MS MARCO(Microsoft Machine Reading Comprehension) is a large scale dataset focused on machine reading comprehension and question answering
Home Page: https://microsoft.github.io/MSMARCO-Question-Answering/
License: MIT License
![microsoft-github-policy-service[bot] avatar](https://avatars.githubusercontent.com/in/95686?v=4 "microsoft-github-policy-service[bot]")
Hi,
the link to the Keyphrase Extraction is dead as of now.
Was it moved somewhere, changed or deleted?
There are important files that Microsoft projects should all have that are not present in this repository. A pull request has been opened to add the missing file(s). When the pr is merged this issue will be closed automatically.
Microsoft teams can learn more about this effort and share feedback within the open source guidance available internally.
It seems the dev_as_references.json is for the V1? Do you have the file from V2?
There is a error when I run script train.py, log as following:
Preparing to train...
Loading data...
Tokenizing data...
Creating model...
/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/torch/nn/modules/rnn.py:46: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.2 and num_layers=1
"num_layers={}".format(dropout, num_layers))
Starting epoch 0
Traceback (most recent call last):
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/train.py", line 230, in <module>
main()
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/train.py", line 224, in main
checkpoint, args.exp_folder)
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/checkpointing.py", line 61, in checkpoint
save_model(model, dest)
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/checkpointing.py", line 45, in save_model
save_params(destination, 'model/'+name, value)
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/checkpointing.py", line 31, in save_params
destination.create_dataset(path, data=params, compression='gzip')
File "/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/h5py/_hl/group.py", line 136, in create_dataset
dsid = dataset.make_new_dset(self, shape, dtype, data, **kwds)
File "/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/h5py/_hl/dataset.py", line 83, in make_new_dset
else base.guess_dtype(data)))
File "/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/numpy/core/numeric.py", line 538, in asarray
return array(a, dtype, copy=False, order=order)
File "/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/torch/tensor.py", line 450, in __array__
return self.numpy()
TypeError: can't convert CUDA tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.
The running code is following:
#!/bin/bash
export CUDA_VISIBLE_DEVICES=0,1
export MS_PATH=/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment
python3 $MS_PATH/train.py $MS_PATH $MS_PATH/data/train_v2.1.json \
--force_restart \
--cuda=True
The environments used python3.6, CUDA9.0.
Hi all,
The MS MARCO files were recently moved from msmarco.blob.core.windows.net to msmarco.z22.web.core.windows.net. However, it looks like now the QnA data are no longer available anywhere. (At least, nowhere I can find...). I know the dataset is retired, but the data is still helpful/important for some experiments, reproducing existing results, etc.
Are there plans to continue hosting the QnA data somewhere?
Thanks,
sean
Please check if this web site works.
http://www.msmarco.org/dataset.aspx
It only shows this error:
404 Web Site not found.
You may be seeing this error due to one of the reasons listed below :
Custom domain has not been configured inside Azure. See how to map an existing domain to resolve this.
Client cache is still pointing the domain to old IP address. Clear the cache by running the command ipconfig/flushdns.
Checkout App Service Domain FAQ for more questions.
Both the two task use the same dataset (same training set, same devset and same evalset).
If I work on the NLGEN task, I should only take the QA pairs with wellFormedAnswers, am I right?
If I submit to the leaderboard on NLGEN task, is every QA pair in testset provided with the wellFormedAnswers?
Another error occurs when I run predict.py, the error information as following:
Loading Model...
/home/wangyongbo/anaconda3/envs/py36/lib/python3.6/site-packages/torch/nn/modules/rnn.py:46: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.2 and num_layers=1
"num_layers={}".format(dropout, num_layers))
Augmenting with random embeddings...
Augmenting with random char embeddings...
Traceback (most recent call last):
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/predict.py", line 202, in <module>
main()
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/predict.py", line 186, in main
toks = regex_multi_space.sub(' ', regex_drop_char.sub(' ', ' '.join(id_to_token[tok] for tok in toks).lower())).strip()
File "/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment/predict.py", line 186, in <genexpr>
toks = regex_multi_space.sub(' ', regex_drop_char.sub(' ', ' '.join(id_to_token[tok] for tok in toks).lower())).strip()
KeyError: tensor(507764)
The running script is following:
#!/bin/bash
export CUDA_VISIBLE_DEVICES=1
export MS_PATH=/DATA2/wangyongbo/ms_marco/MSMARCO-Question-Answering/Baseline/experiment
python3 $MS_PATH/predict.py $MS_PATH/ $MS_PATH/data/dev_v2.1.json prediction.json --cuda=True
may you help to get link to download data for question - answer
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.