from transformers import WhisperProcessor, WhisperForConditionalGeneration
from datasets import load_dataset
import librosa
import time

processor = WhisperProcessor.from_pretrained("manifestasi/whisper-tiny-manifestasi-indo-v2")
model = WhisperForConditionalGeneration.from_pretrained("manifestasi/whisper-tiny-manifestasi-indo-v2")
model.config.forced_decoder_ids = None

a = time.time()
audio, rate = librosa.load("/kaggle/input/berhasil-konek/berhasil_retina.mp3", sr = 16000)
input_features = processor(audio, sampling_rate=16000, return_tensors="pt").input_features 

predicted_ids = model.generate(input_features)
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=False)
print(transcription)
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
b = time.time() - a
print(f"{b} detik")