Translation not available in google and aws

sul-dlss · Apr 13, 2024 · 5a1fd1b · 5a1fd1b
1 parent f246a3d
commit 5a1fd1b
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 0 deletions.
diff --git a/transcribe/aws.py b/transcribe/aws.py
@@ -24,6 +24,10 @@ def run(output_dir):
         file_metadata["run_count"] = len(results) + 1
         file = file_metadata["media_filename"]
 
+        if file_metadata["media_language"] != file_metadata["transcript_language"]:
+            logging.info("skipping since google doesn't support translation")
+            continue
+
         logging.info("transcribing with aws %s", file)
 
         start_time = datetime.datetime.now()

diff --git a/transcribe/google.py b/transcribe/google.py
@@ -18,6 +18,11 @@ def run(output_dir):
     for file_metadata in tqdm.tqdm(utils.get_data_files(), desc="google".ljust(10)):
         file_metadata["run_count"] = len(results) + 1
         file = file_metadata["media_filename"]
+
+        if file_metadata["media_language"] != file_metadata["transcript_language"]:
+            logging.info("skipping since google doesn't support translation")
+            continue
+
         logging.info(f"running google speech-to-text with {file}")
 
         start_time = datetime.datetime.now()