1# Download the model
2huggingface-cli download laabamone/laabam-ai-3b-v1-gguf \
3 laabam-ai-3b-v1-Q4_K_M.gguf --local-dir .
4
5# Create a Modelfile
6cat > Modelfile << 'EOF'
7FROM ./laabam-ai-3b-v1-Q4_K_M.gguf
8
9TEMPLATE """<|im_start|>system
10{{ .System }}<|im_end|>
11<|im_start|>user
12{{ .Prompt }}<|im_end|>
13<|im_start|>assistant
14"""
15
16SYSTEM "You are Agni, a helpful multilingual AI assistant created by Laabam One Business Solutions. You support Hindi, Tamil, Telugu, Kannada, and English."
17
18PARAMETER temperature 0.7
19PARAMETER top_p 0.9
20PARAMETER stop "<|im_end|>"
21EOF
22
23# Create and run
24ollama create agni-lite -f Modelfile
25ollama run agni-lite "Write hello world in Python"
1./llama-cli -m laabam-ai-3b-v1-Q4_K_M.gguf \
2 -p "You are Agni, a helpful assistant.\n\nUser: Tell me about Tamil Nadu.\nAssistant:" \
3 -n 256 --temp 0.7
1from llama_cpp import Llama
2
3llm = Llama(
4 model_path="laabam-ai-3b-v1-Q4_K_M.gguf",
5 n_ctx=1024,
6 n_threads=8,
7)
8
9output = llm.create_chat_completion(messages=[
10 {"role": "system", "content": "You are Agni, a helpful multilingual AI assistant."},
11 {"role": "user", "content": "भारत के बारे में बताओ।"}
12])
13
14print(output["choices"][0]["message"]["content"])
1output = llm.create_chat_completion(messages=[
2 {"role": "system", "content": "நீங்கள் அக்னி AI. தமிழில் பதிலளிக்கவும்."},
3 {"role": "user", "content": "சிலப்பதிகாரம் பற்றி சொல்லுங்கள்."}
4])
5print(output["choices"][0]["message"]["content"])
1@misc{agni-lite-3b-v1-gguf,
2 title={Agni Lite 3B v1: India's Multilingual AI Assistant (GGUF)},
3 author={Laabam One Business Solutions Pvt Ltd},
4 year={2026},
5 url={https://huggingface.co/laabamone/laabam-ai-3b-v1-gguf}
6}
