Update README with latest bm25 metrics and tokenization strategies

README.md

@@ -204,12 +204,9 @@ A lightweight, evaluation-ready subset of [CodeSearchNet](https://huggingface.co
 
 ## BM25 tokenization strategy
 
-- Auto selection chooses the best strategy per split using retrieval metrics.
-- Strategy types:
-  - `transformer`: `Qwen/Qwen3-0.6B` tokenizer
-  - `wordseg`: language-specific word segmentation (`ja`, `zh`, `th`, `ko`)
-  - `stemmer`: `PyStemmer`
-  - `whitespace`: `str.split()`
+- `bm25` was generated with `--auto-select-best-splitter`.
+- Candidate strategies: `transformer`, `stemmer`, `wordseg`, `nltk_stem`, `nltk_stem_stop`, `english_regex`, `english_porter`, `english_porter_stop`, `whitespace`.
+- Selection metric: best `nDCG@100` per split.
 
 Selected strategy by split:
 
@@ -229,4 +226,4 @@ Selected strategy by split:
 
 ## License
 
-Other. This dataset is derived from CodeSearchNet and follows upstream repository licensing and attribution requirements.
+Other. This dataset is derived from CodeSearchNet and follows upstream repository licensing and attribution requirements.