krzysztof/landing_page
1
0
Fork 0
mirror of https://github.com/qdrant/landing_page.git synced 2026-09-30 00:18:32 +02:00
Code Issues Packages Projects Releases Wiki Activity

Apply suggestions from code review

Browse Source 
Co-authored-by: Jojii <15957865+JojiiOfficial@users.noreply.github.com>
This commit is contained in:
Tim Visée
2026-05-08 17:03:33 +02:00
committed by GitHub
co-authored by Jojii
parent 7048d0cd54
commit 2746db8762
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
qdrant-landing/content/blog/qdrant-1.18.x.md
+2 -2
View File
@@ -42,7 +42,7 @@ Qdrant's implementation of TurboQuant extends the original algorithm to close th
#### TurboQuant vs Scalar Quantization
The following table shows recall@10 for 4-bit TurboQuant (TQ4) compared to uncompressed vectors (F32) and scalar quantization (SQ) across four benchmarked datasets. Benchmarks were run with HNSW configured with `m=16` and `ef_construct=128`.
The following table shows recall for 4-bit TurboQuant (TQ4) compared to uncompressed vectors (F32) and scalar quantization (SQ) across four benchmarked datasets. Benchmarks were run with HNSW configured with `m=16` and `ef_construct=128`.
| Dataset | F32 | SQ | TQ4 |
|---|---|---|---|
@@ -55,7 +55,7 @@ Compared to scalar quantization, TurboQuant delivers similar recall **at double
#### TurboQuant vs Binary Quantization
The following table shows recall@10 for 1-bit TurboQuant (TQ1) compared to uncompressed vectors (F32) and 1-bit binary quantization (BQ1) across four benchmarked datasets. Benchmarks were run with HNSW configured with `m=16` and `ef_construct=128`.
The following table shows recall for 1-bit TurboQuant (TQ1) compared to uncompressed vectors (F32) and 1-bit binary quantization (BQ1) across four benchmarked datasets. Benchmarks were run with HNSW configured with `m=16` and `ef_construct=128`.
| Dataset | F32 | BQ1 | TQ1 |
|---|---|---|---|