GPU availability check before deployment

[burnpiro]

Is your feature request related to a problem? Please describe. When I want to deploy the pod in the specific location I cannot get information about availability in that location. Current gpu_query allows only to specify gpu_id and gpu_count to get the list of GPUs.

Describe the solution you'd like I would like to be able to send a get request to check if there is a GPU with given parameters (the same as I latter provide to create_pod()).

Describe alternatives you've considered Maybe an external API to request data from (beside python lib)?

Additional context None

[justinmerrell]

I can add a general availability check; however, this would not help when the GPU becomes unavailable between checking and creating the pod. Is that what you are looking for?

[dvejmz]

I've just raised #307 to retrieve the machine information necessary from the runpod GraphQL API to enable this.