Fix queries schema: drop stray `topic` column so load_dataset works

The `queries` config fails `load_dataset` with a `CastError`:

```
datasets.table.CastError: Couldn't cast ... topic ... to {...no topic...} because column names don't match
```

The `queries` parquet carries an extra **`topic`** column that is not in the config's declared `features` (the standard 8: `original_query, query, answer, score, source_url, frag_start_char, frag_end_char, n_gold`, same as `QasperChunkRetrieval`). `datasets` casts the parquet to the declared schema and errors because `topic` isn't listed. The `documents-*` configs also carry `topic` but declare it consistently, so only `queries` is affected.

This PR drops the stray `topic` column from `queries` so the data matches the declared features and the rest of the q2c family (row count unchanged: 672). Alternatively `topic` could be kept by adding it to the queries features — happy to switch if you'd rather preserve it.

queries/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4baee289386fa6f6d422d22673f483cf9ddbd52dcff964dc8c25bcbf91991241
-size 30252166
+oid sha256:7b3dc696494372c274dcaeba85c6d72f1956b21e6ada0dd5e819d9977d966915
+size 11362372