Issues with running main.py

[paniquex]

When I try to run main.py from root directory with code python main.py, I get this error:

  File "main.py", line 34, in <module>
    main()
  File "main.py", line 28, in main
    settings=settings['logging'])
KeyError: 'logging'

Also when I try to run python processes/dataset.py, I get this error:

KeyError                                  Traceback (most recent call last)
~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/dataset.py in <module>
    173 
    174 if __name__ == '__main__':
--> 175     main()
    176 
    177 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/dataset.py in main()
    151 
    152     init_loggers(verbose=verbose,
--> 153                  settings=settings['dirs_and_files']['logging'])
    154 
    155     logger_main = logger.bind(is_caption=False, indent=1)

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/tools/printing.py in init_loggers(verbose, settings)
     59             record['extra']['indent'] == i and not record['extra']['is_caption'])
     60 
---> 61     logging_path = Path(settings['root_dir'],
     62                         settings['logger_dir'])
     63 

KeyError: 'root_dir'

Could you tell me, please, how to solve this problem. Thanks!

[dr-costas]

Hi,

Sorry for the problem. There is an issue of parsing the settings at the moment. It will be solved during this week.

[dr-costas]

Hi,

The issues are most likely resolved.

Please, re-run the whole pipeline (i.e. starting from dataset creation) and let me know if there are any problems that I did not detect in my tests.

[paniquex]

Hello,

In the first case error was resolved, but now I am waiting approximately for an hour at this phase: INFO | [23:13:21] processes.dataset -- Checking the development split Is it ok?

In the second case (python processes/dataset.py) I get same error:

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/dataset.py in <module>
    173 
    174 if __name__ == '__main__':
--> 175     main()
    176 
    177 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/dataset.py in main()
    151 
    152     init_loggers(verbose=verbose,
--> 153                  settings=settings['dirs_and_files']['logging'])
    154 
    155     logger_main = logger.bind(is_caption=False, indent=1)

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/tools/printing.py in init_loggers(verbose, settings)
     59             record['extra']['indent'] == i and not record['extra']['is_caption'])
     60 
---> 61     logging_path = Path(settings['root_dir'],
     62                         settings['logger_dir'])
     63 

KeyError: 'root_dir'```

[paniquex]

About first case: I've waited a while and got this error:

INFO | [00:42:27] processes.method --   Getting training data
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in <module>
     56 
     57 if __name__ == '__main__':
---> 58     main()
     59 
     60 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in main()
     52     if settings['workflow']['dnn_training'] or \
     53             settings['workflow']['dnn_evaluation']:
---> 54         method.method(settings)
     55 
     56 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in method(settings)
    426             model_file_name=model_file_name,
    427             model_dir=model_dir,
--> 428             indices_list=indices_list)
    429         logger_inner.info('Training done')
    430 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in _do_training(model, settings_training, settings_data, settings_io, model_file_name, model_dir, indices_list)
    226         is_training=True,
    227         settings_data=settings_training,
--> 228         settings_io=settings_io)
    229 
    230     logger_main.info('Done')

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/data_handlers/clotho_loader.py in get_clotho_loader(split, is_training, settings_data, settings_io)
     87         data_dir=data_dir,
     88         split=split,
---> 89         input_field_name=settings_data['input_field_name'],
     90         output_field_name=settings_data['output_field_name'],
     91         load_into_memory=settings_data['load_into_memory'])

KeyError: 'input_field_name'

[dr-costas]

Hi,

Waiting that much is typical, if you choose to validate the creation of the dataset. In the file settings/dataset_creation.yaml (line 9) you can choose to validate or not the created examples. This process takes quite a long time. I guess I can make it by default to not be performed.

I will push today a fix on that, so by default you will not have to wait that much (edit: changed already pushed).

Regarding the problem(s) that you got, they are on the process of training the DNN. So, there is no need for you to re-create the dataset. You can indicate whether you want or not to creat the dataset at the file settings/main_settings.yaml (line 9).

So, at your case, you can safely set

workflow:
  dataset_creation: No

I will check the issues that you encountered today and push a fix. Thank you for your time.

[dr-costas]

Hi,

Now the above are resolved.

Please let me know if you encounter any other bugs.

I will leave this issue open in case that there is something more unresolved.

[paniquex]

Hi,

Thank you for fixes! Everything works well before "text_output_every_nb_epochs"-th epoch in training loop. After "text_output_every_nb_epochs"-th epoch I get this error: (I've changed text_output_every_nb_epochs setting in method_baseline.yaml to !!int 1 to reproduce error faster, but I also get this error if text_output_every_nb_epochs == !!int 10)

INFO | [20:10:24] processes.method --   Epoch: 00000 -- Training loss:  4.1841 | Time: 467.728
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in <module>
     56 
     57 if name == 'main':
---> 58     main()
     59 
     60 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in main()
     52     if settings['workflow']['dnn_training'] or \
     53             settings['workflow']['dnn_evaluation']:
---> 54         method.method(settings)
     55 
     56 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in method(settings)
    429             model_file_name=model_file_name,
    430             model_dir=model_dir,
--> 431             indices_list=indices_list)
    432         logger_inner.info('Training done')
    433 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in _do_training(model, settings_training, settings_data, settings_io, model_file_name, model_dir, indices_list)
    279                             file_names=[f_names[i_f_name]
    280                                         for i_f_name in sampling_indices],
--> 281                             eos_token=settings_data['eos_token'],
    282                             print_to_console=False)
    283 

KeyError: 'eos_token'

[dr-costas]

Ah! Forgotten query to the settings file!

Will push the fix today the soonest.

[dr-costas]

Fixed.

Please, let me know if it worked.

[paniquex]

Hi

Unfortunately, I got a new error:

INFO | [14:51:42] processes.method --   Epoch: 00000 -- Training loss:  4.1357 | Time: 263.318
INFO | [14:51:42] processes.method -- Starting decoding of captions
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in <module>
     56 
     57 if __name__ == '__main__':
---> 58     main()
     59 
     60 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in main()
     52     if settings['workflow']['dnn_training'] or \
     53             settings['workflow']['dnn_evaluation']:
---> 54         method.method(settings)
     55 
     56 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in method(settings)
    429             model_file_name=model_file_name,
    430             model_dir=model_dir,
--> 431             indices_list=indices_list)
    432         logger_inner.info('Training done')
    433 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in _do_training(model, settings_training, settings_data, settings_io, model_file_name, model_dir, indices_list)
    280                                         for i_f_name in sampling_indices],
    281                             eos_token='<eos>',
--> 282                             print_to_console=False)
    283 
    284         # Check improvement of loss

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in _decode_outputs(predicted_outputs, ground_truth_outputs, indices_object, file_names, eos_token, print_to_console)
     84         gt_caption = ' '.join(gt_caption)
     85 
---> 86         f_n = f_name.stem.split('.')[0]
     87 
     88         if f_n not in f_names:

AttributeError: 'numpy.str_' object has no attribute 'stem'

[dr-costas]

Hm.

I did not had any of these errors and I run the process multiple times. Can you please let me know what exactly you did?

[dr-costas]

One thing that I can imagine is that your directories under data are not correct.

Please try re-creating the dataset (without validating it).

The f_name should not be type of numpy. but a pathlib.Path.

[dr-costas]

Found. Fixed. Thnx!

[dr-costas]

Let me know if find any other bug.

I will let this issue open for couple of days more.

[paniquex]

Thank you for fixes! Everything works well in training loop, but in evaluation stage I got this error:

---------------------------------------------------------------------------
FileNotFoundError                         Traceback (most recent call last)
~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in <module>
     56 
     57 if __name__ == '__main__':
---> 58     main()
     59 
     60 # EOF

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/main.py in main()
     52     if settings['workflow']['dnn_training'] or \
     53             settings['workflow']['dnn_evaluation']:
---> 54         method.method(settings)
     55 
     56 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in method(settings)
    451             settings_data=settings['dnn_training_settings']['data'],
    452             settings_io=settings['dirs_and_files'],
--> 453             indices_list=indices_list)
    454         logger_inner.info('Evaluation done')
    455 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/processes/method.py in _do_evaluation(model, settings_data, settings_io, indices_list)
    177     logger_main.info('Evaluation done')
    178 
--> 179     metrics = evaluate_metrics(captions_pred, captions_gt)
    180 
    181     for metric, values in metrics.items():

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/eval_metrics.py in evaluate_metrics(prediction_file, reference_file, nb_reference_captions)
    287         ground_truths.append([reference_dict[file_name][cap] for cap in cap_names])
    288 
--> 289     metrics, per_file_metrics = evaluate_metrics_from_lists(predictions, ground_truths)
    290 
    291     total_metrics = combine_single_and_per_file_metrics(

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/eval_metrics.py in evaluate_metrics_from_lists(predictions, ground_truths, ids)
    159     write_json(pred, pred_file)
    160 
--> 161     metrics, per_file_metrics = evaluate_metrics_from_files(pred_file, ref_file)
    162 
    163     # Delete temporary files

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/eval_metrics.py in evaluate_metrics_from_files(pred_file, ref_file)
    108     cocoEval = COCOEvalCap(coco, cocoRes)
    109     cocoEval.params['audio_id'] = cocoRes.getAudioIds()
--> 110     cocoEval.evaluate()
    111 
    112     # Make dict from metrics

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/coco_caption/pycocoevalcap/eval.py in evaluate(self, verbose)
     38             print('tokenization...')
     39         tokenizer = PTBTokenizer()
---> 40         gts = tokenizer.tokenize(gts)
     41         res = tokenizer.tokenize(res)
     42 

~/Documents/Automated_Audio_Captioning_DCASE2020/dcase-2020-baseline/coco_caption/pycocoevalcap/tokenizer/ptbtokenizer.py in tokenize(self, captions_for_audio)
     56         cmd.append(os.path.basename(tmp_file.name))
     57         p_tokenizer = subprocess.Popen(cmd, cwd=path_to_jar_dirname, \
---> 58                 stdout=subprocess.PIPE)
     59         token_lines = p_tokenizer.communicate(input=sentences.rstrip())[0]
     60         token_lines = token_lines.decode()

~/anaconda3/envs/audio-captioning-baseline/lib/python3.7/subprocess.py in __init__(self, args, bufsize, executable, stdin, stdout, stderr, preexec_fn, close_fds, shell, cwd, env, universal_newlines, startupinfo, creationflags, restore_signals, start_new_session, pass_fds, encoding, errors, text)
    754                                 c2pread, c2pwrite,
    755                                 errread, errwrite,
--> 756                                 restore_signals, start_new_session)
    757         except:
    758             # Cleanup if the child failed starting.

~/anaconda3/envs/audio-captioning-baseline/lib/python3.7/subprocess.py in _execute_child(self, args, executable, preexec_fn, close_fds, pass_fds, cwd, env, startupinfo, creationflags, shell, p2cread, p2cwrite, c2pread, c2pwrite, errread, errwrite, restore_signals, start_new_session)
   1497                         if errno_num == errno.ENOENT:
   1498                             err_msg += ': ' + repr(err_filename)
-> 1499                     raise child_exception_type(errno_num, err_msg, err_filename)
   1500                 raise child_exception_type(err_msg)
   1501 

FileNotFoundError: [Errno 2] No such file or directory: 'java': 'java'

[dr-costas]

Hi,

Thank you for your time.

Unfortunately, this error is not associated with the code of the baseline system. If you observe, the problem that you have initiates from the coco-caption tools. Basically, the problem that you have relates to the prerequisites that the calculation of SPICE has. Though, I definitely would try to help you.

Thus, would you be OK if I close this issue and you open another one about this last problem that you have?

This way would be easier for others to find solutions if they encounter similar problems.

[paniquex]

Hi,

Thank you for help. Yes, you could close this issue and open another one.

[dr-costas]

I'm closing this. Please open a second issue about the last problem :)

Feel free to post again here (or open a new one) if you encounter similar problems!