Skip to content

Latest commit

 

History

History
25 lines (24 loc) · 1.79 KB

rag_eval_multiple_domains_summary_en.md

File metadata and controls

25 lines (24 loc) · 1.79 KB

RAG Evaluations in LlamaIndex

Multiple Domains Scenarios in ["en"]

Embedding Models WithoutReranker
[hit_rate/mrr]
CohereRerank
[hit_rate/mrr]
bge-reranker-large
[hit_rate/mrr]
bge-reranker-v2-m3
[hit_rate/mrr]
bce-reranker-base_v1
[hit_rate/mrr]
OpenAI-ada-2 85.05/62.29 91.72/72.77 91.83/72.17 89.57/70.33 92.90/77.17
OpenAI-embed-3-small 87.20/61.64 91.72/72.60 92.15/72.22 90.65/70.63 92.80/76.46
OpenAI-embed-3-large 85.38/61.79 90.97/72.25 91.29/71.97 89.89/70.71 91.72/76.37
bge-large-en-v1.5 84.62/61.22 91.51/72.71 91.94/72.35 89.35/70.80 92.47/76.61
bge-m3-large 86.67/64.22 92.15/73.19 92.69/72.04 89.68/70.52 93.33/77.24
llm-embedder 77.53/56.10 86.34/69.36 86.56/68.78 89.35/70.61 87.42/73.44
CohereV3-en 80.65/58.33 87.96/70.09 88.71/69.61 86.56/68.63 89.03/74.06
CohereV3-multilingual 83.33/60.70 90.54/72.41 90.43/72.11 88.60/70.34 90.97/76.26
JinaAI-v2-Base-en 81.94/58.03 90.32/71.65 90.75/71.11 88.06/68.87 91.29/75.53
gte-large-en 83.44/59.18 90.97/72.24 91.61/72.38 89.25/70.57 92.26/76.52
e5-large-v2-en 85.05/61.90 91.18/71.45 91.18/70.56 89.68/69.59 92.37/75.81
e5-large-multilingual 85.91/61.87 93.01/73.38 92.80/72.45 90.97/70.94 93.44/77.31
bce-embedding-base_v1 87.42/63.93 92.69/73.34 93.33/73.06 91.08/71.31 93.87/77.88