Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
Shuffling train ids
See original GitHub issueI am a bit confused about the line: https://github.com/jiesutd/NCRFpp/blob/ab82b6868bb81bedbac3e231d8e09f7341321e6f/main.py#L393
Doesn’t shuffling the data throw away useful information about the sequence? For example, the data in sample_data consists of sequences of sentences which are delimited by document delimiter tokens (-DOCSTART-). If the sentences are shuffled the model cannot learn from sequential relationships between sentences.
I would imagine that for sentence classification tasks, shuffling in this way is probably worse still?
Issue Analytics
- State:
- Created 5 years ago
- Comments:6 (2 by maintainers)
Top Results From Across the Web
Top Related Medium Post
Top Related StackOverflow Question
Troubleshoot Live Code
Top Related Reddit Thread
Top Related Hackernoon Post
Top Related Tweet
Top Related Dev.to Post
Top Related Hashnode Post
This is a good question. The model only takes a single sentence into consideration. It will not use the information between sentences. So this shuffle does not affect the model.
I’m confused about this question，sepecially in the sequence labeling task ，can you give me more hint?