Add Qwen2.5 GPU example (#12101)
* Add Qwen2.5 GPU example * fix end line * fix description
This commit is contained in:
Jin, Qiao
GitHub
parent
b36359e2ab
commit
db7500bfd4
4 changed files with 257 additions and 2 deletions
|
|
@ -275,6 +275,7 @@ Over 50 models have been optimized/verified on `ipex-llm`, including *LLaMA/LLaM
|
|||
| Qwen | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/qwen) | [link](python/llm/example/GPU/HuggingFace/LLM/qwen) |
|
||||
| Qwen1.5 | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/qwen1.5) | [link](python/llm/example/GPU/HuggingFace/LLM/qwen1.5) |
|
||||
| Qwen2 | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/qwen2) | [link](python/llm/example/GPU/HuggingFace/LLM/qwen2) |
|
||||
| Qwen2.5 | | [link](python/llm/example/GPU/HuggingFace/LLM/qwen2.5) |
|
||||
| Qwen-VL | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/qwen-vl) | [link](python/llm/example/GPU/HuggingFace/Multimodal/qwen-vl) |
|
||||
| Qwen2-Audio | | [link](python/llm/example/GPU/HuggingFace/Multimodal/qwen2-audio) |
|
||||
| Aquila | [link](python/llm/example/CPU/HF-Transformers-AutoModels/Model/aquila) | [link](python/llm/example/GPU/HuggingFace/LLM/aquila) |
|
||||
|
|
|
|||
164
python/llm/example/GPU/HuggingFace/LLM/qwen2.5/README.md
Normal file
164
python/llm/example/GPU/HuggingFace/LLM/qwen2.5/README.md
Normal file
|
|
@ -0,0 +1,164 @@
|
|||
# Qwen2.5
|
||||
In this directory, you will find examples on how you could apply IPEX-LLM INT4 optimizations on Qwen2.5 models on [Intel GPUs](../../../README.md). For illustration purposes, we utilize [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct), [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) and [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct) as reference Qwen2.5 models.
|
||||
|
||||
## 0. Requirements
|
||||
To run these examples with IPEX-LLM on Intel GPUs, we have some recommended requirements for your machine, please refer to [here](../../../README.md#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 Qwen2.5 model to predict the next N tokens using `generate()` API, with IPEX-LLM INT4 optimizations on Intel GPUs.
|
||||
### 1. Install
|
||||
#### 1.1 Installation on Linux
|
||||
We suggest using conda to manage environment:
|
||||
```bash
|
||||
conda create -n llm python=3.11
|
||||
conda activate llm
|
||||
# below command will install intel_extension_for_pytorch==2.1.10+xpu as default
|
||||
pip install --pre --upgrade ipex-llm[xpu] --extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/
|
||||
```
|
||||
|
||||
#### 1.2 Installation on Windows
|
||||
We suggest using conda to manage environment:
|
||||
```bash
|
||||
conda create -n llm python=3.11 libuv
|
||||
conda activate llm
|
||||
|
||||
# below command will install intel_extension_for_pytorch==2.1.10+xpu as default
|
||||
pip install --pre --upgrade ipex-llm[xpu] --extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/
|
||||
```
|
||||
|
||||
### 2. Configures OneAPI environment variables for Linux
|
||||
|
||||
> [!NOTE]
|
||||
> Skip this step if you are running on Windows.
|
||||
|
||||
This is a required step on Linux for APT or offline installed oneAPI. Skip this step for PIP-installed oneAPI.
|
||||
|
||||
```bash
|
||||
source /opt/intel/oneapi/setvars.sh
|
||||
```
|
||||
|
||||
### 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
|
||||
export SYCL_CACHE_PERSISTENT=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 SYCL_CACHE_PERSISTENT=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>
|
||||
|
||||
<details>
|
||||
|
||||
<summary>For Intel iGPU</summary>
|
||||
|
||||
```bash
|
||||
export SYCL_CACHE_PERSISTENT=1
|
||||
export BIGDL_LLM_XMX_DISABLED=1
|
||||
```
|
||||
|
||||
</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™ A-Series Graphics</summary>
|
||||
|
||||
```cmd
|
||||
set SYCL_CACHE_PERSISTENT=1
|
||||
```
|
||||
|
||||
</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 Qwen2.5 model (e.g. `Qwen/Qwen2.5-7B-Instruct`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'Qwen/Qwen2.5-7B-Instruct'`.
|
||||
- `--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`.
|
||||
|
||||
#### Sample Output
|
||||
##### [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
AI是什么？
|
||||
-------------------- Output --------------------
|
||||
AI是Artificial Intelligence的缩写，意为“人工智能”，是指由人制造出来的系统，能够进行类似于人类智慧的行为，如学习、推理
|
||||
```
|
||||
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
What is AI?
|
||||
-------------------- Output --------------------
|
||||
AI, or Artificial Intelligence, refers to the ability exhibited by machines to imitate human behavior and intelligence. This includes learning, problem-solving, perception, understanding language
|
||||
```
|
||||
|
||||
##### [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
AI是什么？
|
||||
-------------------- Output --------------------
|
||||
AI是“人工智能”（Artificial Intelligence）的缩写。它是一门研究如何创建智能机器的学科，这些机器能够执行通常需要人类
|
||||
```
|
||||
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
What is AI?
|
||||
-------------------- Output --------------------
|
||||
Artificial Intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think, learn, and perform tasks that typically require human intelligence.
|
||||
```
|
||||
|
||||
##### [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct)
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
AI是什么？
|
||||
-------------------- Output --------------------
|
||||
AI是“人工智能”的简称，是指由人结合科学原理设计，并通过工程实践创造的能够完成特定任务的软件或硬件系统。这些系统
|
||||
```
|
||||
|
||||
```log
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
What is AI?
|
||||
-------------------- Output --------------------
|
||||
Artificial Intelligence (AI) refers to the development of computer systems that can perform tasks that would typically require human intelligence. These tasks can include things like visual perception
|
||||
```
|
||||
90
python/llm/example/GPU/HuggingFace/LLM/qwen2.5/generate.py
Normal file
90
python/llm/example/GPU/HuggingFace/LLM/qwen2.5/generate.py
Normal file
|
|
@ -0,0 +1,90 @@
|
|||
#
|
||||
# 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
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser(description='Predict Tokens using generate() API for Qwen2.5 model')
|
||||
parser.add_argument('--repo-id-or-model-path', type=str, default="Qwen/Qwen2.5-7B-Instruct",
|
||||
help='The huggingface repo id for the Qwen2.5 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
|
||||
|
||||
|
||||
from ipex_llm.transformers import AutoModelForCausalLM
|
||||
# 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,
|
||||
optimize_model=True,
|
||||
trust_remote_code=True,
|
||||
use_cache=True)
|
||||
model = model.half().to("xpu")
|
||||
|
||||
# Load tokenizer
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_path,
|
||||
trust_remote_code=True)
|
||||
|
||||
prompt = args.prompt
|
||||
|
||||
# Generate predicted tokens
|
||||
with torch.inference_mode():
|
||||
# The following code for generation is adapted from https://huggingface.co/Qwen/Qwen2.5-7B-Instruct#quickstart
|
||||
messages = [
|
||||
{"role": "system", "content": "You are a helpful assistant."},
|
||||
{"role": "user", "content": prompt}
|
||||
]
|
||||
text = tokenizer.apply_chat_template(
|
||||
messages,
|
||||
tokenize=False,
|
||||
add_generation_prompt=True
|
||||
)
|
||||
model_inputs = tokenizer([text], return_tensors="pt").to("xpu")
|
||||
# warmup
|
||||
generated_ids = model.generate(
|
||||
model_inputs.input_ids,
|
||||
max_new_tokens=args.n_predict
|
||||
)
|
||||
|
||||
st = time.time()
|
||||
generated_ids = model.generate(
|
||||
model_inputs.input_ids,
|
||||
max_new_tokens=args.n_predict
|
||||
)
|
||||
torch.xpu.synchronize()
|
||||
end = time.time()
|
||||
generated_ids = generated_ids.cpu()
|
||||
generated_ids = [
|
||||
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
||||
]
|
||||
|
||||
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
||||
print(f'Inference time: {end-st} s')
|
||||
print('-'*20, 'Prompt', '-'*20)
|
||||
print(prompt)
|
||||
print('-'*20, 'Output', '-'*20)
|
||||
print(response)
|
||||
|
|
@ -135,7 +135,7 @@ AI, or Artificial Intelligence, refers to the simulation of human intelligence i
|
|||
|
||||
##### [Qwen/Qwen2-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2-1.5B-Instruct)
|
||||
```log
|
||||
Inference time: 0.33887791633605957 s
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
AI是什么？
|
||||
-------------------- Output --------------------
|
||||
|
|
@ -143,7 +143,7 @@ AI是人工智能的简称，是一种计算机科学和技术领域，旨在使
|
|||
```
|
||||
|
||||
```log
|
||||
Inference time: 0.340407133102417 s
|
||||
Inference time: xxxx s
|
||||
-------------------- Prompt --------------------
|
||||
What is AI?
|
||||
-------------------- Output --------------------
|
||||
|
|
|
|||
Loading…
Reference in a new issue