https://wandb.ai/gustiwinata/text-recognizer/overview
- LineCNNSimple: Reading multiple characters at once
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0 --max_overlap=0 --model_class=line_cnn_simple.LineCNNSimple --window_width=28 --window_stride=28
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0 --max_overlap=0 --model_class=line_cnn_simple.LineCNNSimple --window_width=28 --window_stride=20
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0.25 --max_overlap=0.25 --model_class=line_cnn_simple.LineCNNSimple --window_width=28 --window_stride=20 --limit_output_length
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0 --max_overlap=0.33 --model_class=line_cnn_simple.LineCNNSimple --window_width=28 --window_stride=20 --limit_output_length
- Loss Function
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0.25 --max_overlap=0.25 --model_class=line_cnn.LineCNN --window_width=28 --window_stride=20 --loss=ctc
- CNN LSTM
python training/run_experiment.py --max_epochs=10 --gpus=1 --num_workers=4 --data_class=emnist_lines.EMNISTLines --min_overlap=0 --max_overlap=0.33 --model_class=line_cnn_lstm.LineCNNLSTM --window_width=28 --window_stride=18 --loss=ctc