440cfe18ed
* modify aquila * modify aquila2 * add baichuan * modify baichuan2 * modify blue-lm * modify chatglm3 * modify chinese-llama2 * modiy codellama * modify distil-whisper * modify dolly-v1 * modify dolly-v2 * modify falcon * modify flan-t5 * modify gpt-j * modify internlm * modify llama2 * modify mistral * modify mixtral * modify mpt * modify phi-1_5 * modify qwen * modify qwen-vl * modify replit * modify solar * modify starcoder * modify vicuna * modify voiceassistant * modify whisper * modify yi * modify aquila2 * modify baichuan * modify baichuan2 * modify blue-lm * modify chatglm2 * modify chatglm3 * modify codellama * modify distil-whisper * modify dolly-v1 * modify dolly-v2 * modify flan-t5 * modify llama2 * modify llava * modify mistral * modify mixtral * modify phi-1_5 * modify qwen-vl * modify replit * modify solar * modify starcoder * modify yi * correct the comments * remove cpu_embedding in code for whisper and distil-whisper * remove comment * remove cpu_embedding for voice assistant * revert modify voice assistant * modify for voice assistant * add comment for voice assistant * fix comments * fix comments
79 lines
3.2 KiB
Python
79 lines
3.2 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 transformers import AutoModelForCausalLM, AutoTokenizer
|
|
from bigdl.llm import optimize_model
|
|
|
|
# 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="meta-llama/Llama-2-7b-chat-hf",
|
|
help='The huggingface repo id for the Llama2 (e.g. `meta-llama/Llama-2-7b-chat-hf` and `meta-llama/Llama-2-13b-chat-hf`) to be downloaded'
|
|
', 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
|
|
model = AutoModelForCausalLM.from_pretrained(model_path,
|
|
trust_remote_code=True,
|
|
torch_dtype='auto',
|
|
low_cpu_mem_usage=True)
|
|
|
|
# With only one line to enable BigDL-LLM optimization on model
|
|
# When running LLMs on Intel iGPUs for Windows users, we recommend setting `cpu_embedding=True` in the optimize_model function.
|
|
# This will allow the memory-intensive embedding layer to utilize the CPU instead of iGPU.
|
|
model = optimize_model(model)
|
|
|
|
model = model.to('xpu')
|
|
|
|
# Load tokenizer
|
|
tokenizer = AutoTokenizer.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')
|
|
# ipex model needs a warmup, then inference time can be accurate
|
|
output = model.generate(input_ids,
|
|
max_new_tokens=args.n_predict)
|
|
|
|
# start inference
|
|
st = time.time()
|
|
output = model.generate(input_ids,
|
|
max_new_tokens=args.n_predict)
|
|
torch.xpu.synchronize()
|
|
end = time.time()
|
|
output = output.cpu()
|
|
output_str = tokenizer.decode(output[0], skip_special_tokens=True)
|
|
print(f'Inference time: {end-st} s')
|
|
print('-'*20, 'Output', '-'*20)
|
|
print(output_str)
|