From 46e37e22ebb206bbb2e1d487f53645d9ed823b36 Mon Sep 17 00:00:00 2001 From: Hunor Laczko Date: Wed, 25 Jan 2023 21:00:47 +0000 Subject: [PATCH] fixed feature sizes, cvae working --- aaa_helpers.txt | 6 +++--- auto_encoder_static.py | 4 ++-- configs/config_ae_static.py | 2 +- configs/review/config_abl_all.py | 6 +++--- configs/review/config_ae_body.py | 2 +- configs/review/config_ae_curv.py | 2 +- configs/review/config_ae_static.py | 4 ++-- cvae.py | 2 -- models/base.py | 30 ++++++++++++++---------------- models/cond_encoder.py | 4 ++-- models/cvae_model.py | 12 ++++++------ models/resnet.py | 8 ++++---- 12 files changed, 39 insertions(+), 43 deletions(-) diff --git a/aaa_helpers.txt b/aaa_helpers.txt index e83d12c..c3f6529 100644 --- a/aaa_helpers.txt +++ b/aaa_helpers.txt @@ -1,6 +1,6 @@ -nohup python auto_encoder_static.py --name="lhunor_ae_static_r7" > nohup/review/ae_static_r7.out 2>&1 & -nohup python auto_encoder_body.py --name="lhunor_ae_body_r7" > nohup/review/ae_body_r7.out 2>&1 & -nohup python auto_encoder_curv.py --name="lhunor_ae_curv_r7" > nohup/review/ae_curv_r7.out 2>&1 & +nohup python auto_encoder_static.py --name="lhunor_ae_static_r7_2" > nohup/review/ae_static_r7_2.out 2>&1 & +nohup python auto_encoder_body.py --name="lhunor_ae_body_r7_2" > nohup/review/ae_body_r7_2.out 2>&1 & +nohup python auto_encoder_curv.py --name="lhunor_ae_curv_r7_2" > nohup/review/ae_curv_r7_2.out 2>&1 & one final convolutional layer for smoothing, on the final coordinates diff --git a/auto_encoder_static.py b/auto_encoder_static.py index 4227d19..a9fcbec 100644 --- a/auto_encoder_static.py +++ b/auto_encoder_static.py @@ -54,7 +54,7 @@ uv2mesh = UV2Mesh(res=256, is_skirt=config['is_skirt'], is_smpl_orig=config['is_ -@tf.function +# @tf.function def train_step(batch): with tf.GradientTape() as tape: pred = model([batch['uv_static'], batch['uv_mask']], training=True) @@ -187,7 +187,7 @@ def test(): if __name__ == '__main__': - init_wandb() + # init_wandb() # allow_memory_growth() parser = argparse.ArgumentParser(description='') parser.add_argument('--name', type=str) diff --git a/configs/config_ae_static.py b/configs/config_ae_static.py index c027578..82e8f73 100644 --- a/configs/config_ae_static.py +++ b/configs/config_ae_static.py @@ -31,7 +31,7 @@ config['cond_ae']['decoder_dims'] = config['cond_ae']['encoder_dims'] config['cond_ae']['activation'] = 'linear' -config['gpu'] = '5' +config['gpu'] = '0' config['lr'] = 0.0005 config['name'] = 'static_post_4' config['notes'] = 'convnext + convnext, | \ diff --git a/configs/review/config_abl_all.py b/configs/review/config_abl_all.py index 11ad950..2ae6f4c 100644 --- a/configs/review/config_abl_all.py +++ b/configs/review/config_abl_all.py @@ -43,9 +43,9 @@ config['test_dir'] = '/home/workspace/data/dataset/test_ready' # config['frames_file_valid'] = 'frames_dress_30k.txt' # config['frames_file_test'] = 'frames_dress_30k.txt' -config['static_model'] = 'checkpoints/post/static_r7' -config['body_model'] = 'checkpoints/post/body_r7' -config['curvature_model'] = 'checkpoints/post/curv_r7' +config['static_model'] = 'checkpoints/post/static_r7_2' +config['body_model'] = 'checkpoints/post/body_r7_2' +config['curvature_model'] = 'checkpoints/post/curv_r7_2' config['smpl_res'] = 14475 diff --git a/configs/review/config_ae_body.py b/configs/review/config_ae_body.py index 210dc6b..1f6c837 100644 --- a/configs/review/config_ae_body.py +++ b/configs/review/config_ae_body.py @@ -33,7 +33,7 @@ config['cond_ae']['activation'] = 'tanh' config['gpu'] = '4' config['lr'] = 0.0001 -config['name'] = 'body_r7' +config['name'] = 'body_r7_2' config['notes'] = 'resnet18' config['tags'] = ['review_7'] diff --git a/configs/review/config_ae_curv.py b/configs/review/config_ae_curv.py index 07418dd..dc0f310 100644 --- a/configs/review/config_ae_curv.py +++ b/configs/review/config_ae_curv.py @@ -33,7 +33,7 @@ config['cond_ae']['activation'] = 'tanh' config['gpu'] = '3' config['lr'] = 0.001 -config['name'] = 'curv_r7' +config['name'] = 'curv_r7_2' config['notes'] = 'resnet18' config['tags'] = ['review_7'] diff --git a/configs/review/config_ae_static.py b/configs/review/config_ae_static.py index 5885a72..f44907e 100644 --- a/configs/review/config_ae_static.py +++ b/configs/review/config_ae_static.py @@ -31,9 +31,9 @@ config['cond_ae']['decoder_dims'] = config['cond_ae']['encoder_dims'] config['cond_ae']['activation'] = 'linear' -config['gpu'] = '6' +config['gpu'] = '0' config['lr'] = 0.0005 -config['name'] = 'static_r7' +config['name'] = 'static_r7_2' config['notes'] = 'resnet18 based' config['tags'] = ['review_7'] diff --git a/cvae.py b/cvae.py index 8d701ea..cd46b2e 100644 --- a/cvae.py +++ b/cvae.py @@ -425,8 +425,6 @@ def sample(folder): - - def test(): str_split = config['name'].split('_') test_idx = str_split.index('test') diff --git a/models/base.py b/models/base.py index 523b2b3..d0feffb 100644 --- a/models/base.py +++ b/models/base.py @@ -1,6 +1,6 @@ import tensorflow as tf from convnext.models.convnext_tf import ConvNeXt, ConvNeXtReverse, Block, ConvNeXtReverseLabels -from models.resnet import resnet_18 +from models.resnet import resnet_18, make_basic_block_layer class Bottleneck(tf.keras.Model): @@ -113,20 +113,18 @@ class Decoder(tf.keras.Model): class VariationalEncoderNew(tf.keras.Model): - def __init__(self, depths, dims, kernel_size=7): + def __init__(self, features): super(VariationalEncoderNew, self).__init__() - self.encoder = ConvNeXt(depths=depths, dims=dims, include_top=False, kernel_size=kernel_size) + self.encoder = resnet_18(features) - self.mean_branch = tf.keras.Sequential( - [Block(dim=64, drop_path=0, kernel_size=kernel_size, - layer_scale_init_value=1e-6, prefix=f'bottleneck{j}') for j in range(5)] - ) + self.mean_branch = make_basic_block_layer(filter_num=features, + blocks=2, + stride=1) - self.var_branch = tf.keras.Sequential( - [Block(dim=64, drop_path=0, kernel_size=kernel_size, - layer_scale_init_value=1e-6, prefix=f'bottleneck{j}') for j in range(5)] - ) + self.var_branch = make_basic_block_layer(filter_num=features, + blocks=2, + stride=1) self.flatten_mean = tf.keras.layers.Flatten() self.flatten_var = tf.keras.layers.Flatten() @@ -149,10 +147,10 @@ class VariationalEncoderNew(tf.keras.Model): class AEEncoderNew(tf.keras.Model): - def __init__(self, depths, dims, kernel_size=7, activation='tanh'): + def __init__(self, features, activation='tanh'): super(AEEncoderNew, self).__init__() - self.encoder = resnet_18() + self.encoder = resnet_18(features) self.act = tf.keras.layers.Activation(activation) @@ -173,15 +171,15 @@ class AEEncoderNew(tf.keras.Model): class DecoderNew(tf.keras.Model): - def __init__(self, depths, dims, kernel_size=7, is_label=False): + def __init__(self, features, is_label=False, is_cvae=False): super(DecoderNew, self).__init__() self.is_label = is_label if not is_label: - self.decoder = resnet_18(reversed=True) + self.decoder = resnet_18(features, reversed=True, is_cvae=is_cvae) else: - self.decoder = resnet_18(reversed=True, labels=6) + self.decoder = resnet_18(features, reversed=True, labels=6, is_cvae=is_cvae) def call(self, inputs, training): diff --git a/models/cond_encoder.py b/models/cond_encoder.py index c86f54c..076357b 100644 --- a/models/cond_encoder.py +++ b/models/cond_encoder.py @@ -25,8 +25,8 @@ class AutoEncoderNew(tf.keras.Model): def __init__(self, config): super(AutoEncoderNew, self).__init__() - self.encoder : AEEncoderNew = AEEncoderNew(config['encoder_depths'], config['encoder_dims'], config['kernel_size'], config['activation']) - self.decoder : DecoderNew = DecoderNew(config['decoder_depths'], config['decoder_dims'], kernel_size=config['kernel_size'], is_label=config['is_label']) + self.encoder : AEEncoderNew = AEEncoderNew(config['bottleneck_size'], config['activation']) + self.decoder : DecoderNew = DecoderNew(config['bottleneck_size'], is_label=config['is_label']) def call(self, inputs, training): diff --git a/models/cvae_model.py b/models/cvae_model.py index 3ed4571..ef2193f 100644 --- a/models/cvae_model.py +++ b/models/cvae_model.py @@ -159,16 +159,16 @@ class CVAENew(tf.keras.Model): self.size = config['resolution'] if not('only_curvatures' in config.keys() and config['only_curvatures'] == True): - self.static_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_static']['encoder_depths'], config['cond_ae_static']['encoder_dims'], config['cond_ae_static']['kernel_size'], config['cond_ae_static']['activation']) + self.static_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_static']['bottleneck_size'], config['cond_ae_static']['activation']) if config['static_model'] != '': load_model(self.static_enc, config['static_model'] + '_enc') - self.body_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_body']['encoder_depths'], config['cond_ae_body']['encoder_dims'], config['cond_ae_body']['kernel_size']) + self.body_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_body']['bottleneck_size']) if config['body_model'] != '': load_model(self.body_enc, config['body_model'] + '_enc') if config['use_curvature']: - self.curv_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_curv']['encoder_depths'], config['cond_ae_curv']['encoder_dims'], config['cond_ae_curv']['kernel_size'], config['cond_ae_curv']['activation']) + self.curv_enc : AEEncoderNew = AEEncoderNew(config['cond_ae_curv']['bottleneck_size'], config['cond_ae_curv']['activation']) if config['curvature_model'] != '': load_model(self.curv_enc, config['curvature_model'] + '_enc') @@ -179,10 +179,10 @@ class CVAENew(tf.keras.Model): if config['use_curvature']: self.curv_enc.trainable = False - self.encoder = VariationalEncoderNew(config['cvae']['encoder_depths'], config['cvae']['encoder_dims'], kernel_size=config['cvae']['kernel_size']) - self.decoder = DecoderNew(config['cvae']['decoder_depths'], config['cvae']['decoder_dims'], config['cvae']['kernel_size']) + self.encoder = VariationalEncoderNew(config['cvae']['encoder_dims'][-1]) + self.decoder = DecoderNew(config['cvae']['decoder_dims'][-1], is_cvae=True) - self.reshape = tf.keras.layers.Reshape((8, 4, 64)) + self.reshape = tf.keras.layers.Reshape((16, 8, config['cvae']['decoder_dims'][-1])) if config['avg_garment_path'] != '': self.avg_garment = np.load(config['avg_garment_path']) diff --git a/models/resnet.py b/models/resnet.py index 40d64fd..102e3f3 100644 --- a/models/resnet.py +++ b/models/resnet.py @@ -116,7 +116,7 @@ def make_bottleneck_layer(filter_num, blocks, stride=1): class ResNetTypeI(tf.keras.Model): - def __init__(self, layer_params, reversed=False, labels=None): + def __init__(self, layer_params, features, reversed=False, labels=None, is_cvae=False): super(ResNetTypeI, self).__init__() self.reversed = reversed self.labels = labels @@ -153,7 +153,7 @@ class ResNetTypeI(tf.keras.Model): blocks=layer_params[2], stride=2, upsample=reversed) - self.layer4 = make_basic_block_layer(filter_num=512, + self.layer4 = make_basic_block_layer(filter_num=features if not is_cvae else features*2, blocks=layer_params[3], stride=2 if not reversed else 1, upsample=reversed) @@ -228,8 +228,8 @@ class ResNetTypeII(tf.keras.Model): return output -def resnet_18(reversed=False, labels=None): - return ResNetTypeI(layer_params=[2, 2, 2, 2], reversed=reversed, labels=labels) +def resnet_18(features, reversed=False, labels=None, is_cvae=False): + return ResNetTypeI(layer_params=[2, 2, 2, 2], features=features, reversed=reversed, labels=labels, is_cvae=is_cvae) def resnet_34():