Celeste Imperia | Hardware-Aware AI Forge
Bridging the gap between frontier AI architectures and consumer silicon.
Celeste Imperia specializes in the precision optimization of LLMs, VLMs, and Diffusion architectures. We provide hardware-validated weights optimized for private, zero-latency execution across Qualcomm, Intel, and NVIDIA ecosystems.
🏗️ The Development Forge: Workstation V2.0
All models are forged on a specialized dual-GPU pipeline to ensure stability across both "Masses-Ready" and "Professional" hardware.
| Component | Legacy Validation Rig | Current AI Workstation |
|---|
| Processor | Intel Core i5-11400 | Intel Core i5-11400 (6C/12T) |
| Primary GPU | NVIDIA RTX A4000 (16GB) | NVIDIA RTX 3090 (24GB GDDR6X) |
| Secondary GPU | - | NVIDIA RTX A4000 (16GB GDDR6) |
| Memory | 16GB RAM | 64GB DDR4 RAM |
| Storage | Standard SSD | High-Speed NVMe, and HDD Storage Pool |
📂 Active Repositories
🐉 Snapdragon & Mobile NPU (QNN Native)
- SDXL-QNN - Optimized for Qualcomm Hexagon NPU. Includes Python and C# automation tools.
🧠 Domain-Specific Reasoning
- Llama-3-Indian-Finance - Specialized reasoning for Indian financial and regulatory sectors.
🎨 Cross-Platform Optimization (OpenVINO / GGUF)
- SDXL-OpenVINO-Trinity - 4-step generation for Intel i5/i7.
- Whisper-Large-V3-Turbo - High-speed speech-to-text.
- Qwen2-VL-2B-INT4 - Vision-language reasoning.
- Llama-3.2-1B-GGUF / Phi-3.5-GGUF - Mobile-ready LLMs.
☕ Support the Forge
Maintaining a dual-GPU AI workstation and hosting high-bandwidth models requires significant resources. If our open-source tools power your projects, consider supporting our development:
Scan to support via UPI (India Only):
Connect with the architect: Abhishek Jaiswal on LinkedIn 
