add diacritizer

poetry_diacritizer/__init__.py

@@ -0,0 +1 @@
+from poetry_diacritizer import predict

poetry_diacritizer/config/ashaar.yml

@@ -0,0 +1,52 @@
+session_name: base
+
+data_directory: "data"
+data_type: "ashaar_proc"
+log_directory: "log_dir_ashaar"
+load_training_data: true
+load_test_data: false
+load_validation_data: true
+n_training_examples: null # null load all training examples, good for fast loading
+n_test_examples: null  # null load all test examples
+n_validation_examples: null # null load all validation examples
+test_file_name: "test.csv"
+is_data_preprocessed: false # The data file is organized as (original text | text | diacritics)
+data_separator: '|' # Required if the data already processed
+diacritics_separator: '*'  # Required if the data already processed
+text_encoder: ArabicEncoderWithStartSymbol
+text_cleaner: valid_arabic_cleaners # a white list that uses only Arabic letters, punctuations, and a space
+max_len: 600 # sentences larger than this size will not be used
+max_sen_len: null
+
+max_steps: 10000
+learning_rate: 0.001
+batch_size: 32
+adam_beta1: 0.9
+adam_beta2: 0.999
+use_decay: true
+weight_decay: 0.0
+embedding_dim: 256
+use_prenet: false
+prenet_sizes: [512, 256]
+cbhg_projections: [128, 256]
+cbhg_filters: 16
+cbhg_gru_units: 256
+post_cbhg_layers_units: [256, 256]
+post_cbhg_use_batch_norm: true
+
+use_mixed_precision: false
+optimizer_type: Adam
+device: cuda
+
+# LOGGING
+evaluate_frequency: 50000000
+max_eval_batches: 100
+evaluate_with_error_rates_frequency: 1000
+n_predicted_text_tensorboard: 10 # To be written to the tensorboard
+model_save_frequency: 1000
+train_plotting_frequency: 50000000 # No plotting for this model
+n_steps_avg_losses: [100, 500, 1_000, 5_000] # command line display of average loss values for the last n steps
+error_rates_n_batches: 10000 # if calculating error rate is slow, then you can specify the number of batches to be calculated
+
+test_model_path: null # load the last saved model
+train_resume_model_path: null # load last saved model

poetry_diacritizer/config/baseline.yml

@@ -0,0 +1,47 @@
+session_name: base
+
+data_directory: "data"
+data_type: "CA_MSA"
+log_directory: "log_dir"
+load_training_data: true
+load_test_data: false
+load_validation_data: true
+n_training_examples: null # null load all training examples, good for fast loading
+n_test_examples: null  # null load all test examples
+n_validation_examples: null # null load all validation examples
+test_file_name: "test.csv"
+is_data_preprocessed: false # The data file is organized as (original text | text | diacritics)
+data_separator: '|' # Required if the data already processed
+diacritics_separator: '*'  # Required if the data already processed
+text_encoder: ArabicEncoderWithStartSymbol
+text_cleaner: valid_arabic_cleaners # a white list that uses only Arabic letters, punctuations, and a space
+max_len: 600 # sentences larger than this size will not be used
+
+
+max_steps: 2_000_000
+learning_rate: 0.001
+batch_size: 64
+adam_beta1: 0.9
+adam_beta2: 0.999
+use_decay: true
+weight_decay: 0.0
+embedding_dim: 512
+n_layers: 3
+layers_units: [256, 256, 256]
+use_mixed_precision: false
+optimizer_type: Adam
+use_batch_norm: False
+device: cuda
+max_sen_len: 256
+
+# LOGGING
+evaluate_frequency: 5000
+evaluate_with_error_rates_frequency: 5000
+n_predicted_text_tensorboard: 10 # To be written to the tensorboard
+model_save_frequency: 5000
+train_plotting_frequency: 50000000 # No plotting for this model
+n_steps_avg_losses: [100, 500, 1_000, 5_000] # command line display of average loss values for the last n steps
+error_rates_n_batches: 10000 # if calculating error rate is slow, then you can specify the number of batches to be calculated
+
+test_model_path: null # load the last saved model
+train_resume_model_path: null # load last saved model

poetry_diacritizer/config/cbhg.yml

@@ -0,0 +1,52 @@
+session_name: base
+
+data_directory: "data"
+data_type: "CA_MSA"
+log_directory: "log_dir_cbhg"
+load_training_data: true
+load_test_data: false
+load_validation_data: true
+n_training_examples: null # null load all training examples, good for fast loading
+n_test_examples: null  # null load all test examples
+n_validation_examples: null # null load all validation examples
+test_file_name: "test.csv"
+is_data_preprocessed: false # The data file is organized as (original text | text | diacritics)
+data_separator: '|' # Required if the data already processed
+diacritics_separator: '*'  # Required if the data already processed
+text_encoder: ArabicEncoderWithStartSymbol
+text_cleaner: valid_arabic_cleaners # a white list that uses only Arabic letters, punctuations, and a space
+max_len: 600 # sentences larger than this size will not be used
+max_sen_len: null
+
+max_steps: 5000
+learning_rate: 0.001
+batch_size: 32
+adam_beta1: 0.9
+adam_beta2: 0.999
+use_decay: true
+weight_decay: 0.0
+embedding_dim: 256
+use_prenet: false
+prenet_sizes: [512, 256]
+cbhg_projections: [128, 256]
+cbhg_filters: 16
+cbhg_gru_units: 256
+post_cbhg_layers_units: [256, 256]
+post_cbhg_use_batch_norm: true
+
+use_mixed_precision: false
+optimizer_type: Adam
+device: cuda
+
+# LOGGING
+evaluate_frequency: 50000000
+max_eval_batches: 100
+evaluate_with_error_rates_frequency: 1000
+n_predicted_text_tensorboard: 10 # To be written to the tensorboard
+model_save_frequency: 5000
+train_plotting_frequency: 50000000 # No plotting for this model
+n_steps_avg_losses: [100, 500, 1_000, 5_000] # command line display of average loss values for the last n steps
+error_rates_n_batches: 10000 # if calculating error rate is slow, then you can specify the number of batches to be calculated
+
+test_model_path: null # load the last saved model
+train_resume_model_path: null # load last saved model

poetry_diacritizer/config/cbhg2.yml

@@ -0,0 +1,51 @@
+session_name: base
+
+data_directory: "ashaar"
+data_type: "CA_MSA"
+log_directory: "/content/drive/MyDrive/Research/Barmajan/Diacritization/log_ashaar_dir"
+load_training_data: true
+load_test_data: false
+load_validation_data: true
+n_training_examples: null # null load all training examples, good for fast loading
+n_test_examples: null  # null load all test examples
+n_validation_examples: null # null load all validation examples
+test_file_name: "test.csv"
+is_data_preprocessed: false # The data file is organized as (original text | text | diacritics)
+data_separator: '|' # Required if the data already processed
+diacritics_separator: '*'  # Required if the data already processed
+text_encoder: ArabicEncoderWithStartSymbol
+text_cleaner: valid_arabic_cleaners # a white list that uses only Arabic letters, punctuations, and a space
+max_len: 600 # sentences larger than this size will not be used
+
+
+max_steps: 25_000
+learning_rate: 0.001
+batch_size: 32
+adam_beta1: 0.9
+adam_beta2: 0.999
+use_decay: true
+weight_decay: 0.0
+embedding_dim: 256
+use_prenet: false
+prenet_sizes: [512, 256]
+cbhg_projections: [128, 256]
+cbhg_filters: 16
+cbhg_gru_units: 256
+post_cbhg_layers_units: [256, 256]
+post_cbhg_use_batch_norm: true
+
+use_mixed_precision: false
+optimizer_type: Adam
+device: cuda
+
+# LOGGING
+evaluate_frequency: 1000
+evaluate_with_error_rates_frequency: 1000
+n_predicted_text_tensorboard: 10 # To be written to the tensorboard
+model_save_frequency: 1000
+train_plotting_frequency: 50000000 # No plotting for this model
+n_steps_avg_losses: [100, 500, 1_000, 5_000] # command line display of average loss values for the last n steps
+error_rates_n_batches: 10000 # if calculating error rate is slow, then you can specify the number of batches to be calculated
+
+test_model_path: null # load the last saved model
+train_resume_model_path: "/content/drive/MyDrive/Research/Barmajan/Diacritization/log_cleaned_dir/CA_MSA.base.cbhg/models/20000-snapshot.pt" # load last saved model

poetry_diacritizer/config/gpt-0.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_0
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 0
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-1.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_1
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 1
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-2.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_2
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 2
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-3.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_3
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 3
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-4.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_4
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 4
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-5.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_5
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 5
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-6.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_6
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 6
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-7.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_7
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 7
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-8.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_8
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 8
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-9.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_9
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 9
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: true
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-0-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_0_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 0
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-0-test.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_0_test
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 0
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-0.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_0
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 0
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-1-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_1_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 1
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-1.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_1
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 1
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-2-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_2_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 2
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-2.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_2
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 2
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-3-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_3_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 3
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-3.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_3
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 3
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-4-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_4_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 4
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-4.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_4
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 4
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-5-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_5_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 5
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-5-test.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: logs/log_dir_cls_5_test
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 5
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-5.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_5
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 5
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-6-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_6_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 6
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-6.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_6
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 6
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-7-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_7_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 7
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-7.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_7
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 7
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-8-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_8_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 8
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-8.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_8
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 8
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-9-tash-proc.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: tash_proc
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: log_dir_cls_9_tash_proc
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 9
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0

poetry_diacritizer/config/gpt-cls-9-test.yml

@@ -0,0 +1,46 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+base_model_path: ashaar-from-scratch-with-spaces-no-tatweel-epochs-75
+batch_size: 64
+data_directory: data
+data_separator: '|'
+data_type: CA_MSA
+device: cuda
+diacritics_separator: '*'
+error_rates_n_batches: 10000
+evaluate_frequency: 50000000
+evaluate_with_error_rates_frequency: 1000
+freeze: true
+is_data_preprocessed: false
+learning_rate: 0.001
+load_test_data: false
+load_training_data: true
+load_validation_data: true
+log_directory: logs/log_dir_cls_9_test
+max_eval_batches: -1
+max_len: 600
+max_sen_len: 256
+max_steps: 5000
+model_save_frequency: 5000
+n_layer: 9
+n_predicted_text_tensorboard: 10
+n_steps_avg_losses:
+- 100
+- 500
+- 1000
+- 5000
+n_test_examples: null
+n_training_examples: null
+n_validation_examples: null
+optimizer_type: Adam
+session_name: base
+test_file_name: test.csv
+test_model_path: null
+text_cleaner: valid_arabic_cleaners
+text_encoder: ArabicEncoderWithStartSymbol
+train_plotting_frequency: 50000000
+train_resume_model_path: null
+use_decay: true
+use_lstm: false
+use_mixed_precision: false
+weight_decay: 0.0