span-parser

span-parser copied to clipboard

Run !python src/main.py --train data/02-21.10way.clean --dev data/22.auto.clean --vocab data/vocab.json --model data/my_model and result like this: [dynet] random seed: 3637392249 [dynet] allocating memory: 2000MB [dynet] memory allocation done. L2 regularization: 0...

lanhtruonggiang-k63a5

[dynet] random seed: 3637392249 [dynet] allocating memory: 2000MB [dynet] memory allocation done. L2 regularization: 0 Hidden units: 200, per-LSTM units: 200 Embeddings: word=(44392, 50) tag=(48, 20) Dropout rate: 0.5 Parameters...

lanhtruonggiang-k63a5

What script did you use to generate data files?

3

What script did you use to generate the files matching `data/*.clean`? In particular, did you do any preprocessing to the data?

timvieira

Can't load saved model

1

I'm having trouble loading the trained model that ships with the repo. ```bash $ python src/main.py --test data/23.auto.clean --vocab data/vocab.json --model data/model ... Loaded test trees from data/23.auto.clean Traceback (most...

timvieira

Easy trick to avoid overflow in softmax

1

I see that you are using an "unsafe" softmax distribution on [line 308 of parser.py](https://github.com/jhcross/span-parser/blob/master/src/parser.py#L308). This line is likely to overflow if `scores * alpha` is large. There is [a...

timvieira