app.py

import gradio as gr
import soundfile as sf
from scipy import signal
import numpy as np
import torch, torchaudio
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor, pipeline

MODEL_IS="language-and-voice-lab/wav2vec2-large-xlsr-53-icelandic-ep30-967h"
MODEL_FO="carlosdanielhernandezmena/wav2vec2-large-xlsr-53-faroese-100h"

torch.random.manual_seed(0)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

model_is = Wav2Vec2ForCTC.from_pretrained(MODEL_IS).to(device)
processor_is = Wav2Vec2Processor.from_pretrained(MODEL_IS) 
model_fo = Wav2Vec2ForCTC.from_pretrained(MODEL_FO).to(device)
processor_fo = Wav2Vec2Processor.from_pretrained(MODEL_FO) 

pipe_is = pipeline(model=MODEL_IS)
pipe_fo = pipeline(model=MODEL_FO)



def readwav(a_f):
    wav, sr = sf.read(a_f, dtype=np.float32)
    if len(wav.shape) == 2:
        wav = wav.mean(1)
    if sr != 16000:
        wlen = int(wav.shape[0] / sr * 16000)
        wav = signal.resample(wav, wlen)
    return wav

def recc(audio_file,model,processor):
    wav = readwav(audio_file)
    with torch.inference_mode():
        input_values = processor(wav,sampling_rate=16000).input_values[0]
        input_values = torch.tensor(input_values, device=device).unsqueeze(0)
        logits = model(input_values).logits
        pred_ids = torch.argmax(logits, dim=-1)
        xcp = processor.batch_decode(pred_ids)
        return xcp[0]

    
def recis(audio_file):
    single_output = recc(audio_file,model_is,processor_is)
    chunk_output = pipe_is(audio_file, chunk_length_s=4)['text']
    return (single_output, chunk_output)

def recfo(audio_file):
    single_output = recc(audio_file,model_fo,processor_fo)
    chunk_output = pipe_fo(audio_file, chunk_length_s=4)['text']
    return (single_output, chunk_output)


bl = gr.Blocks()
with bl:

    gr.Markdown(
        """
    # Speech recognition

    ### Users logged in to a Huggingface account can use each model's normal hosted inference API instead.
    ## * * * * * * * *
    
    Upload a file for recognition with 
    https://huggingface.co/language-and-voice-lab/wav2vec2-large-xlsr-53-icelandic-ep30-967h
    or https://huggingface.co/carlosdanielhernandezmena/wav2vec2-large-xlsr-53-faroese-100h  

    - Wav2Vec2 models have no language model (yet), so it can generate non-words.
    - Whisper can hallucinate.
    - Send errors/bugs to [email protected]
    """
    )

    with gr.Tabs():
        with gr.TabItem("Icelandic"):
            with gr.Row():
                audio_file = gr.Audio(sources=["upload", "microphone"],type="filepath")
                with gr.Column():
                    #whole_output = gr.Textbox(label="whole-file recognition")
                    chunk_output = gr.Textbox(label="Wav2Vec2 recognition")# with chunking")
                    whisper_output = gr.Textbox(label="Whisper recognition")
            text_button = gr.Button("Recognise Icelandic")
            #text_button.click(recis, inputs=audio_file, outputs=[whole_output,chunk_output])
            text_button.click(recis, inputs=audio_file, outputs=[chunk_output])

        with gr.TabItem("Faroese"):
            with gr.Row():
                audio_file = gr.Audio(sources=["upload", "microphone"],type="filepath")
                with gr.Column():
                    #whole_output = gr.Textbox(label="whole-file recognition")
                    chunk_output = gr.Textbox(label="Wav2Vec2 recognition")# with chunking")
            text_button = gr.Button("Recognise Faroese")
            #text_button.click(recfo, inputs=audio_file, outputs=[whole_output,chunk_output])
            text_button.click(recfo, inputs=audio_file, outputs=[chunk_output])

bl.launch()