lanyunshi / conversationalkbqa Goto Github PK

Link to download the processed data CONVEX found that the data is incomplete, there is no processed_q.json under dev_set.can upload data?

hi!
when I run python SPARQL_test.py, then the error:
TimeoutError: [WinError 10060]
urllib.error.URLError: <urlopen error [WinError 10060]

I changed the AGENT,
origin: # AGENT = 'Chrome/77.0.3865.90'
changed: # AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.60 Safari/537.36 Edg/100.0.1185.29'
but it also bad,how can I resolve it? thank you

Sep-21: Can you please share the script for ConvCSQA data creation?

Hello,
Can you please share the download link for the ConvCSQA data or the script to create the data from original CSQA dataset?

Thanks
Sudip Chowdhury

The Conversation History Encoder embeds the questions by feeding glove word embeddings to a Bi-LSTM. How do you encode the answers?

Hi !
when I test SPARQL_test.py, it has a error:

SPARQL_test.py:20 (test)
self = <urllib.request.HTTPSHandler object at 0x000001ECC1970888>
http_class = <class 'http.client.HTTPSConnection'>
req = <urllib.request.Request object at 0x000001ECC1967788>
http_conn_args = {'check_hostname': None, 'context': None}
host = 'query.wikidata.org'
h = <http.client.HTTPSConnection object at 0x000001ECC1970348>

how can I solve it? thank you for your reply!

Hello
what is the use of the sorting of sorted_path = sorted(batch.path2ans.keys()) there?

Can you clear up my confusion? Thank you.

Hi!
I want to train anew model, but there is a error:
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [1200]] is at version 3; expected version 2 instead.

def update_policy_immediately(adjust_loss, optimizer, te_losses=0):
optimizer.zero_grad()
adjust_loss.backward(retain_graph=True) # -----------------------------the error line-----------------------------------------------
optimizer.step()
return adjust_loss.item()#, te_losses

I print the adjust_loss, the result is the following:
tensor(5.4996e-06, device='cuda:0', grad_fn=)
tensor(0., device='cuda:0', grad_fn=)
tensor(4.5414, device='cuda:0', grad_fn=)
tensor(0.9992, device='cuda:0', grad_fn=)
[W ..\torch\csrc\autograd\python_anomaly_mode.cpp:104] Warning: Error detected in ThnnFusedLstmCellBackward.

how can I solve it? thank you

Hi

When I train according to what you said, all the indicators are 0, as shown below,

Test: 100%|██████████| 10/10 [00:00<00:00, 10024.63it/s] 06/03/2022 10:21:54 - INFO - __main__ - ***** Eval results (99)***** 06/03/2022 10:21:54 - INFO - __main__ - dev reward=(0.0, 0.0) 06/03/2022 10:21:54 - INFO - __main__ - dev te reward=(0.0, 0.0) 06/03/2022 10:21:54 - INFO - __main__ - test reward=(0.0, 0.0) 06/03/2022 10:21:54 - INFO - __main__ - training loss=0.0 06/03/2022 10:21:54 - INFO - __main__ - training reward=(0.0, 0.0) 06/03/2022 10:21:54 - INFO - __main__ - training te loss=0.0 06/03/2022 10:21:54 - INFO - __main__ - training te reward=(0.0, 0.0) Epoch: 100%|██████████| 100/100 [03:53<00:00, 2.33s/it]

and the training ends in a few minutes, and the files in the cache are almost empty.
Do you have a solution?

I took the liberty to ask if anyone has created the ConvCSQA dataset？

I can access query.wikidata but why can't I count the loss. The model runs very fast, has anyone run it?
dev reward=(0.0, 0.0)
dev te reward=(0.0, 0.0)
test reward=(0.0, 0.0)
training loss=0.0
training reward=(0.0, 0.0)
training te loss=0.0
training te reward=(0.0, 0.0)

Hi, I am trying to reproduce the convCSQA dataset but I have one question about it. Could you please answer it?

If the question type description did not explicitly mention that this question contains an "indirect" subject, we removed it.

What does it mean? Could you please give an example?

Hi, could you please tell me where is the code for constructing the query generator?

My evaluation results of the new model are always higher than the article report data. Does anyone know why?

when computing focal scores, do parameters w(t) and b(t) compute unnormalized focal score for each entity and softmax computes normalized focal scores with inputs of concatenation of unnormalized focal scores?