Use model_settings.py
Browse files
train.py
CHANGED
@@ -5,6 +5,7 @@ from keras.models import Sequential
|
|
5 |
from keras.layers import Embedding, Dense, Dropout, Flatten, PReLU
|
6 |
from keras.preprocessing.text import Tokenizer
|
7 |
from keras_self_attention import SeqSelfAttention, SeqWeightedAttention
|
|
|
8 |
|
9 |
|
10 |
with open("dataset.json", "r") as f: # TODO: move the outputs into a separate file, so it would be "key": 0, "key2": 1 etc
|
@@ -14,7 +15,6 @@ dset_size = len(dset)
|
|
14 |
tokenizer = Tokenizer() # a tokenizer is a thing to split text into words, it might have some other stuff like making all the letters lowercase, etc.
|
15 |
tokenizer.fit_on_texts(list(dset.keys()))
|
16 |
|
17 |
-
emb_size = 128 # how big are the word vectors in the input (how much information can be fit into one word)
|
18 |
vocab_size = len(tokenizer.get_vocabulary())
|
19 |
inp_len = 10 # limit of the input length, after 10 words the
|
20 |
|
|
|
5 |
from keras.layers import Embedding, Dense, Dropout, Flatten, PReLU
|
6 |
from keras.preprocessing.text import Tokenizer
|
7 |
from keras_self_attention import SeqSelfAttention, SeqWeightedAttention
|
8 |
+
from model_settings import *
|
9 |
|
10 |
|
11 |
with open("dataset.json", "r") as f: # TODO: move the outputs into a separate file, so it would be "key": 0, "key2": 1 etc
|
|
|
15 |
tokenizer = Tokenizer() # a tokenizer is a thing to split text into words, it might have some other stuff like making all the letters lowercase, etc.
|
16 |
tokenizer.fit_on_texts(list(dset.keys()))
|
17 |
|
|
|
18 |
vocab_size = len(tokenizer.get_vocabulary())
|
19 |
inp_len = 10 # limit of the input length, after 10 words the
|
20 |
|