The community member who posted the original post is exporting multiple channels within a server, and they are finding it tricky to create separate indices as the time range of the conversations can be over 2 years in some cases. They express disappointment that GPT is not good with dates.
In the comments, another community member suggests that it might be possible to get the language model to code a Python filter using the datetime library, and wonders if this could be added as a feature to Llama Index. They also think the "tool approach" of using common date ranges like all time, last year, last 6 months, etc. is viable, as it would only require 6 indexes/tools, and it would be easy to filter the messages before creating the index by parsing the timestamps.
The original poster responds by saying they will try the approach suggested by the other community member.
Hmm this would work but we are exporting multiple channels within a server. it might be a little tricky to create seperate indices as the time range of the conversations can be over 2 years in some cases. It's so sad that GPT is not really good with dates
It might be possible to get the LLM to code a python filter using the datetime library. I wonder if this could be added as a feature to llama index 🤔 GPTDateTimeIndex or something lol
Anyways though, I still think the tool approach is viable. Using common date ranges like all time, last year, last 6 month, last month, last week, last 24 hours, is only 6 indexes/tools, not too bad. Easy to filter the messages before creating the index since you can parse the timestamps
