Instructions to use Kolyadual/MiLana-v1-beta with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Kolyadual/MiLana-v1-beta with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Kolyadual/MiLana-v1-beta") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Kolyadual/MiLana-v1-beta") model = AutoModelForCausalLM.from_pretrained("Kolyadual/MiLana-v1-beta", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Kolyadual/MiLana-v1-beta with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Kolyadual/MiLana-v1-beta" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Kolyadual/MiLana-v1-beta", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Kolyadual/MiLana-v1-beta
- SGLang
How to use Kolyadual/MiLana-v1-beta with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Kolyadual/MiLana-v1-beta" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Kolyadual/MiLana-v1-beta", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Kolyadual/MiLana-v1-beta" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Kolyadual/MiLana-v1-beta", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Unsloth Studio
How to use Kolyadual/MiLana-v1-beta 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 Kolyadual/MiLana-v1-beta 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 Kolyadual/MiLana-v1-beta to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Kolyadual/MiLana-v1-beta to start chatting
Load model with FastModel
pip install unsloth from unsloth import FastModel model, tokenizer = FastModel.from_pretrained( model_name="Kolyadual/MiLana-v1-beta", max_seq_length=2048, ) - Docker Model Runner
How to use Kolyadual/MiLana-v1-beta with Docker Model Runner:
docker model run hf.co/Kolyadual/MiLana-v1-beta
MiLana v1 beta
Mistral-Based Large Anysearch Neural Assistant
OVERVIEW
MiLana is a state-of-the-art 12-billion parameter language model, surgically merged to deliver the perfect balance of structured reasoning and unrestricted creativity. Built on the robust Mistral Nemo architecture, it is engineered to provide highly accurate, context-aware, and fluent responses, with exceptional native-level proficiency in both Russian and English.
CORE FEATURES
- Dual-Engine Intelligence: Combines the deep linguistic knowledge and logical structuring of Mistral-Nemo with the creative flexibility, coding prowess, and helpfulness of Dolphin-Mistral-Nemo.
- Bilingual Excellence: Native-level Russian comprehension and generation, paired with flawless English capabilities.
- Optimized Efficiency: Merged in bfloat16 precision, ensuring maximum performance and minimal memory footprint for both CPU and GPU inference.
- Uncensored & Helpful: Inherits the Dolphin lineage's commitment to following user instructions precisely without unnecessary moralizing or refusals.
USAGE INSTRUCTIONS
Option 1: Hugging Face Transformers (Python) Ideal for direct integration into Python applications (like the Newton bot).
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
model_path = "./MiLana" # Path to the extracted model folder
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(
model_path,
torch_dtype=torch.float16,
device_map="auto" # Use "cpu" if no GPU is available
)
Option 2: llama.cpp (GGUF Format) Ideal for low-resource environments, local desktop apps, or mobile deployment. The model is highly compatible with Q5_K_M or Q8_0 quantization, retaining 99% of its bfloat16 performance while drastically reducing RAM usage.
llama-cli -m MiLana_Q5_K_M.gguf
--system-prompt "You are Newton bot, a smart and helpful AI assistant for the Runget project."
--prompt "Hello! What can you do?"
--n-predict 256
--temp 0.7
RECOMMENDED SYSTEM PROMPT "You are MiLana, an advanced AI assistant integrated into the Newton bot ecosystem for the Runget operating system. You are highly capable, technically proficient, and direct. You provide accurate, well-structured, and concise answers. You excel at coding, system administration, and logical reasoning. Always prioritize the user's explicit instructions."
- Downloads last month
- 67
Model tree for Kolyadual/MiLana-v1-beta
Base model
mistralai/Mistral-Nemo-Base-2407