README.md

---
license: apache-2.0
datasets:
- Sweaterdog/Andy-3.5-MASSIVE
- Sweaterdog/Andy-3.5
- Sweaterdog/Andy-3.5-reasoning
language:
- en
base_model:
- unsloth/DeepSeek-R1-Distill-Qwen-7B-unsloth-bnb-4bit
tags:
- minecraft
- Mindcraft
- Minecraft
- MindCraft
---

# 🚀 Welcome to Next Generation Minecraft with Andy 3.6 🚀

## Andy 3.6 is a **LOCAL** model beating Andy-3.5 in performance
*Andy 3.6 is designed to be used with MindCraft, and is not designed nor intended to be used for any other applications*


>  # Please note! [!WARNING]
> 
> Andy-3.6 was trained on older data, and not the newest and latest versions of Mindcraft.
> 
> I **cannot** guarantee that Andy-3.6 will work on future versions as the model was tuned to play MindCraft with a specific version!
> 
> For the rest of the Andy-3.6 generation, this model will **ONLY** be supported on the version of Mindcraft in [this github repo!](https://github.com/Sweaterdog/Mindcraft-for-Andy-3.5)
>
> For more info, as well as the supported version of Mindcraft, please follow [this link to github](https://github.com/Sweaterdog/Mindcraft-for-Andy-3.5)

# How to Install / Setup

**Installing Andy-3.6 is much easier and Andy-3.5!**

1. In the top right of this repo, click "Use This Model"
2. Next, click Ollama
3. Pick your quantization
4. Run the command in your terminal
5. Now you have Andy-3.6 installed!

# How was model trained?

The model was trained on the [MindCraft dataset](https://huggingface.co/datasets/Sweaterdog/Andy-3.5-MASSIVE) for Andy-3.6, a curated dataset for Q & A, reasoning, and playing, which includes ~22,000 prompts.

# What are capabilities and Limitations?

Andy-3.6 was trained on EVERYTHING regarding Minecraft and MindCraft, it knows how to use commands natively without a system prompt.
Andy-3.6 also knows how to build / use !newAction to perform commands, it was trained on lots of building, as well as, using !newAction to do tasks like manually making something or strip mining.

# What models can I choose?

There are going to be 3 model sizes avaliable, Regular, Large, and Small
* Large will be a 32B parameter model, tuned from [Deepseek-R1 Distilled](https://huggingface.co/unsloth/DeepSeek-R1-Distill-Qwen-32B-bnb-4bit)
* Regular is a 7B parameter model, tuned from [Deepseek-R1 Distilled](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B)
* Small is a 3B parameter model, tuned from [Qwen2.5 3B](Qwen/Qwen2.5-3B-Instruct)

Both models will have **case-by-case reasoning** baked **into** the model, meaning when it encounters a hard task, it will reason.

You can also *prompt* Andy-3.6 to reason for better performance

# Safety and FAQ

Q: Is this model safe to use?

A. Yes, this model is non-volatile, and cannot generate malicous content

Q. Can this model be used on a server?

A. Yes, In theory and practice the model is only capable of building and performing manual tasks via newAction

Q. Who is responsible if this model does generate malicous content?

A. You are responsible, even though the model was never trained to be able to make malicous content, there is a ***very very slight chance*** it still generates malicous code.

Q. If I make media based on this model, like photos / videos, do I have to mention the Creator?

A. No, if you are making a post about MindCraft, and using this model, you only have to mention the creator if you mention the model being used.

# 🔥UPDATE🔥

## **Andy-3.6 Release!**
Andy-3.6 is our Next Generation model, feature more capabilties, and stronger performance over ANY other local LLM in Mindcraft!

> # I want to thank all supporters! [!NOTE]
> I would love to thank everyone who supported this project, there is a list of supporters in the files section.
> 
> You can find all of the supporters [here](https://huggingface.co/Sweaterdog/Andy-3.5/blob/main/Supporters.txt)

# Performance Metrics

These benchmarks are a-typical, since most standard benchmarks don't apply to Minecraft

The benchmarks below include models via API that are cheap, and other fine-tuned local models

## Zero info Prompting
*How fast can a model collect 16 oak logs, and convert them all into sticks*

![image/png](https://cdn-uploads.huggingface.co/production/uploads/66960602f0ffd8e3a381106a/IEw1Gydg943qVSNGAL3RW.png)


As shown, the only models that are capable of play without information, is Andy-3.6, and all Andy-3.5 models

You can test this demo out for yourself using [this profile](https://huggingface.co/Sweaterdog/Andy-3.5/blob/main/local_demo.json)


## Time to get a stone pickaxe


![image/png](https://cdn-uploads.huggingface.co/production/uploads/66960602f0ffd8e3a381106a/r3AoHHlmQuPdpt3WEFcyp.png)

## *For Andy-3.6, I used the Q4_K_M quantization*

*For Andy-3.5-mini, I used the FP16 model, I had enough VRAM to do so*

*For Andy-3.5, I used the Q4_K_M quantization*

*For Andy-3.5-small, I used the Q8_0 quantization*

*Andy-3.5-reasoning-small was able to be the most efficient model producing the lowest amount of messages, but took a whopping 34.5 minutes to get a stone pickaxe.*

*For Andy-3.5-Teensy, I used the FP16 quantization*

*For Mineslayerv1 and Mineslayerv2, I used the default (and only) quantization, Q4_K_M*

## Notes about the benchmarks

**Zero Info Prompting**

Andy-3.5-Mini collected 32 oak_log instead of 16 oak_log

Andy-3.5-small *No notes*

Andy-3.5 attempted to continue playing, and make a wooden_pickaxe after the goal was done.

Both Mineslayerv1 and Mineslayerv2 hallucinated commands, like !chop or !grab

**Time to get a stone pickaxe**

## Andy-3.6 performed the best, beating gpt-4o-mini and claude-3.5-haiku

Andy-3.5-Mini was unable to make itself a stone pickaxe, however it collected enough wood, but then got stuck on converting logs to planks, it kept trying "!craftRecipe("wooden_planks", 6) instead of oak_planks

Andy-3.5-small kept trying to make a stone_pickaxe first

Andy-3.5 Made a stone pickaxe the faster than GPT-4o-mini and Claude-3.5-Haiku

Mineslayerv1 Was unable to use !collectBlocks, instead kept trying !collectBlock

Mineslayerv2 Was unable to play, it kept hallucinating on the first command