Release/rc0.1 #15
Draft
Release/rc0.1 #15
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
added 9 commits
November 10, 2020 18:51
…mall refactoring of project files
…g' into release/rc0.1
…-refactoring' into release/rc0.1
mrTahion
requested changes
Dec 17, 2020
Member
mrTahion
left a comment
mrTahion
left a comment
There was a problem hiding this comment.
Вроде местами не хватает запуска linter.sh
| DATA: | ||
| FEAT_EMB: | ||
| PATH: "data/CUB/resnet101/" | ||
| PATH: "datasets/CUB/resnet101/" |
Member
There was a problem hiding this comment.
Все ещё сомневаюсь насчет этого переименования
Comment on lines
+4
to
+6
| ZeroShotEval is provided with 2 build-in datasest: | ||
| - AWA2 | ||
| - CUB No newline at end of file |
Member
There was a problem hiding this comment.
Что мешает использовать, например, SUN?
Comment on lines
11
to
13
| Please note, that this script is designed for CUB, SUN, AWA1, AWA2 | ||
| datasets only, and moreover these datasets must be downloaded from: | ||
| https://www.dropbox.com/sh/btoc495ytfbnbat/AAAaurkoKnnk0uV-swgF-gdSa?dl=0 |
Member
There was a problem hiding this comment.
Suggested change
| Please note, that this script is designed for CUB, SUN, AWA1, AWA2 | |
| datasets only, and moreover these datasets must be downloaded from: | |
| https://www.dropbox.com/sh/btoc495ytfbnbat/AAAaurkoKnnk0uV-swgF-gdSa?dl=0 | |
| Please note, that this script is designed for CUB, SUN, AWA1, AWA2 | |
| .. _datasets: | |
| https://www.dropbox.com/sh/btoc495ytfbnbat/AAAaurkoKnnk0uV-swgF-gdSa?dl=0 |
Comment on lines
105
to
107
| # NOTE: there are two more fields in mat files that are not in use now, but can be used | ||
| # train_loc = mat_data['train_loc'].squeeze() - 1 #--> train_feature = TRAIN SEEN | ||
| # val_unseen_loc = mat_data['val_loc'].squeeze() - 1 #--> test_unseen_feature = TEST UNSEEN |
tools/single_experiment.py
Outdated
| from zeroshoteval.utils.parser import load_config, parse_args | ||
| from zeroshoteval.zeroshotnets.build import build_zsl | ||
|
|
||
| os.chdir("../") |
Comment on lines
+1
to
+3
| import torch | ||
| from fvcore.common.config import CfgNode | ||
| from torch.utils.data.dataloader import DataLoader |
Member
There was a problem hiding this comment.
Suggested change
| import torch | |
| from fvcore.common.config import CfgNode | |
| from torch.utils.data.dataloader import DataLoader | |
| from fvcore.common.config import CfgNode | |
| from torch.utils.data.dataloader import DataLoader |
Comment on lines
+78
to
+83
| loader = torch.utils.data.DataLoader( | ||
| dataset, | ||
| batch_size=cfg.ZSL.BATCH_SIZE, | ||
| num_workers=cfg.DATA_LOADER.NUM_WORKERS, | ||
| pin_memory=cfg.DATA_LOADER.PIN_MEMORY, | ||
| ) |
Member
There was a problem hiding this comment.
Suggested change
| loader = torch.utils.data.DataLoader( | |
| dataset, | |
| batch_size=cfg.ZSL.BATCH_SIZE, | |
| num_workers=cfg.DATA_LOADER.NUM_WORKERS, | |
| pin_memory=cfg.DATA_LOADER.PIN_MEMORY, | |
| ) | |
| loader = DataLoader( | |
| dataset, | |
| batch_size=cfg.ZSL.BATCH_SIZE, | |
| num_workers=cfg.DATA_LOADER.NUM_WORKERS, | |
| pin_memory=cfg.DATA_LOADER.PIN_MEMORY, | |
| ) |
| Args: | ||
| idx: Index of sample to return. | ||
| Returns: Tuple2 with dictionary of `modalities names: item from each modality` and `sample label` |
Member
There was a problem hiding this comment.
Suggested change
| Returns: Tuple2 with dictionary of `modalities names: item from each modality` and `sample label` | |
| Returns: | |
| sample_data ... | |
| sample_label ... |
| return len(self.labels) | ||
|
|
||
| def __getitem__(self, idx): | ||
|
|
| self._data_loader_iter = iter(data_loader) | ||
|
|
||
| def evaluate(self) -> None: | ||
| logger = logging.getLogger(__name__) |
termit209
reviewed
Dec 18, 2020
| return args | ||
|
|
||
|
|
||
| def read_data(images_mat_file: str, |
There was a problem hiding this comment.
может более понятнее будет название read_data_mat или что-то похожее?
termit209
reviewed
Dec 18, 2020
tools/single_experiment.py
Outdated
| # -------------------------------------------------------------------------- | ||
|
|
||
| # TODO: extract gen_synthetic_data from CADA-VAE and move it to inference section | ||
| pass |
There was a problem hiding this comment.
В этом ПР пока отказываемся от идеи вынести емб экстрактор из тренировки?
termit209
reviewed
Dec 20, 2020
| from zeroshoteval.utils.exceptions import ExtractorTypeError | ||
|
|
||
|
|
||
| class ImageEmbeddingExtractor(EmbeddingExtractor): |
There was a problem hiding this comment.
Это окончательная версия экстрактора или будет переделываться?
termit209
reviewed
Dec 20, 2020
|
|
||
|
|
||
| class DecoderTemplate(nn.Module): | ||
| """Decoder part for CADA-VAE model. |
There was a problem hiding this comment.
Может в докстринге написать что это просто декодер, так как и лизгане по сути такойже декодер?
termit209
reviewed
Dec 20, 2020
| self.labels = None | ||
|
|
||
| assert split in ["trainval", "test", "test_unseen"] | ||
| assert mod in ["CLS_ATTR", "IMG"] |
There was a problem hiding this comment.
разве везде не изменили CLS_ATTR на CLSATTR?
mrTahion
reviewed
Dec 22, 2020
zeroshoteval/config/defaults.py
Outdated
Comment on lines
108
to
109
| _C.ZSL.EPOCH = 100 | ||
| # _C.ZSL.EPOCH = 100 | ||
| _C.ZSL.EPOCH = 10 |
Member
There was a problem hiding this comment.
Для чего это изменение? В любом случае не стоит, кмк, оставлять закоменченные строки.
Comment on lines
+31
to
+32
| train_loader = construct_loader(cfg, split="train") | ||
| test_loader = construct_loader(cfg, split="test") |
Member
There was a problem hiding this comment.
Кмк train_loader и test_loader в инференсе совсем неочевидные названия.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Massive changes in folders structure, main classes, call logic, input/output data formats.
Therefore, please make a full review of the entire project.
In case of any questions please contact me @nikolaev-n