0284801fbd
* Set BIGDL_IMPORT_IPEX default to true, i.e., auto import IPEX for XPU. * Remove import intel_extension_for_pytorch as ipex from GPU example. * Add support for bigdl-core-xe-21.
71 lines
3 KiB
Python
71 lines
3 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 bigdl.llm.transformers import AutoModelForCausalLM
|
|
from transformers import LlamaTokenizer, GPTQConfig
|
|
|
|
# you could tune the prompt based on your own model,
|
|
# here the prompt tuning refers to https://huggingface.co/georgesung/llama2_7b_chat_uncensored#prompt-style
|
|
LLAMA2_PROMPT_FORMAT = """### HUMAN:
|
|
{prompt}
|
|
|
|
### RESPONSE:
|
|
"""
|
|
|
|
if __name__ == '__main__':
|
|
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for Llama2 model')
|
|
parser.add_argument('--repo-id-or-model-path', type=str, default="TheBloke/Llama-2-7B-GPTQ",
|
|
help='The huggingface repo id'
|
|
', or the path to the huggingface checkpoint folder')
|
|
parser.add_argument('--prompt', type=str, default="What is AI?",
|
|
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 in 4 bit,
|
|
# which convert the relevant layers in the model into INT4 format
|
|
model = AutoModelForCausalLM.from_pretrained(model_path,
|
|
load_in_4bit=True,
|
|
torch_dtype=torch.float,
|
|
trust_remote_code=True,).to("xpu")
|
|
|
|
# Load tokenizer
|
|
tokenizer = LlamaTokenizer.from_pretrained(model_path, trust_remote_code=True)
|
|
|
|
# Generate predicted tokens
|
|
with torch.inference_mode():
|
|
prompt = LLAMA2_PROMPT_FORMAT.format(prompt=args.prompt)
|
|
input_ids = tokenizer.encode(prompt, return_tensors="pt").to("xpu")
|
|
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 BigDL-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)
|