word level embedding

App/Transcription/Utils/audio_transcription.py

@@ -33,17 +33,18 @@ def transcribe_file(state, file_path, model_size="tiny"):
         with tqdm(total=total_duration, unit=" seconds") as pbar:
             for segment in segments:
                 segment_duration = segment.end - segment.start
-                temp = {
-                    "start": segment.start,
-                    "end": segment.end,
-                    "text": segment.text,
-                }
-                # time_stamp = "[%.2fs -> %.2fs]" % (segment.start, segment.end)
-                result.append(temp)
-                metadata.logs = "Transcribing.."
-                metadata.percentage = f"{((segment.end / total_duration)*100)}"
-                state.update_state(state="PROGRESS", meta=metadata.dict())
-                pbar.update(segment_duration)
+                for word in segment.words:
+                    temp = {
+                        "start": word.start,
+                        "end": word.end,
+                        "text": word.text,
+                    }
+                    # time_stamp = "[%.2fs -> %.2fs]" % (segment.start, segment.end)
+                    result.append(temp)
+            metadata.logs = "Transcribing.."
+            metadata.percentage = f"{((segment.end / total_duration)*100)}"
+            state.update_state(state="PROGRESS", meta=metadata.dict())
+            pbar.update(segment_duration)
     except Exception as e:
         metadata.logs = f"Falied error {e}"
         state.update_state(