ArniXS

Model Details

  • Architecture: Transformer-based (6 layers, 6 attention heads)
  • Parameters: ~50M
  • Context Length: 64 tokens
  • Vocabulary Size: 14,704 tokens

Hardware Requirements

Minimum requirements:

  • CUDA-capable GPU with 4GB VRAM
  • 8GB RAM
  • Python 3.8+

Recommended:

  • NVIDIA GPU with 8GB+ VRAM (tested on RTX 3070)
  • 16GB+ RAM
  • Python 3.10+
Downloads last month
4
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support