yagnasrinath commented on issue #5005: [WIP]TVM android camera demo URL: https://github.com/apache/incubator-tvm/pull/5005#issuecomment-596143301 Also, I get the following warnings while building the model, I am new to tvm, and I am not sure if they are harmless, tried searching in the forums, which did not lead to any proper answer. Also if you observe the screenshot, the predictions are zero. I don't know, if that has something to with warnings from model export as shown below: ``` ~/projects/incubator-tvm/apps/android_camera/models $ python3 prepare_model.py mobilenet_v2 getting model... Using TensorFlow backend. File /Users/srinath/.tvm_test_data/keras/mobilenet_v2_weights.h5 exists, skip. WARNING:tensorflow:From /Users/srinath/Library/Python/3.7/lib/python/site-packages/tensorflow_core/python/ops/resource_variable_ops.py:1630: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version. Instructions for updating: If using Keras pass *_constraint arguments to layers. 2020-03-07 14:17:56.396579: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2020-03-07 14:17:56.421909: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7ff60a160230 initialized for platform Host (this does not guarantee that XLA will be used). Devices: 2020-03-07 14:17:56.421931: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version building... Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 3, 225, 225), 'float32'), ('TENSOR', (16, 3, 3, 3), 'float32'), (2, 2), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 16, 112, 112), 'float32'), ('TENSOR', (16, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 16, 112, 112), 'float32'), ('TENSOR', (8, 16, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 8, 112, 112), 'float32'), ('TENSOR', (48, 8, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 48, 113, 113), 'float32'), ('TENSOR', (48, 1, 3, 3), 'float32'), (2, 2), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 48, 56, 56), 'float32'), ('TENSOR', (16, 48, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 16, 56, 56), 'float32'), ('TENSOR', (96, 16, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 96, 56, 56), 'float32'), ('TENSOR', (96, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 96, 56, 56), 'float32'), ('TENSOR', (16, 96, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 96, 57, 57), 'float32'), ('TENSOR', (96, 1, 3, 3), 'float32'), (2, 2), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 96, 28, 28), 'float32'), ('TENSOR', (16, 96, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 16, 28, 28), 'float32'), ('TENSOR', (96, 16, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 96, 28, 28), 'float32'), ('TENSOR', (96, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 96, 29, 29), 'float32'), ('TENSOR', (96, 1, 3, 3), 'float32'), (2, 2), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 96, 14, 14), 'float32'), ('TENSOR', (32, 96, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 32, 14, 14), 'float32'), ('TENSOR', (192, 32, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 192, 14, 14), 'float32'), ('TENSOR', (192, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 192, 14, 14), 'float32'), ('TENSOR', (32, 192, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 192, 14, 14), 'float32'), ('TENSOR', (48, 192, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 48, 14, 14), 'float32'), ('TENSOR', (288, 48, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 288, 14, 14), 'float32'), ('TENSOR', (288, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 288, 14, 14), 'float32'), ('TENSOR', (48, 288, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 288, 15, 15), 'float32'), ('TENSOR', (288, 1, 3, 3), 'float32'), (2, 2), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 288, 7, 7), 'float32'), ('TENSOR', (80, 288, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 80, 7, 7), 'float32'), ('TENSOR', (480, 80, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('depthwise_conv2d_NCHWc.x86', ('TENSOR', (1, 480, 7, 7), 'float32'), ('TENSOR', (480, 1, 3, 3), 'float32'), (1, 1), (1, 1, 1, 1), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 480, 7, 7), 'float32'), ('TENSOR', (80, 480, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 480, 7, 7), 'float32'), ('TENSOR', (160, 480, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('conv2d_NCHWc.x86', ('TENSOR', (1, 160, 7, 7), 'float32'), ('TENSOR', (1280, 160, 1, 1), 'float32'), (1, 1), (0, 0, 0, 0), (1, 1), 'NCHW', 'NCHW', 'float32'). A fallback configuration is used, which may bring great performance regression. Cannot find config for target=llvm -target=aarch64-linux-android, workload=('dense_nopack.x86', ('TENSOR', (1, 1280), 'float32'), ('TENSOR', (1000, 1280), 'float32'), None, 'float32'). A fallback configuration is used, which may bring great performance regression. dumping lib... dumping graph... dumping params... dumping labels... Downloading from url https://raw.githubusercontent.com/dmlc/web-data/master/mxnet/doc/tutorials/onnx/image_net_labels.json to /Users/srinath/projects/incubator-tvm/apps/android_camera/app/src/main/assets/models/mobilenet_v2/image_net_labels.json ...100%, 0.03 MB, 88 KB/s, 0 seconds passed resnet18_v1 getting model... building... Cannot find config for target=llvm -target=aarch64-linux-android, workload=('dense_nopack.x86', ('TENSOR', (1, 512), 'float32'), ('TENSOR', (1000, 512), 'float32'), None, 'float32'). A fallback configuration is used, which may bring great performance regression. dumping lib... dumping graph... dumping params... dumping labels... Downloading from url https://raw.githubusercontent.com/dmlc/web-data/master/mxnet/doc/tutorials/onnx/image_net_labels.json to /Users/srinath/projects/incubator-tvm/apps/android_camera/app/src/main/assets/models/resnet18_v1/image_net_labels.json ...100%, 0.03 MB, 360 KB/s, 0 seconds passed ```
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
