diff --git a/oceanai/__init__.py b/oceanai/__init__.py index d57654e..6b16ffe 100644 --- a/oceanai/__init__.py +++ b/oceanai/__init__.py @@ -18,7 +18,7 @@ __uri__ = "https://github.com/DmitryRyumin/oceanai" __version__ = "1.0" -__release__ = __version__ + ".0a41" +__release__ = __version__ + ".0a42" __author__ru__ = "Рюмина Елена, Рюмин Дмитрий, Карпов Алексей" __author__en__ = "Elena Ryumina, Dmitry Ryumin, Alexey Karpov" diff --git a/oceanai/modules/lab/video.py b/oceanai/modules/lab/video.py index b6f7a80..04ca2c2 100644 --- a/oceanai/modules/lab/video.py +++ b/oceanai/modules/lab/video.py @@ -353,6 +353,7 @@ def video_model_hc_(self) -> Optional[nn.Module]: video = Video() video.load_video_model_hc( + lang = 'en', show_summary = False, out = True, runtime = True, run = True ) @@ -363,11 +364,17 @@ def video_model_hc_(self) -> Optional[nn.Module]: :execution-count: 1 :linenos: - [2022-10-26 12:37:42] Формирование нейросетевой архитектуры модели для получения оценок по экспертным признакам (видео модальность) ... + [2024-10-09 13:19:24] Формирование нейросетевой архитектуры модели для получения оценок по экспертным признакам (видео модальность) ... - --- Время выполнения: 1.112 сек. --- + --- Время выполнения: 0.005 сек. --- - + video_model_hc( + (lstm1): LSTM(115, 64, batch_first=True) + (dropout1): Dropout(p=0.2, inplace=False) + (lstm2): LSTM(64, 128, batch_first=True) + (dropout2): Dropout(p=0.2, inplace=False) + (fc): Linear(in_features=128, out_features=5, bias=True) + ) :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -423,11 +430,15 @@ def video_model_nn_(self) -> Optional[nn.Module]: :execution-count: 1 :linenos: - [2022-10-27 14:49:00] Формирование нейросетевой архитектуры для получения оценок по нейросетевым признакам (видео модальность) ... + [2024-10-09 13:20:47] Формирование нейросетевой архитектуры для получения оценок по нейросетевым признакам (видео модальность) ... - --- Время выполнения: 1.986 сек. --- + --- Время выполнения: 0.055 сек. --- - + video_model_nn( + (lstm1): LSTM(512, 1024, batch_first=True) + (dropout1): Dropout(p=0.2, inplace=False) + (fc): Linear(in_features=1024, out_features=5, bias=True) + ) :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -483,11 +494,188 @@ def video_model_deep_fe_(self) -> Optional[nn.Module]: :execution-count: 1 :linenos: - [2022-11-01 12:12:35] Формирование нейросетевой архитектуры для получения нейросетевых признаков (видео модальность) ... - - --- Время выполнения: 1.468 сек. --- - - + [2024-10-09 13:17:09] Формирование нейросетевой архитектуры для получения нейросетевых признаков (видео модальность) ... + + --- Время выполнения: 0.228 сек. --- + + ResNet( + (conv_layer_s2_same): Conv2dSame(3, 64, kernel_size=(7, 7), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + (max_pool): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False) + (layer1): Sequential( + (0): Bottleneck( + (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer2): Sequential( + (0): Bottleneck( + (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (3): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer3): Sequential( + (0): Bottleneck( + (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (3): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (4): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (5): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer4): Sequential( + (0): Bottleneck( + (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (avgpool): AdaptiveAvgPool2d(output_size=(1, 1)) + (fc1): Linear(in_features=2048, out_features=512, bias=True) + (relu1): ReLU() + (fc2): Linear(in_features=512, out_features=7, bias=True) + ) :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -543,17 +731,30 @@ def video_models_b5_(self) -> Dict[str, Optional[nn.Module]]: :execution-count: 1 :linenos: - [2022-10-19 15:45:35] Формирование нейросетевых архитектур моделей для получения результатов оценки персональных качеств ... - - --- Время выполнения: 0.07 сек. --- - - { - 'openness': , - 'conscientiousness': , - 'extraversion': , - 'agreeableness': , - 'non_neuroticism': - } + [2024-10-09 13:21:52] Формирование нейросетевых архитектур моделей для получения результатов оценки персональных качеств (видео модальность) ... + + --- Время выполнения: 0.004 сек. --- + + {'openness': video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ), + 'conscientiousness': video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ), + 'extraversion': video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ), + 'agreeableness': video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ), + 'non-neuroticism': video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + )} :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -631,7 +832,7 @@ def __load_model_weights( video.chunk_size_ = 2000000 video._Video__load_model_weights( - url = 'https://download.sberdisk.ru/download/file/412059444?token=JXerCfAjJZg6crD&filename=weights_2022-08-27_18-53-35.h5', + url = 'https://drive.usercontent.google.com/download?id=1QF7ReDQXpCciF7aWjbEt4Q-x06hwDrMZ&export=download&authuser=2&confirm=t&uuid=c2fd5a21-7af7-4b7f-8419-d7d628847768&at=AO7h07eilj-Bm5RIk0HwQBEr37ri:1727175670133', force_reload = True, info_text = 'Загрузка весов нейросетевой модели', out = True, runtime = True, run = True @@ -641,11 +842,11 @@ def __load_model_weights( :execution-count: 1 :linenos: - [2022-10-27 12:46:55] Загрузка весов нейросетевой модели + [2024-10-09 11:42:21] Загрузка весов нейросетевой модели - [2022-10-27 12:46:55] Загрузка файла "weights_2022-08-27_18-53-35.h5" (100.0%) ... + [2024-10-09 11:42:25] Загрузка файла "weights_2022-03-22_16-31-48.pth" 100.0% ... - --- Время выполнения: 0.626 сек. --- + --- Время выполнения: 4.117 сек. --- True @@ -664,7 +865,7 @@ def __load_model_weights( video.chunk_size_ = 2000000 video._Video__load_model_weights( - url = './models/weights_2022-08-27_18-53-35.h5', + url = './models/weights_2022-03-22_16-31-48.pth', force_reload = True, info_text = 'Загрузка весов нейросетевой модели', out = True, runtime = True, run = True @@ -674,9 +875,9 @@ def __load_model_weights( :execution-count: 2 :linenos: - [2022-10-27 12:47:52] Загрузка весов нейросетевой модели + [2024-10-09 11:46:15] Загрузка весов нейросетевой модели - --- Время выполнения: 0.002 сек. --- + --- Время выполнения: 0.005 сек. --- True @@ -695,7 +896,7 @@ def __load_model_weights( video.chunk_size_ = 2000000 video._Video__load_model_weights( - url = 'https://download.sberdisk.ru/download/file/412059444?token=JXerCfAjJZg6crD&filename=weights_2022-08-27_18-53-35.h5', + url = 'https://drive.usercontent.google.com/download?id=1QF7ReDQXpCciF7aWjbEt4Q-x06hwDrMZ&export=download&authuser=2&confirm=t&uuid=c2fd5a21-7af7-4b7f-8419-d7d628847768&at=AO7h07eilj-Bm5RIk0HwQBEr37ri:1727175670133', force_reload = True, info_text = '', out = True, runtime = True, run = True ) @@ -1147,8 +1348,7 @@ def __concat_pred(self, pred_hc: np.ndarray, pred_nn: np.ndarray, out: bool = Tr :execution-count: 3 :linenos: - [2022-10-20 22:33:31] Ой! Что-то пошло не так ... конкатенация оценок по экспертным и нейросетевым - признакам не произведена (видео модальность) ... + [2024-10-09 11:34:39] Что-то пошло не так ... конкатенация оценок по экспертным и нейросетевым признакам не произведена (видео модальность) ... [] """ @@ -1227,22 +1427,14 @@ def __load_video_model_b5(self, show_summary: bool = False, out: bool = True) -> :execution-count: 1 :linenos: - Model: "model" - _________________________________________________________________ - Layer (type) Output Shape Param # - ================================================================= - input_1 (InputLayer) [(None, 32)] 0 - - dense_1 (Dense) (None, 1) 33 - - activ_1 (Activation) (None, 1) 0 - - ================================================================= - Total params: 33 - Trainable params: 33 - Non-trainable params: 0 - _________________________________________________________________ - + video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ) + video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ) :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -1357,7 +1549,7 @@ def _get_visual_features( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['fe']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['fe']['googledisk'] res_load_video_model_weights_deep_fe = video.load_video_model_weights_deep_fe( url = url, @@ -1369,11 +1561,11 @@ def _get_visual_features( :execution-count: 2 :linenos: - [2022-11-03 16:39:10] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... + [2024-10-09 12:19:15] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... - [2022-11-03 16:39:14] Загрузка файла "weights_2022-11-01_12-27-07.h5" (100.0%) ... + [2024-10-09 12:19:20] Загрузка файла "weights_2022-11-01_12-27-07.pth" 100.0% ... - --- Время выполнения: 4.874 сек. --- + --- Время выполнения: 5.445 сек. --- .. code-cell:: python :execution-count: 3 @@ -1392,17 +1584,17 @@ def _get_visual_features( :execution-count: 3 :linenos: - [2022-11-03 16:56:52] Извлечение признаков (экспертных и нейросетевых) из визуального сигнала ... + [2024-10-09 12:20:39] Извлечение признаков (экспертных и нейросетевых) из визуального сигнала ... - [2022-11-03 16:56:58] Статистика извлеченных признаков из визуального сигнала: + [2024-10-09 12:20:46] Статистика извлеченных признаков из визуального сигнала: Общее количество сегментов с: 1. экспертными признаками: 12 2. нейросетевыми признаками: 12 - Размерность матрицы экспертных признаков одного сегмента: 10 ✕ 115 - Размерность тензора с нейросетевыми признаками одного сегмента: 10 ✕ 512 + Размерность матрицы экспертных признаков одного сегмента: 10 ✕ 109 + Размерность матрицы с нейросетевыми признаками одного сегмента: 10 ✕ 512 Понижение кадровой частоты: с 30 до 5 - --- Время выполнения: 6.109 сек. --- + --- Время выполнения: 7.123 сек. --- :bdg-danger:`Ошибка` :bdg-light:`-- 1 --` @@ -1427,11 +1619,11 @@ def _get_visual_features( :execution-count: 4 :linenos: - [2022-11-03 16:59:45] Извлечение признаков (экспертных и нейросетевых) из визуального сигнала ... + [2024-10-09 12:21:55] Извлечение признаков (экспертных и нейросетевых) из визуального сигнала ... - [2022-11-03 16:59:46] Ой! Что-то пошло не так ... нейросетевая архитектура модели для получения нейросетевых признаков не сформирована (видео модальность) ... + [2024-10-09 12:21:57] Что-то пошло не так ... нейросетевая архитектура модели для получения нейросетевых признаков не сформирована (видео модальность) ... - --- Время выполнения: 1.358 сек. --- + --- Время выполнения: 1.202 сек. --- """ try: @@ -1765,9 +1957,9 @@ def alignment_procedure(left_eye: List[int], right_eye: List[int]) -> float: if num_images > batch_size_limit: all_extract_deep_fe = [] all_pred_emo = [] - + for i in range(0, num_images, batch_size_limit): - bndbox_subbatch = bndbox_faces[i:i + batch_size_limit].to(self._device) + bndbox_subbatch = bndbox_faces[i : i + batch_size_limit].to(self._device) pred_emo_sub, extract_deep_fe_sub = self._video_model_deep_fe(bndbox_subbatch) extract_deep_fe_sub = extract_deep_fe_sub.detach().cpu() pred_emo_sub = pred_emo_sub.detach().cpu() @@ -1834,7 +2026,7 @@ def alignment_procedure(left_eye: List[int], right_eye: List[int]) -> float: # ------------------------------------------------------------------------------------------------------------------ def load_video_model_hc( - self, lang: str, show_summary: bool = False, out: bool = True, runtime: bool = True, run: bool = True + self, lang: str = "ru", show_summary: bool = False, out: bool = True, runtime: bool = True, run: bool = True ) -> bool: """Формирование нейросетевой архитектуры модели для получения оценок по экспертным признакам @@ -1862,6 +2054,7 @@ def load_video_model_hc( video = Video() video.load_video_model_hc( + lang = 'en', show_summary = False, out = True, runtime = True, run = True ) @@ -1887,6 +2080,7 @@ def load_video_model_hc( video = Video() video.load_video_model_hc( + lang = 'en', show_summary = 1, out = True, runtime = True, run = True ) @@ -1976,463 +2170,187 @@ def load_video_model_deep_fe( :execution-count: 1 :linenos: - [2022-11-01 12:18:14] Формирование нейросетевой архитектуры для получения нейросетевых признаков (видео модальность) ... - - Model: "model_1" - __________________________________________________________________________________________________ - Layer (type) Output Shape Param # Connected to - ================================================================================================== - input_2 (InputLayer) [(None, 224, 224, 3 0 [] - )] - - conv1/7x7_s2 (Conv2D) (None, 112, 112, 64 9408 ['input_2[0][0]'] - ) - - conv1/7x7_s2/bn (BatchNormaliz (None, 112, 112, 64 256 ['conv1/7x7_s2[0][0]'] - ation) ) - - activation_49 (Activation) (None, 112, 112, 64 0 ['conv1/7x7_s2/bn[0][0]'] - ) - - max_pooling2d_1 (MaxPooling2D) (None, 55, 55, 64) 0 ['activation_49[0][0]'] - - conv2_1_1x1_reduce (Conv2D) (None, 55, 55, 64) 4096 ['max_pooling2d_1[0][0]'] - - conv2_1_1x1_reduce/bn (BatchNo (None, 55, 55, 64) 256 ['conv2_1_1x1_reduce[0][0]'] - rmalization) - - activation_50 (Activation) (None, 55, 55, 64) 0 ['conv2_1_1x1_reduce/bn[0][0]'] - - conv2_1_3x3 (Conv2D) (None, 55, 55, 64) 36864 ['activation_50[0][0]'] - - conv2_1_3x3/bn (BatchNormaliza (None, 55, 55, 64) 256 ['conv2_1_3x3[0][0]'] - tion) - - activation_51 (Activation) (None, 55, 55, 64) 0 ['conv2_1_3x3/bn[0][0]'] - - conv2_1_1x1_increase (Conv2D) (None, 55, 55, 256) 16384 ['activation_51[0][0]'] - - conv2_1_1x1_proj (Conv2D) (None, 55, 55, 256) 16384 ['max_pooling2d_1[0][0]'] - - conv2_1_1x1_increase/bn (Batch (None, 55, 55, 256) 1024 ['conv2_1_1x1_increase[0][0]'] - Normalization) - - conv2_1_1x1_proj/bn (BatchNorm (None, 55, 55, 256) 1024 ['conv2_1_1x1_proj[0][0]'] - alization) - - add_16 (Add) (None, 55, 55, 256) 0 ['conv2_1_1x1_increase/bn[0][0]', - 'conv2_1_1x1_proj/bn[0][0]'] - - activation_52 (Activation) (None, 55, 55, 256) 0 ['add_16[0][0]'] - - conv2_2_1x1_reduce (Conv2D) (None, 55, 55, 64) 16384 ['activation_52[0][0]'] - - conv2_2_1x1_reduce/bn (BatchNo (None, 55, 55, 64) 256 ['conv2_2_1x1_reduce[0][0]'] - rmalization) - - activation_53 (Activation) (None, 55, 55, 64) 0 ['conv2_2_1x1_reduce/bn[0][0]'] - - conv2_2_3x3 (Conv2D) (None, 55, 55, 64) 36864 ['activation_53[0][0]'] - - conv2_2_3x3/bn (BatchNormaliza (None, 55, 55, 64) 256 ['conv2_2_3x3[0][0]'] - tion) - - activation_54 (Activation) (None, 55, 55, 64) 0 ['conv2_2_3x3/bn[0][0]'] - - conv2_2_1x1_increase (Conv2D) (None, 55, 55, 256) 16384 ['activation_54[0][0]'] - - conv2_2_1x1_increase/bn (Batch (None, 55, 55, 256) 1024 ['conv2_2_1x1_increase[0][0]'] - Normalization) - - add_17 (Add) (None, 55, 55, 256) 0 ['conv2_2_1x1_increase/bn[0][0]', - 'activation_52[0][0]'] - - activation_55 (Activation) (None, 55, 55, 256) 0 ['add_17[0][0]'] - - conv2_3_1x1_reduce (Conv2D) (None, 55, 55, 64) 16384 ['activation_55[0][0]'] - - conv2_3_1x1_reduce/bn (BatchNo (None, 55, 55, 64) 256 ['conv2_3_1x1_reduce[0][0]'] - rmalization) - - activation_56 (Activation) (None, 55, 55, 64) 0 ['conv2_3_1x1_reduce/bn[0][0]'] - - conv2_3_3x3 (Conv2D) (None, 55, 55, 64) 36864 ['activation_56[0][0]'] - - conv2_3_3x3/bn (BatchNormaliza (None, 55, 55, 64) 256 ['conv2_3_3x3[0][0]'] - tion) - - activation_57 (Activation) (None, 55, 55, 64) 0 ['conv2_3_3x3/bn[0][0]'] - - conv2_3_1x1_increase (Conv2D) (None, 55, 55, 256) 16384 ['activation_57[0][0]'] - - conv2_3_1x1_increase/bn (Batch (None, 55, 55, 256) 1024 ['conv2_3_1x1_increase[0][0]'] - Normalization) - - add_18 (Add) (None, 55, 55, 256) 0 ['conv2_3_1x1_increase/bn[0][0]', - 'activation_55[0][0]'] - - activation_58 (Activation) (None, 55, 55, 256) 0 ['add_18[0][0]'] - - conv3_1_1x1_reduce (Conv2D) (None, 28, 28, 128) 32768 ['activation_58[0][0]'] - - conv3_1_1x1_reduce/bn (BatchNo (None, 28, 28, 128) 512 ['conv3_1_1x1_reduce[0][0]'] - rmalization) - - activation_59 (Activation) (None, 28, 28, 128) 0 ['conv3_1_1x1_reduce/bn[0][0]'] - - conv3_1_3x3 (Conv2D) (None, 28, 28, 128) 147456 ['activation_59[0][0]'] - - conv3_1_3x3/bn (BatchNormaliza (None, 28, 28, 128) 512 ['conv3_1_3x3[0][0]'] - tion) - - activation_60 (Activation) (None, 28, 28, 128) 0 ['conv3_1_3x3/bn[0][0]'] - - conv3_1_1x1_increase (Conv2D) (None, 28, 28, 512) 65536 ['activation_60[0][0]'] - - conv3_1_1x1_proj (Conv2D) (None, 28, 28, 512) 131072 ['activation_58[0][0]'] - - conv3_1_1x1_increase/bn (Batch (None, 28, 28, 512) 2048 ['conv3_1_1x1_increase[0][0]'] - Normalization) - - conv3_1_1x1_proj/bn (BatchNorm (None, 28, 28, 512) 2048 ['conv3_1_1x1_proj[0][0]'] - alization) - - add_19 (Add) (None, 28, 28, 512) 0 ['conv3_1_1x1_increase/bn[0][0]', - 'conv3_1_1x1_proj/bn[0][0]'] - - activation_61 (Activation) (None, 28, 28, 512) 0 ['add_19[0][0]'] - - conv3_2_1x1_reduce (Conv2D) (None, 28, 28, 128) 65536 ['activation_61[0][0]'] - - conv3_2_1x1_reduce/bn (BatchNo (None, 28, 28, 128) 512 ['conv3_2_1x1_reduce[0][0]'] - rmalization) - - activation_62 (Activation) (None, 28, 28, 128) 0 ['conv3_2_1x1_reduce/bn[0][0]'] - - conv3_2_3x3 (Conv2D) (None, 28, 28, 128) 147456 ['activation_62[0][0]'] - - conv3_2_3x3/bn (BatchNormaliza (None, 28, 28, 128) 512 ['conv3_2_3x3[0][0]'] - tion) - - activation_63 (Activation) (None, 28, 28, 128) 0 ['conv3_2_3x3/bn[0][0]'] - - conv3_2_1x1_increase (Conv2D) (None, 28, 28, 512) 65536 ['activation_63[0][0]'] - - conv3_2_1x1_increase/bn (Batch (None, 28, 28, 512) 2048 ['conv3_2_1x1_increase[0][0]'] - Normalization) - - add_20 (Add) (None, 28, 28, 512) 0 ['conv3_2_1x1_increase/bn[0][0]', - 'activation_61[0][0]'] - - activation_64 (Activation) (None, 28, 28, 512) 0 ['add_20[0][0]'] - - conv3_3_1x1_reduce (Conv2D) (None, 28, 28, 128) 65536 ['activation_64[0][0]'] - - conv3_3_1x1_reduce/bn (BatchNo (None, 28, 28, 128) 512 ['conv3_3_1x1_reduce[0][0]'] - rmalization) - - activation_65 (Activation) (None, 28, 28, 128) 0 ['conv3_3_1x1_reduce/bn[0][0]'] - - conv3_3_3x3 (Conv2D) (None, 28, 28, 128) 147456 ['activation_65[0][0]'] - - conv3_3_3x3/bn (BatchNormaliza (None, 28, 28, 128) 512 ['conv3_3_3x3[0][0]'] - tion) - - activation_66 (Activation) (None, 28, 28, 128) 0 ['conv3_3_3x3/bn[0][0]'] - - conv3_3_1x1_increase (Conv2D) (None, 28, 28, 512) 65536 ['activation_66[0][0]'] - - conv3_3_1x1_increase/bn (Batch (None, 28, 28, 512) 2048 ['conv3_3_1x1_increase[0][0]'] - Normalization) - - add_21 (Add) (None, 28, 28, 512) 0 ['conv3_3_1x1_increase/bn[0][0]', - 'activation_64[0][0]'] - - activation_67 (Activation) (None, 28, 28, 512) 0 ['add_21[0][0]'] - - conv3_4_1x1_reduce (Conv2D) (None, 28, 28, 128) 65536 ['activation_67[0][0]'] - - conv3_4_1x1_reduce/bn (BatchNo (None, 28, 28, 128) 512 ['conv3_4_1x1_reduce[0][0]'] - rmalization) - - activation_68 (Activation) (None, 28, 28, 128) 0 ['conv3_4_1x1_reduce/bn[0][0]'] - - conv3_4_3x3 (Conv2D) (None, 28, 28, 128) 147456 ['activation_68[0][0]'] - - conv3_4_3x3/bn (BatchNormaliza (None, 28, 28, 128) 512 ['conv3_4_3x3[0][0]'] - tion) - - activation_69 (Activation) (None, 28, 28, 128) 0 ['conv3_4_3x3/bn[0][0]'] - - conv3_4_1x1_increase (Conv2D) (None, 28, 28, 512) 65536 ['activation_69[0][0]'] - - conv3_4_1x1_increase/bn (Batch (None, 28, 28, 512) 2048 ['conv3_4_1x1_increase[0][0]'] - Normalization) - - add_22 (Add) (None, 28, 28, 512) 0 ['conv3_4_1x1_increase/bn[0][0]', - 'activation_67[0][0]'] - - activation_70 (Activation) (None, 28, 28, 512) 0 ['add_22[0][0]'] - - conv4_1_1x1_reduce (Conv2D) (None, 14, 14, 256) 131072 ['activation_70[0][0]'] - - conv4_1_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_1_1x1_reduce[0][0]'] - rmalization) - - activation_71 (Activation) (None, 14, 14, 256) 0 ['conv4_1_1x1_reduce/bn[0][0]'] - - conv4_1_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_71[0][0]'] - - conv4_1_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_1_3x3[0][0]'] - tion) - - activation_72 (Activation) (None, 14, 14, 256) 0 ['conv4_1_3x3/bn[0][0]'] - - conv4_1_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_72[0][0]'] - ) - - conv4_1_1x1_proj (Conv2D) (None, 14, 14, 1024 524288 ['activation_70[0][0]'] - ) - - conv4_1_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_1_1x1_increase[0][0]'] - Normalization) ) - - conv4_1_1x1_proj/bn (BatchNorm (None, 14, 14, 1024 4096 ['conv4_1_1x1_proj[0][0]'] - alization) ) - - add_23 (Add) (None, 14, 14, 1024 0 ['conv4_1_1x1_increase/bn[0][0]', - ) 'conv4_1_1x1_proj/bn[0][0]'] - - activation_73 (Activation) (None, 14, 14, 1024 0 ['add_23[0][0]'] - ) - - conv4_2_1x1_reduce (Conv2D) (None, 14, 14, 256) 262144 ['activation_73[0][0]'] - - conv4_2_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_2_1x1_reduce[0][0]'] - rmalization) - - activation_74 (Activation) (None, 14, 14, 256) 0 ['conv4_2_1x1_reduce/bn[0][0]'] - - conv4_2_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_74[0][0]'] - - conv4_2_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_2_3x3[0][0]'] - tion) - - activation_75 (Activation) (None, 14, 14, 256) 0 ['conv4_2_3x3/bn[0][0]'] - - conv4_2_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_75[0][0]'] - ) - - conv4_2_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_2_1x1_increase[0][0]'] - Normalization) ) - - add_24 (Add) (None, 14, 14, 1024 0 ['conv4_2_1x1_increase/bn[0][0]', - ) 'activation_73[0][0]'] - - activation_76 (Activation) (None, 14, 14, 1024 0 ['add_24[0][0]'] - ) - - conv4_3_1x1_reduce (Conv2D) (None, 14, 14, 256) 262144 ['activation_76[0][0]'] - - conv4_3_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_3_1x1_reduce[0][0]'] - rmalization) - - activation_77 (Activation) (None, 14, 14, 256) 0 ['conv4_3_1x1_reduce/bn[0][0]'] - - conv4_3_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_77[0][0]'] - - conv4_3_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_3_3x3[0][0]'] - tion) - - activation_78 (Activation) (None, 14, 14, 256) 0 ['conv4_3_3x3/bn[0][0]'] - - conv4_3_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_78[0][0]'] - ) - - conv4_3_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_3_1x1_increase[0][0]'] - Normalization) ) - - add_25 (Add) (None, 14, 14, 1024 0 ['conv4_3_1x1_increase/bn[0][0]', - ) 'activation_76[0][0]'] - - activation_79 (Activation) (None, 14, 14, 1024 0 ['add_25[0][0]'] - ) - - conv4_4_1x1_reduce (Conv2D) (None, 14, 14, 256) 262144 ['activation_79[0][0]'] - - conv4_4_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_4_1x1_reduce[0][0]'] - rmalization) - - activation_80 (Activation) (None, 14, 14, 256) 0 ['conv4_4_1x1_reduce/bn[0][0]'] - - conv4_4_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_80[0][0]'] - - conv4_4_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_4_3x3[0][0]'] - tion) - - activation_81 (Activation) (None, 14, 14, 256) 0 ['conv4_4_3x3/bn[0][0]'] - - conv4_4_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_81[0][0]'] - ) - - conv4_4_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_4_1x1_increase[0][0]'] - Normalization) ) - - add_26 (Add) (None, 14, 14, 1024 0 ['conv4_4_1x1_increase/bn[0][0]', - ) 'activation_79[0][0]'] - - activation_82 (Activation) (None, 14, 14, 1024 0 ['add_26[0][0]'] - ) - - conv4_5_1x1_reduce (Conv2D) (None, 14, 14, 256) 262144 ['activation_82[0][0]'] - - conv4_5_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_5_1x1_reduce[0][0]'] - rmalization) - - activation_83 (Activation) (None, 14, 14, 256) 0 ['conv4_5_1x1_reduce/bn[0][0]'] - - conv4_5_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_83[0][0]'] - - conv4_5_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_5_3x3[0][0]'] - tion) - - activation_84 (Activation) (None, 14, 14, 256) 0 ['conv4_5_3x3/bn[0][0]'] - - conv4_5_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_84[0][0]'] - ) - - conv4_5_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_5_1x1_increase[0][0]'] - Normalization) ) - - add_27 (Add) (None, 14, 14, 1024 0 ['conv4_5_1x1_increase/bn[0][0]', - ) 'activation_82[0][0]'] - - activation_85 (Activation) (None, 14, 14, 1024 0 ['add_27[0][0]'] - ) - - conv4_6_1x1_reduce (Conv2D) (None, 14, 14, 256) 262144 ['activation_85[0][0]'] - - conv4_6_1x1_reduce/bn (BatchNo (None, 14, 14, 256) 1024 ['conv4_6_1x1_reduce[0][0]'] - rmalization) - - activation_86 (Activation) (None, 14, 14, 256) 0 ['conv4_6_1x1_reduce/bn[0][0]'] - - conv4_6_3x3 (Conv2D) (None, 14, 14, 256) 589824 ['activation_86[0][0]'] - - conv4_6_3x3/bn (BatchNormaliza (None, 14, 14, 256) 1024 ['conv4_6_3x3[0][0]'] - tion) - - activation_87 (Activation) (None, 14, 14, 256) 0 ['conv4_6_3x3/bn[0][0]'] - - conv4_6_1x1_increase (Conv2D) (None, 14, 14, 1024 262144 ['activation_87[0][0]'] - ) - - conv4_6_1x1_increase/bn (Batch (None, 14, 14, 1024 4096 ['conv4_6_1x1_increase[0][0]'] - Normalization) ) - - add_28 (Add) (None, 14, 14, 1024 0 ['conv4_6_1x1_increase/bn[0][0]', - ) 'activation_85[0][0]'] - - activation_88 (Activation) (None, 14, 14, 1024 0 ['add_28[0][0]'] - ) - - conv5_1_1x1_reduce (Conv2D) (None, 7, 7, 512) 524288 ['activation_88[0][0]'] - - conv5_1_1x1_reduce/bn (BatchNo (None, 7, 7, 512) 2048 ['conv5_1_1x1_reduce[0][0]'] - rmalization) - - activation_89 (Activation) (None, 7, 7, 512) 0 ['conv5_1_1x1_reduce/bn[0][0]'] - - conv5_1_3x3 (Conv2D) (None, 7, 7, 512) 2359296 ['activation_89[0][0]'] - - conv5_1_3x3/bn (BatchNormaliza (None, 7, 7, 512) 2048 ['conv5_1_3x3[0][0]'] - tion) - - activation_90 (Activation) (None, 7, 7, 512) 0 ['conv5_1_3x3/bn[0][0]'] - - conv5_1_1x1_increase (Conv2D) (None, 7, 7, 2048) 1048576 ['activation_90[0][0]'] - - conv5_1_1x1_proj (Conv2D) (None, 7, 7, 2048) 2097152 ['activation_88[0][0]'] - - conv5_1_1x1_increase/bn (Batch (None, 7, 7, 2048) 8192 ['conv5_1_1x1_increase[0][0]'] - Normalization) - - conv5_1_1x1_proj/bn (BatchNorm (None, 7, 7, 2048) 8192 ['conv5_1_1x1_proj[0][0]'] - alization) - - add_29 (Add) (None, 7, 7, 2048) 0 ['conv5_1_1x1_increase/bn[0][0]', - 'conv5_1_1x1_proj/bn[0][0]'] - - activation_91 (Activation) (None, 7, 7, 2048) 0 ['add_29[0][0]'] - - conv5_2_1x1_reduce (Conv2D) (None, 7, 7, 512) 1048576 ['activation_91[0][0]'] - - conv5_2_1x1_reduce/bn (BatchNo (None, 7, 7, 512) 2048 ['conv5_2_1x1_reduce[0][0]'] - rmalization) - - activation_92 (Activation) (None, 7, 7, 512) 0 ['conv5_2_1x1_reduce/bn[0][0]'] - - conv5_2_3x3 (Conv2D) (None, 7, 7, 512) 2359296 ['activation_92[0][0]'] - - conv5_2_3x3/bn (BatchNormaliza (None, 7, 7, 512) 2048 ['conv5_2_3x3[0][0]'] - tion) - - activation_93 (Activation) (None, 7, 7, 512) 0 ['conv5_2_3x3/bn[0][0]'] - - conv5_2_1x1_increase (Conv2D) (None, 7, 7, 2048) 1048576 ['activation_93[0][0]'] - - conv5_2_1x1_increase/bn (Batch (None, 7, 7, 2048) 8192 ['conv5_2_1x1_increase[0][0]'] - Normalization) - - add_30 (Add) (None, 7, 7, 2048) 0 ['conv5_2_1x1_increase/bn[0][0]', - 'activation_91[0][0]'] - - activation_94 (Activation) (None, 7, 7, 2048) 0 ['add_30[0][0]'] - - conv5_3_1x1_reduce (Conv2D) (None, 7, 7, 512) 1048576 ['activation_94[0][0]'] - - conv5_3_1x1_reduce/bn (BatchNo (None, 7, 7, 512) 2048 ['conv5_3_1x1_reduce[0][0]'] - rmalization) - - activation_95 (Activation) (None, 7, 7, 512) 0 ['conv5_3_1x1_reduce/bn[0][0]'] - - conv5_3_3x3 (Conv2D) (None, 7, 7, 512) 2359296 ['activation_95[0][0]'] - - conv5_3_3x3/bn (BatchNormaliza (None, 7, 7, 512) 2048 ['conv5_3_3x3[0][0]'] - tion) - - activation_96 (Activation) (None, 7, 7, 512) 0 ['conv5_3_3x3/bn[0][0]'] - - conv5_3_1x1_increase (Conv2D) (None, 7, 7, 2048) 1048576 ['activation_96[0][0]'] - - conv5_3_1x1_increase/bn (Batch (None, 7, 7, 2048) 8192 ['conv5_3_1x1_increase[0][0]'] - Normalization) - - add_31 (Add) (None, 7, 7, 2048) 0 ['conv5_3_1x1_increase/bn[0][0]', - 'activation_94[0][0]'] - - activation_97 (Activation) (None, 7, 7, 2048) 0 ['add_31[0][0]'] - - avg_pool (AveragePooling2D) (None, 1, 1, 2048) 0 ['activation_97[0][0]'] - - global_average_pooling2d_1 (Gl (None, 2048) 0 ['avg_pool[0][0]'] - obalAveragePooling2D) - - gaussian_noise_1 (GaussianNois (None, 2048) 0 ['global_average_pooling2d_1[0][0 - e) ]'] - - dense_x (Dense) (None, 512) 1049088 ['gaussian_noise_1[0][0]'] - - dropout_1 (Dropout) (None, 512) 0 ['dense_x[0][0]'] - - dense_1 (Dense) (None, 7) 3591 ['dropout_1[0][0]'] - - ================================================================================================== - Total params: 24,613,831 - Trainable params: 24,560,711 - Non-trainable params: 53,120 - __________________________________________________________________________________________________ - --- Время выполнения: 2.222 сек. --- + [2024-10-09 12:22:54] Формирование нейросетевой архитектуры для получения нейросетевых признаков (видео модальность) ... + + ResNet( + (conv_layer_s2_same): Conv2dSame(3, 64, kernel_size=(7, 7), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + (max_pool): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False) + (layer1): Sequential( + (0): Bottleneck( + (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(64, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer2): Sequential( + (0): Bottleneck( + (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (3): Bottleneck( + (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(128, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer3): Sequential( + (0): Bottleneck( + (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (3): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (4): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (5): Bottleneck( + (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(256, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(1024, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (layer4): Sequential( + (0): Bottleneck( + (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (i_downsample): Sequential( + (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False) + (1): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + ) + (relu): ReLU() + ) + (1): Bottleneck( + (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + (2): Bottleneck( + (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm1): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=same, bias=False) + (batch_norm2): BatchNorm2d(512, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) + (batch_norm3): BatchNorm2d(2048, eps=0.001, momentum=0.99, affine=True, track_running_stats=True) + (relu): ReLU() + ) + ) + (avgpool): AdaptiveAvgPool2d(output_size=(1, 1)) + (fc1): Linear(in_features=2048, out_features=512, bias=True) + (relu1): ReLU() + (fc2): Linear(in_features=512, out_features=7, bias=True) + ) + --- Время выполнения: 0.222 сек. --- True @@ -2534,28 +2452,14 @@ def load_video_model_nn( :execution-count: 1 :linenos: - [2022-10-27 14:46:11] Формирование нейросетевой архитектуры для получения оценок по нейросетевым признакам (видео модальность) ... - - Model: "model" - _________________________________________________________________ - Layer (type) Output Shape Param # - ================================================================= - input_1 (InputLayer) [(None, 10, 512)] 0 - - lstm (LSTM) (None, 1024) 6295552 + [2024-10-09 12:49:36] Формирование нейросетевой архитектуры для получения оценок по нейросетевым признакам (видео модальность) ... - dropout (Dropout) (None, 1024) 0 - - dense (Dense) (None, 5) 5125 - - activation (Activation) (None, 5) 0 - - ================================================================= - Total params: 6,300,677 - Trainable params: 6,300,677 - Non-trainable params: 0 - _________________________________________________________________ - --- Время выполнения: 2.018 сек. --- + video_model_nn( + (lstm1): LSTM(512, 1024, batch_first=True) + (dropout1): Dropout(p=0.2, inplace=False) + (fc): Linear(in_features=1024, out_features=5, bias=True) + ) + --- Время выполнения: 0.052 сек. --- True @@ -2657,24 +2561,13 @@ def load_video_models_b5( :execution-count: 1 :linenos: - [2022-11-04 15:29:26] Формирование нейросетевых архитектур моделей для получения результатов оценки персональных качеств (видео модальность) ... + [2024-10-09 13:12:19] Формирование нейросетевых архитектур моделей для получения результатов оценки персональных качеств (видео модальность) ... - Model: "model_4" - _________________________________________________________________ - Layer (type) Output Shape Param # - ================================================================= - input_1 (InputLayer) [(None, 32)] 0 - - dense_1 (Dense) (None, 1) 33 - - activ_1 (Activation) (None, 1) 0 - - ================================================================= - Total params: 33 - Trainable params: 33 - Non-trainable params: 0 - _________________________________________________________________ - --- Время выполнения: 0.116 сек. --- + video_model_b5( + (fc): Linear(in_features=32, out_features=1, bias=True) + (sigmoid): Sigmoid() + ) + --- Время выполнения: 0.009 сек. --- True @@ -2771,6 +2664,7 @@ def load_video_model_weights_hc( video = Video() video.load_video_model_hc( + lang = 'en', show_summary = False, out = True, runtime = True, run = True ) @@ -2793,7 +2687,7 @@ def load_video_model_weights_hc( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['hc']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['hc']['googledisk'] video.load_video_model_weights_hc( url = url, @@ -2807,11 +2701,11 @@ def load_video_model_weights_hc( :execution-count: 2 :linenos: - [2022-10-27 13:08:04] Загрузка весов нейросетевой модели для получения оценок по экспертным признакам (видео модальность) ... + [2024-10-09 13:06:56] Загрузка весов нейросетевой модели для получения оценок по экспертным признакам (видео модальность) ... - [2022-10-27 13:08:05] Загрузка файла "weights_2022-08-27_18-53-35.h5" (100.0%) ... + [2024-10-09 13:06:58] Загрузка файла "weights_2022-08-27_18-53-35.pth" 100.0% ... - --- Время выполнения: 0.493 сек. --- + --- Время выполнения: 2.49 сек. --- True @@ -2829,7 +2723,7 @@ def load_video_model_weights_hc( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['hc']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['hc']['googledisk'] video.load_video_model_weights_hc( url = url, @@ -2843,13 +2737,13 @@ def load_video_model_weights_hc( :execution-count: 3 :linenos: - [2022-10-27 13:09:54] Загрузка весов нейросетевой модели для получения оценок по экспертным признакам (видео модальность) ... + [2024-10-09 13:07:56] Загрузка весов нейросетевой модели для получения оценок по экспертным признакам (видео модальность) ... - [2022-10-27 13:09:54] Загрузка файла "weights_2022-08-27_18-53-35.h5" (100.0%) ... + [2024-10-09 13:07:59] Загрузка файла "weights_2022-08-27_18-53-35.pth" 100.0% ... - [2022-10-27 13:09:54] Ой! Что-то пошло не так ... нейросетевая архитектура модели для получения оценок по экспертным признакам не сформирована (видео модальность) ... + [2024-10-09 13:07:59] Что-то пошло не так ... нейросетевая архитектура модели для получения оценок по экспертным признакам не сформирована (видео модальность) ... - --- Время выполнения: 0.424 сек. --- + --- Время выполнения: 2.381 сек. --- False """ @@ -2924,7 +2818,7 @@ def load_video_model_weights_deep_fe( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['fe']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['fe']['googledisk'] video.load_video_model_weights_deep_fe( url = url, @@ -2938,11 +2832,11 @@ def load_video_model_weights_deep_fe( :execution-count: 2 :linenos: - [2022-11-01 12:42:51] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... + [2024-10-09 13:00:35] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... - [2022-11-01 12:43:06] Загрузка файла "weights_2022-11-01_12-27-07.h5" (100.0%) ... + [2024-10-09 13:00:41] Загрузка файла "weights_2022-11-01_12-27-07.pth" 100.0% ... - --- Время выполнения: 14.781 сек. --- + --- Время выполнения: 5.557 сек. --- True @@ -2960,7 +2854,7 @@ def load_video_model_weights_deep_fe( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['fe']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['fe']['googledisk'] video.load_video_model_weights_deep_fe( url = url, @@ -2974,13 +2868,13 @@ def load_video_model_weights_deep_fe( :execution-count: 3 :linenos: - [2022-11-01 12:44:14] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... + [2024-10-09 13:01:48] Загрузка весов нейросетевой модели для получения нейросетевых признаков (видео модальность) ... - [2022-11-01 12:44:28] Загрузка файла "weights_2022-11-01_12-27-07.h5" (100.0%) ... + [2024-10-09 13:01:53] Загрузка файла "weights_2022-11-01_12-27-07.pth" 100.0% ... - [2022-11-01 12:44:28] Ой! Что-то пошло не так ... нейросетевая архитектура модели для получения нейросетевых признаков не сформирована (видео модальность) ... + [2024-10-09 13:01:53] Что-то пошло не так ... нейросетевая архитектура модели для получения нейросетевых признаков не сформирована (видео модальность) ... - --- Время выполнения: 13.926 сек. --- + --- Время выполнения: 4.712 сек. --- False """ @@ -3060,7 +2954,7 @@ def load_video_model_weights_nn( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['nn']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['nn']['googledisk'] video.load_video_model_weights_nn( url = url, @@ -3074,11 +2968,11 @@ def load_video_model_weights_nn( :execution-count: 2 :linenos: - [2022-10-27 15:19:08] Загрузка весов нейросетевой модели для получения оценок по нейросетевым признакам (видео модальность) ... + [2024-10-09 13:09:03] Загрузка весов нейросетевой модели для получения оценок по нейросетевым признакам (видео модальность) ... - [2022-10-27 15:19:11] Загрузка файла "weights_2022-03-22_16-31-48.h5" (100.0%) ... + [2024-10-09 13:09:08] Загрузка файла "weights_2022-03-22_16-31-48.pth" 100.0% ... - --- Время выполнения: 3.423 сек. --- + --- Время выполнения: 5.798 сек. --- True @@ -3096,7 +2990,7 @@ def load_video_model_weights_nn( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url = video.weights_for_big5_['video']['nn']['sberdisk'] + url = video.weights_for_big5_['video']['fi']['nn']['googledisk'] video.load_video_model_weights_nn( url = url, @@ -3110,13 +3004,13 @@ def load_video_model_weights_nn( :execution-count: 3 :linenos: - [2022-10-27 15:19:40] Загрузка весов нейросетевой модели для получения оценок по нейросетевым признакам (видео модальность) ... + [2024-10-09 13:09:56] Загрузка весов нейросетевой модели для получения оценок по нейросетевым признакам (видео модальность) ... - [2022-10-27 15:19:43] Загрузка файла "weights_2022-03-22_16-31-48.h5" (100.0%) ... + [2024-10-09 13:10:02] Загрузка файла "weights_2022-03-22_16-31-48.pth" 100.0% ... - [2022-10-27 15:19:43] Ой! Что-то пошло не так ... нейросетевая архитектура модели для получения оценок по нейросетевым признакам не сформирована (видео модальность) ... + [2024-10-09 13:10:02] Что-то пошло не так ... нейросетевая архитектура модели для получения оценок по нейросетевым признакам не сформирована (видео модальность) ... - --- Время выполнения: 3.469 сек. --- + --- Время выполнения: 5.9 сек. --- False """ @@ -3206,11 +3100,11 @@ def load_video_models_weights_b5( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url_openness = video.weights_for_big5_['video']['b5']['openness']['sberdisk'] - url_conscientiousness = video.weights_for_big5_['video']['b5']['conscientiousness']['sberdisk'] - url_extraversion = video.weights_for_big5_['video']['b5']['extraversion']['sberdisk'] - url_agreeableness = video.weights_for_big5_['video']['b5']['agreeableness']['sberdisk'] - url_non_neuroticism = video.weights_for_big5_['video']['b5']['non_neuroticism']['sberdisk'] + url_openness = video.weights_for_big5_['video']['fi']['b5']['openness']['googledisk'] + url_conscientiousness = video.weights_for_big5_['video']['fi']['b5']['conscientiousness']['googledisk'] + url_extraversion = video.weights_for_big5_['video']['fi']['b5']['extraversion']['googledisk'] + url_agreeableness = video.weights_for_big5_['video']['fi']['b5']['agreeableness']['googledisk'] + url_non_neuroticism = video.weights_for_big5_['video']['fi']['b5']['non_neuroticism']['googledisk'] video.load_video_models_weights_b5( url_openness = url_openness, @@ -3228,19 +3122,19 @@ def load_video_models_weights_b5( :execution-count: 2 :linenos: - [2022-11-04 18:58:59] Загрузка весов нейросетевых моделей для получения результатов оценки персональных качеств (видео модальность) ... + [2024-10-09 13:14:48] Загрузка весов нейросетевых моделей для получения результатов оценки персональных качеств (видео модальность) ... - [2022-11-04 18:59:00] Загрузка файла "weights_2022-06-15_16-46-30.h5" (100.0%) ... Открытость опыту + [2024-10-09 13:14:50] Загрузка файла "weights_2022-06-15_16-46-30.pth" 100.0% ... Открытость опыту - [2022-11-04 18:59:00] Загрузка файла "weights_2022-06-15_16-48-50.h5" (100.0%) ... Добросовестность + [2024-10-09 13:14:52] Загрузка файла "weights_2022-06-15_16-48-50.pth" 100.0% ... Добросовестность - [2022-11-04 18:59:00] Загрузка файла "weights_2022-06-15_16-54-06.h5" (100.0%) ... Экстраверсия + [2024-10-09 13:14:55] Загрузка файла "weights_2022-06-15_16-54-06.pth" 100.0% ... Экстраверсия - [2022-11-04 18:59:01] Загрузка файла "weights_2022-06-15_17-02-03.h5" (100.0%) ... Доброжелательность + [2024-10-09 13:14:57] Загрузка файла "weights_2022-06-15_17-02-03.pth" 100.0% ... Доброжелательность - [2022-11-04 18:59:01] Загрузка файла "weights_2022-06-15_17-06-15.h5" (100.0%) ... Эмоциональная стабильность + [2024-10-09 13:15:00] Загрузка файла "weights_2022-06-15_17-06-15.pth" 100.0% ... Эмоциональная стабильность - --- Время выполнения: 1.827 сек. --- + --- Время выполнения: 11.832 сек. --- True @@ -3258,11 +3152,11 @@ def load_video_models_weights_b5( video.path_to_save_ = './models' video.chunk_size_ = 2000000 - url_openness = video.weights_for_big5_['video']['b5']['openness']['sberdisk'] - url_conscientiousness = video.weights_for_big5_['video']['b5']['conscientiousness']['sberdisk'] - url_extraversion = video.weights_for_big5_['video']['b5']['extraversion']['sberdisk'] - url_agreeableness = video.weights_for_big5_['video']['b5']['agreeableness']['sberdisk'] - url_non_neuroticism = video.weights_for_big5_['video']['b5']['non_neuroticism']['sberdisk'] + url_openness = video.weights_for_big5_['video']['fi']['b5']['openness']['googledisk'] + url_conscientiousness = video.weights_for_big5_['video']['fi']['b5']['conscientiousness']['googledisk'] + url_extraversion = video.weights_for_big5_['video']['fi']['b5']['extraversion']['googledisk'] + url_agreeableness = video.weights_for_big5_['video']['fi']['b5']['agreeableness']['googledisk'] + url_non_neuroticism = video.weights_for_big5_['video']['fi']['b5']['non_neuroticism']['googledisk'] video.load_video_models_weights_b5( url_openness = url_openness, @@ -3280,54 +3174,54 @@ def load_video_models_weights_b5( :execution-count: 3 :linenos: - [2022-11-04 19:02:32] Загрузка весов нейросетевых моделей для получения результатов оценки персональных качеств (видео модальность) ... + [2024-10-09 13:16:08] Загрузка весов нейросетевых моделей для получения результатов оценки персональных качеств (видео модальность) ... - [2022-11-04 19:02:32] Загрузка файла "weights_2022-06-15_16-46-30.h5" (100.0%) ... + [2024-10-09 13:16:10] Загрузка файла "weights_2022-06-15_16-46-30.pth" 100.0% ... - [2022-11-04 19:02:32] Ой! Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Открытость опыту + [2024-10-09 13:16:10] Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Открытость опыту - Файл: /Users/dl/GitHub/oceanai/oceanai/modules/lab/video.py - Линия: 2833 + Файл: /Users/dl/@DmitryRyumin/Python/envs/OCEANAI/lib/python3.9/site-packages/oceanai/modules/lab/video.py + Линия: 3144 Метод: load_video_models_weights_b5 Тип ошибки: AttributeError - [2022-11-04 19:02:32] Загрузка файла "weights_2022-06-15_16-48-50.h5" (100.0%) ... + [2024-10-09 13:16:13] Загрузка файла "weights_2022-06-15_16-48-50.pth" 100.0% ... - [2022-11-04 19:02:32] Ой! Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Добросовестность + [2024-10-09 13:16:13] Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Добросовестность - Файл: /Users/dl/GitHub/oceanai/oceanai/modules/lab/video.py - Линия: 2833 + Файл: /Users/dl/@DmitryRyumin/Python/envs/OCEANAI/lib/python3.9/site-packages/oceanai/modules/lab/video.py + Линия: 3144 Метод: load_video_models_weights_b5 Тип ошибки: AttributeError - [2022-11-04 19:02:33] Загрузка файла "weights_2022-06-15_16-54-06.h5" (100.0%) ... + [2024-10-09 13:16:16] Загрузка файла "weights_2022-06-15_16-54-06.pth" 100.0% ... - [2022-11-04 19:02:33] Ой! Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Экстраверсия + [2024-10-09 13:16:16] Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Экстраверсия - Файл: /Users/dl/GitHub/oceanai/oceanai/modules/lab/video.py - Линия: 2833 + Файл: /Users/dl/@DmitryRyumin/Python/envs/OCEANAI/lib/python3.9/site-packages/oceanai/modules/lab/video.py + Линия: 3144 Метод: load_video_models_weights_b5 Тип ошибки: AttributeError - [2022-11-04 19:02:33] Загрузка файла "weights_2022-06-15_17-02-03.h5" (100.0%) ... + [2024-10-09 13:16:19] Загрузка файла "weights_2022-06-15_17-02-03.pth" 100.0% ... - [2022-11-04 19:02:33] Ой! Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Доброжелательность + [2024-10-09 13:16:19] Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Доброжелательность - Файл: /Users/dl/GitHub/oceanai/oceanai/modules/lab/video.py - Линия: 2833 + Файл: /Users/dl/@DmitryRyumin/Python/envs/OCEANAI/lib/python3.9/site-packages/oceanai/modules/lab/video.py + Линия: 3144 Метод: load_video_models_weights_b5 Тип ошибки: AttributeError - [2022-11-04 19:02:34] Загрузка файла "weights_2022-06-15_17-06-15.h5" (100.0%) ... + [2024-10-09 13:16:21] Загрузка файла "weights_2022-06-15_17-06-15.pth" 100.0% ... - [2022-11-04 19:02:34] Ой! Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Эмоциональная стабильность + [2024-10-09 13:16:21] Что-то пошло не так ... не удалось загрузить веса нейросетевой модели ... Эмоциональная стабильность - Файл: /Users/dl/GitHub/oceanai/oceanai/modules/lab/video.py - Линия: 2833 + Файл: /Users/dl/@DmitryRyumin/Python/envs/OCEANAI/lib/python3.9/site-packages/oceanai/modules/lab/video.py + Линия: 3144 Метод: load_video_models_weights_b5 Тип ошибки: AttributeError - --- Время выполнения: 1.831 сек. --- + --- Время выполнения: 13.055 сек. --- False """