config.py

import texar as tx
dcoder_config = {
    'dim': 768,
    'num_blocks': 12,
    'multihead_attention': {
        'num_heads': 8,
        'output_dim': 768
        # See documentation for more optional hyperparameters
    },
    'position_embedder_hparams': {
        'dim': 768
    },
    'initializer': {
        'type': 'variance_scaling_initializer',
        'kwargs': {
            'scale': 1.0,
            'mode': 'fan_avg',
            'distribution': 'uniform',
        },
    },
    'poswise_feedforward': tx.modules.default_transformer_poswise_net_hparams(
        output_dim=768)
}

loss_label_confidence = 0.9

random_seed = 1234
beam_width = 5
alpha = 0.6
hidden_dim = 768


opt = {
    'optimizer': {
        'type': 'AdamOptimizer',
        'kwargs': {
            'beta1': 0.9,
            'beta2': 0.999,
            'epsilon': 1e-8
        }
    }
}

# opt = {
#     'optimizer': {
#         'type': 'AdamOptimizer',
#         'kwargs': {
#             'beta1': 0.9,
#             'beta2': 0.997,
#             'epsilon': 1e-9
#         }
#     }
# }

#warmup steps must be 0.1% of number of iterations
lr = {
    'learning_rate_schedule': 'constant.linear_warmup.rsqrt_decay.rsqrt_depth',
    'lr_constant': 2 * (hidden_dim ** -0.5),
    'static_lr': 1e-3,
    'warmup_steps': 10000,
}

bos_token_id =101
eos_token_id = 102

model_dir= "./models"
run_mode= "train_and_evaluate"

# 训练、验证、预测的batch大小
batch_size = 32
eval_batch_size = 32
test_batch_size = 32
# batch_size = 16
# eval_batch_size = 16
# test_batch_size = 16

max_train_steps = 100000

display_steps = 10
checkpoint_steps = 2000
eval_steps = 50000
test_steps = 5000

max_decoding_length = 400

max_seq_length_src = 150
max_seq_length_tgt = 30

epochs =10

is_distributed = False


data_dir = "data/"

train_out_file = "data/train.tf_record"
eval_out_file = "data/eval.tf_record"
test_out_file = "data/test.tf_record"

bert_pretrain_dir="./bert_uncased_model"

train_story = "data/train_story.txt"
train_summ = "data/train_summ.txt"

eval_story = "data/eval_story.txt"
eval_summ = "data/eval_summ.txt"

test_story = "data/test_story.txt"
test_summ = "data/test_summ.txt"

# data_dir = "data1/"

# train_out_file = "data1/train.tf_record"
# eval_out_file = "data1/eval.tf_record"
# test_out_file = "data1/test.tf_record"

# bert_pretrain_dir="./bert_uncased_model"

# train_story = "data1/train_story.txt"
# train_summ = "data1/train_summ.txt"

# eval_story = "data1/eval_story.txt"
# eval_summ = "data1/eval_summ.txt"

# test_story = "data1/test_story.txt"
# test_summ = "data1/test_summ.txt"

bert_pretrain_dir = "./uncased_L-12_H-768_A-12"