Team Ai
Apppublic

obvAI/ai-character-chat

sourceHugging Faceupdated 9d agoView on Hugging Face
0likes
App README

AI Character Chat

A small, self-contained chat Space running Dolphin 3.0 Llama 3.1 8B on the Space's own GPU (ZeroGPU). It exists as the "owned capacity" counterpart to the generator's borrowed free lanes: one model, one Space, no third-party pool in the loop.

  • —Model: `dphn/Dolphin3.0-Llama3.1-8B` — an unrestricted instruction-tuned Llama 3.1 8B (~16 GB in bf16, fits the ZeroGPU large slice).
  • —UI: Gradio ChatInterface with an editable system prompt, max-new-tokens and temperature controls.
  • —Hardware: ZeroGPU (zero-a10g). Hardware is set at Space-creation time, not in this front-matter.

Run it yourself

bash
pip install -U huggingface_hub
hf upload <your-namespace>/ai-character-chat . --repo-type space

See NOTES.md for how to point the perchance generator at this Space, and for the exact hf commands used to create it.