Instructions to use Jershone/Echo-CodeEX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use Jershone/Echo-CodeEX with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="Jershone/Echo-CodeEX", filename="EchoAI-CodeEX.gguf", )
llm.create_chat_completion( messages = [ { "role": "user", "content": "What is the capital of France?" } ] ) - Notebooks
- Google Colab
- Kaggle
- Local Apps
- llama.cpp
How to use Jershone/Echo-CodeEX with llama.cpp:
Install from brew
brew install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Jershone/Echo-CodeEX # Run inference directly in the terminal: llama-cli -hf Jershone/Echo-CodeEX
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Jershone/Echo-CodeEX # Run inference directly in the terminal: llama-cli -hf Jershone/Echo-CodeEX
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 Jershone/Echo-CodeEX # Run inference directly in the terminal: ./llama-cli -hf Jershone/Echo-CodeEX
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 Jershone/Echo-CodeEX # Run inference directly in the terminal: ./build/bin/llama-cli -hf Jershone/Echo-CodeEX
Use Docker
docker model run hf.co/Jershone/Echo-CodeEX
- LM Studio
- Jan
- vLLM
How to use Jershone/Echo-CodeEX with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Jershone/Echo-CodeEX" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Jershone/Echo-CodeEX", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Jershone/Echo-CodeEX
- Ollama
How to use Jershone/Echo-CodeEX with Ollama:
ollama run hf.co/Jershone/Echo-CodeEX
- Unsloth Studio new
How to use Jershone/Echo-CodeEX 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 Jershone/Echo-CodeEX 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 Jershone/Echo-CodeEX to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Jershone/Echo-CodeEX to start chatting
- Pi new
How to use Jershone/Echo-CodeEX with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama-server -hf Jershone/Echo-CodeEX
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": "Jershone/Echo-CodeEX" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use Jershone/Echo-CodeEX with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama-server -hf Jershone/Echo-CodeEX
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 Jershone/Echo-CodeEX
Run Hermes
hermes
- Docker Model Runner
How to use Jershone/Echo-CodeEX with Docker Model Runner:
docker model run hf.co/Jershone/Echo-CodeEX
- Lemonade
How to use Jershone/Echo-CodeEX with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Jershone/Echo-CodeEX
Run and chat with the model
lemonade run user.Echo-CodeEX-{{QUANT_TAG}}List all available models
lemonade list
Jerry commited on
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,66 @@
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
---
|
| 3 |
license: apache-2.0
|
| 4 |
+
base_model: Qwen/Qwen2.5-0.5B-Instruct
|
| 5 |
+
tags:
|
| 6 |
+
- gguf
|
| 7 |
+
- code
|
| 8 |
+
- text-generation
|
| 9 |
+
- edge-ai
|
| 10 |
+
- qwen
|
| 11 |
+
model_creator: MLM8372984732947
|
| 12 |
+
model_name: Echo-CodeEX-GGUF
|
| 13 |
+
pipeline_tag: text-generation
|
| 14 |
+
language:
|
| 15 |
+
- en
|
| 16 |
---
|
| 17 |
+
|
| 18 |
+
# 💻 Echo-CodeEX (0.5B Parameters - GGUF)
|
| 19 |
+
|
| 20 |
+
**Echo-CodeEX** is a specialized, edge-optimized 0.5B parameter variant engineered explicitly for offline programming assistance, code execution logic, and structured syntax manipulation. Built upon a fine-tuned **Qwen-2.5-Instruct** architecture and fully merged into a standalone GGUF binary, it balances lightning-fast syntax completion with low-resource hardware execution.
|
| 21 |
+
|
| 22 |
+
## ✨ Key Features
|
| 23 |
+
|
| 24 |
+
* **Syntax Grounded:** Fine-tuned specifically to prioritize code construction, structural scripting loops, and algorithmic optimizations over open-ended narrative generation.
|
| 25 |
+
* **Unified GGUF Engine:** Zero dependencies on external floating adapter weights or complex Python multi-layer environments. Loadable instantly across standard local runtimes (`llama.cpp`, `node-llama-cpp`, `Ollama`).
|
| 26 |
+
* **Fill-in-the-Middle (FIM) Ready:** Inherits raw structural token patterns from the Qwen architecture, enabling seamless inline logic insertions and multi-line code predictions.
|
| 27 |
+
|
| 28 |
+
---
|
| 29 |
+
|
| 30 |
+
## 🧠 Code Prompt Engineering Structure
|
| 31 |
+
|
| 32 |
+
To bypass open-ended conversational filler and force direct code output, structure your inputs strictly within the **ChatML layout**. Define the system parameters explicitly to receive clean code blocks:
|
| 33 |
+
|
| 34 |
+
```text
|
| 35 |
+
<|im_start|>system
|
| 36 |
+
You are Echo-CodeEX, an expert code generation assistant. Respond only with structured code blocks and clean syntax commentaries.<|im_end|>
|
| 37 |
+
<|im_start|>user
|
| 38 |
+
Write a clean Python function to parse JSON strings safely.<|im_end|>
|
| 39 |
+
<|im_start|>assistant
|
| 40 |
+
```
|
| 41 |
+
|
| 42 |
+
## 💻 Sample Implementation (Node.js)
|
| 43 |
+
You can spin this specialized model up locally inside your developer environment using node-llama-cpp:
|
| 44 |
+
|
| 45 |
+
```JavaScript
|
| 46 |
+
import {LlamaModel, LlamaContext, LlamaSequence} from "node-llama-cpp";
|
| 47 |
+
import path from "path";
|
| 48 |
+
|
| 49 |
+
const model = new LlamaModel({
|
| 50 |
+
modelPath: path.join(__dirname, "echo-codeex.gguf")
|
| 51 |
+
});
|
| 52 |
+
|
| 53 |
+
const context = new LlamaContext({model});
|
| 54 |
+
const sequence = new LlamaSequence({context});
|
| 55 |
+
|
| 56 |
+
const prompt = `<|im_start|>system\nYou are Echo-CodeEX.<|im_end|>\n<|im_start|>user\nWrite a basic bash script to check if a file exists.\n<|im_end|>\n<|im_start|>assistant\n`;
|
| 57 |
+
const tokens = model.tokenize(prompt);
|
| 58 |
+
|
| 59 |
+
console.log("Generating script output...");
|
| 60 |
+
const response = await sequence.evaluate(tokens, {
|
| 61 |
+
temperature: 0.1 // Kept low to enforce syntax consistency over creativity
|
| 62 |
+
});
|
| 63 |
+
print(model.detokenize(response));
|
| 64 |
+
```
|
| 65 |
+
## 📄 License
|
| 66 |
+
This model's merged weights are distributed under the `Apache 2.0 License`, fully compliant with the core permissions and commercial deployment conditions set by the original Qwen development team.
|