https://github.com/qvink/SillyTavern-MessageSummarize
Nothing changed but 2 things:
- Summarize requests now can be configured to send parallel requests (boosts speed of summarizing if llm supports many requests per time)
- Now you can choose from what side summarize starts (from start of chat or from end)
