Update README.md

README.md

@@ -61,10 +61,11 @@ configs:
 | Gigaword   | test      | 4,000     | 80        | 12.5             | 11.2            | 24.1   |
 | DUC2003    | duc2003   | 624       | 80        | 2.1              | 12.2            | 27.5   |
 
-*CR (compression rate, %) = #words in summary / #words in transcription * 100. Lower is shorter.
+*CR (compression rate, %) = #words in summary / #words in transcription * 100. Lower is shorter summary.
 
 # Notes
 - The core set is identical to the first 50k samples of the train split.
+  - You may train your model and report the results only with the core set because the train split is too large. Using the entire train split is generally not recommended unless there are special reasons (e.g., to investigate the upper bound).
 - The duc2003 split has four reference summaries for each speech. You can report the best score from 4 scores.
 - Spoken sentences were generated using VITS [Kim+2021](https://proceedings.mlr.press/v139/kim21f.html) trained with LibriTTS-R [Koizumi+2023](https://www.isca-archive.org/interspeech_2023/koizumi23_interspeech.html).
 - More details and some experiments on this dataset can be found [here](https://www.isca-archive.org/interspeech_2024/matsuura24_interspeech.html#).