unify onnx examples prepare model scripts (#1187)

Signed-off-by: Sun, Xuehao <[email protected]>

examples/onnxrt/body_analysis/onnx_model_zoo/arcface/quantization/ptq_static/README.md

@@ -1,35 +1,26 @@
-Step-by-Step
-============
+# Step-by-Step
 
 This example load a face recognition model from [ONNX Model Zoo](https://github.com/onnx/models) and confirm its accuracy and speed based on [Refined MS-Celeb-1M](https://s3.amazonaws.com/onnx-model-zoo/arcface/dataset/faces_ms1m_112x112.zip).
 
 # Prerequisite
 
 ## 1. Environment
+
 ```shell
 pip install neural-compressor
 pip install -r requirements.txt
 ```
+
 > Note: Validated ONNX Runtime [Version](/docs/source/installation_guide.md#validated-software-environment).
 
 ## 2. Prepare Model
-Download model from [ONNX Model Zoo](https://github.com/onnx/models).
 
 ```shell
-wget https://github.com/onnx/models/raw/main/vision/body_analysis/arcface/model/arcfaceresnet100-8.onnx
-```
-
-Convert opset version to 11 for more quantization capability.
-
-```python
-import onnx
-from onnx import version_converter
-model = onnx.load('arcfaceresnet100-8.onnx')
-model = version_converter.convert_version(model, 11)
-onnx.save_model(model, 'arcfaceresnet100-11.onnx')
+python prepare_model.py --input_model='arcfaceresnet100-8.onnx' --output_model='arcfaceresnet100-11.onnx'
 ```
 
 ## 3. Prepare Dataset
+
 Download dataset [Refined MS-Celeb-1M](https://s3.amazonaws.com/onnx-model-zoo/arcface/dataset/faces_ms1m_112x112.zip).
 
 # Run

examples/onnxrt/body_analysis/onnx_model_zoo/arcface/quantization/ptq_static/prepare_model.py

@@ -0,0 +1,71 @@
+import argparse
+import os
+import sys
+from urllib import request
+
+import onnx
+from onnx import version_converter
+
+MODEL_URL = "https://github.com/onnx/models/raw/main/vision/body_analysis/arcface/model/arcfaceresnet100-8.onnx"
+MAX_TIMES_RETRY_DOWNLOAD = 5
+
+
+def parse_arguments():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--input_model", type=str, required=False, default='arcfaceresnet100-8.onnx')
+    parser.add_argument("--output_model", type=str, required=True)
+    return parser.parse_args()
+
+
+def progressbar(cur, total=100):
+    percent = '{:.2%}'.format(cur / total)
+    sys.stdout.write("\r[%-100s] %s" % ('#' * int(cur), percent))
+    sys.stdout.flush()
+
+
+def schedule(blocknum, blocksize, totalsize):
+    if totalsize == 0:
+        percent = 0
+    else:
+        percent = min(1.0, blocknum * blocksize / totalsize) * 100
+    progressbar(percent)
+
+
+def download_model(url, model_name, retry_times=5):
+    if os.path.isfile(model_name):
+        print(f"{model_name} exists, skip download")
+        return True
+
+    print("download model...")
+    retries = 0
+    while retries < retry_times:
+        try:
+            request.urlretrieve(url, model_name, schedule)
+            break
+        except KeyboardInterrupt:
+            return False
+        except:
+            retries += 1
+            print(f"Download failed{', Retry downloading...' if retries < retry_times else '!'}")
+    return retries < retry_times
+
+
+def export_model(input_model, output_model):
+    # Convert opset version to 14 for more quantization capability.
+    print("\nexport model...")
+    model = onnx.load(input_model)
+    model = version_converter.convert_version(model, 14)
+    onnx.save_model(model, output_model)
+    assert os.path.exists(output_model), f"Export failed! {output_model} doesn't exist!"
+
+
+def prepare_model(input_model, output_model):
+    # Download model from [ONNX Model Zoo](https://github.com/onnx/models).
+    is_download_successful = download_model(MODEL_URL, input_model, MAX_TIMES_RETRY_DOWNLOAD)
+    if is_download_successful:
+        export_model(input_model, output_model)
+
+
+if __name__ == "__main__":
+    args = parse_arguments()
+    prepare_model(args.input_model, args.output_model)

examples/onnxrt/body_analysis/onnx_model_zoo/emotion_ferplus/quantization/ptq_static/README.md

@@ -1,35 +1,26 @@
-Step-by-Step
-============
+# Step-by-Step
 
 This example load an model converted from [ONNX Model Zoo](https://github.com/onnx/models) and confirm its accuracy and speed based on [Emotion FER dataset](https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data).
 
 # Prerequisite
 
 ## 1. Environment
+
 ```shell
 pip install neural-compressor
 pip install -r requirements.txt
 ```
+
 > Note: Validated ONNX Runtime [Version](/docs/source/installation_guide.md#validated-software-environment).
 
 ## 2. Prepare Model
-Download model from [ONNX Model Zoo](https://github.com/onnx/models).
 
 ```shell
-wget https://github.com/onnx/models/raw/main/vision/body_analysis/emotion_ferplus/model/emotion-ferplus-8.onnx
-```
-
-Convert opset version to 12 for more quantization capability.
-
-```python
-import onnx
-from onnx import version_converter
-model = onnx.load('emotion-ferplus-8.onnx')
-model = version_converter.convert_version(model, 12)
-onnx.save_model(model, 'emotion-ferplus-12.onnx')
+python prepare_model.py --input_model='emotion-ferplus-8.onnx' --output_model='emotion-ferplus-12.onnx'
 ```
 
 ## 3. Prepare Dataset
+
 Download dataset [Emotion FER dataset](https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data).
 
 # Run

examples/onnxrt/body_analysis/onnx_model_zoo/emotion_ferplus/quantization/ptq_static/prepare_model.py

@@ -0,0 +1,71 @@
+import argparse
+import os
+import sys
+from urllib import request
+
+import onnx
+from onnx import version_converter
+
+MODEL_URL = "https://github.com/onnx/models/raw/main/vision/body_analysis/emotion_ferplus/model/emotion-ferplus-8.onnx"
+MAX_TIMES_RETRY_DOWNLOAD = 5
+
+
+def parse_arguments():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--input_model", type=str, required=False, default='emotion-ferplus-8.onnx')
+    parser.add_argument("--output_model", type=str, required=True)
+    return parser.parse_args()
+
+
+def progressbar(cur, total=100):
+    percent = '{:.2%}'.format(cur / total)
+    sys.stdout.write("\r[%-100s] %s" % ('#' * int(cur), percent))
+    sys.stdout.flush()
+
+
+def schedule(blocknum, blocksize, totalsize):
+    if totalsize == 0:
+        percent = 0
+    else:
+        percent = min(1.0, blocknum * blocksize / totalsize) * 100
+    progressbar(percent)
+
+
+def download_model(url, model_name, retry_times=5):
+    if os.path.isfile(model_name):
+        print(f"{model_name} exists, skip download")
+        return True
+
+    print("download model...")
+    retries = 0
+    while retries < retry_times:
+        try:
+            request.urlretrieve(url, model_name, schedule)
+            break
+        except KeyboardInterrupt:
+            return False
+        except:
+            retries += 1
+            print(f"Download failed{', Retry downloading...' if retries < retry_times else '!'}")
+    return retries < retry_times
+
+
+def export_model(input_model, output_model):
+    # Convert opset version to 14 for more quantization capability.
+    print("\nexport model...")
+    model = onnx.load(input_model)
+    model = version_converter.convert_version(model, 14)
+    onnx.save_model(model, output_model)
+    assert os.path.exists(output_model), f"Export failed! {output_model} doesn't exist!"
+
+
+def prepare_model(input_model, output_model):
+    # Download model from [ONNX Model Zoo](https://github.com/onnx/models).
+    is_download_successful = download_model(MODEL_URL, input_model, MAX_TIMES_RETRY_DOWNLOAD)
+    if is_download_successful:
+        export_model(input_model, output_model)
+
+
+if __name__ == "__main__":
+    args = parse_arguments()
+    prepare_model(args.input_model, args.output_model)

examples/onnxrt/body_analysis/onnx_model_zoo/ultraface/quantization/ptq_static/README.md

@@ -1,35 +1,26 @@
-Step-by-Step
-============
+# Step-by-Step
 
 This example load an model converted from [ONNX Model Zoo](https://github.com/onnx/models) and confirm its accuracy and speed based on [WIDER FACE dataset (Validation Images)](http://shuoyang1213.me/WIDERFACE/).
 
 # Prerequisite
 
 ## 1. Environment
+
 ```shell
 pip install neural-compressor
 pip install -r requirements.txt
 ```
+
 > Note: Validated ONNX Runtime [Version](/docs/source/installation_guide.md#validated-software-environment).
 
 ## 2. Prepare Model
-Download model from [ONNX Model Zoo](https://github.com/onnx/models).
 
 ```shell
-wget https://github.com/onnx/models/raw/main/vision/body_analysis/ultraface/models/version-RFB-640.onnx
-```
-
-Convert opset version to 12 for more quantization capability.
-
-```python
-import onnx
-from onnx import version_converter
-model = onnx.load('version-RFB-640.onnx')
-model = version_converter.convert_version(model, 12)
-onnx.save_model(model, 'version-RFB-640-12.onnx')
+python prepare_model.py --input_model='version-RFB-640.onnx' --output_model='version-RFB-640-12.onnx'
 ```
 
 ## 3. Prepare Dataset
+
 Download dataset [WIDER FACE dataset (Validation Images)](http://shuoyang1213.me/WIDERFACE/).
 
 # Run

examples/onnxrt/body_analysis/onnx_model_zoo/ultraface/quantization/ptq_static/prepare_model.py

@@ -0,0 +1,71 @@
+import argparse
+import os
+import sys
+from urllib import request
+
+import onnx
+from onnx import version_converter
+
+MODEL_URL = "https://github.com/onnx/models/raw/main/vision/body_analysis/ultraface/models/version-RFB-320.onnx"
+MAX_TIMES_RETRY_DOWNLOAD = 5
+
+
+def parse_arguments():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--input_model", type=str, required=False, default='version-RFB-640.onnx')
+    parser.add_argument("--output_model", type=str, required=True)
+    return parser.parse_args()
+
+
+def progressbar(cur, total=100):
+    percent = '{:.2%}'.format(cur / total)
+    sys.stdout.write("\r[%-100s] %s" % ('#' * int(cur), percent))
+    sys.stdout.flush()
+
+
+def schedule(blocknum, blocksize, totalsize):
+    if totalsize == 0:
+        percent = 0
+    else:
+        percent = min(1.0, blocknum * blocksize / totalsize) * 100
+    progressbar(percent)
+
+
+def download_model(url, model_name, retry_times=5):
+    if os.path.isfile(model_name):
+        print(f"{model_name} exists, skip download")
+        return True
+
+    print("download model...")
+    retries = 0
+    while retries < retry_times:
+        try:
+            request.urlretrieve(url, model_name, schedule)
+            break
+        except KeyboardInterrupt:
+            return False
+        except:
+            retries += 1
+            print(f"Download failed{', Retry downloading...' if retries < retry_times else '!'}")
+    return retries < retry_times
+
+
+def export_model(input_model, output_model):
+    # Convert opset version to 14 for more quantization capability.
+    print("\nexport model...")
+    model = onnx.load(input_model)
+    model = version_converter.convert_version(model, 14)
+    onnx.save_model(model, output_model)
+    assert os.path.exists(output_model), f"Export failed! {output_model} doesn't exist!"
+
+
+def prepare_model(input_model, output_model):
+    # Download model from [ONNX Model Zoo](https://github.com/onnx/models).
+    is_download_successful = download_model(MODEL_URL, input_model, MAX_TIMES_RETRY_DOWNLOAD)
+    if is_download_successful:
+        export_model(input_model, output_model)
+
+
+if __name__ == "__main__":
+    args = parse_arguments()
+    prepare_model(args.input_model, args.output_model)

examples/onnxrt/image_recognition/mobilenet_v2/quantization/ptq_static/README.md

@@ -1,38 +1,22 @@
-Step-by-Step
-============
+# Step-by-Step
 
 This example load an image classification model exported from PyTorch and confirm its accuracy and speed based on [ILSVR2012 validation Imagenet dataset](http://www.image-net.org/challenges/LSVRC/2012/downloads). You need to download this dataset yourself.
 
 # Prerequisite
 
 ## 1. Environment
+
 ```shell
 pip install neural-compressor
 pip install -r requirements.txt
 ```
+
 > Note: Validated ONNX Runtime [Version](/docs/source/installation_guide.md#validated-software-environment).
 
 ## 2. Prepare Model
-Please refer to [pytorch official guide](https://pytorch.org/docs/stable/onnx.html) for detailed model export. The following is a simple example:
-
-```python
-import torch
-import torchvision
-batch_size = 1
-model = torchvision.models.mobilenet_v2(pretrained=True)
-x = torch.randn(batch_size, 3, 224, 224)
-
-# Export the model
-torch.onnx.export(model,               # model being run
-                  x,                         # model input (or a tuple for multiple inputs)
-                  "mobilenet_v2.onnx",           # where to save the model (can be a file or file-like object)
-                  export_params=True,        # store the trained parameter weights inside the model file
-                  opset_version=13,          # the ONNX version to export the model to, please ensure at least 11.
-                  do_constant_folding=True,  # whether to execute constant folding for optimization
-                  input_names = ['input'],   # the model's input names
-                  output_names = ['output'], # the model's output names
-                  dynamic_axes={'input' : {0 : 'batch_size'},    # variable length axes
-                                'output' : {0 : 'batch_size'}})
+
+```shell
+python prepare_model.py --output_model='mobilenet_v2.onnx'
 ```
 
 ## 3. Prepare Dataset
@@ -69,7 +53,6 @@ bash run_quant.sh --input_model=path/to/model \  # model path as *.onnx
                    --quant_format=QDQ
 ```
 
-
 ## 2. Benchmark
 
 ```bash
@@ -79,4 +62,3 @@ bash run_benchmark.sh --input_model=path/to/model \  # model path as *.onnx
                       --batch_size=batch_size \
                       --mode=performance # or accuracy
 ```
-