Merge pull request #91 from Wordcab/90-fix-word_timestamps

Implement word_timestamps for batching
Wordcab · Jun 7, 2023 · c231039 · c231039
2 parents c99a717 + 6e2a59a
commit c231039
Show file tree

Hide file tree

Showing 4 changed files with 298 additions and 29 deletions.
diff --git a/wordcab_transcribe/services/align_service.py b/wordcab_transcribe/services/align_service.py
@@ -419,9 +419,9 @@ def align(
                 start, end, score = None, None, None
                 if cdx in segment["clean_cdx"]:
                     char_seg = char_segments[segment["clean_cdx"].index(cdx)]
-                    start = round(char_seg.start * ratio + t1, 3)
-                    end = round(char_seg.end * ratio + t1, 3)
-                    score = round(char_seg.score, 3)
+                    start = round(char_seg.start * ratio + t1, 2)
+                    end = round(char_seg.end * ratio + t1, 2)
+                    score = round(char_seg.score, 2)
 
                 char_segments_arr.append(
                     {
@@ -468,7 +468,7 @@ def align(
                     word_chars = word_chars[word_chars["char"] != " "]
                     word_start = word_chars["start"].min()
                     word_end = word_chars["end"].max()
-                    word_score = round(word_chars["score"].mean(), 3)
+                    word_score = round(word_chars["score"].mean(), 2)
 
                     # -1 indicates unalignable
                     word_segment = {"word": word_text}

diff --git a/wordcab_transcribe/services/asr_service.py b/wordcab_transcribe/services/asr_service.py
@@ -201,7 +201,7 @@ async def process_input(
             "dual_channel": dual_channel,
             "source_lang": source_lang,
             "timestamps_format": timestamps_format,
-            "word_timestamps": False,  # TODO: Implement word timestamps, False for now
+            "word_timestamps": word_timestamps,
             "post_processed": False,
             "transcription_result": None,
             "transcription_done": asyncio.Event(),