add

README.md

@@ -12,11 +12,11 @@ short_description: 'Language Extractor: Polyglot Tagger'
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
-## Offline FLEURS cache
+## Offline caches
 
-The demo can now pull examples from a local, text-only FLEURS parquet cache instead of relying on Tatoeba.
+The demo now uses local parquet caches for both FLEURS and Tatoeba.
 
-Build the cache once with:
+Build the FLEURS cache once with:
 
 ```bash
 ./.venv/bin/python fleurs_cache.py
@@ -24,3 +24,11 @@ Build the cache once with:
 
 That downloads the FLEURS TSV metadata, dedupes repeated sentences, drops unused columns, and writes a reusable lean parquet file at `data/fleurs/fleurs_text_only.parquet`.
 Run it once while online; after that, the app reads only the local parquet and does not need the network.
+
+Build the Tatoeba cache once with:
+
+```bash
+./.venv/bin/python convert_tatoeba_sentences.py
+```
+
+That converts `sentences.csv` into `data/tatoeba/tatoeba_text.parquet` and keeps only the lean inference columns.

data/tatoeba/tatoeba_text.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb6078ec4bc24cde0b84d9637f9ae1c6cbe10125ab3ad188b57cf8834a090a22
+size 389231463