re-adapted process_libritts.py to a 'better' way (better because it processed without needing to shuffle a bunch of things and adapt to cope or something)

scripts/cleanup_dataset.py

@@ -1,3 +1,7 @@
+"""
+# Helper script to clean up transcription metadata, whatever that entailed.
+"""
+
 import os
 import json
 import torch

scripts/deduplicate_librilight_libritts.py

@@ -1,3 +1,7 @@
+"""
+# Helper script to try and detect any duplications between LibriLight and LibriTTS (I don't think there were any)
+"""
+
 import os
 import json
 

scripts/parse_ppp.py

@@ -1,3 +1,7 @@
+"""
+# Helper script to parse PPP dataset into a friendlier hierarchy
+"""
+
 import os
 import json
 import torch
@@ -7,8 +11,6 @@ from pathlib import Path
 from vall_e.emb.g2p import encode as valle_phonemize
 from vall_e.emb.qnt import encode_from_file as valle_quantize, _replace_file_extension
 
-device = "cuda"
-
 target = "in"
 
 audio_map = {}
@@ -86,6 +88,10 @@ for key, entry in audio_map.items():
 for name in data.keys():
 	open(f"./training/{name}/whisper.json", "w", encoding="utf-8").write( json.dumps( data[name], indent='\t' ) )
 
+# to-do: update to "The Proper Way"
+# for now it can just be fed back into "The Proper Way""
+"""
+device = "cuda"
 for key, text in tqdm(txts.items(), desc="Phonemizing..."):
 	path = Path(key)
 	phones = valle_phonemize(text)
@@ -94,3 +100,4 @@ for key, text in tqdm(txts.items(), desc="Phonemizing..."):
 for path in tqdm(wavs, desc="Quantizing..."):
 	qnt = valle_quantize(path, device=device)
 	torch.save(qnt.cpu(), _replace_file_extension(path, ".qnt.pt"))
+"""

scripts/prepare_librilight.py

@@ -1,15 +1,24 @@
+"""
+# Handles processing `facebookresearch/libri-light`'s unlabeled audio into a friendlier hierarchy
+"""
+
 import os
 import json
 
-input_dataset = "duplicate"
+datasets = ["small", "medium", "large", "duplicate"]
 output_dataset = "LibriLight-4K"
 
+for input_dataset in datasets:
+	if not os.path.isdir(f'./{input_dataset}/'):
+		continue
+
 	for speaker_id in os.listdir(f'./{input_dataset}/'):
 		if not os.path.isdir(f'./{input_dataset}/{speaker_id}/'):
 			continue
-	for book_name in os.listdir(f'./{input_dataset}/{speaker_id}/'):
 		
+		for book_name in os.listdir(f'./{input_dataset}/{speaker_id}/'):
 			subid = 0
+
 			for filename in os.listdir(f'./{input_dataset}/{speaker_id}/{book_name}'):
 				if filename[-5:] != ".json":
 					continue

scripts/prepare_libritts.py

@@ -1,21 +0,0 @@
-import os
-import json
-
-input_dataset = "LibriTTS_R"
-output_dataset = "LibriTTS-Train"
-
-for dataset_name in os.listdir(f'./{input_dataset}/'):
-	if not os.path.isdir(f'./{input_dataset}/{dataset_name}/'):
-		continue
-	for speaker_id in os.listdir(f'./{input_dataset}/{dataset_name}/'):
-		if not os.path.isdir(f'./{input_dataset}/{dataset_name}/{speaker_id}'):
-			continue
-		for book_id in os.listdir(f'./{input_dataset}/{dataset_name}/{speaker_id}'):
-			if not os.path.isdir(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}'):
-				continue
-			for filename in os.listdir(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}'):
-				if filename[-4:] != ".wav":
-					continue
-
-				os.makedirs(f'./{output_dataset}/{speaker_id}/', exist_ok=True)
-				os.rename(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}/{filename}', f'./{output_dataset}/{speaker_id}/{filename}')

scripts/process_libritts.py

@@ -1,6 +1,13 @@
+"""
+# Handles processing audio provided through --input-audio of adequately annotated transcriptions provided through --input-metadata (through transcribe.py)
+# Outputs NumPy objects containing quantized audio and adequate metadata for use of loading in the trainer through --output-dataset
+"""
+
 import os
 import json
+import argparse
 import torch
+import torchaudio
 import numpy as np
 
 from tqdm.auto import tqdm
@@ -8,74 +15,146 @@ from pathlib import Path
 
 from vall_e.config import cfg
 
-# things that could be args
+def pad(num, zeroes):
-cfg.sample_rate = 24_000
+	return str(num).zfill(zeroes+1)
-cfg.audio_backend = "encodec"
-"""
-cfg.inference.weight_dtype = "bfloat16"
-cfg.inference.dtype = torch.bfloat16
-cfg.inference.amp = True
-"""
 
-from vall_e.emb.g2p import encode as valle_phonemize
+def process_items( items, stride=0, stride_offset=0 ):
-from vall_e.emb.qnt import encode_from_file as valle_quantize, _replace_file_extension
+	items = sorted( items )
+	return items if stride == 0 else [ item for i, item in enumerate( items ) if (i+stride_offset) % stride == 0 ]
 
+def process(
+		audio_backend="encodec",
+		input_audio="LibriTTS_R",
+		output_dataset="training",
+		raise_exceptions=False,
+		stride=0,
+		stride_offset=0,
+		slice="auto",
+
+		device="cuda",
+		dtype="float16",
+		amp=False,
+	):
+	# encodec / vocos
+
+	if audio_backend in ["encodec", "vocos"]:
 		audio_extension = ".enc"
-if cfg.audio_backend == "dac":
+		cfg.sample_rate = 24_000
+		cfg.model.resp_levels = 8
+	elif audio_backend == "dac":
 		audio_extension = ".dac"
+		cfg.sample_rate = 44_100
+		cfg.model.resp_levels = 9
 	elif cfg.audio_backend == "audiodec":
+		sample_rate = 48_000
 		audio_extension = ".dec"
+		cfg.model.resp_levels = 8 # ?
+	else:
+		raise Exception(f"Unknown audio backend: {audio_backend}")
 
-input_dataset = "LibriTTS_R"
+	# prepare from args
-output_dataset = f"LibriTTS-Train-{'2' if cfg.sample_rate == 24_000 else '4'}{'8' if cfg.sample_rate == 48_000 else '4'}KHz-{cfg.audio_backend}"
+	cfg.audio_backend = audio_backend # "encodec"
-device = "cuda"
+	cfg.inference.weight_dtype = dtype # "bfloat16"
+	cfg.inference.amp = amp # False
+
+	# import after because we've overriden the config above
+	# need to validate if this is even necessary anymore
+	from vall_e.emb.g2p import encode as phonemize
+	from vall_e.emb.qnt import encode as quantize, _replace_file_extension
+
+	output_dataset = f"{output_dataset}/{'2' if cfg.sample_rate == 24_000 else '4'}{'8' if cfg.sample_rate == 48_000 else '4'}KHz-{cfg.audio_backend}" # "training"
+
+	language_map = {} # k = group, v = language
+
+	ignore_groups = [] # skip these groups
+	ignore_speakers = [] # skip these speakers
+
+	only_groups = [] # only process these groups
+	only_speakers = [] # only process these speakers
+
+	always_slice_groups = [] # always slice from this group
+
+	missing = {
+		"transcription": [],
+		"audio": []
+	}
+	dataset = []
+
+	# Layout: ./LibriTTS_R/train-clean-100/103/1241
+	for group_name in sorted(os.listdir(f'./{input_audio}/')):
+		if not os.path.isdir(f'./{input_audio}/{group_name}/'):
+			print("Is not dir:", f'./{input_audio}/{group_name}/')
+			continue
+
+		if group_name in ignore_groups:
+			continue
+		if only_groups and group_name not in only_groups:
+			continue
+
+		for speaker_id in tqdm(process_items(os.listdir(f'./{input_audio}/{group_name}/'), stride=stride, stride_offset=stride_offset), desc=f"Processing speaker in {group_name}"):
+			if not os.path.isdir(f'./{input_audio}/{group_name}/{speaker_id}'):
+				print("Is not dir:", f'./{input_audio}/{group_name}/{speaker_id}')
+				continue
+			
+			if speaker_id in ignore_speakers:
+				continue
+			if only_speakers and speaker_id not in only_speakers:
+				continue
+
+			os.makedirs(f'./{output_dataset}/{group_name}/{speaker_id}/', exist_ok=True)
+
+			if f'{group_name}/{speaker_id}' not in dataset:
+				dataset.append(f'{group_name}/{speaker_id}')
 
 			txts = []
 			wavs = []
 
-for dataset_name in os.listdir(f'./{input_dataset}/'):
+			for book_id in os.listdir(f'./{input_audio}/{dataset_name}/{speaker_id}'):
-	if not os.path.isdir(f'./{input_dataset}/{dataset_name}/'):
+				if not os.path.isdir(f'./{input_audio}/{group_name}/{speaker_id}/{book_id}'):
+					print("Is not dir:", f'./{input_audio}/{group_name}/{speaker_id}/{book_id}')
 					continue
 
-	for speaker_id in tqdm(os.listdir(f'./{input_dataset}/{dataset_name}/'), desc="Processing speaker"):
+				for filename in os.listdir(f'./{input_audio}/{dataset_name}/{speaker_id}/{book_id}'):
-		if not os.path.isdir(f'./{input_dataset}/{dataset_name}/{speaker_id}'):
+					inpath = Path(f'./{input_audio}/{group_name}/{speaker_id}/{book_id}/{filename}')
-			continue
+					if not inpath.exists():
+						missing["audio"].append(str(inpath))
 				
-		os.makedirs(f'./{output_dataset}/{speaker_id}/', exist_ok=True)
+					extension = os.path.splitext(filename)[-1][1:]
-		for book_id in os.listdir(f'./{input_dataset}/{dataset_name}/{speaker_id}'):
+					fname = filename.replace(f'.{extension}', "")
-			if not os.path.isdir(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}'):
-				continue
-			for filename in os.listdir(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}'):
-				# os.rename(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}/{filename}', f'./{output_dataset}/{speaker_id}/{filename}')
 
-				inpath = Path(f'./{input_dataset}/{dataset_name}/{speaker_id}/{book_id}/{filename}')
+					waveform, sample_rate = None, None
-				outpath = Path(f'./{output_dataset}/{speaker_id}/{filename}')
+					language = "en"
 
-				if ".wav" in filename: # and not _replace_file_extension(outpath, ".dac").exists():
+					outpath = Path(f'./{output_dataset}/{group_name}/{speaker_id}/{fname}.{extension}')
-					txts.append((
-						inpath,
-						outpath
-					))
-
-for paths in tqdm(txts, desc="Processing..."):
-	inpath, outpath = paths
-	try:
-		if _replace_file_extension(outpath, ".dac").exists() and _replace_file_extension(outpath, ".json").exists():
-			data = json.loads(open(_replace_file_extension(outpath, ".json"), 'r', encoding='utf-8').read())
-			qnt = np.load(_replace_file_extension(outpath, audio_extension), allow_pickle=True)
-			
-			if not isinstance(data["phonemes"], str):
-				data["phonemes"] = "".join(data["phonemes"])
-
-			for k, v in data.items():
-				qnt[()]['metadata'][k] = v
-
-			np.save(open(_replace_file_extension(outpath, audio_extension), "wb"), qnt)
-		else:
 					text = open(_replace_file_extension(inpath, ".original.txt"), "r", encoding="utf-8").read()
 
-			phones = valle_phonemize(text)
+					if len(text) == 0:
-			qnt = valle_quantize(_replace_file_extension(inpath, ".wav"), device=device)
+						continue
+
+					if _replace_file_extension(outpath, audio_extension).exists():
+						continue
+
+					if waveform is None:
+						waveform, sample_rate = torchaudio.load(inpath)
+						if waveform.shape[0] > 1:
+							waveform = torch.mean(waveform, dim=0, keepdim=True)
+
+					wavs.append((
+						outpath,
+						text,
+						language,
+						waveform,
+						sample_rate
+					))
+
+			if len(wavs) > 0:
+				for job in tqdm(wavs, desc=f"Quantizing: {speaker_id}"):
+					try:
+						outpath, text, language, waveform, sample_rate = job
+
+						phones = phonemize(text, language=language)
+						qnt = quantize(waveform, sr=sample_rate, device=device)
+
 
 						if cfg.audio_backend == "dac":
 							np.save(open(_replace_file_extension(outpath, audio_extension), "wb"), {
@@ -92,20 +171,59 @@ for paths in tqdm(txts, desc="Processing..."):
 
 									"text": text.strip(),
 									"phonemes": "".join(phones),
-						"language": "en",
+									"language": language,
 								},
 							})
 						else:						
 							np.save(open(_replace_file_extension(outpath, audio_extension), "wb"), {
 								"codes": qnt.cpu().numpy().astype(np.uint16),
 								"metadata": {
-						"original_length": qnt.shape[-1] / 75.0,
+									"original_length": waveform.shape[-1],
-						"sample_rate": cfg.sample_rate,
+									"sample_rate": sample_rate,
 
 									"text": text.strip(),
 									"phonemes": "".join(phones),
-						"language": "en",
+									"language": language,
 								},
 							})
 					except Exception as e:
-		tqdm.write(f"Failed to process: {paths}: {e}")
+						print(f"Failed to quantize: {outpath}:", e)
+						if raise_exceptions:
+							raise e
+						continue
+
+	open(f"./{output_dataset}/missing.json", 'w', encoding='utf-8').write(json.dumps(missing))
+	open(f"./{output_dataset}/dataset.json", 'w', encoding='utf-8').write(json.dumps(dataset))
+
+def main():
+	parser = argparse.ArgumentParser()
+
+	parser.add_argument("--audio-backend", type=str, default="encodec")
+	parser.add_argument("--dtype", type=str, default="bfloat16")
+	parser.add_argument("--amp", action="store_true")
+	parser.add_argument("--input-audio", type=str, default="LibriTTS_R")
+	parser.add_argument("--output-dataset", type=str, default="training/dataset")
+	parser.add_argument("--device", type=str, default="cuda")
+	parser.add_argument("--raise-exceptions", action="store_true")
+	parser.add_argument("--stride", type=int, default=0)
+	parser.add_argument("--stride-offset", type=int, default=0)
+	parser.add_argument("--slice", type=str, default="auto")
+	
+	args = parser.parse_args()
+
+	process(
+		audio_backend=args.audio_backend,
+		input_audio=args.input_audio,
+		output_dataset=args.output_dataset,
+		raise_exceptions=args.raise_exceptions,
+		stride=args.stride,
+		stride_offset=args.stride_offset,
+		slice=args.slice,
+
+		device=args.device,
+		dtype=args.dtype,
+		amp=args.amp,
+	)
+
+if __name__ == "__main__":
+	main()

scripts/train_tokenizer.py

@@ -1,3 +1,7 @@
+"""
+# Helper script to grab all phonemes through parsed dataset metadata to find the "best" tokenizer dict
+"""
+
 import os
 import json
 import torch

vall_e/emb/process.py

@@ -59,6 +59,7 @@ def process(
 	cfg.inference.amp = amp # False
 
 	# import after because we've overriden the config above
+	# need to validate if this is even necessary anymore
 	from .g2p import encode as phonemize
 	from .qnt import encode as quantize, _replace_file_extension
 
@@ -275,8 +276,8 @@ def process(
 							raise e
 						continue
 
-	open("./missing.json", 'w', encoding='utf-8').write(json.dumps(missing))
+	open(f"./{output_dataset}/missing.json", 'w', encoding='utf-8').write(json.dumps(missing))
-	open("./dataset_list.json", 'w', encoding='utf-8').write(json.dumps(dataset))
+	open(f"./{output_dataset}/dataset.json", 'w', encoding='utf-8').write(json.dumps(dataset))
 
 def main():
 	parser = argparse.ArgumentParser()