Benchmark
VN-MTEB (vie, v1) - Classification
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 1.22
Score distribution
- 57.9%
- Mean
- 63.1%
- Median
- 10%
- Std dev
- 44.2%
- 10th percentile
- 67.4%
- 90th percentile
- 43.6%
- Min
This benchmark has a weight of 1.22 for embedding models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | BGE M3Classification· MTEB Extra Text Boards data | 67.4% |
| 2 | Multilingual E5 Large InstructClassification· MTEB Extra Text Boards data | 66.5% |
| 3 | Multilingual E5 LargeClassification· MTEB Extra Text Boards data | 63.1% |
| 4 | E5 Mistral 7B InstructClassification· MTEB Extra Text Boards data | 62.3% |
| 5 | All MiniLM L12 V2Classification· MTEB Extra Text Boards data | 44.2% |
| 6 | All MiniLM L6 V2Classification· MTEB Extra Text Boards data | 43.6% |