MiniMax H3 — QuantFunc A4W4 INT4 — A4W4 Showcase
Loading Images
True 4-bit inference — A4W4: 4-bit activations × 4-bit weights on the QuantFunc INT4 inference backend.
AI-generated showcase from the official QuantFunc Hugging Face model card. The source does not specify every example's exact rank, seed or prompt.