feat: Add new theme for docs

Benchmark/distort_images.py

@@ -43,7 +43,8 @@ def distort_image(input_image_path: str, output_image_path: str) -> None:
 def main():
     """This script takes an input directory with images of chemical structure depictions, applies a random
     rotation (-5° or +5°) as well as shearing (angle drawn from [-0.1, 0.1]) to every input image. These
-    distortions are supposed to imitate the image modifications described by Clevert et al (Img2Mol (preprint))"""
+    distortions are supposed to imitate the image modifications described by Clevert et al (Img2Mol (preprint))
+    """
     input_dir = os.path.abspath(sys.argv[1])
     output_dir = os.path.abspath(sys.argv[2])
     if not os.path.exists(output_dir):

DECIMER/DECIMER_EfficinetNetV2_Transfomer_Trainer.py

@@ -220,7 +220,6 @@ def prepare_for_training(lr_config, encoder_config, transformer_config, verbose=
     """
 
     with strategy.scope():
-
         loss_object = tf.keras.losses.SparseCategoricalCrossentropy(
             from_logits=True, reduction=tf.keras.losses.Reduction.NONE
         )

DECIMER/Efficient_Net_encoder.py

@@ -56,7 +56,6 @@ def __init__(
         pretrained_weights=None,
         scale_factor=0,
     ):
-
         super(Encoder, self).__init__()
 
         self.image_embedding_dim = image_embedding_dim

DECIMER/Repack_model.py

@@ -2,10 +2,10 @@
 import tensorflow as tf
 
 import pickle
-import Transformer_decoder
-import Efficient_Net_encoder
-import config
-import utils
+import DECIMER.Efficient_Net_encoder as Efficient_Net_encoder
+import DECIMER.Transformer_decoder as Transformer_decoder
+import DECIMER.config as config
+import DECIMER.utils as utils
 
 print(tf.__version__)
 
@@ -16,8 +16,8 @@
     tf.config.experimental.set_memory_growth(gpu, True)
 
 # load assets
-tokenizer = pickle.load(open("tokenizer_TPU_Stereo.pkl", "rb"))
-max_length = pickle.load(open("max_length_TPU_Stereo.pkl", "rb"))
+tokenizer = pickle.load(open("tokenizer.pkl", "rb"))
+max_length = pickle.load(open("max_length.pkl", "rb"))
 
 # Image parameters
 IMG_EMB_DIM = (16, 16, 232)

DECIMER/Transformer_decoder.py

@@ -105,7 +105,6 @@ def create_look_ahead_mask(size):
 
 
 def create_mask(inp, tar):
-
     # Used in the 1st attention block in the decoder.
     # It is used to pad and mask future tokens in the input received by
     # the decoder.
@@ -209,7 +208,6 @@ def point_wise_feed_forward_network(d_model, dff):
 
 class TransformerEncoderLayer(tf.keras.layers.Layer):
     def __init__(self, d_model, num_heads, dff, dropout_rate=0.1):
-
         super(TransformerEncoderLayer, self).__init__()
 
         self.mha = tf.keras.layers.MultiHeadAttention(
@@ -225,7 +223,6 @@ def __init__(self, d_model, num_heads, dff, dropout_rate=0.1):
         self.dropout2 = tf.keras.layers.Dropout(dropout_rate)
 
     def call(self, x, training, mask=None):
-
         # returns (batch_size, input_seq_len, d_model)
         attn_output, _ = self.mha(x, x, x, mask, return_attention_scores=True)
 
@@ -250,7 +247,6 @@ def call(self, x, training, mask=None):
 
 class TransformerDecoderLayer(tf.keras.layers.Layer):
     def __init__(self, d_model, num_heads, dff, dropout_rate=0.1):
-
         super(TransformerDecoderLayer, self).__init__()
 
         # WE COULD USE A CUSTOM DEFINED MHA MODEL BUT WE WILL USE TFA INSTEAD
@@ -276,7 +272,6 @@ def __init__(self, d_model, num_heads, dff, dropout_rate=0.1):
 
     # enc_output.shape == (batch_size, input_seq_len, d_model)
     def call(self, x, enc_output, training, look_ahead_mask=None, padding_mask=None):
-
         attn1, attn_weights_block1 = self.mha1(x, x, x, look_ahead_mask)
         attn1 = self.dropout1(attn1, training=training)
 
@@ -370,7 +365,6 @@ def __init__(
         self.dropout = tf.keras.layers.Dropout(dropout_rate)
 
     def call(self, x, enc_output, training, look_ahead_mask=None, padding_mask=None):
-
         seq_len = tf.shape(x)[1]
         attention_weights = {}
 
@@ -405,7 +399,6 @@ def __init__(
         pe_target,
         dropout_rate=0.1,
     ):
-
         super(Transformer, self).__init__()
 
         self.t_encoder = TransformerEncoder(

DECIMER/Utils/Create_TFrecord_From_pickles.py

@@ -43,7 +43,6 @@ def _bytes_feature(value):
         return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
 
     def get_train_tfrecord(num_shards, train_captions, img_name_vector, file_index):
-
         print("Total number of TFrecords: ", num_shards, flush=True)
 
         for i in range(num_shards):
@@ -57,7 +56,6 @@ def get_train_tfrecord(num_shards, train_captions, img_name_vector, file_index):
             writer = tf.io.TFRecordWriter(tfrecord_name)
             counter = 0
             for j in range(len(sub_split_img_id)):
-
                 # print(decoded_image.shape)
                 caption_ = sub_split_cap_train[j]
                 # image_id_ = sub_split_img_id[counter]

DECIMER/Utils/Create_image_tokenizer.py

@@ -15,7 +15,6 @@
 
 
 def main():
-
     Smiles_Path = "Path_to_file"
     # Images_Path = "Train_Images/"
 
@@ -61,7 +60,6 @@ def main():
 
 
 def data_loader(Smiles_Path):
-
     # read the Captions file
     with open(Smiles_Path, "r") as txt_file:
         smiles = txt_file.read()

DECIMER/config.py

@@ -363,21 +363,25 @@ def download_trained_weights(model_url: str, model_path: str, verbose=1):
     """This function downloads the trained models and tokenizers to a default location.
     After downloading the zipped file the function unzips the file automatically.
     If the model exists on the default location this function will not work.
+
     Args:
         model_url (str): trained model url for downloading.
         model_path (str): model default path to download.
+
     Returns:
-        downloaded model.
+        path (str): downloaded model.
     """
     # Download trained models
     if verbose > 0:
         print("Downloading trained model to " + str(model_path))
-        model_path = pystow.ensure("DECIMER-V2", url=model_url)
-        print(model_path)
+    model_path = pystow.ensure("DECIMER-V2", url=model_url)
     if verbose > 0:
+        print(model_path)
         print("... done downloading trained model!")
-        with zipfile.ZipFile(model_path.as_posix(), "r") as zip_ref:
-            zip_ref.extractall(model_path.parent.as_posix())
-        # Delete zipfile after downloading
-        if Path(model_path).exists():
-            Path(model_path).unlink()
+
+    with zipfile.ZipFile(model_path.as_posix(), "r") as zip_ref:
+        zip_ref.extractall(model_path.parent.as_posix())
+
+    # Delete zipfile after downloading
+    if Path(model_path).exists():
+        Path(model_path).unlink()

DECIMER/efficientnetv2/utils.py

@@ -543,7 +543,7 @@ def get_ckpt_var_map(
     if not var_map or len(var_map) < 5:
         raise ValueError(f"var_map={var_map} is almost empty, please check logs.")
 
-    for (k, v) in var_map.items():
+    for k, v in var_map.items():
         logging.log_first_n(logging.INFO, f"Init {v.op.name} from ckpt var {k}", 10)
 
     return var_map

docs/conf.py

@@ -17,12 +17,12 @@
 
 # -- Project information -----------------------------------------------------
 
-project = "DECIMER"
-copyright = "2022, Kohulan"
-author = "Kohulan"
+project = "DECIMER-Image_Transformer"
+copyright = "2023, Kohulan Rajan"
+author = "Kohulan Rajan"
 
 # The full version, including alpha/beta/rc tags
-release = "2.0.0"
+release = "2.3.0"
 
 
 # -- General configuration ---------------------------------------------------
@@ -34,7 +34,6 @@
     "sphinx.ext.autodoc",
     "sphinx_autodoc_typehints",
     "sphinx.ext.autosummary",
-    "sphinx_rtd_theme",
     "nbsphinx",
     "sphinx.ext.mathjax",
     "sphinx.ext.githubpages",
@@ -67,13 +66,35 @@
 # The theme to use for HTML and HTML Help pages.  See the documentation for
 # a list of builtin themes.
 #
-html_theme = "sphinx_rtd_theme"
+html_theme = "sphinx_material"
 
 # Theme options are theme-specific and customize the look and feel of a theme
 # further.  For a list of options available for each theme, see the
 # documentation.
 #
-# html_theme_options = {}
+# Material theme options (see theme.conf for more information)
+html_theme_options = {
+    # Set the name of the project to appear in the navigation.
+    "nav_title": "DECIMER-Image_Transformer",
+    # Specify a base_url used to generate sitemap.xml. If not
+    # specified, then no sitemap will be built.
+    "base_url": "https://github.com/Kohulan/DECIMER-Image_Transformer",
+    # Set the color and the accent color
+    "color_primary": "blue",
+    "color_accent": "light-blue",
+    "html_minify": True,
+    "css_minify": True,
+    "logo_icon": "&#xe869",
+    # Set the repo location to get a badge with stats
+    "repo_url": "https://github.com/Kohulan/DECIMER-Image_Transformer/",
+    "repo_name": "DECIMER-Image_Transformer",
+    # Visible levels of the global TOC; -1 means unlimited
+    "globaltoc_depth": 3,
+    # If False, expand all TOC entries
+    "globaltoc_collapse": False,
+    # If True, show hidden TOC entries
+    "globaltoc_includehidden": False,
+}
 
 # Add any paths that contain custom static files (such as style sheets) here,
 # relative to this directory. They are copied after the builtin static files,

docs/requirements.txt

@@ -1,6 +1,8 @@
 nbsphinx
 sphinx-autodoc-typehints
 sphinx_rtd_theme
+furo
+sphinx_material
 numpy>=1.19
 tensorflow>=2.6.3
 opencv-python