Information Retrieval on CoIR (test)
26.6Apps ScoreRevela
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| RevelaModel Size=3B, Supervised=false2025.06 | 26.6 | 29 | 55.9 | 62.6 | 69.1 | 83.2 | 34.5 | 88.3 | 78.8 | 73 | 60.1 | |
| Voyage-2Model Size=N/A, Type=Proprietary API2025.06 | 26.5 | 29.8 | 69.3 | 81.8 | 72.8 | 27.3 | 28.7 | 81.8 | 65.4 | 28.7 | 56.3 | |
| E5-MistralModel Size=7B, Supervised=true, Used code-related pairs during pre-training=true2025.06 | 23.5 | 33.2 | 68 | 67.4 | 64.8 | 80.6 | 31.7 | 91 | 76.4 | 36.4 | 57.3 | |
| RevelaModel Size=0.5B, Supervised=false2025.06 | 20.5 | 27.5 | 53.7 | 57.9 | 68 | 77.6 | 35.4 | 82.5 | 74.5 | 63.6 | 56.1 | |
| RevelaModel Size=1B, Supervised=false2025.06 | 19.4 | 30.2 | 55 | 64 | 70 | 81.1 | 34.2 | 85.7 | 76.2 | 70.4 | 58.6 | |
| REPLUGModel Size=3B, Supervised=false2025.06 | 17.7 | 25.2 | 56.8 | 53.7 | 62.8 | 62.1 | 34.4 | 78.1 | 71.7 | 49.1 | 53.9 | |
| BGEModel Size=0.6B, Supervised=true, Used code-related pairs during pre-training=true2025.06 | 14.7 | 26.4 | 46.9 | 58.3 | 53.7 | 62.6 | 30.2 | 80.6 | 69.3 | 47.9 | 49.1 | |
| REPLUGModel Size=1B, Supervised=false2025.06 | 13.9 | 20.1 | 53.9 | 50.8 | 62.8 | 61.5 | 33.3 | 76.3 | 66 | 42.9 | 48.2 | |
| E5-PTModel Size=0.3B, Supervised=true, Used code-related pairs during pre-training=true2025.06 | 10.6 | 27.1 | 48.9 | 35.2 | 50.5 | 56.3 | 32.2 | 86.9 | 70.4 | 46.2 | 46.4 | |
| Ada-2Model Size=N/A, Type=Proprietary API2025.06 | 8.7 | 28.9 | 58.3 | 74.2 | 53.3 | 26 | 17.7 | 74.2 | 47.1 | 17.7 | 45.6 | |
| RevelaModel Size=0.1B, Supervised=false2025.06 | 8.2 | 26.2 | 45.7 | 49.9 | 63.4 | 70.9 | 34.6 | 69.2 | 63.8 | 51.7 | 48.4 | |
| BM25Model Size=N/A, Supervised=false2025.06 | 4.8 | 15.6 | 25 | 40.9 | 54 | 47.8 | 34.4 | 70.2 | 68.1 | 59.2 | 42 | |
| UniXModel Size=0.1B, Supervised=true2025.06 | 1.4 | 25.1 | 50.5 | 60.2 | 58.4 | 41.8 | 31 | 44.7 | 36 | 24.2 | 37.3 |