app.py

import spaces
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 = 30 # wont try to transcribe if longer than this

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

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

@spaces.GPU
def transcribe(manifest_filepath, audio_duration, duration_limit):
	"""
	Transcribe audio using either model.transcribe or buffered inference.
	Duration limit determines which method to use and what chunk size will 
	be used in the case of buffered inference.

	Note: I have observed that if you try to throw a gr.Error inside a function
	decorated with @spaces.GPU, the error message you specified in gr.Error will
	not be shown, instead it show the message "ZeroGPU worker error".
	"""

	if audio_duration < duration_limit:

		output = model.transcribe(manifest_filepath)

	else:
		frame_asr = FrameBatchMultiTaskAED(
			asr_model=model,
			frame_len=duration_limit,
			total_buffer=duration_limit,
			batch_size=16,
		)
		output = get_buffered_pred_feat_multitaskAED(
			frame_asr,
			model.cfg.preprocessor,
			model_stride_in_secs,
			model.device,
			manifest=manifest_filepath,
			filepaths=None,
		)
	return output


def on_go_btn_click(audio_filepath, src_lang, tgt_lang, pnc, gen_ts):

	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 and gen_ts variables to be "yes" or "no"
		pnc = "yes" if pnc else "no"
		gen_ts = "yes" if gen_ts 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),
			"timestamp": gen_ts,
		}

		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')


		# setup beginning of output html
		output_html = '''
			<!DOCTYPE html>
			<html lang="en">
			<head>
				<style>
					.transcript {
						font-family: Arial, sans-serif;
						line-height: 1.6;
						margin: 20px 0;
					}
					.timestamp {
						color: gray;
						font-size: 0.8em;
						margin-right: 5px;
					}
					.heading {
						color: #2c3e50;
						font-family: Arial, sans-serif;
						font-weight: bold;
						margin: 15px 0 8px 0;
						border-bottom: 1px solid #eee;
					}
				</style>
			</head>
			<body>
			'''
	
  
		if gen_ts == "yes": # if will generate timestamps
			output = transcribe(manifest_filepath, audio_duration=duration, duration_limit=10.0)

			# process output to get word and segment level timestamps
			word_level_timestamps = output[0].timestamp["word"]

			output_html += "<div class='heading'>Transcript with word-level timestamps (in seconds)</div>\n"
			output_html += "<div class='transcript'>\n"
			for entry in word_level_timestamps:
				output_html += f'<span>{entry["word"]} <span class="timestamp">({entry["start"]:.2f}-{entry["end"]:.2f})</span></span>\n'
			output_html += "</div>\n"

			segment_level_timestamps = output[0].timestamp["segment"]
			output_html += "<div class='heading'>Transcript with segment-level timestamps (in seconds)</div>\n"
			output_html += "<div class='transcript'>\n"
			for entry in segment_level_timestamps:
				output_html += f'<span>{entry["segment"]} <span class="timestamp">({entry["start"]:.2f}-{entry["end"]:.2f})</span></span><br>\n'
			output_html += "</div>\n"

		else: # if will not generate timestamps
			output = transcribe(manifest_filepath, audio_duration=duration, duration_limit=40.0)

			if taskname == "asr":
				output_html += "<div class='heading'>Transcript</div>\n"
			else:
				output_html += "<div class='heading'>Translated Text</div>\n"
			output_text = output[0].text
			output_html += f'<div class="transcript">{output_text}</div>\n'

	output_html += '''
		</div>
		</body>
		</html>
		'''				

	return output_html

# add logic to make sure dropdown menus only suggest valid combos
def on_src_or_tgt_lang_change(src_lang_value, tgt_lang_value, pnc_value, gen_ts_value):
	"""Callback function for when src_lang or tgt_lang dropdown menus are changed.

	Args:
		src_lang_value(string), tgt_lang_value (string), pnc_value(bool), gen_ts_value(bool) - the current 
			chosen "values" of each Gradio component
	Returns:
		src_lang, tgt_lang, pnc, gen_ts - these are the new Gradio components that will be displayed
	
	Note: I found the required logic is easier to understand if you think about the possible src & tgt langs as
	a matrix, e.g. with English, Spanish, French, German as the langs, and only transcription in the same language,
	and X -> English and English -> X translation being allowed, the matrix looks like the diagram below ("Y" means it is
	allowed to go into that state).
	It is easier to understand the code if you think about which state you are in, given the current src_lang_value and
	tgt_lang_value, and then which states you can go to from there.

				tgt lang
				- |EN |ES |FR |DE
				------------------
				EN| Y | Y | Y | Y
				------------------
		src 	ES| Y | Y |   |
		lang	------------------
				FR| Y |   | Y |
				------------------
				DE| Y |   |   | Y
		"""

	if src_lang_value == "English" and tgt_lang_value == "English":
		# src_lang and tgt_lang can go anywhere
		src_lang = gr.Dropdown(
			choices=["English", "Spanish", "French", "German"],
			value=src_lang_value,
			label="Input audio is spoken in:"
		)
		tgt_lang = gr.Dropdown(
			choices=["English", "Spanish", "French", "German"],
			value=tgt_lang_value,
			label="Transcribe in language:"
		)
	elif src_lang_value == "English": 
		# src is English & tgt is non-English
		# => src can only be English or current tgt_lang_values
		# & tgt can be anything
		src_lang = gr.Dropdown(
			choices=["English", tgt_lang_value],
			value=src_lang_value,
			label="Input audio is spoken in:"
		)
		tgt_lang = gr.Dropdown(
			choices=["English", "Spanish", "French", "German"],
			value=tgt_lang_value,
			label="Transcribe in language:"
		)
	elif tgt_lang_value == "English": 
		# src is non-English & tgt is English
		# => src can be anything
		# & tgt can only be English or current src_lang_value
		src_lang = gr.Dropdown(
			choices=["English", "Spanish", "French", "German"],
			value=src_lang_value,
			label="Input audio is spoken in:"
		)
		tgt_lang = gr.Dropdown(
			choices=["English", src_lang_value],
			value=tgt_lang_value,
			label="Transcribe in language:"
		)
	else:
		# both src and tgt are non-English
		# => both src and tgt can only be switch to English or themselves
		src_lang = gr.Dropdown(
			choices=["English", src_lang_value],
			value=src_lang_value,
			label="Input audio is spoken in:"
		)
		tgt_lang = gr.Dropdown(
			choices=["English", tgt_lang_value],
			value=tgt_lang_value,
			label="Transcribe in language:"
		)
	# if src_lang_value == tgt_lang_value then pnc and gen_ts can be anything
	# else, fix pnc to True and gen_ts to False
	if src_lang_value == tgt_lang_value:
		pnc = gr.Checkbox(
			value=pnc_value,
			label="Punctuation & Capitalization in model output?",
			interactive=True
		)
		gen_ts = gr.Checkbox(
			value=gen_ts_value,
			label="Generate timestamps?",
			interactive=True
		)
	else:
		pnc = gr.Checkbox(
			value=True,
			label="Punctuation & Capitalization in model output?",
			interactive=False
		)
		gen_ts = gr.Checkbox(
			value=False,
			label="Generate timestamps?",
			interactive=False
		)

	return src_lang, tgt_lang, pnc, gen_ts


with gr.Blocks(
	title="NeMo Canary 1B Flash Model",
	css="""
		textarea { font-size: 18px;}
	""",
	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 1B Flash 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>"

				f"<p style='color: #A0A0A0;'>This demo supports audio files up to {MAX_AUDIO_MINUTES} mins long. "
				"You can transcribe longer files locally with this NeMo "
				"<a href='https://github.com/NVIDIA/NeMo/blob/main/examples/asr/asr_chunked_inference/aed/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>"
				"<p style='color: #A0A0A0;'>If input & output languages are the same, you can also toggle generating punctuation & capitalization and timestamps.</p>"
			)


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

				tgt_lang = gr.Dropdown(
					choices=["English", "Spanish", "French", "German"],
					value="English",
					label="Transcribe in language:"
				)
				pnc = gr.Checkbox(
					value=True,
					label="Punctuation & Capitalization in model output?",
				)
				gen_ts = gr.Checkbox(
					value=False,
					label="Generate timestamps?",
				)

		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_html = gr.HTML(
				# initialize with min-height to ensure "processing" animation will be visible
				value='<div style="min-height: 100px;"></div>',
				label="Model Output",
			)


	with gr.Row():

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

	go_button.click(
		fn=on_go_btn_click, 
		inputs = [audio_file, src_lang, tgt_lang, pnc, gen_ts],
		outputs = [model_output_html]
	)

	# 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, gen_ts],
		outputs=[src_lang, tgt_lang, pnc, gen_ts],
	)
	tgt_lang.change(
		fn=on_src_or_tgt_lang_change,
		inputs=[src_lang, tgt_lang, pnc, gen_ts],
		outputs=[src_lang, tgt_lang, pnc, gen_ts],
	)


demo.queue()
demo.launch()