Great to hear that you're giving inf1 a try! Do you mind providing more details about the model and the errors that are printed when you tried to compile? Is this post the same issue as this one: https://forums.aws.amazon.com/message.jspa?messageID=986172#986172 ?
In case it's helpful, we have a quickstart page dedicated to pytorch on our documentation site (here)https://awsdocs-neuron.readthedocs-hosted.com/en/latest/neuron-guide/neuron-frameworks/pytorch-neuron/index.html. There's links to tutorials, troubleshooting guides, and more from that initial link.
I responded in the "sagemaker" issue referenced above. Please confirm that his is indeed the same issue.
AWS torch-neuron errorasked a year ago
Unable to compile model to Neuron: no error message, no outputAccepted Answerasked 13 days ago
I am not able to ssh to EC2 instancesasked 5 months ago
I want to deploy my model as a Serverless inferenceasked a month ago
I am trying to deploy the EKS cluster using s3 template but getting errors, please help measked 8 months ago
How to deploy pre-trained model?asked 4 years ago
How to register a multi container model to a model registry?asked 3 months ago
Unable to create endpointAccepted Answerasked 4 years ago
Unable to train model: Amazon Rekognition experienced a service issue.asked 3 years ago
I am unable to delete my AWS Amplify projectasked 4 months ago