Spaces:
Runtime error
Runtime error
| import gradio as gr | |
| import librosa | |
| import soundfile as sf | |
| import torch | |
| import warnings | |
| import os | |
| from transformers import Wav2Vec2ProcessorWithLM, Wav2Vec2CTCTokenizer, Wav2Vec2Model | |
| warnings.filterwarnings("ignore") | |
| from speechbrain.pretrained import EncoderDecoderASR | |
| asr_model = EncoderDecoderASR.from_hparams(source="speechbrain/asr-wav2vec2-commonvoice-rw", savedir="pretrained_models/asr-wav2vec2-commonvoice-rw") | |
| #asr_model.transcribe_file("speechbrain/asr-wav2vec2-commonvoice-rw/example.mp3") | |
| # define speech-to-text function | |
| def asr_transcript(audio, audio_microphone, model_params): | |
| audio = audio_microphone if audio_microphone else audio | |
| if audio == None and audio_microphone == None: | |
| return "Please provide audio by uploading a file or by recording audio using microphone by pressing Record (And allow usage of microphone)", "Please provide audio by uploading a file or by recording audio using microphone by pressing Record (And allow usage of microphone)" | |
| text = "" | |
| if audio: | |
| text = asr_model.transcribe_file(audio.name) | |
| return text | |
| else: | |
| return "File not valid" | |
| gradio_ui = gr.Interface( | |
| fn=asr_transcript, | |
| title="Kinyarwanda Speech Recognition", | |
| description="Upload an audio clip or record from browser using microphone, and let AI do the hard work of transcribing.", | |
| article = """ | |
| This demo showcases the pretrained model from deepspeech. | |
| """, | |
| inputs=[gr.inputs.Audio(label="Upload Audio File", type="file", optional=True), gr.inputs.Audio(source="microphone", type="file", optional=True, label="Record from microphone"), gr.inputs.Dropdown(choices=["deepspeech","coqui (soon)"], type="value", default="deepspeech", label="Select speech recognition model ", optional=False)], | |
| outputs=[gr.outputs.Textbox(label="Recognized speech")], | |
| examples = [["sample_1.wav","sample_1.wav","deepspeech"],["sample_2.wav","sample_2.wav","deepspeech"]] | |
| ) | |
| gradio_ui.launch(enable_queue=True) |