Hugging Face's logo

Salesforce
/
blip-image-captioning-base

Image-to-Text
Transformers
PyTorch
TensorFlow
blip
image-text-to-text
image-captioning
Model card Files
xet
 Community
60

Input for calling predictor.predict() on Sagemaker

#19
by Callam - opened
Discussion
Callam

Hi would like to call predictor.predict() on Sagemaker but I haven't been able to figure out or find how the input should be structured. Is there anyone who could provide me with an example?

rnekkanti

+1 I am also looking for the same
anujyadav140

did anyone find a solution for this ?

anujyadav140

/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcalling-image-classification-model-deployed-in-sagemaker-endpoint%2F32200%2F15%3C%2Fa%3E%3C%2Fp%3E

i found this and i followed the steps but still i wasnt able to get a response, it still gives 400 error, but using this same method i tried running the nvidia image segmentation model too and was able to get a response.

Callam

Yo I never found an answer to the simple solution of just calling predict() out of the box but I did get it working by writing my own inference.py. Feel free to use this notebook as is or as an example.

https://github.com/DSCO-Co/SageMaker-image-to-text
rnekkanti

Thank you @Callam , This worked like a charm! Thanks for putting this together.
anujyadav140

Thanks bro.

Sign up or log in to comment