def __init__(self,
generate_num=120,
temperature=0.4,
top_k=None,
top_p=0.9,
penalty=.2):
self.generate_num = generate_num
self.temp = temperature
self.top_k = top_k
self.top_p = top_p
self.penalty = penalty
self.model_name = "model_v5"
self.model_dir = "generator/gpt2/models"
self.checkpoint_path = os.path.join(self.model_dir, self.model_name)
models_dir = os.path.expanduser(os.path.expandvars(self.model_dir))
self.batch_size = 1
self.samples = 1
self.enc = encoder.get_encoder(self.model_name, models_dir)
hparams = model.default_hparams()
with open(os.path.join(models_dir, self.model_name,
'hparams.json')) as f:
hparams.override_from_dict(json.load(f))
config = tf.compat.v1.ConfigProto()
config.gpu_options.allow_growth = True
self.sess = tf.compat.v1.Session(config=config)
self.context = tf.placeholder(tf.int32, [self.batch_size, None])
# np.random.seed(seed)
# tf.set_random_seed(seed)
self.output = sample.sample_sequence(hparams=hparams,
length=self.generate_num,
context=self.context,
batch_size=self.batch_size,
temperature=temperature,
top_k=top_k,
top_p=top_p,
penalty=penalty)
self.top_output = sample.sample_sequence(hparams=hparams,
length=self.generate_num,
context=self.context,
batch_size=self.batch_size,
temperature=0.1,
top_k=top_k,
top_p=top_p)
saver = tf.train.Saver()
ckpt = tf.train.latest_checkpoint(
os.path.join(models_dir, self.model_name))
saver.restore(self.sess, ckpt)
def gen_output(self):
models_dir = os.path.expanduser(os.path.expandvars(self.model_dir))
hparams = model.default_hparams()
with open(os.path.join(models_dir, self.model_name, "hparams.json")) as f:
hparams.override_from_dict(json.load(f))
# seed = np.random.randint(0, 100000)
self.output = sample.sample_sequence(
hparams=hparams,
length=self.generate_num,
context=self.context,
batch_size=self.batch_size,
temperature=self.temp,
#top_k=self.top_k,
top_p=self.top_p,
)
