Instructions to use ganjakha/codium-bd-coder-v2-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ganjakha/codium-bd-coder-v2-lora 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 ganjakha/codium-bd-coder-v2-lora
# Run inference directly in the terminal:
llama cli -hf ganjakha/codium-bd-coder-v2-lora

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ganjakha/codium-bd-coder-v2-lora
# Run inference directly in the terminal:
llama cli -hf ganjakha/codium-bd-coder-v2-lora

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 ganjakha/codium-bd-coder-v2-lora
# Run inference directly in the terminal:
./llama-cli -hf ganjakha/codium-bd-coder-v2-lora

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 ganjakha/codium-bd-coder-v2-lora
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ganjakha/codium-bd-coder-v2-lora

Use Docker

docker model run hf.co/ganjakha/codium-bd-coder-v2-lora

LM Studio
Jan

vLLM

How to use ganjakha/codium-bd-coder-v2-lora with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ganjakha/codium-bd-coder-v2-lora"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ganjakha/codium-bd-coder-v2-lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ganjakha/codium-bd-coder-v2-lora

Ollama
How to use ganjakha/codium-bd-coder-v2-lora with Ollama:
```
ollama run hf.co/ganjakha/codium-bd-coder-v2-lora
```
Unsloth Desktop

How to use ganjakha/codium-bd-coder-v2-lora with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ganjakha/codium-bd-coder-v2-lora

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": "ganjakha/codium-bd-coder-v2-lora"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ganjakha/codium-bd-coder-v2-lora with Docker Model Runner:
```
docker model run hf.co/ganjakha/codium-bd-coder-v2-lora
```

Lemonade

How to use ganjakha/codium-bd-coder-v2-lora with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ganjakha/codium-bd-coder-v2-lora

Run and chat with the model

lemonade run user.codium-bd-coder-v2-lora-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use ganjakha/codium-bd-coder-v2-lora with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ganjakha/codium-bd-coder-v2-lora

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 ganjakha/codium-bd-coder-v2-lora

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ganjakha/codium-bd-coder-v2-lora with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ganjakha/codium-bd-coder-v2-lora

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 "ganjakha/codium-bd-coder-v2-lora" \
  --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"

🌟 CODIUM BD CODER V2 🌟

🚀 The Ultimate Bengali-Sourced Programming & Vision AI Assistant

⭐️ If you like this model, please leave a STAR to show your support! ⭐️

আপনার যদি এই মডেলটি ভালো লেগে থাকে, তবে অবশ্যই একটি STAR (⭐️) দিয়ে আমাদের টিমকে উৎসাহিত করবেন!

🔮 Overview / সংক্ষিপ্ত পরিচিতি

Codium BD Coder v2 is a state-of-the-art, high-performance programming assistant and vision-capable AI model tailored for software engineering, code-generation, and instruction-following tasks. Optimized via advanced 4-bit QLoRA techniques, it offers unparalleled coding capabilities with minimal resource footprints.

📢 Join Our Communities / আমাদের টেলিগ্রাম চ্যানেলসমূহ

মডেলটির পরবর্তী আপডেট, ফ্রি কোডিং সাপোর্ট এবং নিত্যনতুন প্রজেক্ট সম্পর্কে সবার আগে জানতে আমাদের অফিসিয়াল টেলিগ্রাম চ্যানেলগুলোতে যুক্ত হোন:

🌀 Aura Vibes (Itachi)	🔥 Team Developer Force BD

Join Aura Vibes Channel	Join Developer Force Channel

🌟 Key Features / প্রধান বৈশিষ্ট্যসমূহ

🎨 Multimodal Vision Support: Analyze system architectures, database designs, or wireframe UI screenshots, and generate clean, responsive front-end/back-end code instantly!
💻 Advanced Multi-Language Code Generation: Outstanding coding proficiency in Python, JavaScript, TypeScript, C++, SQL, HTML/CSS, Go, and more.
🇧🇩 Multilingual & Native Bengali Support: Perfectly understands programming instructions given in both English and Bengali (বাংলা).
🧠 Extremely Lightweight: Ultra-fast inference with optimal memory usage, optimized for everyday consumer GPUs and lightweight local deployments.

💾 GGUF Local Run Guide / লোকাল ব্যবহারের নির্দেশিকা

আপনি যদি অফলাইনে, কম র‍্যামে বা আপনার পিসিতে অত্যন্ত দ্রুত গতিতে কোনো GPU ছাড়াই মডেলটি চালাতে চান, তবে সরাসরি নিচে দেওয়া বোতাম বা লিঙ্ক থেকে একক GGUF ফাইলটি ডাউনলোড করুন:

📥 Direct Downloads & Links

Direct Link	Files Browser
📥 Download GGUF (Direct Link)	📂 View All Files

⚙️ Setup Instructions / কীভাবে রান করবেন:

1️⃣ Using GUI Tools (LM Studio / Ollama / Jan AI)

উপরের Direct Link-এ ক্লিক করে codium-bd-coder-v2.gguf ফাইলটি ডাউনলোড করে নিন।
LM Studio অথবা Jan AI অ্যাপটি ওপেন করুন।
ডাউনলোড করা ফাইলটি ড্র্যাগ করে অ্যাপের নির্দিষ্ট ফোল্ডারে রাখলেই তা ব্যবহারের জন্য প্রস্তুত!

2️⃣ Using Llama.cpp CLI

# Download the model file
wget https://huggingface.co/ganjakha/codium-bd-coder-v2-lora/resolve/main/codium-bd-coder-v2.gguf

# Run the CLI inference
./llama-cli -m codium-bd-coder-v2.gguf -p "Write a quicksort in Python." -n 512

💻 Python Quickstart / পাইথনে ব্যবহারের সহজ নিয়ম

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

base_model_id = "Qwen/Qwen3.5-4B"
adapter_id = "ganjakha/codium-bd-coder-v2-lora"

tokenizer = AutoTokenizer.from_pretrained(base_model_id)
if tokenizer.pad_token is None:
    tokenizer.pad_token = tokenizer.eos_token

# Load base model
model = AutoModelForCausalLM.from_pretrained(
    base_model_id,
    torch_dtype=torch.float16,
    device_map="auto",
    low_cpu_mem_usage=True
)

# Mount the custom Codium BD Coder v2 weights
model = PeftModel.from_pretrained(model, adapter_id)
model.eval()

prompt = "Write a Python function to check if a number is prime."
messages = [{"role": "user", "content": prompt}]

formatted_prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(formatted_prompt, return_tensors="pt").to("cuda")

with torch.no_inference_mode():
    outputs = model.generate(
        **inputs,
        max_new_tokens=512,
        do_sample=True,
        temperature=0.7,
        pad_token_id=tokenizer.pad_token_id,
        eos_token_id=tokenizer.eos_token_id
    )

response = tokenizer.decode(outputs[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True)
print(response)

🤝 License & Attribution

License: Distributed under the permissive Apache-2.0 License.
Team: Crafted with passion and dedication by Neloy Chandra Deb and Team Aura Vibes (BD).

💖 Support Our Work

If you find this project helpful, please consider leaving a ⭐️ on our Hugging Face repository to help us reach more developers!

Downloads last month: 37

GGUF

Model size

4B params

Architecture

qwen35

Hardware compatibility

We're not able to determine the quantization variants.

View all variants