readme : update hot topics

Author: ggerganov

README.md

@@ -20,7 +20,8 @@ Inference of Meta's [LLaMA](https://arxiv.org/abs/2302.13971) model (and others)
 
 ### Hot topics
 
-- **MoE memory layout has been updated - reconvert models for `mmap` support and regenerate `imatrix` https://github.com/ggerganov/llama.cpp/pull/6387**
+- **BPE pre-tokenization support has been added: https://github.com/ggerganov/llama.cpp/pull/6920**
+- MoE memory layout has been updated - reconvert models for `mmap` support and regenerate `imatrix` https://github.com/ggerganov/llama.cpp/pull/6387
 - Model sharding instructions using `gguf-split` https://github.com/ggerganov/llama.cpp/discussions/6404
 - Fix major bug in Metal batched inference https://github.com/ggerganov/llama.cpp/pull/6225
 - Multi-GPU pipeline parallelism support https://github.com/ggerganov/llama.cpp/pull/6017