search

emedvedev / attention-ocr

A Tensorflow model for text recognition (CNN + seq2seq with visual attention) available as a Python package and compatible with Google Cloud ML Engine.
MIT License
1.08k stars 256 forks source link

2201904051650 #130

Closed kspook closed 5 years ago

kspook commented 5 years ago

I changed 'tensorflow serving execution part' in README.md file. I hope this could help many people.

emedvedev commented 5 years ago

Thank you! The model name and the column format are all correct now.

kspook commented 5 years ago

In my case, I encountered the error "Unable to base64 decode" with current input.

why don't you add more, example for input?

"inputs": { "input": { "b64": "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" } } }'

emedvedev commented 5 years ago

Mostly because I'd like to avoid huge chunks of base64 in the README, especially since everyone will end up using their own images anyway—the example is not supposed to be used as is. Updated a little, this way is probably better. Thanks!