app.py

# import gradio as gr
# import torch
# from transformers import pipeline, AutoTokenizer
# from nemo.collections.asr.models import EncDecMultiTaskModel

# # load model
# canary_model = EncDecMultiTaskModel.from_pretrained('nvidia/canary-1b')

# # update dcode params
# decode_cfg = canary_model.cfg.decoding
# decode_cfg.beam.beam_size = 1
# canary_model.change_decoding_strategy(decode_cfg)

# pipe = pipeline(
#     "automatic-speech-recognition", 
#     model="nvidia/canary-1b"
# )

# # pipe = pipeline(
# #     "text-generation", 
# #     model="QuantFactory/Meta-Llama-3-8B-Instruct-GGUF", 
# #     model_kwargs={"torch_dtype": torch.bfloat16}, 
# #     device_map="auto"
# # )

# gr.Interface.from_pipeline(pipe,
#                            title="ASR",
#                            description="Using pipeline with Canary-1B",
#                            ).launch(inbrowser=True)

import gradio as gr
import json
import librosa
import os
import soundfile as sf
import tempfile
import uuid

import torch

from nemo.collections.asr.models import ASRModel
from nemo.collections.asr.parts.utils.streaming_utils import FrameBatchMultiTaskAED
from nemo.collections.asr.parts.utils.transcribe_utils import get_buffered_pred_feat_multitaskAED

SAMPLE_RATE = 16000 # Hz
MAX_AUDIO_MINUTES = 180 # wont try to transcribe if longer than this

model = ASRModel.from_pretrained("nvidia/canary-1b")
model.eval()

# make sure beam size always 1 for consistency
model.change_decoding_strategy(None)
decoding_cfg = model.cfg.decoding
decoding_cfg.beam.beam_size = 1
model.change_decoding_strategy(decoding_cfg)

# setup for buffered inference
model.cfg.preprocessor.dither = 0.0
model.cfg.preprocessor.pad_to = 0

feature_stride = model.cfg.preprocessor['window_stride']
model_stride_in_secs = feature_stride * 8 # 8 = model stride, which is 8 for FastConformer

frame_asr = FrameBatchMultiTaskAED(
	asr_model=model,
	frame_len=40.0,
	total_buffer=40.0,
	batch_size=16,
)

amp_dtype = torch.float16

def convert_audio(audio_filepath, tmpdir, utt_id):
	"""
	Convert all files to monochannel 16 kHz wav files.
	Do not convert and raise error if audio too long.
	Returns output filename and duration.
	"""

	data, sr = librosa.load(audio_filepath, sr=None, mono=True)

	duration = librosa.get_duration(y=data, sr=sr)

	if duration / 60.0 > MAX_AUDIO_MINUTES:
		raise gr.Error(
			f"This demo can transcribe up to {MAX_AUDIO_MINUTES} minutes of audio. "
			"If you wish, you may trim the audio using the Audio viewer in Step 1 "
			"(click on the scissors icon to start trimming audio)."
		)

	if sr != SAMPLE_RATE:
		data = librosa.resample(data, orig_sr=sr, target_sr=SAMPLE_RATE)

	out_filename = os.path.join(tmpdir, utt_id + '.wav')

	# save output audio
	sf.write(out_filename, data, SAMPLE_RATE)

	return out_filename, duration


def transcribe(audio_filepath, src_lang, tgt_lang, pnc):

	if audio_filepath is None:
		raise gr.Error("Please provide some input audio: either upload an audio file or use the microphone")
	
	utt_id = uuid.uuid4()
	with tempfile.TemporaryDirectory() as tmpdir:
		converted_audio_filepath, duration = convert_audio(audio_filepath, tmpdir, str(utt_id))

		# map src_lang and tgt_lang from long versions to short
		LANG_LONG_TO_LANG_SHORT = {
			"English": "en",
			"Spanish": "es",
			"French": "fr",
			"German": "de",
		}
		if src_lang not in LANG_LONG_TO_LANG_SHORT.keys():
			raise ValueError(f"src_lang must be one of {LANG_LONG_TO_LANG_SHORT.keys()}")
		else:
			src_lang = LANG_LONG_TO_LANG_SHORT[src_lang]
		
		if tgt_lang not in LANG_LONG_TO_LANG_SHORT.keys():
			raise ValueError(f"tgt_lang must be one of {LANG_LONG_TO_LANG_SHORT.keys()}")
		else:
			tgt_lang = LANG_LONG_TO_LANG_SHORT[tgt_lang]
		

		# infer taskname from src_lang and tgt_lang
		if src_lang == tgt_lang:
			taskname = "asr"
		else:
			taskname = "s2t_translation"

		# update pnc variable to be "yes" or "no"
		pnc = "yes" if pnc else "no"

		# make manifest file and save
		manifest_data = {
			"audio_filepath": converted_audio_filepath,
			"source_lang": src_lang,
			"target_lang": tgt_lang,
			"taskname": taskname,
			"pnc": pnc,
			"answer": "predict",
			"duration": str(duration),
		}

		manifest_filepath = os.path.join(tmpdir, f'{utt_id}.json')

		with open(manifest_filepath, 'w') as fout:
			line = json.dumps(manifest_data)
			fout.write(line + '\n')

		# call transcribe, passing in manifest filepath
		if duration < 40:
			output_text = model.transcribe(manifest_filepath)[0]
		else: # do buffered inference
			with torch.cuda.amp.autocast(dtype=amp_dtype): # TODO: make it work if no cuda
				with torch.no_grad():
					hyps = get_buffered_pred_feat_multitaskAED(
						frame_asr,
						model.cfg.preprocessor,
						model_stride_in_secs,
						model.device,
						manifest=manifest_filepath,
						filepaths=None,
					)

					output_text = hyps[0].text

	return output_text

with gr.Blocks(
	title="NeMo Canary Model",
	css="""
		textarea { font-size: 18px;}
		#model_output_text_box span {
			font-size: 18px;
			font-weight: bold;
		}
	""",
	theme=gr.themes.Default(text_size=gr.themes.sizes.text_lg) # make text slightly bigger (default is text_md )
) as demo:

	gr.HTML("<h1 style='text-align: center'>NeMo Canary model: Transcribe & Translate audio</h1>")

	with gr.Row():
		with gr.Column():
			gr.HTML(
				"<p><b>Step 1:</b> Upload an audio file or record with your microphone.</p>"

				"<p style='color: #A0A0A0;'>This demo supports audio files up to 10 mins long. "
				"You can transcribe longer files locally with this NeMo "
				"<a href='https://github.com/NVIDIA/NeMo/blob/main/examples/asr/speech_multitask/speech_to_text_aed_chunked_infer.py'>script</a>.</p>"
			)

			audio_file = gr.Audio(sources=["microphone", "upload"], type="filepath")

			gr.HTML("<p><b>Step 2:</b> Choose the input and output language.</p>")

			src_lang = gr.Dropdown(
				choices=["English", "Spanish", "French", "German"],
				value="English",
				label="Input audio is spoken in:"
			)

			with gr.Column():
				tgt_lang = gr.Dropdown(
					choices=["English", "Spanish", "French", "German"],
					value="English",
					label="Transcribe in language:"
				)
				pnc = gr.Checkbox(
					value=True,
					label="Punctuation & Capitalization in transcript?",
				)

		with gr.Column():

			gr.HTML("<p><b>Step 3:</b> Run the model.</p>")

			go_button = gr.Button(
				value="Run model",
				variant="primary", # make "primary" so it stands out (default is "secondary")
			)

			model_output_text_box = gr.Textbox(
				label="Model Output",
				elem_id="model_output_text_box",
			)

        with gr.Row():

    		gr.HTML(
    			"<p style='text-align: center'>"
    				"🐤 <a href='https://huggingface.co/nvidia/canary-1b' target='_blank'>Canary model</a> | "
    				"🧑‍💻 <a href='https://github.com/NVIDIA/NeMo' target='_blank'>NeMo Repository</a>"
    			"</p>"
    		)

    	go_button.click(
    		fn=transcribe, 
    		inputs = [audio_file, src_lang, tgt_lang, pnc],
    		outputs = [model_output_text_box]
    	)
    
    	# call on_src_or_tgt_lang_change whenever src_lang or tgt_lang dropdown menus are changed	
    	src_lang.change(
    		fn=on_src_or_tgt_lang_change,
    		inputs=[src_lang, tgt_lang, pnc],
    		outputs=[src_lang, tgt_lang, pnc],
    	)
    	tgt_lang.change(
    		fn=on_src_or_tgt_lang_change,
    		inputs=[src_lang, tgt_lang, pnc],
    		outputs=[src_lang, tgt_lang, pnc],
    	)


demo.queue()
demo.launch(share=True)