nvidia

/

parakeet-tdt_ctc-0.6b-ja

like 12

Follow

NVIDIA 4,859

I am not getting the recognition results. Please find the code used for testing.

import nemo
# Import Speech Recognition collection
import nemo.collections.asr as nemo_asr
# Import Natural Language Processing collection
import nemo.collections.nlp as nemo_nlp
# We'll use this to listen to audio
import IPython

# Load the Speech Recognition model (Citrinet) trained on a multilingual dataset
asr_model = nemo_asr.models.ASRModel.from_pretrained(model_name="nvidia/parakeet-tdt_ctc-0.6b-ja")

# Load the Neural Machine Translation model (Chinese to English)
nmt_model = nemo_nlp.models.MTEncDecModel.from_pretrained(model_name='nmt_zh_en_transformer6x6').cuda()

# Define the audio sample path
audio_sample = 'D:/Audio_end2end/audio_samples/male_test.wav'

# Transcribe the audio sample
transcribed_text = asr_model.transcribe([audio_sample])
print(transcribed_text)

results:
(['私は、2ギラボ音声チームによって新たにリリースされた生成系音声大規模モデルで快適で自然な音声合成能力を提供します。'], ['私は、2ギラボ音声チームによって新たにリリースされた生成系音声大規模モデルで快適で自然な音声合成能力を提 供します。'])