Update onnxwrappger samples README

Author: WeidongFeng916

tools/onnxwrapper/README.md

@@ -17,23 +17,18 @@ This tool provides a transparent runtime replacement that allows existing **ONNX
 - Rapid validation of QNN models generated from ONNX or downloaded from **AI Hub**.
 
 ## Usage Overview
-There are two usages for this tool:  
-**Usage 1:**
-1. Copy `onnxexec.py` and `onnxwrapper.py` into your ONNX sample code directory.
-2. Run your original ONNX inference script via:
-
-```bash
-python onnxexec.py your_onnx_sample.py
-```
-**Usage 2:**  
 Import the wrapper explicitly at the beginning of your script:
 
 ```python
 from qai_appbuilder import onnxwrapper
 ```
-
 No other changes to your ONNX inference code are required.
 
+```bash
+python your_onnx_sample.py
+```
+Note: If your onnx sample needs some options, please add them when running above command. 
+
 ## Technical Implementation
 
 ### Overall Architecture

tools/onnxwrapper/Samples/facemap_3dmm_onnx/README.md

@@ -4,21 +4,23 @@
 This is onnx sample code for using QAI AppBuilder to load facemap_3dmm QNN model to HTP and execute 3D face reconstruction inference to generate parametric 3D face models from 2D facial images. The facemap_3dmm_onnx_infer.py is onnx sample code.
 
 ## Run the sample code
-
-```
 If you want to run the sample code with onnx models.
+```bash
 python prepare_facemap_3dmm_onnx_models.py
 python facemap_3dmm_onnx_infer.py --model models-onnx\facemap_3dmm-onnx-float\facemap_3dmm.onnx --image input.jpg
-
+```
 If you want to run the sample code with qnn models.
-python prepare_facemap_3dmm_qnn_models.py
-python onnxexec.py facemap_3dmm_onnx_infer.py --model models-qnn\facemap_3dmm.bin --image input.jpg
-
-You also can add the following code at beginning of facemap_3dmm_onnx_infer.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python facemap_3dmm_onnx_infer.py --model models-qnn\facemap_3dmm.bin --image input.jpg
-
+1. Run the following command.
+```bash
+   python prepare_facemap_3dmm_qnn_models.py
+```
+2. Add the following code at beginning of facemap_3dmm_onnx_infer.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python facemap_3dmm_onnx_infer.py --model models-qnn\facemap_3dmm.bin --image input.jpg
 ```
 ## Output
 You can see output.jpg in out folder.

tools/onnxwrapper/Samples/real_esrgan_x4plus_onnx/README.md

@@ -4,20 +4,24 @@
 This is onnx sample code for using AppBuilder to load real_esrgan_x4plus QNN model to HTP and execute inference to generate image. 
 
 ## Run the sample code
-```
 If you want to run the sample code with onnx models.
+```bash
 python prepare_real_esrgan_x4plus_onnx_models.py
 python real_esrgan_x4plus_onnx_inference.py --model models-onnx\real_esrgan_x4plus-onnx-float\real_esrgan_x4plus.onnx
+```
 
 If you want to run the sample code with qnn models.
-python prepare_real_esrgan_x4plus_qnn_models.py
-python onnxexec.py real_esrgan_x4plus_onnx_inference.py --model models-qnn\real_esrgan_x4plus.bin --tile 512
-
-You also can add the following code at beginning of real_esrgan_x4plus_onnx_inference.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python real_esrgan_x4plus_onnx_inference.py --model models-qnn\real_esrgan_x4plus.bin --tile 512
-
+1. Run the following command.
+```bash
+   python prepare_real_esrgan_x4plus_qnn_models.py
+```
+2. Add the following code at beginning of real_esrgan_x4plus_onnx_inference.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python real_esrgan_x4plus_onnx_inference.py --model models-qnn\real_esrgan_x4plus.bin --tile 512
 ```
 ## Output
 The output image will be saved to output_x4.png

tools/onnxwrapper/Samples/stable_diffusion_1_5_onnx/README.md

@@ -4,20 +4,23 @@
 This is sample code for using AppBuilder to load Stable Diffusion 1.5 QNN models to HTP and execute inference to generate image.
 
 ## Run the sample code
-```
 If you want to run the sample code with onnx models.
+```bash
 python prepeare_stable_diffusion_onnx_models.py
 python stable_diffusion_1_5_onnx_infer.py --model_root models-onnx\modularai_stable-diffusion-1-5-onnx --provider cpu --out sd15_out.png
-
+```
 If you want to run the sample code with qnn models.
-python prepeare_stable_diffusion_qnn_models.py
-python onnxexec.py stable_diffusion_1_5_onnx_infer.py --model_root models-qnn --vae_scale 1.0
-
-You also can add the following code at beginning of stable_diffusion_1_5_onnx_infer.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python stable_diffusion_1_5_onnx_infer.py --model_root models-qnn --vae_scale 1.0
-
+1. Run the following command.
+```bash
+   python prepeare_stable_diffusion_qnn_models.py
+```
+2. Add the following code at beginning of stable_diffusion_1_5_onnx_infer.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python stable_diffusion_1_5_onnx_infer.py --model_root models-qnn --vae_scale 1.0
 ```
 ## Output
 The output image will be saved to sd15_out.png

tools/onnxwrapper/Samples/stable_diffusion_2_1_onnx/README.md

@@ -4,20 +4,23 @@
 This is sample code for using AppBuilder to load Stable Diffusion 2.1 QNN models to HTP and execute inference to generate image.  
 
 ## Run the sample code
-```
 If you want to run the sample code with onnx models.
+```bash
 python prepeare_stable_diffusion_onnx_models.py
 python stable_diffusion_2_1_onnx_infer.py --model_root ./models-onnx/aislamov_stable-diffusion-2-1-base-onnx --provider dml
-
+```
 If you want to run the sample code with qnn models.
-python prepeare_stable_diffusion_qnn_models.py
-python onnxexec.py stable_diffusion_2_1_onnx_infer.py --model_root models-qnn --vae_scale 1.0
-
-You also can add the following code at beginning of stable_diffusion_2_1_onnx_infer.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python stable_diffusion_2_1_onnx_infer.py --model_root models-qnn --vae_scale 1.0
-
+1. Run the following command.
+```bash
+   python prepeare_stable_diffusion_qnn_models.py
+```
+2. Add the following code at beginning of stable_diffusion_2_1_onnx_infer.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python stable_diffusion_2_1_onnx_infer.py --model_root models-qnn --vae_scale 1.0
 ```
 ## Output
 The output image will be saved to sd21_out.png

tools/onnxwrapper/Samples/whisper_base_en_onnx/README.md

@@ -4,20 +4,24 @@
 This is onnx sample code for using AppBuilder to load whisper_base_en QNN model to HTP and execute inference. 
 
 ## Run the sample code
-```
 If you want to run the sample code with onnx models.
+```bash
 python prepare_whisper_onnx_models.py
 python whisper_base_en_onnx_infer.py --audio_file jfk.wav --encoder_onnx models-onnx\base.en-encoder.onnx --decoder_onnx models-onnx\base.en-decoder.onnx --mel_filters mel_filters.npz
+```
 
 If you want to run the sample code with qnn models.
-python prepare_whisper_qnn_models.py
-python onnxexec.py whisper_base_en_onnx_infer.py --audio_file jfk.wav --encoder_onnx models-qnn\whisper_base_en-whisperencoder-snapdragon_x_elite.bin --decoder_onnx models-qnn\whisper_base_en-whisperdecoder-snapdragon_x_elite.bin --mel_filters mel_filters.npz
-
-You also can add the following code at beginning of whisper_base_en_onnx_infer.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python whisper_base_en_onnx_infer.py --audio_file jfk.wav --encoder_onnx models-qnn\whisper_base_en-whisperencoder-snapdragon_x_elite.bin --decoder_onnx models-qnn\whisper_base_en-whisperdecoder-snapdragon_x_elite.bin --mel_filters mel_filters.npz
-
+1. Run the following command.
+```bash
+   python prepare_whisper_qnn_models.py
+```
+2. Add the following code at beginning of whisper_base_en_onnx_infer.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python whisper_base_en_onnx_infer.py --audio_file jfk.wav --encoder_onnx models-qnn\whisper_base_en-whisperencoder-snapdragon_x_elite.bin --decoder_onnx models-qnn\whisper_base_en-whisperdecoder-snapdragon_x_elite.bin --mel_filters mel_filters.npz
 ```
 ## Output
 You can see "Transcription: And so my fellow Americans, ask not what your country can do for you, ask what you can do for your country." in log.

tools/onnxwrapper/Samples/yolov8_det_onnx/README.md

@@ -4,19 +4,24 @@
 This is onnx sample code for using AppBuilder to load yolov8_det QNN model to HTP and execute inference to predicts bounding boxes and classes of objects in an image.
 
 ## Run the following command:
-
 If you want to run the sample code with onnx models.
+```bash
 python prepare_yolov8_det_onnx_models.py
 python yolov8_det_onnx_inference.py --model models-onnx\yolov8n.onnx
+```
 
 If you want to run the sample code with qnn models.
-python prepare_yolov8_det_qnn_models.py
-python onnxexec.py yolov8_det_onnx_inference.py --model models-qnn\yolov8_det.bin
-
-You also can add the following code at beginning of yolov8_det_onnx_inference.py.
-from qai_appbuilder import onnxwrapper
-Then run the following command.
-python yolov8_det_onnx_inference.py --model models-qnn\yolov8_det.bin
-
+1. Run the following command.
+```bash
+   python prepare_yolov8_det_qnn_models.py
+```
+2. Add the following code at beginning of yolov8_det_onnx_inference.py.
+```python
+   from qai_appbuilder import onnxwrapper
+```
+3. Then run the following command.
+```bash
+   python yolov8_det_onnx_inference.py --model models-qnn\yolov8_det.bin
+```
 ## Output
 The output image will be saved to output.jpg