Add Stable Diffusion examples on GPU and CPU (#11166)

* add sdxl and lcm-lora * readme * modify * add cpu * add license * modify * add file
intel-analytics · Jun 12, 2024 · e7a4e22 · e7a4e22
1 parent f224e98
commit e7a4e22
Show file tree

Hide file tree

Showing 8 changed files with 371 additions and 2 deletions.
diff --git a/python/llm/example/CPU/README.md b/python/llm/example/CPU/README.md
@@ -12,7 +12,7 @@ This folder contains examples of running IPEX-LLM on Intel CPU:
 - [Native-Models](Native-Models): converting & running LLM in `llama`/`chatglm`/`bloom`/`gptneox`/`starcoder` model family using native (cpp) implementation
 - [Speculative-Decoding](Speculative-Decoding): running any ***Hugging Face Transformers*** model with ***self-speculative decoding*** on Intel CPUs
 - [ModelScope-Models](ModelScope-Models): running ***ModelScope*** model with IPEX-LLM on Intel CPUs
-
+- [StableDiffusion-Models](StableDiffusion): running **stable diffusion** models on Intel CPUs. 
 
 ## System Support
 **Hardware**:

diff --git a/python/llm/example/CPU/StableDiffusion/README.md b/python/llm/example/CPU/StableDiffusion/README.md
@@ -0,0 +1,45 @@
+# Stable Diffusion
+In this directory, you will find examples on how to run StableDiffusion models on CPU.
+
+### 1. Installation
+#### 1.1. Install IPEX-LLM
+Follow the instructions in [IPEX-LLM CPU installation guide](https://ipex-llm.readthedocs.io/en/latest/doc/LLM/Overview/install_cpu.html) to install ipex-llm. We recommend to use miniconda to manage your python environment.
+
+#### 1.2 Install dependencies for Stable Diffusion
+Assume you have created a conda environment named diffusion with ipex-llm installed. Run below commands to install dependencies for running Stable Diffusion.
+```bash
+conda activate diffusion
+pip install diffusers["torch"] transformers
+pip install -U PEFT transformers
+pip install setuptools==69.5.1
+```
+
+### 2. Examples
+
+#### 2.1 StableDiffusion XL Example
+The example shows how to run StableDiffusion XL example on Intel CPU.
+```bash
+python ./sdxl.py
+```
+
+Arguments info:
+- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the stable diffusion xl model (e.g. `stabilityai/stable-diffusion-xl-base-1.0`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'stabilityai/stable-diffusion-xl-base-1.0'`.
+- `--prompt PROMPT`: argument defining the prompt to be infered. It is default to be `'An astronaut in the forest, detailed, 8k'`.
+- `--save-path`: argument defining the path to save the generated figure. It is default to be `sdxl-cpu.png`.
+- `--num-steps`: argument defining the number of inference steps. It is default to be `20`. 
+
+The sample output image looks like below. 
+![image](https://llm-assets.readthedocs.io/en/latest/_images/sdxl-cpu.png)
+
+#### 4.2 LCM-LoRA Example
+The example shows how to performing inference with LCM-LoRA on Intel CPU.
+```bash
+python ./lora-lcm.py
+```
+
+Arguments info:
+- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the stable diffusion xl model (e.g. `stabilityai/stable-diffusion-xl-base-1.0`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'stabilityai/stable-diffusion-xl-base-1.0'`.
+- `--lora-weights-path`: argument defining the huggingface repo id for the LCM-LoRA model (e.g. `latent-consistency/lcm-lora-sdxl`) to be downloaded, or the path to huggingface checkpoint folder. It is default to be `'latent-consistency/lcm-lora-sdxl'`. 
+- `--prompt PROMPT`: argument defining the prompt to be infered. It is default to be `'A lovely dog on the table, detailed, 8k'`.
+- `--save-path`: argument defining the path to save the generated figure. It is default to be `lcm-lora-sdxl-cpu.png`.
+- `--num-steps`: argument defining the number of inference steps. It is default to be `4`.
diff --git a/python/llm/example/CPU/StableDiffusion/lora-lcm.py b/python/llm/example/CPU/StableDiffusion/lora-lcm.py
@@ -0,0 +1,55 @@
+#
+# 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.
+#
+# Code is adapted from https://huggingface.co/docs/diffusers/main/en/using-diffusers/inference_with_lcm_lora
+
+import torch
+from diffusers import DiffusionPipeline, LCMScheduler
+import ipex_llm
+import argparse
+
+
+def main(args):
+    pipe = DiffusionPipeline.from_pretrained(
+        args.repo_id_or_model_path,
+        torch_dtype=torch.bfloat16,
+    ).to("cpu")
+
+    # set scheduler
+    pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
+
+    # load LCM-LoRA
+    pipe.load_lora_weights(args.lora_weights_path)
+
+    generator = torch.manual_seed(42)
+    image = pipe(
+        prompt=args.prompt, num_inference_steps=args.num_steps, generator=generator, guidance_scale=1.0
+    ).images[0]
+    image.save(args.save_path)
+
+if __name__=="__main__":
+    parser = argparse.ArgumentParser(description="Stable Diffusion lora-lcm")
+    parser.add_argument('--repo-id-or-model-path', type=str, default="stabilityai/stable-diffusion-xl-base-1.0",
+                        help='The huggingface repo id for the stable diffusion model checkpoint')
+    parser.add_argument('--lora-weights-path',type=str,default="latent-consistency/lcm-lora-sdxl",
+                        help='The huggingface repo id for the lcm lora sdxl checkpoint')
+    parser.add_argument('--prompt', type=str, default="A lovely dog on the table, detailed, 8k",
+                        help='Prompt to infer')
+    parser.add_argument('--save-path',type=str,default="lcm-lora-sdxl-cpu.png",
+                        help="Path to save the generated figure")
+    parser.add_argument('--num-steps',type=int,default=4,
+                        help="Number of inference steps")
+    args = parser.parse_args()
+    main(args)
diff --git a/python/llm/example/CPU/StableDiffusion/sdxl.py b/python/llm/example/CPU/StableDiffusion/sdxl.py
@@ -0,0 +1,47 @@
+#
+# 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.
+#
+# Code is adapted from https://huggingface.co/docs/diffusers/en/using-diffusers/sdxl
+
+from diffusers import AutoPipelineForText2Image
+import torch
+import ipex_llm
+import numpy as np
+from PIL import Image
+import argparse
+
+
+def main(args):
+    pipeline_text2image = AutoPipelineForText2Image.from_pretrained(
+        args.repo_id_or_model_path, 
+        torch_dtype=torch.float16, 
+        use_safetensors=True
+    ).to("cpu")
+
+    image = pipeline_text2image(prompt=args.prompt,num_inference_steps=args.num_steps).images[0]
+    image.save(args.save_path)
+
+if __name__=="__main__":
+    parser = argparse.ArgumentParser(description="Stable Diffusion")
+    parser.add_argument('--repo-id-or-model-path', type=str, default="stabilityai/stable-diffusion-xl-base-1.0",
+                        help='The huggingface repo id for the stable diffusion model checkpoint')
+    parser.add_argument('--prompt', type=str, default="An astronaut in the forest, detailed, 8k",
+                        help='Prompt to infer')
+    parser.add_argument('--save-path',type=str,default="sdxl-cpu.png",
+                        help="Path to save the generated figure")
+    parser.add_argument('--num-steps',type=int,default=20,
+                        help="Number of inference steps")
+    args = parser.parse_args()
+    main(args)
diff --git a/python/llm/example/GPU/README.md b/python/llm/example/GPU/README.md
@@ -14,7 +14,8 @@ This folder contains examples of running IPEX-LLM on Intel GPU:
 - [PyTorch-Models](PyTorch-Models): running any PyTorch model on IPEX-LLM (with "one-line code change")
 - [Speculative-Decoding](Speculative-Decoding): running any ***Hugging Face Transformers*** model with ***self-speculative decoding*** on Intel GPUs
 - [ModelScope-Models](ModelScope-Models): running ***ModelScope*** model with IPEX-LLM on Intel GPUs
-- [Long-Context](Long-Context): running **long-context** generation with IPEX-LLM on Intel Arc™ A770 Graphics
+- [Long-Context](Long-Context): running **long-context** generation with IPEX-LLM on Intel Arc™ A770 Graphics.
+- [StableDiffusion](StableDiffusion): running **stable diffusion** with IPEX-LLM on Intel GPUs. 
 
 
 ## System Support

diff --git a/python/llm/example/GPU/StableDiffusion/README.md b/python/llm/example/GPU/StableDiffusion/README.md
@@ -0,0 +1,119 @@
+# Stable Diffusion
+In this directory, you will find examples on how to run StableDiffusion models on [Intel GPUs](../README.md).
+
+### 1. Installation
+#### 1.1 Install IPEX-LLM
+Follow the instructions in IPEX-GPU installation guides ([Linux Guide](https://ipex-llm.readthedocs.io/en/latest/doc/LLM/Quickstart/install_linux_gpu.html), [Windows Guide](https://ipex-llm.readthedocs.io/en/latest/doc/LLM/Quickstart/install_windows_gpu.html)) according to your system to install IPEX-LLM. After the installation, you should have created a conda environment, named diffusion for instance. 
+
+#### 1.2 Install dependencies for Stable Diffusion
+Assume you have created a conda environment named diffusion with ipex-llm installed. Run below commands to install dependencies for running Stable Diffusion.
+```bash
+conda activate diffusion
+pip install diffusers["torch"] transformers
+pip install -U PEFT transformers
+```
+
+### 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
+```
+</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. Examples
+
+#### 4.1 StableDiffusion XL Example
+The example shows how to run StableDiffusion XL example on Intel GPU.
+```bash
+python ./sdxl.py
+```
+
+Arguments info:
+- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the stable diffusion xl model (e.g. `stabilityai/stable-diffusion-xl-base-1.0`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'stabilityai/stable-diffusion-xl-base-1.0'`.
+- `--prompt PROMPT`: argument defining the prompt to be infered. It is default to be `'An astronaut in the forest, detailed, 8k'`.
+- `--save-path`: argument defining the path to save the generated figure. It is default to be `sdxl-gpu.png`.
+- `--num-steps`: argument defining the number of inference steps. It is default to be `20`. 
+
+
+The sample output image looks like below. 
+![image](https://llm-assets.readthedocs.io/en/latest/_images/sdxl-gpu.png)
+
+#### 4.2 LCM-LoRA Example
+The example shows how to performing inference with LCM-LoRA on Intel GPU.
+```bash
+python ./lora-lcm.py
+```
+
+Arguments info:
+- `--repo-id-or-model-path REPO_ID_OR_MODEL_PATH`: argument defining the huggingface repo id for the stable diffusion xl model (e.g. `stabilityai/stable-diffusion-xl-base-1.0`) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be `'stabilityai/stable-diffusion-xl-base-1.0'`.
+- `--lora-weights-path`: argument defining the huggingface repo id for the LCM-LoRA model (e.g. `latent-consistency/lcm-lora-sdxl`) to be downloaded, or the path to huggingface checkpoint folder. It is default to be `'latent-consistency/lcm-lora-sdxl'`. 
+- `--prompt PROMPT`: argument defining the prompt to be infered. It is default to be `'A lovely dog on the table, detailed, 8k'`.
+- `--save-path`: argument defining the path to save the generated figure. It is default to be `lcm-lora-sdxl-gpu.png`.
+- `--num-steps`: argument defining the number of inference steps. It is default to be `4`.
diff --git a/python/llm/example/GPU/StableDiffusion/lora-lcm.py b/python/llm/example/GPU/StableDiffusion/lora-lcm.py
@@ -0,0 +1,55 @@
+#
+# 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.
+#
+# Code is adapted from https://huggingface.co/docs/diffusers/main/en/using-diffusers/inference_with_lcm_lora
+
+import torch
+from diffusers import DiffusionPipeline, LCMScheduler
+import ipex_llm
+import argparse
+
+
+def main(args):
+    pipe = DiffusionPipeline.from_pretrained(
+        args.repo_id_or_model_path,
+        torch_dtype=torch.bfloat16,
+    ).to("xpu")
+
+    # set scheduler
+    pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
+
+    # load LCM-LoRA
+    pipe.load_lora_weights(args.lora_weights_path)
+
+    generator = torch.manual_seed(42)
+    image = pipe(
+        prompt=args.prompt, num_inference_steps=args.num_steps, generator=generator, guidance_scale=1.0
+    ).images[0]
+    image.save(args.save_path)
+
+if __name__=="__main__":
+    parser = argparse.ArgumentParser(description="Stable Diffusion lora-lcm")
+    parser.add_argument('--repo-id-or-model-path', type=str, default="stabilityai/stable-diffusion-xl-base-1.0",
+                        help='The huggingface repo id for the stable diffusion model checkpoint')
+    parser.add_argument('--lora-weights-path',type=str,default="latent-consistency/lcm-lora-sdxl",
+                        help='The huggingface repo id for the lcm lora sdxl checkpoint')
+    parser.add_argument('--prompt', type=str, default="A lovely dog on the table, detailed, 8k",
+                        help='Prompt to infer')
+    parser.add_argument('--save-path',type=str,default="lcm-lora-sdxl-gpu.png",
+                        help="Path to save the generated figure")
+    parser.add_argument('--num-steps',type=int,default=4,
+                        help="Number of inference steps")
+    args = parser.parse_args()
+    main(args)
diff --git a/python/llm/example/GPU/StableDiffusion/sdxl.py b/python/llm/example/GPU/StableDiffusion/sdxl.py
@@ -0,0 +1,47 @@
+#
+# 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.
+#
+# Code is adapted from https://huggingface.co/docs/diffusers/en/using-diffusers/sdxl
+
+from diffusers import AutoPipelineForText2Image
+import torch
+import ipex_llm
+import numpy as np
+from PIL import Image
+import argparse
+
+
+def main(args):
+    pipeline_text2image = AutoPipelineForText2Image.from_pretrained(
+        args.repo_id_or_model_path, 
+        torch_dtype=torch.bfloat16, 
+        use_safetensors=True
+    ).to("xpu")
+
+    image = pipeline_text2image(prompt=args.prompt,num_inference_steps=args.num_steps).images[0]
+    image.save(args.save_path)
+
+if __name__=="__main__":
+    parser = argparse.ArgumentParser(description="Stable Diffusion")
+    parser.add_argument('--repo-id-or-model-path', type=str, default="stabilityai/stable-diffusion-xl-base-1.0",
+                        help='The huggingface repo id for the stable diffusion model checkpoint')
+    parser.add_argument('--prompt', type=str, default="An astronaut in the forest, detailed, 8k",
+                        help='Prompt to infer')
+    parser.add_argument('--save-path',type=str,default="sdxl-gpu.png",
+                        help="Path to save the generated figure")
+    parser.add_argument('--num-steps',type=int,default=20,
+                        help="Number of inference steps")
+    args = parser.parse_args()
+    main(args)