def train_adhoc_with_reinforce():
hp = hyperparams.HPAdhoc()
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}".format("E")
e_config.num_epoch = 4
e_config.save_interval = 30 * 60 # 30 minutes
e_config.load_names = ['bert']
vocab_size = 30522
vocab_filename = "bert_voca.txt"
data_loader = ws.DataLoader(hp.seq_max, vocab_filename, vocab_size)
load_id = ("uncased_L-12_H-768_A-12", 'bert_model.ckpt')
e.train_adhoc(e_config, data_loader, load_id)
def train_adhoc_on_robust():
hp = hyperparams.HPAdhoc()
hp.batch_size = 16 * 3
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}".format("FAD")
e_config.num_epoch = 4
e_config.save_interval = 10 * 60 # 60 minutes
e_config.load_names = ['bert'] #, 'reg_dense']
vocab_size = 30522
data_loader = data_sampler.DataLoaderFromFile(hp.batch_size, vocab_size)
load_id = ("uncased_L-12_H-768_A-12", 'bert_model.ckpt')
#load_id = ("Adhoc_I2", 'model-290')
e.train_adhoc2(e_config, data_loader, load_id)
def test_ql():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}".format("C")
e_config.num_epoch = 4
e_config.load_names = ['bert', 'cls']
vocab_size = 30522
vocab_filename = "bert_voca.txt"
data_loader = ws.DataLoader(hp.seq_max, vocab_filename, vocab_size)
load_id = ("uncased_L-12_H-768_A-12", 'bert_model.ckpt')
e.test_ql(e_config, data_loader, load_id)
def run_adhoc_rank_on_robust():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}_eval".format("F")
e_config.num_epoch = 4
e_config.load_names = ['bert', 'reg_dense']
vocab_size = 30522
vocab_filename = "bert_voca.txt"
data_loader = data_sampler.DataLoaderFromFile(hp.batch_size, vocab_size)
load_id = ("uncased_L-12_H-768_A-12", 'bert_model.ckpt')
load_id = ("Adhoc_E", 'model-58338')
e.rank_adhoc(e_config, data_loader, load_id)
def predict_adhoc_robust_K():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}_eval".format("K")
e_config.load_names = ['bert', 'dense1', 'dense_reg']
vocab_size = 30522
#payload_path = os.path.join(cpath.data_path, "robust_payload", "payload_B_200.pickle")
payload_path = os.path.join(cpath.data_path, "robust_payload",
"payload_desc.pickle")
task_idx = int(sys.argv[2])
print(task_idx)
load_id = ("Adhoc_K", 'model-6397')
e.predict_robust(e_config, vocab_size, load_id, payload_path, task_idx)
def run_adhoc_rank():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}_eval2".format("E")
e_config.num_epoch = 4
e_config.save_interval = 30 * 60 # 30 minutes
e_config.load_names = ['bert', 'reg_dense']
vocab_size = 30522
vocab_filename = "bert_voca.txt"
data_loader = ws.DataLoader(hp.seq_max, vocab_filename, vocab_size)
load_id = ("uncased_L-12_H-768_A-12", 'bert_model.ckpt')
load_id = ("Adhoc_E", 'model-58338')
e.rank_adhoc(e_config, data_loader, load_id)
def pool_adhoc():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}_pool".format("L")
#e_config.load_names = ['bert', 'reg_dense']
e_config.load_names = ['bert', 'reg_dense', 'aux_q_info']
vocab_size = 30522
task_idx = int(sys.argv[2])
print(task_idx)
payload_path = os.path.join(cpath.data_path, "robust",
"robust_train_merge",
"merger_train_{}.pickle".format(task_idx))
load_id = ("Adhoc_L", 'model-644')
e.predict_for_pooling(e_config, vocab_size, load_id, payload_path)
def predict_adhoc_robust_L():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e_config = ExperimentConfig()
e_config.name = "Adhoc_{}_eval".format("L")
#e_config.load_names = ['bert', 'reg_dense']
e_config.load_names = ['bert', 'reg_dense', 'aux_q_info']
vocab_size = 30522
payload_path = os.path.join(cpath.data_path, "robust_payload",
"payload_B_200.pickle")
task_idx = int(sys.argv[2])
print(task_idx)
q_id_list = [
(301, 325),
(326, 350),
(351, 375),
(376, 400),
(401, 425),
(426, 450),
(601, 625),
(626, 650),
(651, 675),
(676, 700),
]
st, ed = q_id_list[task_idx]
load_id = ("Adhoc_L", 'model-644')
middle_result = e.predict_robust_L_part1(e_config, vocab_size, load_id,
payload_path, (st, ed))
preload_id2 = ("MergerE_C (copy)", 'model-3075')
e.predict_robust_L_part2(e_config, middle_result, preload_id2, (st, ed))
def predict_bm25_robust():
hp = hyperparams.HPAdhoc()
hp.batch_size = 512
e = Experiment(hp)
e.rank_robust_bm25()