Personal CFO Nemotron v2
LoRA adapter for nvidia/Llama-3.1-Nemotron-Nano-8B-v1, fine-tuned for personal transaction classification.
For base-model architecture, capabilities, limitations, and licensing, see the NVIDIA Nemotron Nano 8B model card.
Fine-tuning
Trained on NVIDIA DGX Spark with Unsloth using QLoRA.
- LoRA rank: 8
- LoRA alpha: 32
- 2 epochs
- 530 training examples
- 134 validation examples
- 100 held-out test examples
The adapter predicts:
{
"category": "...",
"subcategory": "...",
"protected": false,
"one_time_exceptional": false
}
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for shashanksabhlok/personal-cfo-nemotron-v2
Base model
nvidia/Llama-3.1-Nemotron-Nano-8B-v1