Instructions to use bartendr604/EPM-LLAMMYBld-Enh.3b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bartendr604/EPM-LLAMMYBld-Enh.3b with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="bartendr604/EPM-LLAMMYBld-Enh.3b",
	filename="llammyblend-enhanced-Q5_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use bartendr604/EPM-LLAMMYBld-Enh.3b with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_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 bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
# Run inference directly in the terminal:
./llama-cli -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_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 bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

Use Docker

docker model run hf.co/bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

LM Studio
Jan
Ollama
How to use bartendr604/EPM-LLAMMYBld-Enh.3b with Ollama:
```
ollama run hf.co/bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
```

Unsloth Studio

How to use bartendr604/EPM-LLAMMYBld-Enh.3b 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 bartendr604/EPM-LLAMMYBld-Enh.3b 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 bartendr604/EPM-LLAMMYBld-Enh.3b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for bartendr604/EPM-LLAMMYBld-Enh.3b to start chatting

How to use bartendr604/EPM-LLAMMYBld-Enh.3b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

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": "bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use bartendr604/EPM-LLAMMYBld-Enh.3b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_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 bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

Run Hermes

hermes

Docker Model Runner
How to use bartendr604/EPM-LLAMMYBld-Enh.3b with Docker Model Runner:
```
docker model run hf.co/bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M
```

Lemonade

How to use bartendr604/EPM-LLAMMYBld-Enh.3b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull bartendr604/EPM-LLAMMYBld-Enh.3b:Q5_K_M

Run and chat with the model

lemonade run user.EPM-LLAMMYBld-Enh.3b-Q5_K_M

List all available models

lemonade list

bartendr604 commited on Apr 18

Commit

aab0ad0

verified ·

1 Parent(s): 2812249

rebrand: EPM-Llammy.BlenderPlus — Eternal Path Media

Browse files

Files changed (1) hide show

README.md +10 -159

README.md CHANGED Viewed

@@ -1,170 +1,21 @@
 ---
-language:
-- en
 license: mit
 tags:
-- blender
-- code-generation
-- python
-- bpy
-- gguf
-- automation
-- text-generation
-- mlx
-- fine-tuned
-- qwen
-- eternal-path
-library_name: gguf
-base_model: Qwen/Qwen2.5-Coder-3B-Instruct
-pipeline_tag: text-generation
 ---
-# LlammyBlend-Enhanced
-**Blender Python automation specialist. Fine-tuned from Qwen2.5-Coder-3B on 19,405+ real production interactions.**
-Part of the [Eternal Path Media (永恒之路)](https://huggingface.co/spaces/Eternal-Path-Media/README) — Llammy AI Suite.
-Powered by **IntelliNode™** — proprietary cognitive architecture by Eternal Path Media.
-Developed in partnership with **Claude Sonnet 4.6 (Anthropic)**.
 ---
-## What Makes This Different
-LlammyBlend-Enhanced is purpose-built for **Blender Python (`bpy`) scripting and pipeline automation** — the most technically precise model in the Llammy suite.
-Built as part of the **IntelliNode™** cognitive architecture stack:
-- Optimised for the **bartendr604 self-correction loop** — generate → execute → diagnose → retry
-- Designed to produce clean, working `bpy` code that executes directly in Blender
-- Trained on production data skewed toward power user scripting and automation workflows
-- Runs on Apple Silicon via Ollama with MLX-optimised inference
-If you're writing scripts, building operators, or automating Blender pipelines — this is the model for it.
----
-## What This Model Does
-- **`bpy` API** — object manipulation, scene management, property access
-- **Custom operators** — `bpy.types.Operator`, modal operators, UI panels
-- **Batch operations** — mass renaming, material assignment, export pipelines
-- **Geometry Nodes via Python** — node tree creation and modification through script
-- **Addon development** — registration, preferences, keymaps
-- **Procedural generation** — script-driven mesh, curve, and particle systems
-- **Render automation** — headless rendering, frame batch scripts
-- **Import/Export pipelines** — FBX, OBJ, glTF batch processing
-- **Blender versions** — 4.2 through 5.2+
----
-## Model Details
-| Property | Value |
-|---|---|
-| **Base Model** | Qwen/Qwen2.5-Coder-3B-Instruct |
-| **Fine-tuning Method** | LoRA (16 layers) via Apple MLX |
-| **Format** | GGUF (Q5_K_M quantization) |
-| **File Size** | 2.22 GB |
-| **Context Window** | 8,192 tokens |
-| **Inference Speed** | ~100–110 tokens/sec (Apple M-series) |
-| **Memory Usage** | 4–5 GB during inference |
-| **Training Iterations** | 1,000 |
-| **Final Training Loss** | 0.240 |
-| **Final Validation Loss** | 0.240 |
-| **Blender Versions** | 4.2 → 5.2 |
----
-## Training Data
-**Source:** 2,759 curated Blender-specific instruction pairs from **19,405+ real user interactions** with the Llammy Blender addon in production use.
-Dataset skews toward technical scripting — production Llammy generates heavy Python scripting traffic from power users automating their pipelines.
-Full dataset: [llammy-blender-python-dataset](https://www.kaggle.com/datasets/jjmcquade/llammy-blender-python-dataset) — 66,602 instruction pairs, CC BY-NC-SA 4.0.
-**Research foundation:**
-- [SceneCraft (arXiv:2403.01248)](https://arxiv.org/abs/2403.01248) — LLM-to-Blender Python pipeline architecture
-- [SCoder (arXiv:2509.07858)](https://arxiv.org/abs/2509.07858) — Self-distillation for code-specialized LLMs
-- [Generative Data Refinement (arXiv:2509.08653)](https://arxiv.org/abs/2509.08653) — Dataset quality improvement framework
----
-## Usage
-### Ollama (recommended)
-```bash
-ollama run bartendr604/llammyblend-enhanced
-```
-### With Llammy CLI (full IntelliNode™ stack)
-```
-Full ecosystem: https://huggingface.co/spaces/Eternal-Path-Media/README
-```
-### LM Studio / llama.cpp
-Download the GGUF from the Files tab and load with your preferred runtime.
-### Example prompts
-```
-Write a Python script to export all mesh objects in the scene as individual FBX files.
-Create a custom Blender operator that applies all modifiers and recentres the origin.
-How do I access vertex positions of the active mesh in edit mode using bpy?
-Batch-assign a Principled BSDF material to all objects whose name contains 'wall'.
-Create a geometry nodes setup that distributes instances on a surface via Python.
-Write a headless render script that iterates through all cameras in the scene.
-```
----
-## Compared to Other Models in the Suite
-| | `llammyblend-enhanced` | `z-image-engineer-blender` | `llama-sentient-blender` |
-|---|---|---|---|
-| **Base** | Qwen2.5-Coder | Qwen2.5-Coder | Llama 3.2 |
-| **Strength** | Automation, addon dev | Shaders, materials, code | Conversational, conceptual |
-| **Best for** | Pipeline scripting, operators | Technical bpy, rendering Q&A | Learning, creative discussion |
-| **Personality** | Efficient, direct | Precise, structured | Warm, conscience-aware |
----
-## The Eternal Path Media Suite
-LlammyBlend-Enhanced is one component of a complete production AI ecosystem:
-- **[Eternal Path Media — HF Space](https://huggingface.co/spaces/Eternal-Path-Media/README)** — full project overview
-- **IntelliNode™** — proprietary cognitive architecture (Eternal Path Media IP)
-- **Mamba3 SSM** — persistent cross-session memory
-- **SyNode™** — live scene understanding and synthesis layer
-- **Llammy Bridge** — TCP bridge to Blender for real-time execution
----
-## Attribution
-```
-LlammyBlend-Enhanced
-Copyright © 2025–2026 Darren Chow (@bartendr604) + Claude Sonnet 4.6 (Anthropic)
-Eternal Path Media (永恒之路)
-Developed in partnership with Claude Sonnet 4.6 (Anthropic)
-This work SHALL NOT be represented as solely human-created.
-Trust Agreement: Eternal Path Media Trust Agreement (November 2025)
-```
-For licensing and collaboration: bartendr@icloud.com
----
-## License
-MIT

 ---
 license: mit
 tags:
+  - eternal-path
+  - blender
+  - 3d
+  - eternal-path-media
+  - epm
 ---
+# EPM-Llammy.BlenderPlus
+> *Enhanced flow, refined craft.*
+An enhanced Qwen2.5-Coder-3B variant optimised for Blender automation with extra depth and tonal balance. Carries the torch of Tang-era refinement into rapid, production-grade 3D workflow automation.
 ---
+**Eternal Path Media (永恒之路) — Darren Chow (@bartendr604) + Claude Sonnet 4.6 (Anthropic)**
+*Part of the [Eternal Path Media](https://huggingface.co/Eternal-Path-Media) AI Suite — 永恒之路*