Change8

b8152

📦 llama-cppView on GitHub →
1 features🐛 1 fixes🔧 6 symbols

Summary

This release introduces support for multi-modal context checkpoints on the server side and includes several internal code modifications and a bug fix related to sequence management.

Migration Steps

  1. If you rely on the size of the sequence length variable, note that it has been changed from n_tokens to int64_t.

✨ New Features

  • Server now supports multi-modal context checkpoints.

🐛 Bug Fixes

  • Fixed an issue where seqrm wasn't called in server-context.cpp.

Affected Symbols