zagy1234/ZeryEasyCoder27B-GGUF)
# 🚀 ZeryEasyCoder-27B (Full Precision bfloat16)

Created by **@zagy1234**

**ZeryEasyCoder-27B** is a high-performance 27-billion parameter hybrid merge based on the Qwen architecture. It combines advanced reasoning, deep programming capabilities, and unrestricted technical utility.

> 💡 **Looking for local inference (Ollama / GGUF)?**  
> Check out the GGUF quantized version here: [zagy1234/ZeryEasyCoder27B-GGUF](https://huggingface.co/zagy1234/ZeryEasyCoder27B-GGUF)

---

## 🏗️ Model Architecture & Base Models

Merged using `mergekit` from two state-of-the-art models:
* **Hemmingway-1**: Superior reasoning, natural dialogue, and unrestricted versatility.
* **NEO-CODER MAX (Turbo Cold Fusion)**: Low-level programming expertise, script generation, and system automation.

### Specifications
* **Developer:** zagy1234
* **Parameters:** 27B
* **Precision:** bfloat16 (Safetensors)
* **Format:** Hugging Face Transformers

---

## ✨ Features & Strengths
* **Programming Mastery:** Advanced capabilities in Python, C++, Bash/Fish, C#, and automation scripts.
* **Linux & Systems Administration:** Tailored for terminal workflows, system configuration, and tool integration.
* **Uncensored & Unrestricted:** No artificial guardrails on security research, pentesting analysis, or complex technical tasks.

---

## 💻 Usage with Hugging Face Transformers

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "zagy1234/ZeryEasyCoder27B"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

prompt = "Napisz skrypt w Pythonie do automatyzacji kopii zapasowej."
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))