ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

Add CPU and GPU examples for DeciLM-7B (#9867)

Browse source 
* Add cpu and gpu examples for DeciLM-7B

* Add cpu and gpu examples for DeciLM-7B

* Add DeciLM-7B to README table

* modify deciLM

* modify deciLM

* modify deciLM

* Add verified model in README

* Add cpu_embedding=True
This commit is contained in:
Keyan (Kyrie) Zhang 2024-02-27 13:15:49 +08:00 committed by GitHub
parent 04a6b0040c
commit 843fe546b0
10 changed files with 702 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
README.md
View file

@ -193,6 +193,7 @@ Over 40 models have been optimized/verified on `bigdl-llm`, including *LLaMA/LLa
| Phi-2 | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/phi-2) | [link](python/llm/example/GPU/HF-Transformers-AutoModels/Model/phi-2) |
| Yuan2 | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/yuan2) | [link](python/llm/example/GPU/HF-Transformers-AutoModels/Model/yuan2) |
| Gemma | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/gemma) | [link](python/llm/example/GPU/HF-Transformers-AutoModels/Model/gemma) |
| DeciLM-7B | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/deciLM-7b) | [link](python/llm/example/GPU/HF-Transformers-AutoModels/Model/deciLM-7b) |
***For more details, please refer to the `bigdl-llm` [Document](https://test-bigdl-llm.readthedocs.io/en/main/doc/LLM/index.html), [Readme](python/llm), [Tutorial](https://github.com/intel-analytics/bigdl-llm-tutorial) and [API Doc](https://bigdl.readthedocs.io/en/latest/doc/PythonAPI/LLM/index.html).***

2
python/llm/README.md
View file

@ -84,7 +84,7 @@ Over 20 models have been optimized/verified on `bigdl-llm`, including *LLaMA/LLa
| Ziya-Coding-34B-v1.0 | [link](example/CPU/HF-Transformers-AutoModels/Model/ziya) | |
| Phi-2 | [link](example/CPU/HF-Transformers-AutoModels/Model/phi-2) | [link](example/GPU/HF-Transformers-AutoModels/Model/phi-2) |
| Yuan2 | [link](example/CPU/HF-Transformers-AutoModels/Model/yuan2) | [link](example/GPU/HF-Transformers-AutoModels/Model/yuan2) |
| DeciLM-7B | [link](example/CPU/HF-Transformers-AutoModels/Model/deciLM-7b) | [link](example/GPU/HF-Transformers-AutoModels/Model/deciLM-7b) |
### Working with `bigdl-llm`
<details><summary>Table of Contents</summary>

69
python/llm/example/CPU/HF-Transformers-AutoModels/Model/deciLM-7b/README.md Normal file
View file

@ -0,0 +1,69 @@
# DeciLM-7B
In this directory, you will find examples on how you could apply BigDL-LLM INT4 optimizations on DeciLM-7B models. For illustration purposes, we utilize the [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct) as a reference DeciLM-7B model.
## 0. Requirements
To run these examples with BigDL-LLM, 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 DeciLM-7B model to predict the next N tokens using `generate()` API, with BigDL-LLM INT4 optimizations.
### 1. Install
We suggest using conda to manage environment:
```bash
conda create -n llm python=3.9
conda activate llm
pip install --pre --upgrade bigdl-llm[all] # install the latest bigdl-llm nightly build with 'all' option
pip install transformers==4.35.2 # required by DeciLM-7B
```
### 2. Run
```
python ./generate.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH --prompt PROMPT --n-predict N_PREDICT
```
Arguments info:
- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the DeciLM-7B model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'Deci/DeciLM-7B-instruct'`.
- `--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`.
> **Note**: When loading the model in 4-bit, BigDL-LLM converts linear layers in the model into INT4 format. In theory, a *X*B model saved in 16-bit will requires approximately 2*X* GB of memory for loading, and ~0.5*X* GB memory for further inference.
>
> Please select the appropriate size of the DeciLM-7B model based on the capabilities of your machine.
#### 2.1 Client
On client Windows machine, it is recommended to run directly with full utilization of all cores:
```powershell
python ./generate.py
```
#### 2.2 Server
For optimal performance on server, it is recommended to set several environment variables (refer to [here](../README.md#best-known-configuration-on-linux) for more information), and run the example with all the physical cores of a single socket.
E.g. on Linux,
```bash
# set BigDL-LLM env variables
source bigdl-llm-init
# e.g. for a server with 48 cores per socket
export OMP_NUM_THREADS=48
numactl -C 0-47 -m 0 python ./generate.py
```
#### 2.3 Sample Output
#### [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct)
```log
Inference time: XXXX s
-------------------- Prompt --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
-------------------- Output --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
AI stands for Artificial Intelligence, which refers to the development of computer systems and software that can perform tasks that typically require human intelligence, such as recognizing patterns
```

71
python/llm/example/CPU/HF-Transformers-AutoModels/Model/deciLM-7b/generate.py Normal file
View file

@ -0,0 +1,71 @@
#
# 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 AutoTokenizer
from bigdl.llm.transformers import AutoModelForCausalLM
# you could tune the prompt based on your own model,
# here the prompt tuning refers to https://huggingface.co/Deci/DeciLM-7B-instruct#prompt-template
SYSTEM_PROMPT_TEMPLATE ="""
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
{prompt}
### Assistant:
"""
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for DeciLM-7B model')
parser.add_argument('--repo-id-or-model-path', type=str, default="Deci/DeciLM-7B-instruct",
help='The huggingface repo id for the DeciLM-7B (e.g. `Deci/DeciLM-7B-instruct`) 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,
load_in_4bit=True,
trust_remote_code=True,
)
# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_path)
tokenizer.pad_token = tokenizer.eos_token
# Generate predicted tokens
with torch.inference_mode():
prompt = SYSTEM_PROMPT_TEMPLATE.format(prompt=args.prompt)
input_ids = tokenizer.encode(prompt, return_tensors="pt")
st = time.time()
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(prompt)
print('-'*20, 'Output', '-'*20)
print(output_str)

69
python/llm/example/CPU/PyTorch-Models/Model/deciLM-7b/README.md Normal file
View file

@ -0,0 +1,69 @@
# DeciLM-7B
In this directory, you will find examples on how you could use BigDL-LLM `optimize_model` API to accelerate DeciLM-7B models. For illustration purposes, we utilize the [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct) as a reference DeciLM-7B model.
## Requirements
To run these examples with BigDL-LLM, 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 DeciLM-7B model to predict the next N tokens using `generate()` API, 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
pip install --pre --upgrade bigdl-llm[all] # install the latest bigdl-llm nightly build with 'all' option
pip install transformers==4.35.2 # required by DeciLM-7B
```
### 2. Run
After setting up the Python environment, you could run the example by following steps.
#### 2.1 Client
On client Windows machines, it is recommended to run directly with full utilization of all cores:
```powershell
python ./generate.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH --prompt PROMPT --n-predict N_PREDICT
```
More information about arguments can be found in [Arguments Info](#23-arguments-info) section. The expected output can be found in [Sample Output](#24-sample-output) section.
#### 2.2 Server
For optimal performance on server, it is recommended to set several environment variables (refer to [here](../README.md#best-known-configuration-on-linux) for more information), and run the example with all the physical cores of a single socket.
E.g. on Linux,
```bash
# set BigDL-LLM env variables
source bigdl-llm-init
# e.g. for a server with 48 cores per socket
export OMP_NUM_THREADS=48
numactl -C 0-47 -m 0 python ./generate.py --prompt 'What is AI?'
```
More information about arguments can be found in [Arguments Info](#23-arguments-info) section. The expected output can be found in [Sample Output](#24-sample-output) section.
#### 2.3 Arguments Info
In the example, several arguments can be passed to satisfy your requirements:
- `--repo-id-or-model-path`: str, argument defining the huggingface repo id for the DeciLM-7B model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'Deci/DeciLM-7B-instruct'`.
- `--prompt`: str, argument defining the prompt to be inferred (with integrated prompt format for chat). It is default to be `'What is AI?'`.
- `--n-predict`: int, argument defining the max number of tokens to predict. It is default to be `32`.
#### 2.4 Sample Output
#### [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct)
```log
Inference time: XXXX s
-------------------- Prompt --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
-------------------- Output --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
AI stands for Artificial Intelligence, which refers to the development of computer systems and software that can perform tasks that typically require human intelligence, such as recognizing patterns
```

72
python/llm/example/CPU/PyTorch-Models/Model/deciLM-7b/generate.py Normal file
View file

@ -0,0 +1,72 @@
#
# 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/Deci/DeciLM-7B-instruct#prompt-template
SYSTEM_PROMPT_TEMPLATE ="""
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
{prompt}
### Assistant:
"""
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for DeciLM-7B model')
parser.add_argument('--repo-id-or-model-path', type=str, default="Deci/DeciLM-7B-instruct",
help='The huggingface repo id for the DeciLM-7B (e.g. `Deci/DeciLM-7B-instruct`) 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,
)
# With only one line to enable BigDL-LLM optimization on model
model = optimize_model(model)
# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_path)
tokenizer.pad_token = tokenizer.eos_token
# Generate predicted tokens
with torch.inference_mode():
prompt = SYSTEM_PROMPT_TEMPLATE.format(prompt=args.prompt)
input_ids = tokenizer.encode(prompt, return_tensors="pt")
st = time.time()
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, 'Output', '-'*20)
print(output_str)

130
python/llm/example/GPU/HF-Transformers-AutoModels/Model/deciLM-7b/README.md Normal file
View file

@ -0,0 +1,130 @@
# DeciLM-7B
In this directory, you will find examples on how you could apply BigDL-LLM INT4 optimizations on DeciLM-7B models on [Intel GPUs](../README.md). For illustration purposes, we utilize the [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct) as a reference DeciLM-7B model.
## 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 DeciLM-7B model to predict the next N tokens using `generate()` API, with BigDL-LLM INT4 optimizations on Intel GPUs.
### 1. Install
#### 1.1 Installation on Linux
We suggest using conda to manage 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
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
pip install transformers==4.35.2 # required by DeciLM-7B
```
#### 1.2 Installation on Windows
We suggest using conda to manage environment:
```bash
conda create -n llm python=3.9 libuv
conda activate llm
# below command will install intel_extension_for_pytorch==2.1.10+xpu as default
pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
pip install transformers==4.35.2 # required by DeciLM-7B
```
### 2. Configures OneAPI environment variables
#### 2.1 Configurations for Linux
```bash
source /opt/intel/oneapi/setvars.sh
```
#### 2.2 Configurations for Windows
```cmd
call "C:\Program Files (x86)\Intel\oneAPI\setvars.bat"
```
> Note: Please make sure you are using **CMD** (**Anaconda Prompt** if using conda) to run the command as PowerShell is not supported.
### 3. Runtime Configurations
For optimal performance, it is recommended to set several environment variables. Please check out the suggestions based on your device.
#### 3.1 Configurations for Linux
<details>
<summary>For Intel Arc™ A-Series Graphics and Intel Data Center GPU Flex Series</summary>
```bash
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
```
</details>
<details>
<summary>For Intel Data Center GPU Max Series</summary>
```bash
export LD_PRELOAD=${LD_PRELOAD}:${CONDA_PREFIX}/lib/libtcmalloc.so
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
export ENABLE_SDP_FUSION=1
```
> Note: Please note that `libtcmalloc.so` can be installed by `conda install -c conda-forge -y gperftools=2.10`.
</details>
#### 3.2 Configurations for Windows
<details>
<summary>For Intel iGPU</summary>
```cmd
set SYCL_CACHE_PERSISTENT=1
set BIGDL_LLM_XMX_DISABLED=1
```
</details>
<details>
<summary>For Intel Arc™ A300-Series or Pro A60</summary>
```cmd
set SYCL_CACHE_PERSISTENT=1
```
</details>
<details>
<summary>For other Intel dGPU Series</summary>
There is no need to set further environment variables.
</details>
> Note: For the first time that each model runs on Intel iGPU/Intel Arc™ A300-Series or Pro A60, it may take several minutes to compile.
### 4. Running examples
```
python ./generate.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH --prompt PROMPT --n-predict N_PREDICT
```
Arguments info:
- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the DeciLM-7B model (e.g `Deci/DeciLM-7B-instruct`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'Deci/DeciLM-7B-instruct'`.
- `--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`.
#### Sample Output
#### [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct)
```log
Inference time: XXXX s
-------------------- Prompt --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
-------------------- Output --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
AI stands for Artificial Intelligence, which refers to the development of computer systems and software that can perform tasks that typically require human intelligence, such as recognizing patterns
```

78
python/llm/example/GPU/HF-Transformers-AutoModels/Model/deciLM-7b/generate.py Normal file
View file

@ -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 time
import argparse
from transformers import AutoTokenizer
from bigdl.llm.transformers import AutoModelForCausalLM
# you could tune the prompt based on your own model,
# here the prompt tuning refers to https://huggingface.co/Deci/DeciLM-7B-instruct#prompt-template
SYSTEM_PROMPT_TEMPLATE ="""
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
{prompt}
### Assistant:
"""
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for DeciLM-7B model')
parser.add_argument('--repo-id-or-model-path', type=str, default="Deci/DeciLM-7B-instruct",
help='The huggingface repo id for the DeciLM-7B (e.g. `Deci/DeciLM-7B-instruct`) 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 in 4 bit,
# which convert the relevant layers in the model into INT4 format
# When running LLMs on Intel iGPUs for Windows users, we recommend setting `cpu_embedding=True` in the from_pretrained function.
# This will allow the memory-intensive embedding layer to utilize the CPU instead of iGPU.
model = AutoModelForCausalLM.from_pretrained(
model_path,
load_in_4bit=True,
trust_remote_code=True,
cpu_embedding=True
)
# With only one line to enable BigDL-LLM optimization on model
model = model.to('xpu')
# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_path)
tokenizer.pad_token = tokenizer.eos_token
# Generate predicted tokens
with torch.inference_mode():
prompt = SYSTEM_PROMPT_TEMPLATE.format(prompt=args.prompt)
input_ids = tokenizer.encode(prompt, return_tensors="pt").to('xpu')
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)

132
python/llm/example/GPU/PyTorch-Models/Model/deciLM-7b/README.md Normal file
View file

@ -0,0 +1,132 @@
# DeciLM-7B
In this directory, you will find examples on how you could use BigDL-LLM `optimize_model` API to accelerate DeciLM-7B models. For illustration purposes, we utilize the [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct) as a reference DeciLM-7B model.
## 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 DeciLM-7B model to predict the next N tokens using `generate()` API, with BigDL-LLM INT4 optimizations on Intel GPUs.
### 1. Install
#### 1.1 Installation on Linux
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
pip install transformers==4.35.2 # required by DeciLM-7B
```
#### 1.2 Installation on Windows
We suggest using conda to manage environment:
```bash
conda create -n llm python=3.9 libuv
conda activate llm
# below command will install intel_extension_for_pytorch==2.1.10+xpu as default
pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
```
### 2. Configures OneAPI environment variables
#### 2.1 Configurations for Linux
```bash
source /opt/intel/oneapi/setvars.sh
```
#### 2.2 Configurations for Windows
```cmd
call "C:\Program Files (x86)\Intel\oneAPI\setvars.bat"
```
> Note: Please make sure you are using **CMD** (**Anaconda Prompt** if using conda) to run the command as PowerShell is not supported.
### 3. Runtime Configurations
For optimal performance, it is recommended to set several environment variables. Please check out the suggestions based on your device.
#### 3.1 Configurations for Linux
<details>
<summary>For Intel Arc™ A-Series Graphics and Intel Data Center GPU Flex Series</summary>
```bash
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
```
</details>
<details>
<summary>For Intel Data Center GPU Max Series</summary>
```bash
export LD_PRELOAD=${LD_PRELOAD}:${CONDA_PREFIX}/lib/libtcmalloc.so
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
export ENABLE_SDP_FUSION=1
```
> Note: Please note that `libtcmalloc.so` can be installed by `conda install -c conda-forge -y gperftools=2.10`.
</details>
#### 3.2 Configurations for Windows
<details>
<summary>For Intel iGPU</summary>
```cmd
set SYCL_CACHE_PERSISTENT=1
set BIGDL_LLM_XMX_DISABLED=1
```
</details>
<details>
<summary>For Intel Arc™ A300-Series or Pro A60</summary>
```cmd
set SYCL_CACHE_PERSISTENT=1
```
</details>
<details>
<summary>For other Intel dGPU Series</summary>
There is no need to set further environment variables.
</details>
> Note: For the first time that each model runs on Intel iGPU/Intel Arc™ A300-Series or Pro A60, it may take several minutes to compile.
### 4. Running examples
```bash
python ./generate.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH --prompt PROMPT --n-predict N_PREDICT
```
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 DeciLM-7B model (e.g `Deci/DeciLM-7B-instruct`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'Deci/DeciLM-7B-instruct'`.
- `--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`.
#### Sample Output
#### [Deci/DeciLM-7B-instruct](https://huggingface.co/Deci/DeciLM-7B-instruct)
```log
Inference time: XXXX s
-------------------- Prompt --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
-------------------- Output --------------------
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
What is AI?
### Assistant:
AI stands for Artificial Intelligence, which refers to the development of computer systems and software that can perform tasks that typically require human intelligence, such as recognizing patterns
```

79
python/llm/example/GPU/PyTorch-Models/Model/deciLM-7b/generate.py Normal file
View file

@ -0,0 +1,79 @@
#
# 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/Deci/DeciLM-7B-instruct#prompt-template
SYSTEM_PROMPT_TEMPLATE ="""
### System:
You are an AI assistant that follows instruction extremely well. Help as much as you can.
### User:
{prompt}
### Assistant:
"""
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Predict Tokens using `generate()` API for DeciLM-7B model')
parser.add_argument('--repo-id-or-model-path', type=str, default="Deci/DeciLM-7B-instruct",
help='The huggingface repo id for the DeciLM-7B (e.g. `Deci/DeciLM-7B-instruct`) 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,
)
# 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 from_pretrained function.
# This will allow the memory-intensive embedding layer to utilize the CPU instead of iGPU.
model = optimize_model(
model,
cpu_embedding=True
)
model = model.to('xpu')
# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_path)
tokenizer.pad_token = tokenizer.eos_token
# Generate predicted tokens
with torch.inference_mode():
prompt = SYSTEM_PROMPT_TEMPLATE.format(prompt=args.prompt)
input_ids = tokenizer.encode(prompt, return_tensors="pt").to('xpu')
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)