The audio is recorded in mp4 format.
```
transcript = client.audio.transcriptions.create(
model="whisper-1",
file=audio_file
)
```
#Whisper API only transcribing first few seconds for audio recorded in safari browser in mac device.
wild shard
visual solar
Have you confirmed that your audio file indeed contains all of the audio that you need transcribed? Is the audio file in the expected correct format for the API? @wild shard
wild shard
@visual solar yes. I'm storing the audio in AWS S3 after sending it to the whisper API for transcription. I downloaded the mp4 from AWS S3 to verify this. What I found was the audio in mp4 contains all the expected content. But the whisper API only returns transcription of just an initial 1 or 2 seconds.
visual solar
Can you share the surrouding code before and after the whisper transcription request?
wild shard
Sure
```
audio_file = serializer.validated_data['audio_file'] ## blod data from frontend
stringified_messages = request.data.get('messages')
file_name = request.data.get('file_name')
messages = json.loads(stringified_messages)
try:
temp_file_path = handle_temporary_storage(
audio_file, 'audio', file_name)
except Exception as e:
return Response({"detail": str(e)}, status=status.HTTP_500_INTERNAL_SERVER_ERROR)
try:
user_message = AssessmentsAPI().transcribe_audio(temp_file_path)
```
```
def handle_temporary_storage(file, folder, key):
try:
directory_path = os.path.join(MEDIA_ROOT, folder)
if not os.path.exists(directory_path):
os.makedirs(directory_path)
file_path = os.path.join(directory_path, key)
with open(file_path, 'wb+') as destination:
for chunk in file.chunks():
destination.write(chunk)
return file_path
except Exception as e:
logger.error(f"Error handling temporary storage for file: {e}")
raise
```
```
@staticmethod
def transcribe_audio(file_path):
with open(file_path, 'rb') as audio_file:
transcript = client.audio.transcriptions.create(
model="whisper-1",
file=audio_file
)
return transcript.text
```