class Solver(BaseSolver):
''' Solver for training'''
def __init__(self, config, paras, mode):
super().__init__(config, paras, mode)
# ToDo : support tr/eval on different corpus
assert self.config['data']['corpus']['name'] == self.src_config['data']['corpus']['name']
self.config['data']['corpus']['path'] = self.src_config['data']['corpus']['path']
self.config['data']['corpus']['bucketing'] = False
# The follow attribute should be identical to training config
self.config['data']['audio'] = self.src_config['data']['audio']
self.config['data']['corpus']['train_split'] = self.src_config['data']['corpus']['train_split']
self.config['data']['text'] = self.src_config['data']['text']
self.tokenizer = load_text_encoder(**self.config['data']['text'])
self.config['model'] = self.src_config['model']
self.finetune_first = 5
self.best_wer = {'att': 3.0, 'ctc': 3.0}
# Output file
self.output_file = str(self.ckpdir)+'_{}_{}.csv'
# Override batch size for beam decoding
self.greedy = self.config['decode']['beam_size'] == 1
self.dealer = Datadealer(self.config['data']['audio'])
self.ctc = self.config['decode']['ctc_weight'] == 1.0
if not self.greedy:
self.config['data']['corpus']['batch_size'] = 1
else:
# ToDo : implement greedy
raise NotImplementedError
# Logger settings
self.logdir = os.path.join(paras.logdir, self.exp_name)
self.log = SummaryWriter(
self.logdir, flush_secs=self.TB_FLUSH_FREQ)
self.timer = Timer()
def fetch_data(self, data):
''' Move data to device and compute text seq. length'''
_, feat, feat_len, txt = data
feat = feat.to(self.device)
feat_len = feat_len.to(self.device)
txt = txt.to(self.device)
txt_len = torch.sum(txt != 0, dim=-1)
return feat, feat_len, txt, txt_len
def load_data(self, batch_size=7):
''' Load data for training/validation, store tokenizer and input/output shape'''
prev_batch_size = self.config['data']['corpus']['batch_size']
self.config['data']['corpus']['batch_size'] = batch_size
self.tr_set, self.dv_set, self.feat_dim, self.vocab_size, self.tokenizer, msg = \
load_dataset(self.paras.njobs, self.paras.gpu,
self.paras.pin_memory, False, **self.config['data'])
self.config['data']['corpus']['batch_size'] = prev_batch_size
self.verbose(msg)
def set_model(self):
''' Setup ASR model '''
# Model
self.feat_dim = 120
self.vocab_size = 46
init_adadelta = True
''' Setup ASR model and optimizer '''
# Model
# init_adadelta = self.config['hparas']['optimizer'] == 'Adadelta'
self.model = ASR(self.feat_dim, self.vocab_size, init_adadelta, **
self.src_config['model']).to(self.device)
self.verbose(self.model.create_msg())
if self.finetune_first>0:
names = ["encoder.layers.%d"%i for i in range(self.finetune_first)]
model_paras = [{"params": [p for n, p in self.model.named_parameters() if any(nd in n for nd in names)]}]
else:
model_paras = [{'params': self.model.parameters()}]
# Losses
self.seq_loss = torch.nn.CrossEntropyLoss(ignore_index=0)
# Note: zero_infinity=False is unstable?
self.ctc_loss = torch.nn.CTCLoss(blank=0, zero_infinity=False)
# Plug-ins
self.emb_fuse = False
self.emb_reg = ('emb' in self.config) and (
self.config['emb']['enable'])
if self.emb_reg:
from src.plugin import EmbeddingRegularizer
self.emb_decoder = EmbeddingRegularizer(
self.tokenizer, self.model.dec_dim, **self.config['emb']).to(self.device)
model_paras.append({'params': self.emb_decoder.parameters()})
self.emb_fuse = self.emb_decoder.apply_fuse
if self.emb_fuse:
self.seq_loss = torch.nn.NLLLoss(ignore_index=0)
self.verbose(self.emb_decoder.create_msg())
# Optimizer
self.optimizer = Optimizer(model_paras, **self.src_config['hparas'])
self.verbose(self.optimizer.create_msg())
# Enable AMP if needed
self.enable_apex()
# Automatically load pre-trained model if self.paras.load is given
self.load_ckpt()
# Beam decoder
self.decoder = BeamDecoder(
self.model, self.emb_decoder, **self.config['decode'])
self.verbose(self.decoder.create_msg())
# del self.model
# del self.emb_decoder
self.decoder.to(self.device)
def exec(self):
''' Testing End-to-end ASR system '''
while True:
try:
filename = input("Input wav file name: ")
if filename == "exit":
return
feat, feat_len = self.dealer(filename)
feat = feat.to(self.device)
feat_len = feat_len.to(self.device)
# Decode
with torch.no_grad():
hyps = self.decoder(feat, feat_len)
hyp_seqs = [hyp.outIndex for hyp in hyps]
hyp_txts = [self.tokenizer.decode(hyp, ignore_repeat=self.ctc) for hyp in hyp_seqs]
for txt in hyp_txts:
print(txt)
except:
print("Invalid file")
pass
def recognize(self, filename):
try:
feat, feat_len = self.dealer(filename)
feat = feat.to(self.device)
feat_len = feat_len.to(self.device)
# Decode
with torch.no_grad():
hyps = self.decoder(feat, feat_len)
hyp_seqs = [hyp.outIndex for hyp in hyps]
hyp_txts = [self.tokenizer.decode(hyp, ignore_repeat=self.ctc) for hyp in hyp_seqs]
return hyp_txts[0]
except Exception as e:
print(e)
app.logger.debug(e)
return "Invalid file"
def fetch_finetune_data(self, filename, fixed_text):
feat, feat_len = self.dealer(filename)
feat = feat.to(self.device)
feat_len = feat_len.to(self.device)
text = self.tokenizer.encode(fixed_text)
text = torch.tensor(text).to(self.device)
text_len = len(text)
return [feat, feat_len, text, text_len]
def merge_batch(self, main_batch, attach_batch):
max_feat_len = max(main_batch[1])
max_text_len = max(main_batch[3])
if attach_batch[0].shape[1] > max_feat_len:
# reduce extra long example
attach_batch[0] = attach_batch[0][:,:max_feat_len]
attach_batch[1][0] = max_feat_len
else:
# pad to max_feat_len
padding = torch.zeros(1, max_feat_len - attach_batch[0].shape[1], attach_batch[0].shape[2], dtype=attach_batch[0].dtype).to(self.device)
attach_batch[0] = torch.cat([attach_batch[0], padding], dim=1)
if attach_batch[2].shape[0] > max_text_len:
attach_batch[2] = attach_batch[2][:max_text_len]
main_batch[3][0] = max_text_len
else:
padding = torch.zeros(max_text_len - attach_batch[2].shape[0], dtype=attach_batch[2].dtype).to(self.device)
try:
attach_batch[2] = torch.cat([attach_batch[2], padding], dim=0).unsqueeze(0)
except:
pdb.set_trace()
new_batch = (
torch.cat([main_batch[0], attach_batch[0]], dim=0),
torch.cat([main_batch[1], attach_batch[1]], dim=0),
torch.cat([main_batch[2], attach_batch[2]], dim=0),
torch.cat([main_batch[3], torch.tensor([attach_batch[3]]).to(self.device)], dim=0)
)
return new_batch
def finetune(self, filename, fixed_text, max_step=5):
# Load data for finetune
self.verbose('Total training steps {}.'.format(
human_format(max_step)))
ctc_loss, att_loss, emb_loss = None, None, None
n_epochs = 0
accum_count = 0
self.timer.set()
step = 0
for data in self.tr_set:
# Pre-step : update tf_rate/lr_rate and do zero_grad
if max_step == 0:
break
tf_rate = self.optimizer.pre_step(400000)
total_loss = 0
# Fetch data
finetune_data = self.fetch_finetune_data(filename, fixed_text)
main_batch = self.fetch_data(data)
new_batch = self.merge_batch(main_batch, finetune_data)
feat, feat_len, txt, txt_len = new_batch
self.timer.cnt('rd')
# Forward model
# Note: txt should NOT start w/ <sos>
ctc_output, encode_len, att_output, att_align, dec_state = \
self.model(feat, feat_len, max(txt_len), tf_rate=tf_rate,
teacher=txt, get_dec_state=self.emb_reg)
# Plugins
if self.emb_reg:
emb_loss, fuse_output = self.emb_decoder(
dec_state, att_output, label=txt)
total_loss += self.emb_decoder.weight*emb_loss
# Compute all objectives
if ctc_output is not None:
if self.paras.cudnn_ctc:
ctc_loss = self.ctc_loss(ctc_output.transpose(0, 1),
txt.to_sparse().values().to(device='cpu', dtype=torch.int32),
[ctc_output.shape[1]] *
len(ctc_output),
txt_len.cpu().tolist())
else:
ctc_loss = self.ctc_loss(ctc_output.transpose(
0, 1), txt, encode_len, txt_len)
total_loss += ctc_loss*self.model.ctc_weight
if att_output is not None:
b, t, _ = att_output.shape
att_output = fuse_output if self.emb_fuse else att_output
att_loss = self.seq_loss(
att_output.contiguous().view(b*t, -1), txt.contiguous().view(-1))
total_loss += att_loss*(1-self.model.ctc_weight)
self.timer.cnt('fw')
# Backprop
grad_norm = self.backward(total_loss)
step += 1
# Logger
self.progress('Tr stat | Loss - {:.2f} | Grad. Norm - {:.2f} | {}'
.format(total_loss.cpu().item(), grad_norm, self.timer.show()))
self.write_log(
'loss', {'tr_ctc': ctc_loss, 'tr_att': att_loss})
self.write_log('emb_loss', {'tr': emb_loss})
self.write_log('wer', {'tr_att': cal_er(self.tokenizer, att_output, txt),
'tr_ctc': cal_er(self.tokenizer, ctc_output, txt, ctc=True)})
if self.emb_fuse:
if self.emb_decoder.fuse_learnable:
self.write_log('fuse_lambda', {
'emb': self.emb_decoder.get_weight()})
self.write_log(
'fuse_temp', {'temp': self.emb_decoder.get_temp()})
# End of step
# https://github.com/pytorch/pytorch/issues/13246#issuecomment-529185354
torch.cuda.empty_cache()
self.timer.set()
if step > max_step:
break
ret = self.validate()
self.log.close()
return ret
def validate(self):
# Eval mode
self.model.eval()
if self.emb_decoder is not None:
self.emb_decoder.eval()
dev_wer = {'att': [], 'ctc': []}
for i, data in enumerate(self.dv_set):
self.progress('Valid step - {}/{}'.format(i+1, len(self.dv_set)))
# Fetch data
feat, feat_len, txt, txt_len = self.fetch_data(data)
# Forward model
with torch.no_grad():
ctc_output, encode_len, att_output, att_align, dec_state = \
self.model(feat, feat_len, int(max(txt_len)*self.DEV_STEP_RATIO),
emb_decoder=self.emb_decoder)
dev_wer['att'].append(cal_er(self.tokenizer, att_output, txt))
dev_wer['ctc'].append(cal_er(self.tokenizer, ctc_output, txt, ctc=True))
# Show some example on tensorboard
if i == len(self.dv_set)//2:
for i in range(min(len(txt), self.DEV_N_EXAMPLE)):
if True:
self.write_log('true_text{}'.format(
i), self.tokenizer.decode(txt[i].tolist()))
if att_output is not None:
self.write_log('att_align{}'.format(i), feat_to_fig(
att_align[i, 0, :, :].cpu().detach()))
self.write_log('att_text{}'.format(i), self.tokenizer.decode(
att_output[i].argmax(dim=-1).tolist()))
if ctc_output is not None:
self.write_log('ctc_text{}'.format(i), self.tokenizer.decode(ctc_output[i].argmax(dim=-1).tolist(),
ignore_repeat=True))
# Skip save model here
# Ckpt if performance improves
to_prints = []
for task in ['att', 'ctc']:
dev_wer[task] = sum(dev_wer[task]) / len(dev_wer[task])
if dev_wer[task] < self.best_wer[task]:
to_print = f"WER of {task}: {dev_wer[task]} < prev best ({self.best_wer[task]})"
self.best_wer[task] = dev_wer[task]
else:
to_print = f"WER of {task}: {dev_wer[task]} >= prev best ({self.best_wer[task]})"
print(to_print, flush=True)
to_prints.append(to_print)
# self.save_checkpoint('best_{}.pth'.format(task), 'wer', dev_wer[task])
self.write_log('wer', {'dv_'+task: dev_wer[task]})
# self.save_checkpoint('latest.pth', 'wer', dev_wer['att'], show_msg=False)
# Resume training
self.model.train()
if self.emb_decoder is not None:
self.emb_decoder.train()
return '\n'.join(to_prints)
class BaseSolver():
'''
Prototype Solver for all kinds of tasks
Arguments
config - yaml-styled config
paras - argparse outcome
'''
def __init__(self, config, paras, mode):
# General Settings
self.config = config
self.paras = paras
self.mode = mode
for k, v in default_hparas.items():
setattr(self, k, v)
self.device = torch.device(
'cuda') if self.paras.gpu and torch.cuda.is_available(
) else torch.device('cpu')
self.amp = paras.amp
# Name experiment
self.exp_name = paras.name
if self.exp_name is None:
# By default, exp is named after config file
self.exp_name = paras.config.split('/')[-1].replace('.yaml', '')
if mode == 'train':
self.exp_name += '_sd{}'.format(paras.seed)
# Plugin list
self.emb_decoder = None
if mode == 'train':
# Filepath setup
os.makedirs(paras.ckpdir, exist_ok=True)
self.ckpdir = os.path.join(paras.ckpdir, self.exp_name)
os.makedirs(self.ckpdir, exist_ok=True)
# Logger settings
self.logdir = os.path.join(paras.logdir, self.exp_name)
self.log = SummaryWriter(self.logdir,
flush_secs=self.TB_FLUSH_FREQ)
self.timer = Timer()
# Hyperparameters
self.step = 0
self.valid_step = config['hparas']['valid_step']
self.max_step = config['hparas']['max_step']
self.verbose('Exp. name : {}'.format(self.exp_name))
self.verbose('Loading data... large corpus may took a while.')
elif mode == 'test':
# Output path
os.makedirs(paras.outdir, exist_ok=True)
self.ckpdir = os.path.join(paras.outdir, self.exp_name)
# Load training config to get acoustic feat, text encoder and build model
self.src_config = yaml.load(open(config['src']['config'], 'r'),
Loader=yaml.FullLoader)
self.paras.load = config['src']['ckpt']
self.verbose('Evaluating result of tr. config @ {}'.format(
config['src']['config']))
def backward(self, loss):
'''
Standard backward step with self.timer and debugger
Arguments
loss - the loss to perform loss.backward()
'''
self.timer.set()
loss.backward()
grad_norm = torch.nn.utils.clip_grad_norm_(self.model.parameters(),
self.GRAD_CLIP)
if math.isnan(grad_norm):
self.verbose('Error : grad norm is NaN @ step ' + str(self.step))
else:
self.optimizer.step()
self.timer.cnt('bw')
return grad_norm
def load_ckpt(self):
''' Load ckpt if --load option is specified '''
if self.paras.load:
# Load weights
ckpt = torch.load(
self.paras.load,
map_location=self.device if self.mode == 'train' else 'cpu')
self.model.load_state_dict(ckpt['model'])
if self.emb_decoder is not None:
self.emb_decoder.load_state_dict(ckpt['emb_decoder'])
# if self.amp:
# amp.load_state_dict(ckpt['amp'])
# Load task-dependent items
if self.mode == 'train':
self.step = ckpt['global_step']
self.optimizer.load_opt_state_dict(ckpt['optimizer'])
self.verbose('Load ckpt from {}, restarting at step {}'.format(
self.paras.load, self.step))
else:
for k, v in ckpt.items():
if type(v) is float:
metric, score = k, v
self.model.eval()
if self.emb_decoder is not None:
self.emb_decoder.eval()
self.verbose(
'Evaluation target = {} (recorded {} = {:.2f} %)'.format(
self.paras.load, metric, score))
def verbose(self, msg):
''' Verbose function for print information to stdout'''
if self.paras.verbose:
if type(msg) == list:
for m in msg:
print('[INFO]', m.ljust(100))
else:
print('[INFO]', msg.ljust(100))
def progress(self, msg):
''' Verbose function for updating progress on stdout (do not include newline) '''
if self.paras.verbose:
sys.stdout.write("\033[K") # Clear line
print('[{}] {}'.format(human_format(self.step), msg), end='\r')
def write_log(self, log_name, log_dict):
'''
Write log to TensorBoard
log_name - <str> Name of tensorboard variable
log_value - <dict>/<array> Value of variable (e.g. dict of losses), passed if value = None
'''
if type(log_dict) is dict:
log_dict = {
key: val
for key, val in log_dict.items()
if (val is not None and not math.isnan(val))
}
if log_dict is None:
pass
elif len(log_dict) > 0:
if 'align' in log_name or 'spec' in log_name:
img, form = log_dict
self.log.add_image(log_name,
img,
global_step=self.step,
dataformats=form)
elif 'text' in log_name or 'hyp' in log_name:
self.log.add_text(log_name, log_dict, self.step)
else:
self.log.add_scalars(log_name, log_dict, self.step)
def save_checkpoint(self, f_name, metric, score, show_msg=True):
''''
Ckpt saver
f_name - <str> the name phnof ckpt file (w/o prefix) to store, overwrite if existed
score - <float> The value of metric used to evaluate model
'''
ckpt_path = os.path.join(self.ckpdir, f_name)
full_dict = {
"model": self.model.state_dict(),
"optimizer": self.optimizer.get_opt_state_dict(),
"global_step": self.step,
metric: score
}
# Additional modules to save
# if self.amp:
# full_dict['amp'] = self.amp_lib.state_dict()
if self.emb_decoder is not None:
full_dict['emb_decoder'] = self.emb_decoder.state_dict()
torch.save(full_dict, ckpt_path)
if show_msg:
self.verbose(
"Saved checkpoint (step = {}, {} = {:.2f}) and status @ {}".
format(human_format(self.step), metric, score, ckpt_path))
def enable_apex(self):
if self.amp:
# Enable mixed precision computation (ToDo: Save/Load amp)
from apex import amp
self.amp_lib = amp
self.verbose(
"AMP enabled (check https://github.com/NVIDIA/apex for more details)."
)
self.model, self.optimizer.opt = self.amp_lib.initialize(
self.model, self.optimizer.opt, opt_level='O1')
# ----------------------------------- Abtract Methods ------------------------------------------ #
@abc.abstractmethod
def load_data(self):
'''
Called by main to load all data
After this call, data related attributes should be setup (e.g. self.tr_set, self.dev_set)
No return value
'''
raise NotImplementedError
@abc.abstractmethod
def set_model(self):
'''
Called by main to set models
After this call, model related attributes should be setup (e.g. self.l2_loss)
The followings MUST be setup
- self.model (torch.nn.Module)
- self.optimizer (src.Optimizer),
init. w/ self.optimizer = src.Optimizer(self.model.parameters(),**self.config['hparas'])
Loading pre-trained model should also be performed here
No return value
'''
raise NotImplementedError
@abc.abstractmethod
def exec(self):
'''
Called by main to execute training/inference
'''
raise NotImplementedError
class BaseSolver():
'''
Prototype Solver for all kinds of tasks
Arguments
config - yaml-styled config
paras - argparse outcome
mode - string that specifies training/testing
'''
def __init__(self, config, paras):
# General Settings
self.config = config
self.paras = paras
for k, v in default_hparas.items():
setattr(self, k, v)
if self.paras.gpu and torch.cuda.is_available():
self.gpu = True
self.device = torch.device('cuda')
else:
self.gpu = False
self.device = torch.device('cpu')
# Settings for training/testing
self.mode = self.paras.mode # legacy, should be removed
# Name experiment
self.exp_name = paras.name
if self.exp_name is None:
# By default, exp is named after config file
self.exp_name = paras.config.split('/')[-1].split('.y')[0]
self.exp_name += '_sd{}'.format(paras.seed)
# Filepath setup
os.makedirs(paras.ckpdir, exist_ok=True)
self.ckpdir = os.path.join(paras.ckpdir, self.exp_name)
os.makedirs(self.ckpdir, exist_ok=True)
# Logger settings
self.logdir = os.path.join(paras.logdir, self.exp_name)
self.log = SummaryWriter(self.logdir, flush_secs=self.TB_FLUSH_FREQ)
self.timer = Timer()
# Hyperparameters
self.step = 0
self.epoch = config['hparas']['epoch']
self.verbose('Exp. name : {}'.format(self.exp_name))
self.verbose('Loading data...')
def backward(self, loss):
'''
Standard backward step with timer and debugger
Arguments
loss - the loss to perform loss.backward()
'''
self.timer.set()
loss.backward()
grad_norm = torch.nn.utils.clip_grad_norm_(self.model.parameters(),
self.GRAD_CLIP)
if math.isnan(grad_norm):
self.verbose('Error : grad norm is NaN @ step ' + str(self.step))
else:
self.optimizer.step()
self.timer.cnt('bw')
return grad_norm
def load_ckpt(self):
''' Load ckpt if --load option is specified '''
if self.paras.load:
# Load weights
ckpt = torch.load(self.paras.load,
map_location=self.device
if self.paras.mode == 'train' else 'cpu')
ckpt['model'] = {k.replace('module.','',1):v \
for k,v in ckpt['model'].items()}
self.model.load_state_dict(ckpt['model'])
# Load task-dependent items
metric = "None"
score = 0.0
for k, v in ckpt.items():
if type(v) is float:
metric, score = k, v
if self.paras.mode == 'train':
self.cur_epoch = ckpt['epoch']
self.step = ckpt['global_step']
self.optimizer.load_opt_state_dict(ckpt['optimizer'])
msg = \
'Load ckpt from {}, restarting at step {} \
(recorded {} = {:.2f} %)'\
.format(self.paras.load, self.step, metric, score)
self.verbose(msg)
else:
# Inference
msg = 'Evaluation target = {} (recorded {} = {:.2f} %)'\
.format(self.paras.load, metric, score)
self.verbose(msg)
def verbose(self, msg, display_step=False):
''' Verbose function for print information to stdout'''
header = '[' + human_format(
self.step) + ']' if display_step else '[INFO]'
if self.paras.verbose:
if type(msg) == list:
for m in msg:
print(header, m.ljust(100))
else:
print(header, msg.ljust(100))
def progress(self, msg):
''' Verbose function for updating progress on stdout
Do not include newline in msg '''
if self.paras.verbose:
sys.stdout.write("\033[K") # Clear line
print('[Ep {}] {}'.format(human_format(self.cur_epoch), msg),
end='\r')
def write_log(self, log_name, log_dict, bins=None):
''' Write log to TensorBoard
log_name - <str> Name of tensorboard variable
log_dict - <dict>/<array> Value of variable (e.g. dict of losses)
'''
if log_dict is not None:
if type(log_dict) is dict:
log_dict = {
key: val
for key, val in log_dict.items()
if (val is not None and not math.isnan(val))
}
self.log.add_scalars(log_name, log_dict, self.step)
elif 'Hist.' in log_name or 'Spec' in log_name:
img, form = log_dict
self.log.add_image(log_name,
img,
global_step=self.step,
dataformats=form)
else:
raise NotImplementedError
def save_checkpoint(self, f_name, metric, score, show_msg=True):
'''' pt saver
f_name - <str> the name of ckpt (w/o prefix), overwrite if existed
score - <float> The value of metric used to evaluate model
'''
ckpt_path = os.path.join(self.ckpdir, f_name)
full_dict = {
"model": self.model.state_dict(),
"optimizer": self.optimizer.get_opt_state_dict(),
"global_step": self.step,
"epoch": self.cur_epoch,
metric: score
}
torch.save(full_dict, ckpt_path)
if show_msg:
msg = "Saved checkpoint (epoch = {}, {} = {:.2f}) and status @ {}"
self.verbose(
msg.format(human_format(self.cur_epoch), metric, score,
ckpt_path))
return ckpt_path
# ----------------------------------- Abtract Methods ------------------- #
@abc.abstractmethod
def load_data(self):
'''
Called by main to load all data
After this call, data related attributes should be setup
(e.g. self.tr_set, self.dev_set)
No return value
'''
raise NotImplementedError
@abc.abstractmethod
def set_model(self):
'''
Called by main to set models
After this call, model related attributes should be setup
The followings MUST be setup
- self.model (torch.nn.Module)
- self.optimizer (src.Optimizer),
Loading pre-trained model should also be performed here
No return value
'''
raise NotImplementedError
@abc.abstractmethod
def exec(self):
'''
Called by main to execute training/inference
'''
raise NotImplementedError
