Where can I found ans2cat.json?

[beckybai]

Hi, Ranjay

Thanks for releasing the code.

The input parameters of "iq/utils/store_dataset.py" includes a path to the processed "ans2cat.json file". There isn't a processed code for generating this file. Can it be done by exchanging the position of key-value pair in here?

Thanks
Becky

[beckybai]

Some other questions:

1. utils/init.py
   from .nlg_eval.nlgeval import NLGEval
   No file called "nlg_eval". This function is very important for the evaluation.

2. train_iq.py
   data_loader = get_loader(args.dataset, transform, ...
   No get_vae_loader. Should it be get_loader?

3. train_iq.py
   train_weights = json.load(open(args.train_dataset_weights))
   Where can we get "train-dataset-weights"?

4. iq.py
   self.decoder = DecoderRNN(vocab_size, max_len, hidden_size, sos_id=sos_id, eos_id=eos_id, n_layers=num_layers, rnn_cell=rnn_cell, input_dropout_p=input_dropout_p, dropout_p=dropout_p, # use_attention=use_attention, embedding=embedding)
   DecoderRNN doesn't have use_attention keywork argument.

Thanks!

[tobytyx]

Hi, Ranjay

Thanks for releasing the code.

The input parameters of "iq/utils/store_dataset.py" includes a path to the processed "ans2cat.json file". There isn't a processed code for generating this file. Can it be done by exchanging the position of key-value pair in here?

Thanks
Becky

Hi beckybai
I'm facing the same question. If you solve the question, could you tell me the solution?
Thanks a lot!

[beckybai]

Hi, Ranjay
Thanks for releasing the code.
The input parameters of "iq/utils/store_dataset.py" includes a path to the processed "ans2cat.json file". There isn't a processed code for generating this file. Can it be done by exchanging the position of key-value pair in here?
Thanks
Becky

Hi beckybai
I'm facing the same question. If you solve the question, could you tell me the solution?
Thanks a lot!

I didn't find any solution.

[jennifererwangg]

Hi @beckybai and @tobytyx, I solved this issue by substituting cat2ans.json with iq_dataset.json and made some changes in store_dataset.py. You can see the changes that I made in my forked repo. Similarly, although I haven't tried it, I think you could also reverse the key and value pairs in iq_dataset.json to generate cat2ans.json.

I also couldn't find the train_datasets_weight.json file. Have you guys found a solution? Thanks!

[ljsun]

Hi, Ranjay

Thanks for releasing the code.

The input parameters of "iq/utils/store_dataset.py" includes a path to the processed "ans2cat.json file". There isn't a processed code for generating this file. Can it be done by exchanging the position of key-value pair in here?

Thanks
Becky

Hi beckybai
I'm facing the same question. If you solve the question, could you tell me the solution?
Thanks a lot!

[nanguoshun]

I am facing the same issues too.