opensource/PaddleOCR
3
0
Fork 0
mirror of https://github.com/PaddlePaddle/PaddleOCR.git synced 2026-09-24 23:33:08 +08:00
Code Issues Packages Projects Releases Wiki Activity

merge upstream

Browse Source
This commit is contained in:
smilelite
2022-07-10 12:15:27 +08:00
parent 32f85fd92b b6982b7fc7
commit bc1c19c5d2
219 changed files with 13944 additions and 4841 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/program.py
+30 -15
View File
@@ -255,6 +255,8 @@ def train(config,
with paddle.amp.auto_cast():
if model_type == 'table' or extra_input:
preds = model(images, data=batch[1:])
elif model_type in ["kie", 'vqa']:
preds = model(batch)
else:
preds = model(images)
else:
@@ -279,8 +281,11 @@ def train(config,
if cal_metric_during_train and epoch % calc_epoch_interval == 0: # only rec and cls need
batch = [item.numpy() for item in batch]
if model_type in ['table', 'kie']:
if model_type in ['kie']:
eval_class(preds, batch)
elif model_type in ['table']:
post_result = post_process_class(preds, batch)
eval_class(post_result, batch)
else:
if config['Loss']['name'] in ['MultiLoss', 'MultiLoss_v2'
]: # for multi head loss
@@ -307,7 +312,8 @@ def train(config,
train_stats.update(stats)
if log_writer is not None and dist.get_rank() == 0:
log_writer.log_metrics(metrics=train_stats.get(), prefix="TRAIN", step=global_step)
log_writer.log_metrics(
metrics=train_stats.get(), prefix="TRAIN", step=global_step)
if dist.get_rank() == 0 and (
(global_step > 0 and global_step % print_batch_step == 0) or
@@ -354,7 +360,8 @@ def train(config,
# logger metric
if log_writer is not None:
log_writer.log_metrics(metrics=cur_metric, prefix="EVAL", step=global_step)
log_writer.log_metrics(
metrics=cur_metric, prefix="EVAL", step=global_step)
if cur_metric[main_indicator] >= best_model_dict[
main_indicator]:
@@ -377,11 +384,18 @@ def train(config,
logger.info(best_str)
# logger best metric
if log_writer is not None:
log_writer.log_metrics(metrics={
"best_{}".format(main_indicator): best_model_dict[main_indicator]
}, prefix="EVAL", step=global_step)
log_writer.log_model(is_best=True, prefix="best_accuracy", metadata=best_model_dict)
log_writer.log_metrics(
metrics={
"best_{}".format(main_indicator):
best_model_dict[main_indicator]
},
prefix="EVAL",
step=global_step)
log_writer.log_model(
is_best=True,
prefix="best_accuracy",
metadata=best_model_dict)
reader_start = time.time()
if dist.get_rank() == 0:
@@ -413,7 +427,8 @@ def train(config,
epoch=epoch,
global_step=global_step)
if log_writer is not None:
log_writer.log_model(is_best=False, prefix='iter_epoch_{}'.format(epoch))
log_writer.log_model(
is_best=False, prefix='iter_epoch_{}'.format(epoch))
best_str = 'best metric, {}'.format(', '.join(
['{}: {}'.format(k, v) for k, v in best_model_dict.items()]))
@@ -451,7 +466,6 @@ def eval(model,
preds = model(batch)
else:
preds = model(images)
batch_numpy = []
for item in batch:
if isinstance(item, paddle.Tensor):
@@ -461,9 +475,9 @@ def eval(model,
# Obtain usable results from post-processing methods
total_time += time.time() - start
# Evaluate the results of the current batch
if model_type in ['table', 'kie']:
if model_type in ['kie']:
eval_class(preds, batch_numpy)
elif model_type in ['vqa']:
elif model_type in ['table', 'vqa']:
post_result = post_process_class(preds, batch_numpy)
eval_class(post_result, batch_numpy)
else:
@@ -564,8 +578,8 @@ def preprocess(is_train=False):
assert alg in [
'EAST', 'DB', 'SAST', 'Rosetta', 'CRNN', 'STARNet', 'RARE', 'SRN',
'CLS', 'PGNet', 'Distillation', 'NRTR', 'TableAttn', 'SAR', 'PSE',
'SEED', 'SDMGR', 'LayoutXLM', 'LayoutLM', 'PREN', 'FCE', 'SVTR',
'RobustScanner'
'SEED', 'SDMGR', 'LayoutXLM', 'LayoutLM', 'LayoutLMv2', 'PREN', 'FCE',
'SVTR', 'ViTSTR', 'ABINet', 'DB++', 'TableMaster', 'RobustScanner'
]
if use_xpu:
@@ -586,7 +600,8 @@ def preprocess(is_train=False):
vdl_writer_path = '{}/vdl/'.format(save_model_dir)
log_writer = VDLLogger(save_model_dir)
loggers.append(log_writer)
if ('use_wandb' in config['Global'] and config['Global']['use_wandb']) or 'wandb' in config:
if ('use_wandb' in config['Global'] and
config['Global']['use_wandb']) or 'wandb' in config:
save_dir = config['Global']['save_model_dir']
wandb_writer_path = "{}/wandb".format(save_dir)
if "wandb" in config: