Can you provide a real example command for MODEL_WEIGHTS_PATH please ?

i have downloaded the crnn_synth90k.zip

And i tried this:

python ./tools/test_shadownet.py --image_path ./data/test_images/test_01.jpg --weights_path ./model/crnn_synth90k/checkpoint    --char_dict_path ./data/char_dict/char_dict_en.json    --ord_map_dict_path ./data/char_dict/ord_map_en.json

And i got this :

2019-10-21 06:20:26.476474: W tensorflow/core/framework/op_kernel.cc:1651] OP_REQUIRES failed at save_restore_tensor.cc:175 : Data loss: Unable to open table file ./model/crnn_synth90k/checkpoint: Data loss: file is too short to be an sstable: perhaps your file is in a different file format and you need to use a different restore operator?
Traceback (most recent call last):
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1365, in _do_call
    return fn(*args)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1350, in _run_fn
    target_list, run_metadata)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1443, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.DataLossError: Unable to open table file ./model/crnn_synth90k/checkpoint: Data loss: file is too short to be an sstable: perhaps your file is in a different file format and you need to use a different restore operator?
         [[{{node save/RestoreV2}}]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "./tools/test_shadownet.py", line 166, in <module>
    is_vis=args.visualize
  File "./tools/test_shadownet.py", line 131, in recognize
    saver.restore(sess=sess, save_path=weights_path)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 1290, in restore
    {self.saver_def.filename_tensor_name: save_path})
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 956, in run
    run_metadata_ptr)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1180, in _run
    feed_dict_tensor, options, run_metadata)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1359, in _do_run
    run_metadata)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1384, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.DataLossError: Unable to open table file ./model/crnn_synth90k/checkpoint: Data loss: file is too short to be an sstable: perhaps your file is in a different file format and you need to use a different restore operator?
         [[node save/RestoreV2 (defined at /home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py:1748) ]]

Original stack trace for 'save/RestoreV2':
  File "./tools/test_shadownet.py", line 166, in <module>
    is_vis=args.visualize
  File "./tools/test_shadownet.py", line 120, in recognize
    saver = tf.train.Saver()
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 828, in __init__
    self.build()
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 840, in build
    self._build(self._filename, build_save=True, build_restore=True)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 878, in _build
    build_restore=build_restore)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 508, in _build_internal
    restore_sequentially, reshape)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 328, in _AddRestoreOps
    restore_sequentially)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/training/saver.py", line 575, in bulk_restore
    return io_ops.restore_v2(filename_tensor, names, slices, dtypes)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/ops/gen_io_ops.py", line 1696, in restore_v2
    name=name)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/framework/op_def_library.py", line 794, in _apply_op_helper
    op_def=op_def)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/util/deprecation.py", line 507, in new_func
    return func(*args, **kwargs)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 3357, in create_op
    attrs, op_def, compute_device)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 3426, in _create_op_internal
    op_def=op_def)
  File "/home/darker/.conda/envs/tests/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 1748, in __init__
    self._traceback = tf_stack.extract_stack()

[Sanix-Darker]

AAHHH, it’s OK, now, thanks, so the complete command is :

python ./tools/test_shadownet.py 
--image_path ./data/test_images/test_01.jpg 
--weights_path ./model/crnn_synth90k/shadownet.ckpt    
--char_dict_path ./data/char_dict/char_dict_en.json    
--ord_map_dict_path ./data/char_dict/ord_map_en.json

That’s what i was asking 😄 Thank’s for sharing this project !

[MaybeShewill-CV]

@Sanix-Darker Check this https://stackoverflow.com/questions/33759623/tensorflow-how-to-save-restore-a-model