def init_context_word_embs_sentence_model(model_path, model_type, device, force_reload=False,
min_length=100, max_length=300, batch_size=32, temperature=1.0, top_k=50, top_p=0.9,
silence=True, use_custom_api=True):
global CONTEXT_WORD_EMBS_SENTENCE_MODELS
model_name = '_'.join([os.path.basename(model_path), str(device)])
if model_name in CONTEXT_WORD_EMBS_SENTENCE_MODELS and not force_reload:
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].min_length = min_length
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].max_length = max_length
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].temperature = temperature
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].top_k = top_k
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].top_p = top_p
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].batch_size = batch_size
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].silence = silence
return CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name]
if use_custom_api:
if model_type == 'xlnet':
model = nml.XlNet(model_path, device=device, temperature=temperature, top_k=top_k,
optimize=None, silence=True)
elif model_type == 'gpt2':
model = nml.Gpt2(model_path, device=device, temperature=temperature, top_k=top_k,
optimize=None, silence=True)
else:
raise ValueError('Model name value is unexpected. Only support XLNet and GPT2 model.')
else:
model = nml.TextGenTransformers(model_path, device=device, min_length=min_length, max_length=max_length,
temperature=temperature, top_k=top_k, top_p=top_p, batch_size=batch_size)
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name] = model
return model
def init_context_word_embs_sentence_model(model_path, device, force_reload=False, temperature=1.0, top_k=None,
top_p=None, optimize=None, silence=True):
global CONTEXT_WORD_EMBS_SENTENCE_MODELS
model_name = os.path.basename(model_path)
if model_name in CONTEXT_WORD_EMBS_SENTENCE_MODELS and not force_reload:
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].device = device
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].temperature = temperature
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].top_k = top_k
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].top_p = top_p
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].optimize = optimize
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name].silence = silence
return CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name]
if 'xlnet' in model_path:
model = nml.XlNet(model_path, device=device, temperature=temperature, top_k=top_k, top_p=top_p,
optimize=optimize, silence=True)
elif 'gpt2' in model_path:
model = nml.Gpt2(model_path, device=device, temperature=temperature, top_k=top_k, top_p=top_p,
optimize=optimize, silence=True)
else:
raise ValueError('Model name value is unexpected. Only support XLNet and GPT2 model.')
CONTEXT_WORD_EMBS_SENTENCE_MODELS[model_name] = model
return model
def init_gpt2_model(model_path, device, force_reload=False):
# Load model once at runtime
global GPT2_MODEL
if GPT2_MODEL and not force_reload:
return GPT2_MODEL
gpt2_model = nml.Gpt2(model_path, device=device)
gpt2_model.model.eval()
GPT2_MODEL = gpt2_model
return gpt2_model
def init_gpt2_model(model_path,
device,
force_reload=False,
temperature=1.0,
top_k=None,
top_p=None):
# Load model once at runtime
global GPT2_MODEL
if GPT2_MODEL and not force_reload:
GPT2_MODEL.temperature = temperature
GPT2_MODEL.top_k = top_k
GPT2_MODEL.top_p = top_p
return GPT2_MODEL
gpt2_model = nml.Gpt2(model_path,
device=device,
temperature=temperature,
top_k=top_k,
top_p=top_p)
gpt2_model.model.eval()
GPT2_MODEL = gpt2_model
return gpt2_model