Merge pull request #63 from foss-for-synopsys-dwc-arc-processors/cifar10_tut

Cifar10 tutorial-like scripts

Author: ipozdnov

examples/tutorial_cifar10_caffe_deployment/cifar10_small.caffemodel.h5

@@ -0,0 +1,134 @@
+name: "CIFAR10_quick"
+layer {
+  name: "data"
+  type: "Input"
+  top: "data"
+  input_param { shape: { dim: 1 dim: 3 dim: 32 dim: 32 } }
+}
+layer {
+  name: "conv1"
+  type: "Convolution"
+  bottom: "data"
+  top: "conv1"
+  param {
+    lr_mult: 1
+  }
+  param {
+    lr_mult: 2
+  }
+  convolution_param {
+    num_output: 32
+    pad: 2
+    kernel_size: 5
+    stride: 1
+  }
+}
+layer {
+  name: "pool1"
+  type: "Pooling"
+  bottom: "conv1"
+  top: "pool1"
+  pooling_param {
+    pool: MAX
+    kernel_size: 3
+    stride: 2
+  }
+}
+layer {
+  name: "relu1"
+  type: "ReLU"
+  bottom: "pool1"
+  top: "pool1"
+}
+layer {
+  name: "conv2"
+  type: "Convolution"
+  bottom: "pool1"
+  top: "conv2"
+  param {
+    lr_mult: 1
+  }
+  param {
+    lr_mult: 2
+  }
+  convolution_param {
+    num_output: 16
+    pad: 2
+    kernel_size: 5
+    stride: 1
+  }
+}
+layer {
+  name: "relu2"
+  type: "ReLU"
+  bottom: "conv2"
+  top: "conv2"
+}
+layer {
+  name: "pool2"
+  type: "Pooling"
+  bottom: "conv2"
+  top: "pool2"
+  pooling_param {
+    pool: AVE
+    kernel_size: 3
+    stride: 2
+  }
+}
+layer {
+  name: "conv3"
+  type: "Convolution"
+  bottom: "pool2"
+  top: "conv3"
+  param {
+    lr_mult: 1
+  }
+  param {
+    lr_mult: 2
+  }
+  convolution_param {
+    num_output: 32
+    pad: 2
+    kernel_size: 5
+    stride: 1
+  }
+}
+layer {
+  name: "relu3"
+  type: "ReLU"
+  bottom: "conv3"
+  top: "conv3"
+}
+layer {
+  name: "pool3"
+  type: "Pooling"
+  bottom: "conv3"
+  top: "pool3"
+  pooling_param {
+    pool: AVE
+    kernel_size: 3
+    stride: 2
+  }
+}
+layer {
+  name: "ip1"
+  type: "InnerProduct"
+  bottom: "pool3"
+  top: "ip1"
+  param {
+    lr_mult: 1
+  }
+  param {
+    lr_mult: 2
+  }
+  inner_product_param {
+    num_output: 10
+  }
+}
+layer {
+  name: "prob"
+  type: "Softmax"
+  bottom: "ip1"
+  top: "prob"
+}
+

examples/tutorial_cifar10_caffe_deployment/cifar10_small.prototxt

@@ -0,0 +1,79 @@
+# Copyright 2019, Synopsys, Inc.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-3-Clause license found in
+# the LICENSE file in the root directory of this source tree.
+#
+"""
+embarc Machine Learning Inference Library (embarc MLI):
+Model Deployment tutorial for Caffe and CIFAR-10: Launching script
+"""
+
+import argparse
+from deployment_steps import *
+
+
+def main(argn):
+    # Go through the model deployment tutorial for Caffe and CIFAR-10 step-by-step
+    print('MLI Deployment tutorial script.')
+    print('Used parameters:')
+    for param, val in vars(argn).items():
+        print('\t{}: {}'.format(param, val))
+    print("\n")
+
+    print("Step 1: Instrument the input model.")
+    classifier, layers, ir_tensors = instrument_model_step(argn.prototext, argn.model)
+    print("Done\n")
+
+    print("Step 2: Collect data statistic for each layer.")
+    collect_inference_statistic_step(classifier, argn.lmdb_data_dir, ir_tensors, 128.0, 1.0 / 128.0)
+    print("Done\n")
+
+    print("Step 3: Define Qm.n data format.")
+    define_qmn_step(layers, argn.kernel_type)
+    print("Done\n")
+
+    print("Step 4: Quantize and output weights.")
+    quantize_and_output_step(layers, argn.kernel_type, argn.output_prefix, argn.wrap_coefficients)
+    print("Done\n")
+
+    # Next step should be done manually using generated files
+    # write_code_for_me(...)
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description="CIFAR-10 for Caffe Deployment Tutorial Script ",
+                                     formatter_class=argparse.ArgumentDefaultsHelpFormatter)
+    parser.add_argument(
+        '--lmdb_data_dir',
+        type=str,
+        default='cifar10_train_lmdb',
+        help='Path to CIFAR-10 subset in LMDB format (created by caffe cifar-10 example specific tool)')
+    parser.add_argument(
+        '--model',
+        type=str,
+        default='cifar10_small.caffemodel.h5',
+        help='CIFAR-10 caffemodl file')
+    parser.add_argument(
+        '--prototext',
+        type=str,
+        default='cifar10_small.prototxt',
+        help='CIFAR-10 test prototext file')
+    parser.add_argument(
+        '--output_prefix',
+        type=str,
+        default='cifar10_small',
+        help='Prefix for output files including *.inc(coefficients) *.c (tensors) *.h (declarations)')
+    parser.add_argument(
+        '--kernel_type',
+        type=str,
+        default='fx8',
+        choices=['fx8', 'fx16', 'fx8w16d'],
+        help='')
+    parser.add_argument(
+        '--wrap_coefficients',
+        action='store_true',
+        help='Wrap float values into macro for compile-time quantization purpose')
+
+    flags, unparsed = parser.parse_known_args()
+    main(flags)