Aura-1-Lightning ⚡ Request Access Form
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
Please fill out your deployment profile to unlock the model binaries instantly.
Log in or Sign Up to review the conditions and access this model content.
Aura-1-Lightning ⚡
A high-performance, structurally distilled 7B causal language model optimized for rapid local inference and reasoning density. Developed under the Waveforce AI framework, this model was trained using a custom hybrid CPU/GPU cross-device execution pipeline designed to maximize optimization stability under tight hardware resource thresholds.
🔒 Terms & Conditions of Access (Model Gate)
By requesting access to this repository, you explicitly agree to the following binding operational terms:
- Strict Safety Alignment & Operational Restrictions: You are explicitly prohibited from intentionally adversarial prompting, jailbreaking, or altering the model's baseline safety/sampling configurations to trick, force, or compel the architecture to comply with, generate scripts for, or assist in any terrorist activities, violent extremism, or illegal operations.
- Telemetry & Registration Authorization: You grant Waveforce AI explicit permission to log your registration details (Full Name, DOB, Use Case, and Entity telemetry). This user list is compiled and reviewed as an automated report for local model evaluation, error tracking, and architectural improvement in future training runs.
📖 Model Card
📢 Access Notice: If you have not accepted the gate agreement above, please complete the form fields to unlock the structural assets.
:::gated
🔬 Architectural Strategy & Distillation Profile
Traditional quantization often degrades a model's latent logic representation. Aura-1-Lightning avoids this degradation by employing True Cross-Device Probabilistic Logit Alignment.
During training, the student model's base parameters were frozen utilizing a dual 4-bit QLoRA configuration to fit execution within standard hardware limits. Concurrently, the unreduced logit streams from both the student and teacher models were cross-compared natively on their respective devices. By slicing matching vocabulary spaces and routing the heavy matrix operations directly to CPU memory, the training loop successfully optimized structural Kullback-Leibler (KL) divergence alongside hard cross-entropy targets without memory overflow.
Model Ancestry & Credits
To give credit to the engineering teams behind the base architectures:
- Student Base Model: Qwen/Qwen2.5-7B-Instruct — Celebrated for its massive vocabulary size ($151,643$ tokens) and incredible base coding and mathematical reasoning density.
- Teacher Foundation Model: google/gemma-2-9b-it — Providing the high-precision probabilistic curves that guided the student's hidden states.
📈 Performance & Hardware Footprint
By compiling the learned adapter weights directly back into the unquantized base layers using full precision math, the final repository is delivered as a pure standalone model (native .safetensors format). It requires no custom wrappers or loose adapter configurations to run.
VRAM Allocation Metrics
- Base Weight Footprint (FP16/BF16): ~14.5 GB (Fits perfectly within single-card consumer GPUs, including RTX 3090/4090, T4, or L4 systems).
- Quantized Deployment (GGUF / Int4 / NF4): ~4.5 GB to 5.2 GB (Ideal for ultra-low latency execution on local laptops, embedded devices, and local AI workflows like AnythingLLM).
- Context Optimization: Fully configured with native
.jinjachat templates supporting strict multi-turn dialogue boundaries up to its optimized token windows.
📦 Repository Structure
The aura-1-lightning-weights directory contains all components required for a native native execution:
waveforce-ai/Aura-1-Lightning/
└── aura-1-lightning-weights/
├── model.safetensors # Compiled, unquantized standalone weights
├── config.json # Native layer and attention head geometry
├── generation_config.json # Default inference sampling hyper-parameters
├── tokenizer.json # Vocabulary token mappings (151k token capacity)
├── tokenizer_config.json # Chat layout definitions
└── chat_template.jinja # Structural instruction template wrapper