66 lines
2.7 KiB
Python
66 lines
2.7 KiB
Python
#
|
||
# Copyright 2016 The BigDL Authors.
|
||
#
|
||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||
# you may not use this file except in compliance with the License.
|
||
# You may obtain a copy of the License at
|
||
#
|
||
# http://www.apache.org/licenses/LICENSE-2.0
|
||
#
|
||
# Unless required by applicable law or agreed to in writing, software
|
||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||
# See the License for the specific language governing permissions and
|
||
# limitations under the License.
|
||
#
|
||
|
||
import torch
|
||
import time
|
||
import argparse
|
||
|
||
from ipex_llm import optimize_model
|
||
from transformers import AutoModelForCausalLM, AutoTokenizer
|
||
|
||
# you could tune the prompt based on your own model,
|
||
YI_PROMPT_FORMAT = "{prompt}"
|
||
|
||
if __name__ == '__main__':
|
||
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for Yi model')
|
||
parser.add_argument('--repo-id-or-model-path', type=str, default="01-ai/Yi-6B",
|
||
help='The huggingface repo id for the Yi to be downloaded'
|
||
', or the path to the huggingface checkpoint folder')
|
||
parser.add_argument('--prompt', type=str, default="AI是什么？",
|
||
help='Prompt to infer')
|
||
parser.add_argument('--n-predict', type=int, default=32,
|
||
help='Max tokens to predict')
|
||
|
||
args = parser.parse_args()
|
||
model_path = args.repo_id_or_model_path
|
||
|
||
# Load model
|
||
model = AutoModelForCausalLM.from_pretrained(model_path,
|
||
trust_remote_code=True)
|
||
model = optimize_model(model)
|
||
|
||
# Load tokenizer
|
||
tokenizer = AutoTokenizer.from_pretrained(model_path,
|
||
trust_remote_code=True)
|
||
|
||
# Generate predicted tokens
|
||
with torch.inference_mode():
|
||
prompt = YI_PROMPT_FORMAT.format(prompt=args.prompt)
|
||
input_ids = tokenizer.encode(prompt, return_tensors="pt")
|
||
st = time.time()
|
||
# if your selected model is capable of utilizing previous key/value attentions
|
||
# to enhance decoding speed, but has `"use_cache": false` in its model config,
|
||
# it is important to set `use_cache=True` explicitly in the `generate` function
|
||
# to obtain optimal performance with IPEX-LLM INT4 optimizations
|
||
output = model.generate(input_ids,
|
||
max_new_tokens=args.n_predict)
|
||
end = time.time()
|
||
output_str = tokenizer.decode(output[0], skip_special_tokens=True)
|
||
print(f'Inference time: {end-st} s')
|
||
print('-'*20, 'Prompt', '-'*20)
|
||
print(prompt)
|
||
print('-'*20, 'Output', '-'*20)
|
||
print(output_str)
|