Skip to content

Commit

Permalink
Update id column name
Browse files Browse the repository at this point in the history
  • Loading branch information
liPatrick committed Sep 13, 2024
1 parent 050f54a commit 23806fb
Showing 1 changed file with 3 additions and 3 deletions.
6 changes: 3 additions & 3 deletions ultravox/tools/ds_tool/ds_tool.py
Original file line number Diff line number Diff line change
Expand Up @@ -242,7 +242,7 @@ def _map_batch_combine(self, batch):
audios = batch[self.audio_column_name]
sentences = batch[self.asr_column_name]
translations = batch[self.translation_column_name]
ids = batch["id"]
ids = batch[self.id_column_name]

combined_audio = {
"sampling_rate": audios[0]["sampling_rate"],
Expand All @@ -256,7 +256,7 @@ def _map_batch_combine(self, batch):
self.audio_column_name: [combined_audio],
self.asr_column_name: [combined_sentences],
self.translation_column_name: [combined_translations],
"id": combined_ids,
self.id_column_name: [combined_ids],
}
return new_batch

Expand Down Expand Up @@ -396,7 +396,7 @@ def process_and_upload_split_rescursive(
self.chunks_not_uploaded.append((start_index, end_index))
return None
failed_chunk_ranges.append((chunk_start, chunk_end))
successful_chunks = self.args.num_chunks - len(failed_chunk_ranges)
successful_chunks = total_chunks - len(failed_chunk_ranges)
print(
f"Finished processing and uploading {successful_chunks}/{total_chunks} chunks for range [{start_index}, {end_index})"
)
Expand Down

0 comments on commit 23806fb

Please sign in to comment.