File size: 1,300 Bytes
72546dd |
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 |
[core/auto]
task_type = core/auto/supervised_task
cache_dir = ./cache
# model
[core/model/mass]
pretrained_name = mass-base-uncased
num_beams = 5
no_repeat_ngram_size = 0
max_gen_seq_length = 15
repetition_penalty = 1.0
# dataset
[core/dataset]
# data columns: id, num, query, doc, label, score
data_name = fuliucansheng/data_for_test
[core/dataset/train]
preprocess_funcs = ['core/process/mass_for_generation(query, doc)']
[core/dataset/dev]
preprocess_funcs = ['core/process/mass_for_tokens(query)', 'core/process/mass_for_next_tokens(doc)']
[core/dataset/test]
preprocess_funcs = ['core/process/mass_for_tokens(query)']
# process
[core/process/mass]
pretrained_name = mass-base-uncased
max_seq_length = 24
max_gen_seq_length = 15
# task
[core/auto/supervised_task]
model = core/model/mass_for_generation
optim = core/optim/adam
dataset = core/dataset/auto
loss_fn = core/loss/lm
score_fn = core/score/bleu
monitor_fns = ['core/score/bleu', 'core/score/rouge1', 'core/score/rouge2', 'core/score/rougel']
output_header = query,doc
post_process_fn = partial(core/process/mass_for_decode)
opt_fp16 = O1
from_ckpt_dir = ${core/auto:cache_dir}
to_ckpt_dir = ${core/auto:cache_dir}
output_path = ${core/auto:cache_dir}/output.txt
train_batch_size = 128
dev_batch_size = 128
test_batch_size = 256
|