Remove most args from deprecation zone

use_mask to be replaced with a new tokenizer implementation

README.md

@@ -362,7 +362,7 @@ cd src/visualize
 
 ```bash
 python3 main.py 
---existing_model /path/to/existing/model 
+--model /path/to/existing/model 
 --sample_image /path/to/sample/img
 ```
 

configs/default.json

@@ -19,17 +19,6 @@
             "greedy": false,
             "wbs_smoothing": 0.1
         },
-        "depr": {
-            "channels": 3,
-            "config_file_output": null,
-            "do_inference": false,
-            "do_train": true,
-            "height": 64,
-            "no_auto": false,
-            "output_charlist": null,
-            "thaw": false,
-            "use_mask": true
-        },
         "general": {
             "batch_size": 4,
             "charlist": null,

configs/training.json

@@ -17,9 +17,6 @@
             "beam_width": 10,
             "greedy": false
         },
-        "depr": {
-            "channels": 1
-        },
         "general": {
             "batch_size": 64,
             "gpu": "0",

src/main.py

@@ -95,7 +95,7 @@ def main():
             decay_rate=config["decay_rate"],
             decay_steps=config["decay_steps"],
             train_batches=data_manager.get_train_batches(),
-            do_train=config["do_train"],
+            do_train=config["train_list"],
             warmup_ratio=config["warmup_ratio"],
             epochs=config["epochs"],
             decay_per_epoch=config["decay_per_epoch"],

src/model/management.py

@@ -100,14 +100,12 @@ def customize_model(model: tf.keras.Model,
         model = replace_final_layer(model, len(charlist), model.name,
                                     use_mask=config["use_mask"])
 
-    # Freeze or thaw layers if specified
-    if any([config["thaw"], config["freeze_conv_layers"],
-            config["freeze_recurrent_layers"], config["freeze_dense_layers"]]):
+    # Freeze layers if specified
+    if any([config["freeze_conv_layers"],
+            config["freeze_recurrent_layers"],
+            config["freeze_dense_layers"]]):
         for layer in model.layers:
-            if config["thaw"]:
-                layer.trainable = True
-                logging.info("Thawing layer: %s", layer.name)
-            elif config["freeze_conv_layers"] and \
+            if config["freeze_conv_layers"] and \
                 (layer.name.lower().startswith("conv") or
                  layer.name.lower().startswith("residual")):
                 logging.info("Freezing layer: %s", layer.name)

src/setup/arg_parser.py

@@ -238,96 +238,20 @@ def get_arg_parser():
     # Deprecation zone
     depr_args = parser.add_argument_group(
         'Deprecation zone', 'These arguments will be removed in the future')
-    depr_args.add_argument('--do_train', help='enable the training. '
-                           'Use this flag if you want to train.',
-                           action='store_true')
-    depr_args.add_argument('--do_inference', help='inference',
-                           action='store_true')
     depr_args.add_argument('--use_mask', help='whether or not to mask certain '
                            'parts of the data. Defaults to true when '
                            'batch_size > 1', action='store_true')
-    depr_args.add_argument('--no_auto', action='store_true',
-                           help='No Auto disabled automatic "fixing" of '
-                           'certain parameters')
-    depr_args.add_argument('--height', metavar='height', type=int, default=64,
-                           help='rescale everything to this height before '
-                           'training, default 64')
-    depr_args.add_argument('--channels', metavar='channels', type=int,
-                           default=3, help='number of channels to use. 1 for '
-                           'grey-scale/binary images, three for color images, '
-                           '4 for png\'s with transparency')
-    depr_args.add_argument('--output_charlist', metavar='output_charlist',
-                           type=str, default=None, help="Path to save the "
-                           "character list used during training/inference. "
-                           "If not specified, the charlist is saved to"
-                           "'output/charlist.txt'.")
-    depr_args.add_argument('--config_file_output',
-                           metavar='config_file_output', type=str,
-                           default=None, help="Path to save the "
-                           "configuration file. If not specified, the "
-                           "configuration is set to 'output/config.json'.")
-    depr_args.add_argument('--thaw', action='store_true',
-                           help="Unfreeze convolutional layers in an "
-                           "existing model for further training.")
-    depr_args.add_argument('--existing_model', metavar='existing_model',
-                           type=str, default=None, help="Path to an existing "
-                           "model to continue training, validation, testing, "
-                           "or inferencing. Used as a starting point.")
 
     return parser
 
 
-def fix_args(args):
-    if not args.no_auto and args.train_list:
-        logging.warning('--do_train implied by providing a train_list')
-        args.__dict__['do_train'] = True
-    if not args.no_auto and args.batch_size > 1:
-        logging.warning('--batch_size > 1, setting use_mask=True')
-        args.__dict__['use_mask'] = True
-    if not args.no_auto and args.inference_list:
-        logging.warning('--do_inference implied by providing a inference_list')
-        args.__dict__['do_inference'] = True
-    if not args.no_auto and args.existing_model:
-        args.__dict__['model'] = args.existing_model
-
-
 def arg_future_warning(args):
     logger = logging.getLogger(__name__)
 
     # May 2024
-    if args.do_train:
-        logger.warning("Argument will lose support in May 2024: --do_train. "
-                       "Training will be enabled by providing a train_list. ")
-    if args.do_inference:
-        logger.warning("Argument will lose support in May 2024: "
-                       "--do_inference. Inference will be enabled by "
-                       "providing an inference_list. ")
     if args.use_mask:
         logger.warning("Argument will lose support in May 2024: --use_mask. "
                        "Masking will be enabled by default.")
-    if args.no_auto:
-        logger.warning("Argument will lose support in May 2024: --no_auto.")
-    if args.height:
-        logger.warning("Argument will lose support in May 2024: --height. "
-                       "Height will be inferred from the VGSL spec.")
-    if args.channels:
-        logger.warning("Argument will lose support in May 2024: --channels. "
-                       "Channels will be inferred from the VGSL spec.")
-    if args.output_charlist:
-        logger.warning("Argument will lose support in May 2024: "
-                       "--output_charlist. The charlist will be saved to "
-                       "output/charlist.txt by default.")
-    if args.config_file_output:
-        logger.warning("Argument will lose support in May 2024: "
-                       "--config_file_output. The configuration will be saved "
-                       "to output/config.json by default.")
-    if args.thaw:
-        logging.warning("Argument will lose support in May 2024: --thaw. "
-                        "Models are saved with all layers thawed by default.")
-    if args.existing_model:
-        logger.warning("Argument will lose support in May 2024: "
-                       "--existing_model. The --model argument can be used "
-                       "to load or create a model instead.")
 
 
 def check_required_args(args, explicit):
@@ -377,7 +301,6 @@ def get_args():
 
     # TODO: remove after deprecation period
     arg_future_warning(args)
-    fix_args(args)
     check_required_args(args, explicit)
 
     if args.steps_per_epoch:

src/setup/config.py

@@ -129,8 +129,7 @@ def save(self, output_file: str = None) -> None:
         """
 
         if not output_file:
-            output_file = self.args.config_file_output or \
-                f"{self.args.output}/config.json"
+            output_file = f"{self.args.output}/config.json"
         try:
             with open(output_file, "w", encoding="utf-8") as file:
                 json.dump(self.config, file, indent=4, sort_keys=True)
@@ -222,16 +221,7 @@ def organize_args(self, args: argparse.Namespace) -> dict:
                 "deterministic": args.deterministic
             },
             "depr": {
-                "do_train": args.do_train,
-                "do_inference": args.do_inference,
                 "use_mask": args.use_mask,
-                "no_auto": args.no_auto,
-                "height": args.height,
-                "channels": args.channels,
-                "output_charlist": args.output_charlist,
-                "config_file_output": args.config_file_output,
-                "thaw": args.thaw,
-                "existing_model": args.existing_model,
             }
         }
 

src/utils/text.py

@@ -38,7 +38,9 @@ class Tokenizer:
         Decodes the tokenized sequences back into text.
     """
 
-    def __init__(self, chars: list, use_mask: bool = False,
+    def __init__(self,
+                 chars: list,
+                 use_mask: bool = False,
                  num_oov_indices: int = 1):
         """
         Initializes the Tokenizer with a given character list and mask option.