bm-configurator 0.0.0

The configurator allows you to collect knowledge about models in configs and select among them suitable for local deployment on certain devices.

---

Model Configurator

A smart configuration system that collects model specifications in config files and selects suitable options for local deployment on target devices.

---

Key Features

• 🗂 Centralized Model Catalog - Manage all model specifications in YAML config files
• 🖥 Device Compatibility Engine - Automatically detects hardware capabilities:
  • RAM/VRAM availability
  • GPU support and memory
  • Disk space requirements
• 🎯 Smart Selection - Recommends optimal models based on:
  • Device constraints
  • Performance priorities
  • Model ranking system
• 🛠 Extensible Architecture - Supports multiple model types through config inheritance

---

Installation

pip install bm_configurator

From source:

git clone https://github.com/Dmatryus/bm_configurator
cd bm_configurator
pip install .

---

Basic Usage

from pathlib import Path

from bm_configurator import DiffusionModelConfig, LLMModelConfig, ModelConfigurator

configurator = ModelConfigurator(
    model_class=DiffusionModelConfig,
    config_path=Path("./configs/diffusion_models.yaml")
)

configurator.print_available_models()

---

Configuration Example (models.yaml)

models:
  - name: "Tiny Diffusion"
    model_id: "parlance/diffusion_mini"
    description: "Ultra-light CPU model"
    requirements:
      min_ram: 4
      requires_gpu: false
      disk_space: 1
    tags: ["cpu", "low-res", "fast"]
    image_size: [256, 256]
    default_steps: 15
    rank: 3

  - name: "SD 1.5 Optimized"
    model_id: "runwayml/stable-diffusion-v1-5"
    description: "The balance of quality and speed"
    requirements:
      min_ram: 8
      requires_gpu: true
      min_vram: 4
      disk_space: 2
    tags: ["mid-tier-gpu", "general-use"]
    image_size: [512, 512]
    default_steps: 25
    rank: 2

---

Contribution

We welcome contributions! Please see our Contribution Guidelines for:

• Adding new hardware detection methods
• Supporting additional model types
• Improving compatibility checks

---

License

MIT License. See LICENSE for full text.