From 6df941bdf5cad80db38d851dcb23a08a9dc55617 Mon Sep 17 00:00:00 2001
From: Gustaf Rydholm <gustaf.rydholm@gmail.com>
Date: Sat, 11 Jun 2022 23:09:22 +0200
Subject: Fix emnist mapping
---
text_recognizer/data/base_data_module.py | 4 ++--
text_recognizer/data/emnist_lines.py | 2 +-
text_recognizer/data/iam_extended_paragraphs.py | 4 ++--
text_recognizer/data/iam_lines.py | 2 +-
text_recognizer/data/iam_paragraphs.py | 25 +++++++++++++++++++++---
text_recognizer/data/iam_synthetic_paragraphs.py | 25 +++++++++++++++++++++---
text_recognizer/data/mappings/emnist.py | 13 ++++++------
text_recognizer/models/base.py | 4 ++--
text_recognizer/models/conformer.py | 2 +-
text_recognizer/models/transformer.py | 2 +-
10 files changed, 61 insertions(+), 22 deletions(-)
diff --git a/text_recognizer/data/base_data_module.py b/text_recognizer/data/base_data_module.py
index 28ba775..69581a3 100644
--- a/text_recognizer/data/base_data_module.py
+++ b/text_recognizer/data/base_data_module.py
@@ -1,6 +1,6 @@
"""Base lightning DataModule class."""
from pathlib import Path
-from typing import Callable, Dict, Optional, Tuple, Type, TypeVar
+from typing import Callable, Dict, Optional, Tuple, TypeVar
from pytorch_lightning import LightningDataModule
from torch.utils.data import DataLoader
@@ -24,7 +24,7 @@ class BaseDataModule(LightningDataModule):
def __init__(
self,
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
transform: Optional[Callable] = None,
test_transform: Optional[Callable] = None,
target_transform: Optional[Callable] = None,
diff --git a/text_recognizer/data/emnist_lines.py b/text_recognizer/data/emnist_lines.py
index ba1b61c..2107d74 100644
--- a/text_recognizer/data/emnist_lines.py
+++ b/text_recognizer/data/emnist_lines.py
@@ -38,7 +38,7 @@ class EMNISTLines(BaseDataModule):
def __init__(
self,
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
transform: Optional[Callable] = None,
test_transform: Optional[Callable] = None,
target_transform: Optional[Callable] = None,
diff --git a/text_recognizer/data/iam_extended_paragraphs.py b/text_recognizer/data/iam_extended_paragraphs.py
index 61bf6a3..90df5f8 100644
--- a/text_recognizer/data/iam_extended_paragraphs.py
+++ b/text_recognizer/data/iam_extended_paragraphs.py
@@ -1,5 +1,5 @@
"""IAM original and sythetic dataset class."""
-from typing import Callable, Optional, Type
+from typing import Callable, Optional
from torch.utils.data import ConcatDataset
from text_recognizer.data.base_data_module import BaseDataModule, load_and_print_info
@@ -14,7 +14,7 @@ class IAMExtendedParagraphs(BaseDataModule):
def __init__(
self,
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
transform: Optional[Callable] = None,
test_transform: Optional[Callable] = None,
target_transform: Optional[Callable] = None,
diff --git a/text_recognizer/data/iam_lines.py b/text_recognizer/data/iam_lines.py
index cf50b60..4899a48 100644
--- a/text_recognizer/data/iam_lines.py
+++ b/text_recognizer/data/iam_lines.py
@@ -39,7 +39,7 @@ class IAMLines(BaseDataModule):
def __init__(
self,
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
transform: Optional[Callable] = None,
test_transform: Optional[Callable] = None,
target_transform: Optional[Callable] = None,
diff --git a/text_recognizer/data/iam_paragraphs.py b/text_recognizer/data/iam_paragraphs.py
index 9c75129..3bf28ff 100644
--- a/text_recognizer/data/iam_paragraphs.py
+++ b/text_recognizer/data/iam_paragraphs.py
@@ -1,7 +1,7 @@
"""IAM Paragraphs Dataset class."""
import json
from pathlib import Path
-from typing import Dict, List, Optional, Sequence, Tuple
+from typing import Callable, Dict, List, Optional, Sequence, Tuple
from loguru import logger as log
import numpy as np
@@ -35,8 +35,27 @@ MAX_WORD_PIECE_LENGTH = 451
class IAMParagraphs(BaseDataModule):
"""IAM handwriting database paragraphs."""
- def __init__(self) -> None:
- super().__init__()
+ def __init__(
+ self,
+ mapping: EmnistMapping,
+ transform: Optional[Callable] = None,
+ test_transform: Optional[Callable] = None,
+ target_transform: Optional[Callable] = None,
+ train_fraction: float = 0.8,
+ batch_size: int = 16,
+ num_workers: int = 0,
+ pin_memory: bool = True,
+ ) -> None:
+ super().__init__(
+ mapping,
+ transform,
+ test_transform,
+ target_transform,
+ train_fraction,
+ batch_size,
+ num_workers,
+ pin_memory,
+ )
self.dims = (1, IMAGE_HEIGHT, IMAGE_WIDTH)
self.output_dims = (MAX_LABEL_LENGTH, 1)
diff --git a/text_recognizer/data/iam_synthetic_paragraphs.py b/text_recognizer/data/iam_synthetic_paragraphs.py
index 1dc517d..d51e010 100644
--- a/text_recognizer/data/iam_synthetic_paragraphs.py
+++ b/text_recognizer/data/iam_synthetic_paragraphs.py
@@ -1,6 +1,6 @@
"""IAM Synthetic Paragraphs Dataset class."""
import random
-from typing import Any, List, Sequence, Tuple
+from typing import Any, Callable, List, Optional, Sequence, Tuple
from loguru import logger as log
import numpy as np
@@ -36,8 +36,27 @@ PROCESSED_DATA_DIRNAME = (
class IAMSyntheticParagraphs(IAMParagraphs):
"""IAM Handwriting database of synthetic paragraphs."""
- def __init__(self) -> None:
- super().__init__()
+ def __init__(
+ self,
+ mapping: EmnistMapping,
+ transform: Optional[Callable] = None,
+ test_transform: Optional[Callable] = None,
+ target_transform: Optional[Callable] = None,
+ train_fraction: float = 0.8,
+ batch_size: int = 16,
+ num_workers: int = 0,
+ pin_memory: bool = True,
+ ) -> None:
+ super().__init__(
+ mapping,
+ transform,
+ test_transform,
+ target_transform,
+ train_fraction,
+ batch_size,
+ num_workers,
+ pin_memory,
+ )
def prepare_data(self) -> None:
"""Prepare IAM lines to be used to generate paragraphs."""
diff --git a/text_recognizer/data/mappings/emnist.py b/text_recognizer/data/mappings/emnist.py
index 606d200..03465a1 100644
--- a/text_recognizer/data/mappings/emnist.py
+++ b/text_recognizer/data/mappings/emnist.py
@@ -14,20 +14,21 @@ class EmnistMapping:
def __init__(
self,
- input_size: List[int],
- mapping: List[str],
- inverse_mapping: Dict[str, int],
extra_symbols: Optional[Sequence[str]] = None,
lower: bool = True,
) -> None:
- self.input_size = input_size
- self.mapping = mapping
- self.inverse_mapping = inverse_mapping
self.extra_symbols = set(extra_symbols) if extra_symbols is not None else None
self.mapping, self.inverse_mapping, self.input_size = self._load_mapping()
if lower:
self._to_lower()
+ def __len__(self) -> int:
+ return len(self.mapping)
+
+ @property
+ def num_classes(self) -> int:
+ return self.__len__()
+
def _load_mapping(self) -> Tuple[List, Dict[str, int], List[int]]:
"""Return the EMNIST mapping."""
with ESSENTIALS_FILENAME.open() as f:
diff --git a/text_recognizer/models/base.py b/text_recognizer/models/base.py
index 886394d..26cc18c 100644
--- a/text_recognizer/models/base.py
+++ b/text_recognizer/models/base.py
@@ -10,7 +10,7 @@ from torch import nn
from torch import Tensor
from torchmetrics import Accuracy
-from text_recognizer.data.mappings.base import EmnistMapping
+from text_recognizer.data.mappings import EmnistMapping
class LitBase(LightningModule):
@@ -22,7 +22,7 @@ class LitBase(LightningModule):
loss_fn: Type[nn.Module],
optimizer_configs: DictConfig,
lr_scheduler_configs: Optional[DictConfig],
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
) -> None:
super().__init__()
diff --git a/text_recognizer/models/conformer.py b/text_recognizer/models/conformer.py
index 655ebf6..41a9e4d 100644
--- a/text_recognizer/models/conformer.py
+++ b/text_recognizer/models/conformer.py
@@ -21,7 +21,7 @@ class LitConformer(LitBase):
loss_fn: Type[nn.Module],
optimizer_configs: DictConfig,
lr_scheduler_configs: Optional[DictConfig],
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
max_output_len: int = 451,
start_token: str = "<s>",
end_token: str = "<e>",
diff --git a/text_recognizer/models/transformer.py b/text_recognizer/models/transformer.py
index 1ffff60..b511947 100644
--- a/text_recognizer/models/transformer.py
+++ b/text_recognizer/models/transformer.py
@@ -19,7 +19,7 @@ class LitTransformer(LitBase):
loss_fn: Type[nn.Module],
optimizer_configs: DictConfig,
lr_scheduler_configs: Optional[DictConfig],
- mapping: Type[EmnistMapping],
+ mapping: EmnistMapping,
max_output_len: int = 451,
start_token: str = "<s>",
end_token: str = "<e>",
--
cgit v1.2.3-70-g09d2