dodandeniya/panda-model

This is an example of custom object detection using Tensorflow object detection api

panda-model's Issues

Can I convert fine_tuned_models/saved_model.pb to tensorflowjs model ?

Dear sir,

This repo is very useful for me. Thank you very much.
I want to convert the "fine_tuned_model/saved_model" to tensorflowjs model.
I tried it, but does not work..

On Google Colab, after "Validation".

%cd /content/drive/My Drive/panda-model/fine_tuned_model
!mkdir web_model

!pip install tensorflowjs==1.3.1

!tensorflowjs_converter \
    --input_format=tf_saved_model \
    --output_format=tfjs_graph_model \
    --output_node_names='num_detections,detection_boxes,detection_scores,detection_classes,detection_masks' \
    --skip_op_check \
    --saved_model_tags=serve \
    ./saved_model \
    ./web_model

!ls -la ./web_model
-rw-r--r-- 1 root root 4194304 Nov  2 03:24 group1-shard1of5.bin
-rw-r--r-- 1 root root 4194304 Nov  2 03:24 group1-shard2of5.bin
-rw-r--r-- 1 root root 4194304 Nov  2 03:24 group1-shard3of5.bin
-rw-r--r-- 1 root root 4194304 Nov  2 03:24 group1-shard4of5.bin
-rw-r--r-- 1 root root 1753480 Nov  2 03:24 group1-shard5of5.bin
-rw-r--r-- 1 root root  303953 Nov  2 03:24 model.json

!rm -rf web_model.zip
!zip -r web_model.zip ./web_model/

And Javascript.

<script src="https://cdn.jsdelivr.net/npm/@tensorflow/[email protected]"></script>
......
var model = await tf.loadGraphModel('http://xxxxxxxxxx.com/web_model/model.json');

Error.

[email protected]:2 Uncaught (in promise) TypeError: Cannot read property 'producer' of undefined
at t. ([email protected]:2)
at [email protected]:2
at Object.next ([email protected]:2)
at o ([email protected]:2)

I don't know if the cause is the version or the conversion is bad. If you know it, please let me know.

Regards.

Google Collab Not Compatible TF

Thanks for the notebook. The first step throws an error. You probably did not code for TF 2.2

---

AttributeError Traceback (most recent call last)
in ()
1 import tensorflow as tf
----> 2 config = tf.ConfigProto()
3 config.gpu_options.allow_growth = True
4
5 with tf.device('/gpu:0'):

AttributeError: module 'tensorflow' has no attribute 'ConfigProto'

print(tf. version)
2.2.0