MM-NIAH

benchmark

[NeurIPS 2024] Needle In A Multimodal Haystack (MM-NIAH): A comprehensive benchmark designed to systematically evaluate the capability of existing MLLMs to comprehend long multimodal documents.

How to download and setup MM-NIAH

Open terminal and run command

git clone https://github.com/OpenGVLab/MM-NIAH.git

git clone is used to create a copy or clone of MM-NIAH repositories. You pass git clone a repository URL.
it supports a few different network protocols and corresponding URL formats.

Also you may download zip file with MM-NIAH https://github.com/OpenGVLab/MM-NIAH/archive/master.zip

Or simply clone MM-NIAH with SSH

[email protected]:OpenGVLab/MM-NIAH.git

If you have some problems with MM-NIAH

You may open issue on MM-NIAH support forum (system) here: https://github.com/OpenGVLab/MM-NIAH/issues

Similar to MM-NIAH repositories

Here you may see MM-NIAH alternatives and analogs

netdata primesieve fashion-mnist FrameworkBenchmarks BenchmarkDotNet jmeter awesome-semantic-segmentation sysbench hyperfine tsung benchmark_results across web-frameworks php-framework-benchmark jsperf.com go-web-framework-benchmark huststore phoronix-test-suite Attabench ann-benchmarks sbt-jmh caffenet-benchmark chillout IocPerformance prophiler TBCF NBench sympact awesome-http-benchmark BlurTestAndroid