LLM: add gpu pytorch-models example llama2 and chatglm2 (#9142)
This commit is contained in:
JIN Qiao
GitHub
parent
4f34557224
commit
d74834ff4c
6 changed files with 407 additions and 0 deletions
|
|
@ -5,6 +5,8 @@ You can use `optimize_model` API to accelerate general PyTorch models on Intel G
|
|||
| Model | Example |
|
||||
|----------------|----------------------------------------------------------|
|
||||
| Mistral | [link](mistral) |
|
||||
| LLaMA 2 | [link](llama2) |
|
||||
| ChatGLM2 | [link](chatglm2) |
|
||||
|
||||
## Verified Hardware Platforms
|
||||
|
||||
|
|
|
|||
109
python/llm/example/GPU/PyTorch-Models/Model/chatglm2/README.md
Normal file
109
python/llm/example/GPU/PyTorch-Models/Model/chatglm2/README.md
Normal file
|
|
@ -0,0 +1,109 @@
|
|||
# ChatGLM2
|
||||
In this directory, you will find examples on how you could use BigDL-LLM `optimize_model` API to accelerate ChatGLM2 models. For illustration purposes, we utilize the [THUDM/chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b) as reference ChatGLM2 models.
|
||||
|
||||
## Requirements
|
||||
To run these examples with BigDL-LLM on Intel GPUs, we have some recommended requirements for your machine, please refer to [here](../README.md#recommended-requirements) for more information.
|
||||
|
||||
## Example 1: Predict Tokens using `generate()` API
|
||||
In the example [generate.py](./generate.py), we show a basic use case for a ChatGLM2 model to predict the next N tokens using `generate()` API, with BigDL-LLM INT4 optimizations on Intel GPUs.
|
||||
### 1. Install
|
||||
We suggest using conda to manage the Python environment. For more information about conda installation, please refer to [here](https://docs.conda.io/en/latest/miniconda.html#).
|
||||
|
||||
After installing conda, create a Python environment for BigDL-LLM:
|
||||
```bash
|
||||
conda create -n llm python=3.9 # recommend to use Python 3.9
|
||||
conda activate llm
|
||||
|
||||
# below command will install intel_extension_for_pytorch==2.0.110+xpu as default
|
||||
# you can install specific ipex/torch version for your need
|
||||
pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
|
||||
```
|
||||
|
||||
### 2. Configures OneAPI environment variables
|
||||
```bash
|
||||
source /opt/intel/oneapi/setvars.sh
|
||||
```
|
||||
|
||||
### 3. Run
|
||||
|
||||
For optimal performance on Arc, it is recommended to set several environment variables.
|
||||
|
||||
```bash
|
||||
export USE_XETLA=OFF
|
||||
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
|
||||
```
|
||||
|
||||
```bash
|
||||
python ./generate.py --prompt 'AI是什么？'
|
||||
```
|
||||
|
||||
In the example, several arguments can be passed to satisfy your requirements:
|
||||
|
||||
- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the ChatGLM2 model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'THUDM/chatglm2-6b'`.
|
||||
- `--prompt PROMPT`: argument defining the prompt to be infered (with integrated prompt format for chat). It is default to be `'AI是什么？'`.
|
||||
- `--n-predict N_PREDICT`: argument defining the max number of tokens to predict. It is default to be `32`.
|
||||
|
||||
#### 2.3 Sample Output
|
||||
#### [THUDM/chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Output --------------------
|
||||
问：AI是什么？
|
||||
|
||||
答： AI指的是人工智能,是一种能够通过学习和推理来执行任务的计算机程序。AI可以分为弱人工智能和强人工智能。
|
||||
|
||||
弱人工智能(也称为狭
|
||||
```
|
||||
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Output --------------------
|
||||
问：What is AI?
|
||||
|
||||
答： Artificial Intelligence (AI) refers to the ability of a computer or machine to perform tasks that typically require human-like intelligence, such as understanding language, recognizing patterns
|
||||
```
|
||||
|
||||
## Example 2: Stream Chat using `stream_chat()` API
|
||||
In the example [streamchat.py](./streamchat.py), we show a basic use case for a ChatGLM2 model to stream chat, with BigDL-LLM INT4 optimizations.
|
||||
### 1. Install
|
||||
We suggest using conda to manage the Python environment. For more information about conda installation, please refer to [here](https://docs.conda.io/en/latest/miniconda.html#).
|
||||
|
||||
After installing conda, create a Python environment for BigDL-LLM:
|
||||
```bash
|
||||
conda create -n llm python=3.9 # recommend to use Python 3.9
|
||||
conda activate llm
|
||||
|
||||
# below command will install intel_extension_for_pytorch==2.0.110+xpu as default
|
||||
# you can install specific ipex/torch version for your need
|
||||
pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
|
||||
```
|
||||
|
||||
### 2. Configures OneAPI environment variables
|
||||
```bash
|
||||
source /opt/intel/oneapi/setvars.sh
|
||||
```
|
||||
|
||||
### 3. Run
|
||||
|
||||
For optimal performance on Arc, it is recommended to set several environment variables.
|
||||
|
||||
```bash
|
||||
export USE_XETLA=OFF
|
||||
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
|
||||
```
|
||||
|
||||
**Stream Chat using `stream_chat()` API**:
|
||||
```
|
||||
python ./streamchat.py
|
||||
```
|
||||
|
||||
**Chat using `chat()` API**:
|
||||
```
|
||||
python ./streamchat.py --disable-stream
|
||||
```
|
||||
|
||||
In the example, several arguments can be passed to satisfy your requirements:
|
||||
|
||||
- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the ChatGLM2 model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'THUDM/chatglm2-6b'`.
|
||||
- `--question QUESTION`: argument defining the question to ask. It is default to be `"晚上睡不着应该怎么办"`.
|
||||
- `--disable-stream`: argument defining whether to stream chat. If include `--disable-stream` when running the script, the stream chat is disabled and `chat()` API is used.
|
||||
|
|
@ -0,0 +1,74 @@
|
|||
#
|
||||
# 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 intel_extension_for_pytorch as ipex
|
||||
import time
|
||||
import argparse
|
||||
|
||||
from transformers import AutoModel, 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/THUDM/chatglm2-6b/blob/main/modeling_chatglm.py#L1007
|
||||
CHATGLM_V2_PROMPT_FORMAT = "问：{prompt}\n\n答："
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for ChatGLM2 model')
|
||||
parser.add_argument('--repo-id-or-model-path', type=str, default="THUDM/chatglm2-6b",
|
||||
help='The huggingface repo id for the ChatGLM2 model 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 = AutoModel.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
|
||||
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 = CHATGLM_V2_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)
|
||||
|
|
@ -0,0 +1,75 @@
|
|||
#
|
||||
# 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 intel_extension_for_pytorch as ipex
|
||||
import time
|
||||
import argparse
|
||||
import numpy as np
|
||||
|
||||
from transformers import AutoModel, AutoTokenizer
|
||||
from bigdl.llm import optimize_model
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser(description='Stream Chat for ChatGLM2 model')
|
||||
parser.add_argument('--repo-id-or-model-path', type=str, default="THUDM/chatglm2-6b",
|
||||
help='The huggingface repo id for the ChatGLM2 model to be downloaded'
|
||||
', or the path to the huggingface checkpoint folder')
|
||||
parser.add_argument('--question', type=str, default="晚上睡不着应该怎么办",
|
||||
help='Qustion you want to ask')
|
||||
parser.add_argument('--disable-stream', action="store_true",
|
||||
help='Disable stream chat')
|
||||
|
||||
args = parser.parse_args()
|
||||
model_path = args.repo_id_or_model_path
|
||||
disable_stream = args.disable_stream
|
||||
|
||||
# Load model
|
||||
model = AutoModel.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
|
||||
model = optimize_model(model)
|
||||
|
||||
model.to('xpu')
|
||||
|
||||
# Load tokenizer
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_path,
|
||||
trust_remote_code=True)
|
||||
|
||||
with torch.inference_mode():
|
||||
prompt = args.question
|
||||
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=32)
|
||||
|
||||
# start inference
|
||||
if disable_stream:
|
||||
# Chat
|
||||
response, history = model.chat(tokenizer, args.question, history=[])
|
||||
print('-'*20, 'Chat Output', '-'*20)
|
||||
print(response)
|
||||
else:
|
||||
# Stream chat
|
||||
response_ = ""
|
||||
print('-'*20, 'Stream Chat Output', '-'*20)
|
||||
for response, history in model.stream_chat(tokenizer, args.question, history=[]):
|
||||
print(response.replace(response_, ""), end="")
|
||||
response_ = response
|
||||
69
python/llm/example/GPU/PyTorch-Models/Model/llama2/README.md
Normal file
69
python/llm/example/GPU/PyTorch-Models/Model/llama2/README.md
Normal file
|
|
@ -0,0 +1,69 @@
|
|||
# Llama2
|
||||
In this directory, you will find examples on how you could use BigDL-LLM `optimize_model` API to accelerate Llama2 models. For illustration purposes, we utilize the [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) and [meta-llama/Llama-2-13b-chat-hf](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf) as reference Llama2 models.
|
||||
|
||||
## Requirements
|
||||
To run these examples with BigDL-LLM on Intel GPUs, we have some recommended requirements for your machine, please refer to [here](../README.md#recommended-requirements) for more information.
|
||||
|
||||
## Example: Predict Tokens using `generate()` API
|
||||
In the example [generate.py](./generate.py), we show a basic use case for a Llama2 model to predict the next N tokens using `generate()` API, with BigDL-LLM INT4 optimizations on Intel GPUs.
|
||||
### 1. Install
|
||||
We suggest using conda to manage the Python environment. For more information about conda installation, please refer to [here](https://docs.conda.io/en/latest/miniconda.html#).
|
||||
|
||||
After installing conda, create a Python environment for BigDL-LLM:
|
||||
```bash
|
||||
conda create -n llm python=3.9 # recommend to use Python 3.9
|
||||
conda activate llm
|
||||
|
||||
# below command will install intel_extension_for_pytorch==2.0.110+xpu as default
|
||||
# you can install specific ipex/torch version for your need
|
||||
pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
|
||||
```
|
||||
|
||||
### 2. Configures OneAPI environment variables
|
||||
```bash
|
||||
source /opt/intel/oneapi/setvars.sh
|
||||
```
|
||||
|
||||
### 3. Run
|
||||
|
||||
For optimal performance on Arc, it is recommended to set several environment variables.
|
||||
|
||||
```bash
|
||||
export USE_XETLA=OFF
|
||||
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
|
||||
```
|
||||
|
||||
```bash
|
||||
python ./generate.py --prompt 'What is AI?'
|
||||
```
|
||||
|
||||
In the example, several arguments can be passed to satisfy your requirements:
|
||||
|
||||
- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the Llama2 model (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. It is default to be `'meta-llama/Llama-2-7b-chat-hf'`.
|
||||
- `--prompt PROMPT`: argument defining the prompt to be infered (with integrated prompt format for chat). It is default to be `'What is AI?'`.
|
||||
- `--n-predict N_PREDICT`: argument defining the max number of tokens to predict. It is default to be `32`.
|
||||
|
||||
#### 2.3 Sample Output
|
||||
#### [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Output --------------------
|
||||
### HUMAN:
|
||||
What is AI?
|
||||
|
||||
### RESPONSE:
|
||||
|
||||
AI is a field of computer science that focuses on creating intelligent machines that can perform tasks that typically require human intelligence, such as understanding natural language,
|
||||
```
|
||||
|
||||
#### [meta-llama/Llama-2-13b-chat-hf](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Output --------------------
|
||||
### HUMAN:
|
||||
What is AI?
|
||||
|
||||
### RESPONSE:
|
||||
|
||||
AI, or artificial intelligence, refers to the ability of machines to perform tasks that would typically require human intelligence, such as learning, problem-solving,
|
||||
```
|
||||
|
|
@ -0,0 +1,78 @@
|
|||
#
|
||||
# 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 intel_extension_for_pytorch as ipex
|
||||
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
|
||||
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)
|
||||
Loading…
Reference in a new issue