wrong font recognized

[kelogs]

With a fresh docker image:

# rm -rf data
# python3 create_data.py -c 1 -f data
# rm top_model.h5
# python3 train.py -e 1
2024-08-21 15:01:42.604305: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/lib/python3.7/site-packages/cv2/../../lib64:
2024-08-21 15:01:42.604370: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2024-08-21 15:01:47.251728: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/lib/python3.7/site-packages/cv2/../../lib64:
2024-08-21 15:01:47.251788: W tensorflow/stream_executor/cuda/cuda_driver.cc:269] failed call to cuInit: UNKNOWN ERROR (303)
2024-08-21 15:01:47.251828: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (dd5673d4dbfa): /proc/driver/nvidia/version does not exist
2024-08-21 15:01:47.252011: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2024-08-21 15:01:47.456512: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:01:47.706134: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:01:47.843645: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:01:48.388955: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:01:57.717679: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
4/4 [==============================] - ETA: 0s - loss: 0.1631 - accuracy: 0.1880  
Epoch 1: val_loss improved from inf to 0.15841, saving model to top_model.h5
4/4 [==============================] - 160s 40s/step - loss: 0.1631 - accuracy: 0.1880 - val_loss: 0.1584 - val_accuracy: 0.3333
Test loss: 0.15841002762317657
Test accuracy: 0.3333333432674408
# python3 eval.py -d test_data/Times.jpg -m top_model.h5
2024-08-21 15:04:55.851911: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2024-08-21 15:04:55.851952: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2024-08-21 15:04:57.153665: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory
2024-08-21 15:04:57.153724: W tensorflow/stream_executor/cuda/cuda_driver.cc:269] failed call to cuInit: UNKNOWN ERROR (303)
2024-08-21 15:04:57.153743: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (dd5673d4dbfa): /proc/driver/nvidia/version does not exist
2024-08-21 15:04:57.153933: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2024-08-21 15:04:57.395157: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:04:57.755244: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:04:57.908225: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:04:59.734032: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
2024-08-21 15:04:59.927867: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 943718400 exceeds 10% of free system memory.
1/1 [==============================] - 0s 290ms/step
[[0.20492785 0.20434484 0.20602909 0.20206164 0.18263653]]
[2]
test_data/Times.jpg: SFCompactRounded-Bold

I have also had AvenirNext for a match...