search

krai / axs2mlperf

Automated KRAI X workflows for reproducing MLPerf Inference submissions
https://krai.ai
MIT License
1 stars 1 forks source link

Move `num_gpu` and parameters which use it to the base class `nvidia_gpu_support` and add `supported_execution_providers` #8

Closed maria-18-git closed 1 year ago

maria-18-git commented 1 year ago
maria-18-git commented 1 year ago

Updated and check:

maria@chai ~/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py (master *=)$ axs byname object_detection_using_onnxrt_py , get onnxruntime_name
WARNING:root:[kilt-mlperf-dev] parameters file /home/maria/work_collection/kilt-mlperf-dev/data_axs.json did not exist, initializing to empty parameters
WARNING:root:shell.run() about to execute (with env=None, in_dir=None, capture_output=True, errorize_output=False, capture_stderr=False, split_to_lines=False):
        "/usr/bin/nvidia-smi" -L | grep -c GPU
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxruntime-gpu
maria@chai ~/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py (master *=)$ axs byname object_detection_using_onnxrt_py , get num_gpus
WARNING:root:[kilt-mlperf-dev] parameters file /home/maria/work_collection/kilt-mlperf-dev/data_axs.json did not exist, initializing to empty parameters
WARNING:root:shell.run() about to execute (with env=None, in_dir=None, capture_output=True, errorize_output=False, capture_stderr=False, split_to_lines=False):
        "/usr/bin/nvidia-smi" -L | grep -c GPU
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2

Run accuracy:

maria@chai ~/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py (master *=)$ axs byquery program_output,task=object_detection,framework=onnxrt
...
2023-09-25 11:38:22.837671084 [W:onnxruntime:, graph.cc:1283 Graph] Initializer backbone.model.layer2.0.5.conv2.weight appears in graph inputs and will not be treated as constant value/weight. This may prevent some of the graph optimizations, like const folding. Move it out of graph inputs if there is no need to override it, by either re-generating the model with latest exporter/converter or with the tool onnxruntime/tools/python/remove_initializer_from_input.py.                                                                                                                2023-09-25 11:38:29.080452825 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 10:38:29 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.                                                                                                        2023-09-25 11:38:29.080492523 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 10:38:29 WARNING] onnx2trt_utils.cpp:400: One or more weights outside the range of INT32 was clamped                                                                                                                                                                                      2023-09-25 11:38:29.249120983 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 10:38:29 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.                                                                                                        2023-09-25 11:38:29.249156886 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 10:38:29 WARNING] onnx2trt_utils.cpp:400: One or more weights outside the range of INT32 was clamped
2023-09-25 11:38:29.436545121 [E:onnxruntime:, inference_session.cc:1785 operator()] Exception during initialization: /onnxruntime_src/onnxruntime/core/providers/tensorrt/tensorrt_execution_provider.cc:1552 SubGraphCollection_t onnxruntime::TensorrtExecutionProvider::GetSupportedList(SubGraphCollection_t, int, int, const onnxruntime::GraphViewer&, bool*) const [ONNXRuntimeError] : 1 : FAIL : TensorRT input: NonMaxSuppression_683 has no shape specified. Please run shape inference on the onnx model first. Details can be found in https://onnxruntime.ai/docs/execution-providers/TensorRT-ExecutionProvider.html#shape-inference-for-tensorrt-subgraphs

Traceback (most recent call last):
  File "/home/maria/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py/onnx_detect.py", line 257, in <module>                                                main()                                                                                                                                                                                         File "/home/maria/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py/onnx_detect.py", line 110, in main
    sess = rt.InferenceSession(model_path, sess_options, providers= [requested_provider] if execution_device else rt.get_available_providers())
  File "/home/maria/work_collection/onnxruntime-gpu_package_for_python3.9/install/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 419, in __init__           self._create_inference_session(providers, provider_options, disabled_optimizers)
  File "/home/maria/work_collection/onnxruntime-gpu_package_for_python3.9/install/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 471, in _create_inference_session
    sess.initialize_session(providers, provider_options, disabled_optimizers)
onnxruntime.capi.onnxruntime_pybind11_state.RuntimeException: [ONNXRuntimeError] : 6 : RUNTIME_EXCEPTION : Exception during initialization: /onnxruntime_src/onnxruntime/core/providers/tensorrt/tensorrt_execution_provider.cc:1552 SubGraphCollection_t onnxruntime::TensorrtExecutionProvider::GetSupportedList(SubGraphCollection_t, int, int, const onnxruntime::GraphViewer&, bool*) const [ONNXRuntimeError] : 1 : FAIL : TensorRT input: NonMaxSuppression_683 has no shape specified. Please run shape inference on the onnx model first. Details can be found in https://onnxruntime.ai/docs/execution-providers/TensorRT-ExecutionProvider.html#shape-inference-for-tensorrt-subgraphs
maria-18-git commented 1 year ago

On chai we have gpu, so we can you cuda and tensorrt. But now we support only cuda in our programs. So we need to run on chai setting execution_device to cpu, gpu or cuda.

maria@chai ~/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py (master *=)$ axs byquery program_output,task=object_detection,framework=onnxrt,execution_device=cuda , get accuracy
WARNING:root:[kilt-mlperf-dev] parameters file /home/maria/work_collection/kilt-mlperf-dev/data_axs.json did not exist, initializing to empty parameters
WARNING:root:[base_object_detection_experiment] touch _BEFORE_CODE_LOADING=/home/maria/work_collection/pycocotools_package_for_python3.9/install/lib/python3.9/site-packages
loading annotations into memory...
Done (t=0.36s)
creating index...
index created!
Loading and preparing results...
DONE (t=0.00s)
creating index...
index created!
Running per image evaluation...
Evaluate annotation type *bbox*
DONE (t=0.18s).
Accumulating evaluation results...
DONE (t=0.11s).
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.230
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.402
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.232
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.155
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.467
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.316
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.247
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.333
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.347
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.209
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.540
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.408
0.23015432623401041
maria@chai ~/axs/core_collection/workflows_collection/object_detection/object_detection_using_onnxrt_py (master *=)$ axs byquery program_output,task=object_detection,framework=onnxrt,execution_device=gpu , get accuracy
WARNING:root:[kilt-mlperf-dev] parameters file /home/maria/work_collection/kilt-mlperf-dev/data_axs.json did not exist, initializing to empty parameters
WARNING:root:[base_object_detection_experiment] touch _BEFORE_CODE_LOADING=/home/maria/work_collection/pycocotools_package_for_python3.9/install/lib/python3.9/site-packages
loading annotations into memory...
Done (t=0.38s)
creating index...
index created!
Loading and preparing results...
DONE (t=0.00s)
creating index...
index created!
Running per image evaluation...
Evaluate annotation type *bbox*
DONE (t=0.18s).
Accumulating evaluation results...
DONE (t=0.13s).
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.230
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.403
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.232
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.155
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.467
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.316
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.247
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.333
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.347
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.209
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.540
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.408
0.23014983711985593
maria-18-git commented 1 year ago 
maria@chai ~/axs/core_collection/workflows_collection/image_classification/image_classification_using_onnxrt_py (master *=)$ axs byname image_classification_using_onnxrt_py , run $ONNXRUNTIME_QUERY_MOD ---capture_output=false --output_file_path=
...
2023-09-25 13:53:58.637326721 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:53:58 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
2023-09-25 13:53:58.644886943 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:53:58 WARNING] Tensor DataType is determined at build time for tensors not marked as input or output.
2023-09-25 13:54:05.240398741 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:54:05 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
2023-09-25 13:54:05.242841622 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:54:05 WARNING] Tensor DataType is determined at build time for tensors not marked as input or output.
Session execution provider:  ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'AzureExecutionProvider', 'CPUExecutionProvider']
Device: GPU
input_layer_names=['input_tensor:0']
output_layer_name=softmax_tensor:0
model_input_shape=['unk__271', 3, 224, 224]
model_output_shape=['unk__273', 1001]
batch 1/1: (1..20) [65, 795, 230, 809, 516, 67, 334, 415, 674, 332, 109, 286, 370, 757, 595, 147, 327, 23, 478, 517]
0

Accuracy:

maria@chai ~/axs/core_collection/workflows_collection/image_classification/image_classification_using_onnxrt_py (master *=)$ axs byquery program_output,task=image_classification,framework=onnxrt,num_of_images=32 , get accuracy
...
2023-09-25 13:56:50.357322822 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:56:50 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
2023-09-25 13:56:50.360067336 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:56:50 WARNING] Tensor DataType is determined at build time for tensors not marked as input or output.
2023-09-25 13:56:56.544946851 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:56:56 WARNING] onnx2trt_utils.cpp:374: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
2023-09-25 13:56:56.547443802 [W:onnxruntime:Default, tensorrt_execution_provider.h:77 log] [2023-09-25 12:56:56 WARNING] Tensor DataType is determined at build time for tensors not marked as input or output.
Session execution provider:  ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'AzureExecutionProvider', 'CPUExecutionProvider']
Device: GPU
input_layer_names=['input_tensor:0']
output_layer_name=softmax_tensor:0
model_input_shape=['unk__271', 3, 224, 224]
model_output_shape=['unk__273', 1001]
batch 1/2: (1..25) [65, 795, 230, 809, 516, 67, 334, 415, 674, 332, 109, 286, 370, 757, 595, 147, 327, 23, 478, 517, 334, 176, 948, 727, 23]
...
0.84375
maria-18-git commented 1 year ago 
maria@chai ~/axs/core_collection/workflows_collection/image_classification (master *=)$ axs byquery program_output,task=image_classification,framework=pytorch,num_of_images=32 , get accuracy
...
Device: GPU
Using cache found in /home/maria/.cache/torch/hub/pytorch_vision_v0.12.0
batch 1/2: (1..25) [65, 795, 230, 809, 520, 65, 334, 852, 674, 332, 109, 286, 370, 757, 595, 147, 327, 23, 478, 517, 334, 172, 948, 727, 23]
batch 2/2: (26..32) [583, 270, 264, 55, 538, 324, 573]
...
0.71875
maria-18-git commented 1 year ago 
maria@chai ~/axs/core_collection/workflows_collection/bert (master *=)$ echo $BERT_DEMO_OUTPUT
Question_1: Which country has Moscow as the capital? Answer_1: the soviet union Question_2: How old is the capital of the Soviet Union? Answer_2: more than 800 years Question_3: Where is the Bolshoi Theater? Answer_3: moscow Question_4: How many museums are there in the capital of the Soviet Union? Answer_4: 150 Question_5: What is the Kremlin? Answer_5: a fortress surrounded by red stone walls Question_6: Where is the Kremlin? Answer_6: in the heart of moscow Question_7: What is inside the Kremlin? Answer_7: palaces , cathedrals and buildings housing the seat of the soviet government Question_8: What colour are the stones of the Kremlin? Answer_8: red
maria-18-git commented 1 year ago
  1. bert_using_onnxrt_py 
    
maria@chai ~/axs/core_collection/workflows_collection/bert (master *=)$ export ACCURACY_OUTPUT=`axs byquery program_output,task=bert,framework=onnxrt,batch_count=20 , get accuracy_dict`

maria@chai ~/axs/core_collection/workflows_collection/bert (master *=)$ echo "Accuracy: $ACCURACY_OUTPUT" Accuracy: {'exact_match': 85.0, 'f1': 85.0}

maria-18-git commented 1 year ago

axs2mlperf:

  1. image_classification_using_onnxrt_loadgen 
    maria@chai ~/work_collection/axs2mlperf/image_classification_using_onnxrt_loadgen (master *=)$ axs byquery loadgen_output,task=image_classification,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,model_name=resnet50,loadgen_dataset_size=20,loadgen_buffer_size=100,execution_device=cpu
...
Session execution provider:  ['CPUExecutionProvider']
Device: CPU
load_query_samples([8, 1, 17, 12, 11, 10, 5, 15, 19, 9, 18, 6, 7, 2, 13, 14, 16, 4, 0, 3])
B20llllllllllllllllllll
Q20[batch of 1] inference=10.21 ms
...

    Accuracy:

    
maria@chai ~/work_collection/axs2mlperf/image_classification_using_onnxrt_loadgen (master *=)$ export ACCURACY_OUTPUT=`axs byquery loadgen_output,task=image_classification,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,model_name=resnet50,loadgen_dataset_size=20,loadgen_buffer_size=100,execution_device=cpu , get accuracy `

maria@chai ~/work_collection/axs2mlperf/image_classification_using_onnxrt_loadgen (master *=)$ echo "Accuracy: $ACCURACY_OUTPUT" Accuracy: 85.0

GPU:

maria@chai ~/work_collection/axs2mlperf/image_classification_using_onnxrt_loadgen (master *=)$ axs byquery loadgen_output,task=image_classification,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,model_name=resnet50,loadgen_dataset_size=20,loadgen_buffer_size=100,execution_device=gpu , get accuracy ... Session execution provider: ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'AzureExecutionProvider', 'CPUExecutionProvider'] Device: GPU load_query_samples([8, 1, 17, 12, 11, 10, 5, 15, 19, 9, 18, 6, 7, 2, 13, 14, 16, 4, 0, 3]) B20llllllllllllllllllll ... 85.0

maria-18-git commented 1 year ago
  1. image_classification_using_torch_loadgen 
    
maria@chai ~/work_collection/axs2mlperf/image_classification_using_torch_loadgen (master *=)$ export ACCURACY_OUTPUT=`axs byquery loadgen_output,task=image_classification,framework=pytorch,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,model_name=resnet50,loadgen_dataset_size=20,loadgen_buffer_size=100,execution_device=cpu , get accuracy`

maria@chai ~/work_collection/axs2mlperf/image_classification_using_torch_loadgen (master *=)$ echo "Accuracy: $ACCURACY_OUTPUT" Accuracy: 75.0

maria-18-git commented 1 year ago
  1. object_detection_using_onnxrt_loadgen 
    
maria@chai ~/work_collection/axs2mlperf/object_detection_using_onnxrt_loadgen (master *=)$ export ACCURACY_OUTPUT=`axs byquery loadgen_output,task=object_detection,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,model_name=ssd_resnet34,loadgen_dataset_size=20,loadgen_buffer_size=100,execution_device=cpu , get accuracy`

maria@chai ~/work_collection/axs2mlperf/object_detection_using_onnxrt_loadgen (master *=)$ echo "Accuracy: $ACCURACY_OUTPUT" Accuracy: 22.852

maria-18-git commented 1 year ago
  1. object_detection_using_onnxrt_loadgen 
    maria@chai ~/work_collection/axs2mlperf/object_detection_using_onnxrt_loadgen (master *=)$ axs byquery loadgen_output,task=bert,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,execution_device=cpu
...
Session execution provider:  ['CPUExecutionProvider']
Device: CPU
Loading tokenized SQuAD dataset as features from /home/maria/work_collection/preprocessed_squad_v1_1_msl_384_calibration_no/preprocessed_squad_v1.1.pickled ...
2023-09-25 23:56:46.456661: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2023-09-25 23:56:46.490513: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
Example width: 384
...

    Accuracy:

    
maria@chai ~/work_collection/axs2mlperf/object_detection_using_onnxrt_loadgen (master *=)$ export ACCURACY_OUTPUT=`axs byquery loadgen_output,task=bert,framework=onnxrt,loadgen_scenario=Offline,loadgen_mode=AccuracyOnly,execution_device=cpu , get accuracy_dict`

maria@chai ~/work_collection/axs2mlperf/object_detection_using_onnxrt_loadgen (master *=)$ echo "Accuracy: $ACCURACY_OUTPUT" Accuracy: {'exact_match': 85.0, 'f1': 85.0}

maria-18-git commented 1 year ago

Commits:

maria-18-git commented 1 year ago

New updates (supporting of supported_execution_providers):

maria-18-git commented 1 year ago

Check this update for object_detection_using_onnxrt_py:

maria@chai ~(master *=)$ axs byquery program_output,task=object_detection,framework=onnxrt,execution_device=cuda , get accuracy
...
Session execution provider:  ['CUDAExecutionProvider', 'CPUExecutionProvider']
Device: GPU
inp.name=image , inp.shape=[1, 3, 1200, 1200] , inp.type=tensor(float)
output.name=bboxes , output.shape=[1, 'nbox', 4] , output.type=tensor(float)
output.name=labels , output.shape=[1, 'nbox'] , output.type=tensor(int64)
output.name=scores , output.shape=[1, 'nbox'] , output.type=tensor(float)
Data layout: NCHW                                                                                                                                                                           Input layers: ['image']
Output layers: ['bboxes', 'labels', 'scores']
Input layer name: image
Expected input shape: [1, 3, 1200, 1200]
Expected input type: <class 'numpy.float32'>
Output layer names: ['bboxes', 'labels', 'scores']
Background/unlabelled classes to skip: 1

[batch 1 of 20] loading=300.20 ms, inference=4380.72 ms
...
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.230
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.402
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.232
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.155
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.467
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.316
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.247
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.333
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.347
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.209
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.540
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.408
0.23015432623401041
maria-18-git commented 1 year ago
maria-18-git commented 1 year ago

image_classification_using_onnxrt_py

maria-18-git commented 1 year ago

bert_using_onnxrt_py

maria-18-git commented 1 year ago

Commits:

maria-18-git commented 1 year ago

object_detection_using_onnxrt_loadgen

maria-18-git commented 1 year ago

image_classification_using_onnxrt_loadgen

maria-18-git commented 1 year ago

bert_using_onnxrt_loadgen

maria-18-git commented 1 year ago

Commits: in axs2mlperf: Added supported_execution_providers for object_detection_using_onnxrt… Added supported_execution_providers for image_classification_using_on… Added supported_execution_providers for bert_using_onnxrt_loadgen

maria-18-git commented 1 year ago

Status: Done.