priya-dwivedi / cs224n-squad-project Goto Github PK

Hi @priya-dwivedi ,

When i am running python code/main.py --experiment_name=rnet_best --dropout=0.20 --batch_size=20 --hidden_size_encoder=200 --embedding_size=300 --do_char_embed=False --add_highway_layer=False --rnet_attention=True --bidaf_attention=False --answer_pointer_RNET=True --smart_span=True--mode=official_eval --json_in_path=data/tiny-dev.json --json_out_path=predictions_rnet.json --ckpt_load_dir=experiments/rnet_best/best_checkpoint this command getting an error
ImportError: libcublas.so.9.0: cannot open shared object file: No such file or directory

could you please help me in this issue to resolve..

Thanks and Regards,
Manikantha Sekhar.

Happy Codding..

Can any one help on this.. if u faced similar issue?? Got stucked for 2 days :(

Traceback (most recent call last):
File "main.py", line 233, in
tf.app.run()
File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\platform\app.py", line 125, in run
sys.exit(main(argv))
File "main.py", line 191, in main
qa_model.train(sess, train_context_path, train_qn_path, train_ans_path, dev_qn_path, dev_context_path, dev_ans_path)
File "D:\MOM\cs224n-Squad-Project-master\code\qa_model.py", line 855, in train
loss, global_step, param_norm, grad_norm = self.run_train_iter(session, batch, summary_writer)
File "D:\MOM\cs224n-Squad-Project-master\code\qa_model.py", line 449, in run_train_iter
[, summaries, loss, global_step, param_norm, gradient_norm] = session.run(output_feed, input_feed)
File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py", line 900, in run
run_metadata_ptr)
File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py", line 1104, in _run
np_val = np.asarray(subfeed_val, dtype=subfeed_dtype)
File "C:\ProgramData\Anaconda3\lib\site-packages\numpy\core\numeric.py", line 492, in asarray
return array(a, dtype, copy=False, order=order)
TypeError: int() argument must be a string, a bytes-like object or a number, not 'map'

I have two csv First & Second

First is having data

      V1 V2  V3  V4 V5

Row 1: 1 2 3 4 5

Row 2: 1 6 7

Second is having

     V1   V2

Row 1: A 1

Row 2: B 1

Row 3: C X

I want to compare Second[2] with First[1] if matches add Second[1] to column of First

like

First =

        V1  V2   V3   V4  V5  V6  V7

Row 1: 1 2 3 4 5 A B

Row 2: 1 6 7 A B

WHY No module named 'nltk.tokenize.moses'

It is mentioned in the code that it is written in Tensorflow==1.4.1 in Python 2.7
I am unable to install Tensorflow==1.4.1 in Python 2.7 Environment. I am able to install Tensorflow only in python 3.5 and above. When I try to run main.py with I facing several Python 2.7 to 3.6 compatibility issues

Error While installing Tensorflow in Python 2.7

C:\Python27\Scripts>pip install tensorflow==1.4.0 --proxy approxy.rockwellcollins.com:9090
Collecting tensorflow==1.4.0
  Could not find a version that satisfies the requirement tensorflow==1.4.0 (from versions: )
No matching distribution found for tensorflow==1.4.0

My OS: Windows 10
Anaconda 64 bit Environment
Tensorflowversion 1.4
Python 3.6.5

Error with Python 3.6.5

(squad) D:\cs224n-Squad-Project>python D://cs224n-Squad-Project//code//main.py --experiment_name=bidaf_best --dropout=0.15 --batch_size=60 --hidden_size_encoder=150 --embedding_size=100 --do_char_embed=False --add_highway_layer=True --rnet_attention=False --bidaf_attention=True --answer_pointer_RNET=False --smart_span=True --hidden_size_modeling=150 --mode=train
This code was developed and tested on TensorFlow 1.4.1. Your TensorFlow version: 1.4.0
Loading GLoVE vectors from file: .\data\glove.6B.100d.txt
  0%|                                                                                                                          | 0/400000 [00:00<?, ?it/s]Traceback (most recent call last):
  File "D://cs224n-Squad-Project//code//main.py", line 232, in <module>
    tf.app.run()
  File "C:\anaconda3\envs\squad\lib\site-packages\tensorflow\python\platform\app.py", line 48, in run
    _sys.exit(main(_sys.argv[:1] + flags_passthrough))
  File "D://cs224n-Squad-Project//code//main.py", line 150, in main
    emb_matrix, word2id, id2word = get_glove(FLAGS.glove_path, FLAGS.embedding_size)
  File "D:\cs224n-Squad-Project\code\vocab.py", line 67, in get_glove
    for line in tqdm(fh, total=vocab_size):
  File "C:\anaconda3\envs\squad\lib\site-packages\tqdm\_tqdm.py", line 930, in __iter__
    for obj in iterable:
  File "C:\anaconda3\envs\squad\lib\encodings\cp1252.py", line 23, in decode
    return codecs.charmap_decode(input,self.errors,decoding_table)[0]
UnicodeDecodeError: 'charmap' codec can't decode byte 0x9d in position 2776: character maps to <undefined>

Though GPU is been given in the parameter (Either default or particularly specified as a parameter in the command for training --gpu 0).. It doesn't utilizes the GPU while training.

IOError: [Errno socket error] [SSL: WRONG_VERSION_NUMBER] wrong version number (_ssl.c:726)

Hi my name is vipin
run this your project two issue, 1 - not google collab file and TensorFlow model not run.
2 - my local PC does not run, please share the google collab file

I am waiting for your reply

please help

It would be really helpful if you could upload your trained weights/saved checkpoint files. Thank you.

I already try the model for a quite long time, but when I test using show_examples mode, the result is not too good, I get 5 answers correct from 10 questions. I want to ask how good the result when you test the model?

I am trying to run an evaluation on the model built for bidaf, since its missing in the readme on how to run the model after it is built, the code (in main.py) pretty well explains it (thanks :)),

elif FLAGS.mode == "official_eval":
        if FLAGS.json_in_path == "":
            raise Exception("For official_eval mode, you need to specify --json_in_path")

and in generating_predicitions_json file for official_eval mode, --json_in_path=data/tiny-dev.json is used. But the file data/tiny-dev.json does not exists. Is it a smaller file of data/dev-v1.1.json, if so why are we using that to evaluate?

Any help will be appriciated!

Also the repo is missing evaluation and running the model as an application, this I am ready to write and send a pull request :) but need to figure out the rest of the code.

Thanks in advance!

I try to run your code, but it still in epoch one (not go to epoch two), but the iter increasing. I use several argprase when run your code

!python main.py --experiment_name=test --dropout=0.15 --batch_size=10 --hidden_size_encoder=20 --embedding_size=100 --do_char_embed=False --add_highway_layer=True --rnet_attention=False --bidaf_attention=True --answer_pointer_RNET=False --smart_span=True --hidden_size_modeling=30 --mode=train --gpu=1 --train_dir experiments

INFO:root:epoch 1, iter 3939, loss 6.04379, smoothed loss 6.18154, grad norm 4.54307, param norm 53.22952, batch time 1.682 INFO:root:epoch 1, iter 3940, loss 5.82439, smoothed loss 6.17797, grad norm 5.33789, param norm 53.23118, batch time 1.692

My question is, how many iter will do in one epoch? How to make one epoch only do 100 iter? and how long when you train SQUAD dataset using this code?

any response and answer from all of you will very help me. thank you

InvalidArgumentError (see above for traceback): indices[27,5] = 60 is not in [0, 60)

[[node QAModel/char_embeddings/embedding_lookup_1 (defined at /Users/sarza/Documents/Finale/1/code/qa_model.py:163) ]]

How to fix this ?

Hi @priya-dwivedi ,
I am using python 3.6.8 and created a virtual environment for this project and running the code in this environment while running this ./get_started.sh command getting an error Import Error: cannot import name reload i change the version of python this get_started.bh file also but no response getting same error,could you please help me in this...

Thanks and Regards,
Manikantha Sekhar

Happy cooding...

Hi,

I was reading your article on medium about question answering system. Thanks for sharing it.

How to build a Question Answering system that can answer same question with different answers from different perspective?

I was going through predictions_analysis.ipynb and i found that each question has three answers,so please tell me why is it so?
Also i have 8 GB RAM laptop with Intel® Core™ i5-8250U CPU @ 1.60GHz × 8,Os type - 64 bit and hard disk 1TB so will my system be able to support and run your code?

I am new to this filed but ver much impressed with the web demo , I want to start this...please suggest me how to start with given code

Do you have an updated version for Python3, as Python2 is reaching the end of its life?

I am getting this error when i run this in google collab .