The community member is using LlamaIndex and is curious about using a local GPTSimpleIndex vs an external one like MongoDB Atlas or Rocketset. Their total query time is 18 seconds and they are trying to get it down as fast as possible, using a couple of large CSV files.
The comments suggest that external indexes may be faster, but 18 seconds for a full retrieval and completion is not that long. The community members are asked about the size of their dataset, which is around 3 MB of CSV data. They are also asked if they have timed the retrieval part and synthesis, as the retrieval is typically fast even with a local index unless there is a lot of data.
There is no explicitly marked answer in the provided information.
Hello, I have been using LlamaIndex for awhile and love it, but I am curious in terms of using the local GPTSimpleindex vs an external one like mongo atlas, or rocketset, etc. My total query time is 18 sec and i am trying to get that down as fast as possible, using a couple large csv files.
Its about 3 MB of CSV data and and in terms of measuring each stage beyond prints trying to find a better way to benchmark, (though I may be just being a noob here)
