refactor: use transposed centroids in PQ preprocess

[usamoi]

With this PR, on GIST, (nlist = 4096, residual_quantization = true)

---

nprobe = 300

• IVF-RQ 0.9843 2.5s; 1x memory usage
• IVF-PQ 0.9914 3.3s (ratio = 2, bits = 4; pq_rerank_size = 100); 2x memory usage
• IVF-PQ 0.9812 2.7s (ratio = 3, bits = 4; pq_rerank_size = 188); 1.5x memory usage
• IVF-PQ 0.9806 2.5s (ratio = 4, bits = 4; pq_rerank_size = 264); 1x memory usage
• IVF-PQ 0.9526 2.3s (ratio = 8, bits = 4; pq_rerank_size = 720); 0.5x memory usage

---

nprobe = 200

• IVF-RQ 0.9720 1.9s; 1x memory usage
• IVF-PQ 0.9731 2.3s (ratio = 2, bits = 4; pq_rerank_size = 100); 2x memory usage
• IVF-PQ 0.9696 2.0s (ratio = 3, bits = 4; pq_rerank_size = 188); 1.5x memory usage
• IVF-PQ 0.9700 1.9s (ratio = 4, bits = 4; pq_rerank_size = 264); 1x memory usage
• IVF-PQ 0.9452 1.8s (ratio = 8, bits = 4; pq_rerank_size = 720); 0.5x memory usage

---

nprobe = 100

• IVF-RQ 0.9258 1.3s; 1x memory usage
• IVF-PQ 0.9250 1.5s (ratio = 2, bits = 4; pq_rerank_size = 100); 2x memory usage
• IVF-PQ 0.9276 1.3s (ratio = 3, bits = 4; pq_rerank_size = 188); 1.5x memory usage
• IVF-PQ 0.9261 1.3s (ratio = 4, bits = 4; pq_rerank_size = 264); 1x memory usage
• IVF-PQ 0.9117 1.3s (ratio = 8, bits = 4; pq_rerank_size = 720); 0.5x memory usage

---