hellohaptik/multi-task-nlp

multi_task_NLP is a utility toolkit enabling NLP developers to easily train and infer a single model for multiple tasks.

multi-task-nlp's Issues

model are not created

Describe the bug
after run train.py the output is just these info:
INFO - logger created.
INFO - ARGS : {'data_dir': '/ .......
NFO - Task params object created from task file...
INFO - task parameters:
{'our': {'model_type': 'BERT', 'co...................

and model are not created.
could you please help me what is the problem?

(i run it with colab and google drive, for my new dataset)

Is this support for hyperparameters tuning?

I am looking for a multi-task learning model and found this repo is a good start. Thanks for your contribution! I am wondering if there is any easy way to implement hyperparameters tuning in your codes?

"snips_intent_ner_to_tsv" : snips_intent_ner_to_tsv, NameError: name 'snips_intent_ner_to_tsv' is not defined

Describe the bug
A clear and concise description of what the bug is.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.

Docs for inferPipeline is incomplete and difficult to understand how to use

Problem:
The python library multi-task-NLP is built on specific formatted inputs and outputs the entire way through. After training completes, the documentation for how to use the InferPipeline is incomplete. How do we format the inputs for pipe.infer() from the prepared data outputted by the data_preparation.py script? Do we really have to parse the .json files ourselves in order to use the infer() pipeline, or is that functionality already included? I suspect it is already included, but this is not documented either on the multi-task-NLP docs or the ipython notebooks for the specific task code examples.

Describe the solution you'd like
The documentation in the Jupyter notebook for the Inference section of the Sentiment Analysis is incomplete. I would like to see the exact code used in order to evaluate the sentiment analysis task in /examples/sentiment_analysis/IMDb_sentiment_analysis.ipynb.

can you mention the server configuration to train the model more specifically answerability detection

Describe the bug
A clear and concise description of what the bug is.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.

error in evalution for sentiment task

File "train.py", line 315, in main
evaluate(allDataTest, BatchSamplerTest, multiTaskDataLoaderTest, taskParams,
File "/netscratch/sripathi/task2/NLP_Task/multi-task-NLP/models/eval.py", line 25, in evaluate
prediction, scores = model.predict_step(batchMetaData, batchData)
File "/netscratch/sripathi/task2/NLP_Task/multi-task-NLP/models/model.py", line 265, in predict_step
outLogits = self.network(*modelInputs)
File "/netscratch/sripathi/anaconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/netscratch/sripathi/task2/NLP_Task/multi-task-NLP/models/model.py", line 120, in forward
logits = self.allHeaderstaskName
File "/netscratch/sripathi/anaconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/netscratch/sripathi/anaconda3/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 93, in forward
return F.linear(input, self.weight, self.bias)
File "/netscratch/sripathi/anaconda3/lib/python3.8/site-packages/torch/nn/functional.py", line 1688, in linear
if input.dim() == 2 and bias is not None:
AttributeError: 'tuple' object has no attribute 'dim'

how to provide samples in answerability and waht is the output for that

Describe the bug
A clear and concise description of what the bug is.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.

What is format of tsv file

I want to train for multitask learning but couldn't understand how this model interprets which column of tsv file should be used for labeling. I am 3 labeling column for 3 different task in a single tsv file

Creating Multi-label Classification Task

Hi,

How to incorporate multi-label text classification in your framework?

Thanks

answerability_detection model alwasy predict same class

Describe the bug
A clear and concise description of what the bug is.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.

torch==1.2.0 not supported

While Cloning, I got this error:

ERROR: Could not find a version that satisfies the requirement torch==1.2.0 (from -r requirements.txt (line 7)) (from versions: 0.1.2, 0.1.2.post1, 0.1.2.post2, 1.7.1)

torch==1.2.0 is supported now I guess.