File size: 2,925 Bytes
d758c99 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 | # Model details:
# - NL2Code
# - Pretrained, fixed word embeddings
# - glove-42B
# - min_freq 50
# - Spiderv2 encoder
# - question_encoder ['emb', 'bilstm']
# - column_encoder ['emb', 'bilstm-summarize']
# - table_encoder ['emb', 'bilstm-summarize']
# - upd_steps 4
# - Optimization
# - max_steps 40k
# - batch_size 10
# - Adam with lr 1e-3
function(output_from, data_path='data/spider-en/') {
local PREFIX = data_path,
data: {
train: {
name: 'spider',
paths: [
PREFIX + 'train_%s.json' % [s]
for s in ['spider', 'others']],
tables_paths: [
PREFIX + 'tables.json',
],
db_path: PREFIX + 'database',
},
val: {
name: 'spider',
paths: [PREFIX + 'dev.json'],
tables_paths: [PREFIX + 'tables.json'],
db_path: PREFIX + 'database',
},
},
model: {
name: 'EncDec',
encoder: {
name: 'spiderv2',
dropout: 0.2,
word_emb_size: 300,
question_encoder: ['emb', 'bilstm'],
column_encoder: ['emb', 'bilstm-summarize'],
table_encoder: ['emb', 'bilstm-summarize'],
update_config: {
name: 'relational_transformer',
num_layers: 4,
num_heads: 8,
},
},
decoder: {
name: 'NL2Code',
dropout: 0.2,
desc_attn: 'mha',
},
encoder_preproc: {
word_emb: {
name: 'glove',
kind: '42B',
},
count_tokens_in_word_emb_for_vocab: false,
min_freq: 50,
max_count: 5000,
include_table_name_in_column: false,
save_path: PREFIX + 'nl2code-0401,output_from=%s,emb=glove-42B,min_freq=50/' % [output_from],
},
decoder_preproc: self.encoder_preproc {
grammar: {
name: 'spider',
output_from: output_from,
use_table_pointer: output_from,
include_literals: false,
},
use_seq_elem_rules: true,
word_emb:: null,
include_table_name_in_column:: null,
count_tokens_in_word_emb_for_vocab:: null,
},
},
train: {
batch_size: 10,
eval_batch_size: 50,
keep_every_n: 1000,
eval_every_n: 100,
save_every_n: 100,
report_every_n: 10,
max_steps: 51000,
num_eval_items: 50,
},
optimizer: {
name: 'adam',
lr: 0.0,
},
lr_scheduler: {
name: 'warmup_polynomial',
num_warmup_steps: $.train.max_steps / 20,
start_lr: 1e-3,
end_lr: 0,
decay_steps: $.train.max_steps - self.num_warmup_steps,
power: 0.5,
}
}
|