Tensorflow Serving Issue

[delta9]

Hi,

Thanks so much for this repository, I think you're doing a great service by saving everybody a lot of time!

Did you successfully deploy the model on Tensorflow Serving? I've exported it using your tool but I keep getting an error message when trying to test it:

{ code: 3, details: 'Invalid character found in base64.\n\t [[Node: map/while/DecodeBase64 = DecodeBase64[_output_shapes=[[]], _device="/job:localhost/replica:0/task:0/cpu:0"](map/while/TensorArrayReadV3)]]', metadata: Metadata { _internal_repr: {} } }

My client code is here: https://github.com/delta9/tensorflow-serving-node-client

I've setup a test server reachable on: 192.99.0.117:9000

I assume it's an issue with the client (it works fine on Google Cloud ML-Engine) but I've sort of hit a road block here and would really appreciate it, if you have any advice.

Many thanks,

Melvin

[mdietrichstein]

Hey,

Thanks so much for this repository, I think you're doing a great service by saving everybody a lot of time!

Thanks! It's always nice to get some feedback.

Did you successfully deploy the model on Tensorflow Serving?

No, haven't had a chance yet. I was under the impression that it should work just like on ML-Engine.

The error sounds familiar though. Happened to me when I was using the wrong base64 encoding. Try using a different base64 encoder (urlsafe versus standard/non-urlsafe) and see if it helps.

[delta9]

Ah so after going through the code, it turns out it's actually really easy to solve the issue by simply not trying to decode the input - took me a while to figure that out but then again, I've never used Python or TensorFlow before 😂

https://github.com/delta9/tensorflow-open_nsfw/commit/1b9b8849ae3d76bdcc62bcc009825e1273747935

[shaunster0]

hi @delta9 I know this is an old, closed issue, but I am getting exactly this error message! and I can't find any examples discussing it.

So how exactly did u fix it? what do u mean 'not trying to decode the input' ?

Thanks