3 ms·I'm currently using a 32B Qwen model for summarization and it's pretty good.by dimgl 2y agoI'm currently using a 32B Qwen model for summarization and it's pretty good.