Rank1: Test Time Compute in Reranking

NOTE: for demo purposes this is a quantized model limited to a 1024 context length. HF spaces cannot use vLLM so this is significantly slower

📄 Paper Link: https://arxiv.org/abs/2502.18418