ggml/llama.cpp releases: b10642

TL;DR

The ggml/llama.cpp project updated to track token IDs within KV cells, enhancing token management. This update is significant for improving the efficiency and functionality of language models using this framework.

Detailed Summary

The ggml/llama.cpp project released a new version (b10642) that includes an update to track token IDs within KV cells, which involves removing and moving functions related to previous tokens. This update is part of ongoing improvements in the project's codebase and can be tracked on GitHub. The broader impact could enhance the efficiency and functionality of language models built using this framework.

Key Points

• • kv: track token id
• • rm get_prev_tokens, move it to the main pr
• • add get_prev_tokens