How to convert checkpoint file to model .pb file using tf.saved_model.builder.SavedModelBuilder to host on tensorflow serving

[gr8Adakron]

Basically I have models which are already developed in Mxnet and I want to convert them into tensorflow serving format without performing retraining.

So for performing conversion from Mxnet to Tensorflow I have used this Intermediate layer created by Microsoft Mdnn.

Using this tool/library I was able to convert Mxnet model files to Tensorflow checkpoint format:

These are initial Mxnet model files.
And this is my flow of conversion.

Mxnet_model => IR format => Conversion code => Checkpoint => Tensorflow Model, Serving format

IR = intermediate representation
I have successfully converted till checkpoint files using this instruction Mxnet to IR then IR to Tensorflow checkpoint

This final structure of tensorflow checkpoint file are:

The only step remaining is to convert and save these checkpoint files to .pb and variable folder format using this SaveBuilder function
Saving using this function is essentials because its is the only valid format to save the model, inorder to serve the model on tensorflow serving

This is how the final structure of the converted model must look like in order to get served using TF-serving:

This is the accurate structure which TF-serving accept and generate predictions for.

I tried using these scripts freeze_graph.py and incepiton_save_model.py but nothing came out they have some arguments which I don't have files to pass.

Help!!!
Is there a way? I am trying since last 3days but couldn't find anything.
Thanks in advance.

[Johnson-yue]

hi, I have the same question!!
I have checkpoint files like:(data-0000-of-0001, index, .meta), I want to convert them to tf serving use.
the format is (saved_model.pb, variables, variables.data-0000-of-0001, variabels/variables.index)
Do you fix it??

[elvys-zhang]

Use imagenet1k-resnet-50 as example:
mmdownload -f mxnet -n imagenet1k-resnet-50 -o ./
mmtoir -f mxnet -n resnet-50-symbol.json -w resnet-50-0000.params -d resnet50 --inputShape 3 224 224
mmtocode -f tensorflow --IRModelPath resnet50.pb --IRWeightPath resnet50.npy -d tf_resnet50.py
mmtomodel -f tensorflow -in tf_resnet50.py -iw resnet50.npy -o tf_resnet50 --dump_tag SERVING

Final structure:

It worked fine, and no need to convert mxnet to tf ckpt. Hope it will help.

[wt-huang]

Closing as the issue is resolved.

[YB221]

hi, I have the same question!!
I have checkpoint files like:(data-0000-of-0001, index, .meta), I want to convert them to tf serving use.
the format is (saved_model.pb, variables, variables.data-0000-of-0001, variabels/variables.index)
Do you fix it??

Hey could you please tell me if you solved it? if yes how?