Skip to content
This repository has been archived by the owner on Jan 16, 2025. It is now read-only.

[DO NOT MERGE UNTIL RERANK ON AZURE + UPDATES TO BENCHMARKS] added rerank benchmarks #155

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions public_rerank_benchmarks/bm25_with_rerank.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
| Model | First Stage| BEIR Evaluation (14 Datasets)| Code Evaluation (6 Datasets)| Long Context Evaluation (7 Datasets)| Multilingual (18 Datasets)| Semi-Structured Data Evaluation (5 Datasets)
| -------- | -------- | -------- | -------- | -------- | -------- | -------- |
|N/A|BM25|43.7|34.0|54.6|36.5|47.5|
|Rerank v2|BM25|49.4|37.6|63.1|62.8|60.3|
|Rerank v3|BM25|53.0|51.7|69.0|70.8|62.7|
7 changes: 7 additions & 0 deletions public_rerank_benchmarks/embed_with_rerank.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
| Model | First Stage| Code Evaluation (6 Datasets)| Long Context Evaluation (7 Datasets)| Multilingual (18 Datasets)| Semi-Structured Data Evaluation (5 Datasets)
| -------- | -------- | -------- | -------- | -------- | -------- |
|N/A|embed-v3.0|65.0|56.8|TBU|47.8
|Rerank v2|embed-v3.0|77.6|62.8|TBU|TBU
|Rerank v3|embed-v3.0|64.5|69.3|TBU|62.7

Note: For Multilingual, both we used `embed-multilingual-v3.0`, `rerank-multilingual-v2.0`, and `rerank-multilingual-v3.0`. All other evaluations were done with the english model variants
7 changes: 7 additions & 0 deletions public_rerank_benchmarks/miracl.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
| Model | First Stage| Arabic (ar) | Bengali (bn) | English (en) | Spanish (es) | Persian (fa) | Finnish (fi) | French (fr) | Hindi (hi) | Indonesian (id) | Japanese (ja) | Korean (ko) | Russian (ru) | Swahili (sw) | Telugu (te) | Thai (th) | Chinese (zh) | Germany (de) | Yoruba (yo) | Avg (18 datasets) | Avg (excl. de and yo) |
| -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- |
| BM25 | N/A | 48.1 | 50.8 | 35.1 | 31.9 | 33.3 | 55.1 | 18.3 | 45.8 | 44.9 | 36.9 | 41.9 | 33.4 | 38.3 | 49.4 | 48.8 | 18.0 | N/A | N/A | N/A | 39.4 |
| embed-multilingual-v3.0 | N/A | 76.5 | 75.8 | 57.0 | 55.1 | 57.5 | 77.1 | 57.4 | 61.7 | 52.5 | 69.6 | 66.0 | 68.8 | 75.7 | 83.3 | 79.5 | 58.9 | 58.7 | 61.8 | 66.3 | 67.0 |
| Rerank 3 | BM25 | 75.3 | 78.6 | 60.7 | 55.1 | 56.1 | 77.5 | 47.2 | 62.2 | 58.1 | 67.4 | 44.0 | 55.9 | 70.3 | 72.6 | 77.3 | 50.9 | 45.0 | 76.5 | 62.8 | 63.1 |
| Rerank 3 | embed-multilingual-v3.0 | 80.4 | 82.5 | 61.4 | 57.0 | 62.4 | 80.8 | 58.4 | 62.9 | 57.5 | 75.1 | 74.5 | 67.3 | 77.2 | 83.8| 82.8 | 65.7 | 60.8 | 83.1 | 70.8 | 70.6 |