# Using TF-TRT to convert MobileNet / SSDLite model gives errors **URL:** \ **Category:** Jetson TX2\ **Created:** [November 9, 2018, 5:18pm UTC](https://forums.developer.nvidia.com/t/using-tf-trt-to-convert-mobilenet-ssdlite-model-gives-errors/67197 "2018-11-09T17:18:51Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![rsandler00](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@rsandler00](https://forums.developer.nvidia.com/u/rsandler00)\ **Post date:** [November 9, 2018, 5:18pm UTC](https://forums.developer.nvidia.com/t/using-tf-trt-to-convert-mobilenet-ssdlite-model-gives-errors/67197/1 "2018-11-09T17:18:51Z")
Hi, I am trying to convert a Tensorflow MobileNet graph (\*.pb file) into TensorRT via the TF-TRT module. However I get an error ```auto `NotFoundError: No attr named 'shape' in NodeDef:` ``` . I am running the following: Jetson TX2 Jetpack 3.3 TensorRT 4.0.2.0-1+cuda9.0 tensorflow-gpu==1.9.0+nv18.8 Here is my code: ```auto import tensorflow as tf import tensorflow.contrib.tensorrt as trt import numpy as np import PIL from timeit import default_timer as timer from tqdm import tqdm ''' This script performs inference on a pure Tensorflow (TF) model vs a converted TensorRT model In order to run it, it assumes you have a frozen TF graph in the form of a *.pb file. The specific model is MobileNetv2 (https://github.com/tensorflow/models/tree/master/research/slim/nets/mobilenet( Frozen graph can be downloaded from: https://storage.googleapis.com/mobilenet_v2/checkpoints/mobilenet_v2_1.4_224.tgz The image file 'panda.jpg' is downloaded from: https://upload.wikimedia.org/wikipedia/commons/f/fe/Giant_Panda_in_Beijing_Zoo_1.JPG ''' ## File paths (needs to be set) pb_path = "mobilenet_v2_1.0_224_frozen.pb" # frozen TF graph input_name = 'input' # input layer of the graph output_names = ['MobilenetV2/Predictions/Reshape_1'] # output layer of the graph. (Can be multiple) img_path = 'data/panda.jpg' # image to perform inference on (from https://upload.wikimedia.org/wikipedia/commons/f/fe/Giant_Panda_in_Beijing_Zoo_1.JPG) ## Load image img_raw = np.array(PIL.Image.open(img_path).resize((224, 224))).astype(np.float) / 128 - 1 img = img_raw.reshape(1, 224,224, 3) ##### Run w/ TensorRT trt_graph = trt.create_inference_graph( input_graph_def=frozen_graph, outputs=output_names, max_batch_size=1, precision_mode='FP16', # 'INT8'/'FP16' ) tf_config = tf.ConfigProto() tf_config.gpu_options.allow_growth = True tf_sess = tf.Session(config=tf_config) tf.import_graph_def(trt_graph, name='') tf_input = tf_sess.graph.get_tensor_by_name(input_name + ':0') tf_output = tf_sess.graph.get_tensor_by_name(output_names[0] + ':0') output = tf_sess.run(tf_output, feed_dict={tf_input: img + np.random.random(img.shape)/10}) ``` And here is the full error trace: ```auto 2018-11-09 16:43:48.943803: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:864] ARM64 does not support NUMA - returning NUMA node zero 2018-11-09 16:43:48.943953: I tensorflow/core/grappler/devices.cc:51] Number of eligible GPUs (core count >= 8): 0 --------------------------------------------------------------------------- NotFoundError Traceback (most recent call last) /home/nvidia/Projects/SANATA/trt_test.py in () 71 outputs=output_names, 72 max_batch_size=1, ---> 73 precision_mode='FP16', # 'INT8'/'FP16' 74 ) 75 /home/nvidia/Projects/SANATA/.venv/lib/python3.5/site-packages/tensorflow/contrib/tensorrt/python/trt_convert.py in create_inference_graph(input_graph_def, outputs, max_batch_size, max_workspace_size_bytes, precision_mode, minimum_segment_size) 113 # pylint: disable=protected-access 114 raise _impl._make_specific_exception(None, None, ";".join(msg[1:]), --> 115 int(msg[0])) 116 # pylint: enable=protected-access 117 output_graph_def = graph_pb2.GraphDef() NotFoundError: No attr named 'shape' in NodeDef: [[Node: input = Placeholder[dtype=DT_FLOAT]()]] for 'input' (op: 'Placeholder') with input shapes: ``` Note, I get essentially the same error even when using the SSDLite graph from: `http://download.tensorflow.org/models/object_detection/ssdlite_mobilenet_v2_coco_2018_05_09.tar.gz`. How can I fix this? Thanks, Roman ---
**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\ **Post date:** [November 12, 2018, 1:14am UTC](https://forums.developer.nvidia.com/t/using-tf-trt-to-convert-mobilenet-ssdlite-model-gives-errors/67197/2 "2018-11-12T01:14:13Z")
Hi You can check our tutorial for converting _ssd\_mobilenet\_v1\_coco_ to _TF-TRT_ model: [https://github.com/NVIDIA-AI-IOT/tf\_trt\_models#object-detection](https://github.com/NVIDIA-AI-IOT/tf_trt_models#object-detection) The workflow should like this: Build TensorRT / Jetson compatible graph ```auto from tf_trt_models.detection import build_detection_graph frozen_graph, input_names, output_names = build_detection_graph( config=config_path, checkpoint=checkpoint_path ) ``` Optimize with TensorRT ```auto import tensorflow.contrib.tensorrt as trt trt_graph = trt.create_inference_graph( input_graph_def=frozen_graph, outputs=output_names, max_batch_size=1, max_workspace_size_bytes=1 << 25, precision_mode='FP16', minimum_segment_size=50 ) ``` Thanks. ---
**Author:** ![rsandler00](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@rsandler00](https://forums.developer.nvidia.com/u/rsandler00)\ **Post date:** [November 15, 2018, 6:26pm UTC](https://forums.developer.nvidia.com/t/using-tf-trt-to-convert-mobilenet-ssdlite-model-gives-errors/67197/3 "2018-11-15T18:26:55Z")
Thank you @aastall for the reference. That was exactly what I was looking for. For the record, I tried comparing inference speed between the pure Tensorflow vs TF-TRT graphs on the MobileNetV1 and MobileNetV2 networks. I used a 640x480 image for both tests and ran ```auto sudo ~/jetson_clocks.sh ``` prior to the tests. Code for MobileNetV1 benchmark is [here](https://gist.github.com/rsandler00/b139a081f548b2895cd6e6fffad0b781), and for MobileNetV2 benchmark is [here](https://gist.github.com/rsandler00/585b6489011e05a8ba18823745018cde). The results were: MobileNetV1: Pure tensorflow: 89ms TF-TRT: 119ms (~33% speedup) MobileNetV2: Pure tensorflow: 203ms TF-TRT: 203ms (0% speedup) I am not sure why MobileNetV2 is so inefficient here and sees no speedup from TF-TRT. Any help would be welcome. For the record, [this](https://github.com/chuanqi305/MobileNetv2-SSDLite/blob/master/ssdlite/deploy.prototxt) Caffe implmentation of MobileNetv2 took 86ms on the Jetson using NVCaffe (not TRT optimized). ---
**Author:** ![nadeemm](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/nadeemm/32/367020_2.png) [@nadeemm](https://forums.developer.nvidia.com/u/nadeemm)\ **Post date:** [October 18, 2021, 6:30pm UTC](https://forums.developer.nvidia.com/t/using-tf-trt-to-convert-mobilenet-ssdlite-model-gives-errors/67197/4 "2021-10-18T18:30:47Z")