Spaces:
Build error
Build error
| import os | |
| os.system("pip install gradio==2.8.0b2") | |
| import gradio as gr | |
| import librosa | |
| from transformers import AutoFeatureExtractor, AutoTokenizer, SpeechEncoderDecoderModel | |
| model_name = "facebook/wav2vec2-xls-r-2b-21-to-en" | |
| feature_extractor = AutoFeatureExtractor.from_pretrained(model_name) | |
| tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False) | |
| model = SpeechEncoderDecoderModel.from_pretrained(model_name) | |
| def process_audio_file(file): | |
| data, sr = librosa.load(file) | |
| if sr != 16000: | |
| data = librosa.resample(data, sr, 16000) | |
| input_values = feature_extractor(data, return_tensors="pt").input_values | |
| return input_values | |
| def transcribe(file_mic, file_upload): | |
| warn_output = "" | |
| if (file_mic is not None) and (file_upload is not None): | |
| warn_output = "WARNING: You've uploaded an audio file and used the microphone. The recorded file from the microphone will be used and the uploaded audio will be discarded.\n" | |
| file = file_mic | |
| elif (file_mic is None) and (file_upload is None): | |
| return "ERROR: You have to either use the microphone or upload an audio file" | |
| elif file_mic is not None: | |
| file = file_mic | |
| else: | |
| file = file_upload | |
| input_values = process_audio_file(file) | |
| sequences = model.generate(input_values, num_beams=1, max_length=30) | |
| transcription = tokenizer.batch_decode(sequences, skip_special_tokens=True) | |
| return warn_output + transcription[0] | |
| iface = gr.Interface( | |
| fn=transcribe, | |
| inputs=[ | |
| gr.inputs.Audio(source="microphone", type='filepath', optional=True), | |
| gr.inputs.Audio(source="upload", type='filepath', optional=True), | |
| ], | |
| outputs="text", | |
| layout="horizontal", | |
| theme="huggingface", | |
| title="XLS-R 2B 21-to-EN Speech Translation", | |
| description="A simple interface to translate from 21 spoken languages to written English.", | |
| ) | |
| iface.launch() | |