New answers tagged

0

In QA, it's computed over the individual words in the prediction against those in the True Answer. The number of shared words between the prediction and the truth is the basis of the F1 score: precision is the ratio of the number of shared words to the total number of words in the prediction, and recall is the ratio of the number of shared words to the total ...


0

If the text is really that regular, you can do simple pattern matching: Search for the keywords "limit of liability", "deductible", and "sum insured". The take the next numerical value (possibly preceded by "INR"), and assign it to the corresponding value. However, this is very simplistic and brittle, and will fail if ...


3

GPT-2 is a close copy of the basic transformer architecture. GPT-2 does not require the encoder part of the original transformer architecture as it is decoder-only, and there are no encoder attention blocks, so the decoder is equivalent to the encoder, except for the MASKING in the multi-head attention block, the decoder is only allowed to glean information ...


2

There isn't, really. Natural language is way more complex and irregular than algebra, which is far more formalised and unambiguous. So far, in NLP, most success/progress has been made in little toy domains, which exclude most of the complexities of real life, including many ambiguities. When you say the rules of algebra are somewhat like grammar, then that ...


0

You could try to train a recurrent neural net on char level. Basically, you took GRU or LSTM and use a sequence of characters, not tags or words. In the blogpost "The Unreasonable Effectiveness of Recurrent Neural Networks" there are examples for Shakespeare, Linux source code on C and for papers in latex code, and the results are quite valid, ...


1

One of the ways to ask if this two problmes are related is to ask, could we solve math/algebra equations with NLP approaches, and the answer is yes, it's an absolutely valid idea and it was approached by many researchers. For example in the "Deep learning for symbolic mathemathics" paper by facebook researchers, the NLP-based approach was used to ...


0

I think the problem is that you're only training the network on words. Every example in your training data has a desired label of "is a word," and so your network could achieve the lowest possible loss by simply giving a probability of 100% to "is a word" all of the time. The most straightforward way to fix this would be to also include ...


0

LASER creates multilingual contextualized word embeddings, what you do with them is up to you. You can use this as a feature extraction and add whatever you want to the end of the network. I believe the implementation by facebook does not let you change the weights of the LASER model itself, they are froozen to the best of my knowledge. So, yes, you can use ...


1

Both ways are valid. It depends on what you want from the model and expect from the data. Generally though I would use 1 assumption and stick with it (unless there was a specific reason not to), so I would use all lines for test if training done that way, and same for majority. Also note if you ever get more than 3 people, you can choose to do a variance ...


0

while using a neural network for this type of problem is not the ideal use-case, it is a good exercise. In terms of conceptual issues, the most concerning that I see is the loss: $\sum_{i=1}^N L_i$. First issue, is that it validates loss at each time step equivalently. This is probably not ideal because in the example (cat), we dont expect it to know its ...


0

A lot of natural language processing software are (in 2021) using statistical approaches. Read The Deep Learning Revolution (by T.Sejnowski), Artificial Beings, the Conscious of a conscient machine (by J.Pitrat), Introduction to Deep Learning (by E.Charniak). However, mixed approaches (like in RefPerSys) can also be used. Email me to basile@starynkevitch.net ...


Top 50 recent answers are included