Evaluation is crucial for the development of information retrieval models. In recent years, a series of milestone works have been introduced to the community, such as MSMARCO, Natural Question (open-domain QA), MIRACL (multilingual retrieval), BEIR and MTEB (general-domain zero-shot retrieval). However, the existing benchmarks are severely limited in the following perspectives.

• Incapability of dealing with new domains. All of the existing benchmarks are static, which means they are established for the pre-defined domains based on human labeled data. Therefore, they are incapable of dealing with new domains which are interested by the users.
• Potential risk of over-fitting and data leakage. The existing retrievers are intensively fine-tuned in order to achieve strong performances on popular benchmarks, like BEIR and MTEB. Despite that these benchmarks are initially designed for zero-shot evaluation of O.O.D. Evaluation, the in-domain training data is widely used during the fine-tuning process. What is worse, given the public availability of the existing evaluation datasets, the testing data could be falsely mixed into the retrievers' training set by mistake.

• 🤖 Automated. The testing data is automatically generated by large language models without human intervention. Therefore, it is able to instantly support the evaluation of new domains at a very small cost. Besides, the new testing data is almost impossible to be covered by the training sets of any existing retrievers.
• 🔍 Retrieval and RAG-oriented. The new benchmark is dedicated to the evaluation of retrieval performance. In addition to the typical evaluation scenarios, like open-domain question answering or paraphrase retrieval, the new benchmark also incorporates a new setting called inner-document retrieval which is closely related with today's LLM and RAG applications. In this new setting, the model is expected to retrieve the relevant chunks of a very long documents, which contain the critical information to answer the input question.
• 🔄 Heterogeneous and Dynamic. The testing data is generated w.r.t. diverse and constantly augmented domains and languages (i.e. Multi-domain, Multi-lingual). As a result, it is able to provide an increasingly comprehensive evaluation benchmark for the community developers.

We plan to release new test dataset on regular basis. The latest version of is 24.04. You could check out the results at AIR-Bench Leaderboard.

Detailed results are available here.

This repo is used to maintain the codebases for running AIR-Bench evaluation. To run the evaluation, please install air-benchmark.

pip install air-benchmark

Refer to the steps below to run evaluations and submit the results to the leaderboard (refer to here for more detailed information).

1. Run evaluations

   • See the scripts to run evaluations on AIR-Bench for your models.

2. Submit search results

   • Package the output files

     • As for the results without reranking models,

     cd scripts
     python zip_results.py \
     --results_dir search_results \
     --retriever_name [YOUR_RETRIEVAL_MODEL] \
     --save_dir search_results

     • As for the results with reranking models

     cd scripts
     python zip_results.py \
     --results_dir search_results \
     --retriever_name [YOUR_RETRIEVAL_MODEL] \
     --reranker_name [YOUR_RERANKING_MODEL] \
     --save_dir search_results

   • Upload the output .zip and fill in the model information at AIR-Bench Leaderboard

This work is inspired by MTEB and BEIR. Many thanks for the early feedbacks from @tomaarsen, @Muennighoff, @takatost, @chtlp.

TBD