Search results

isolate the clouds, such that running aws, does not depend on all the limitations of gcp, and azure. The spinup should be as fast as possible.

Status: Open.

#55 In slashml/magemaker;

· eff-kay opened on Jan 20, 2025

provide support for other model registries in azure

Status: Open.

#53 In slashml/magemaker;

· eff-kay opened on Jan 12, 2025

investigate packaging clis with the project, some people might have trouble installing them. Compare how heavy it makes the package

Status: Open.

#45 In slashml/magemaker;

· eff-kay opened on Jan 10, 2025

investigate open-ai compatible endpoints

Status: Open.

#42 In slashml/magemaker;

· eff-kay opened on Jan 10, 2025

quit takes a few seconds to quit, it should be instantaneous

Status: Open.

#41 In slashml/magemaker;

· eff-kay opened on Jan 10, 2025

add the ability to attach a schedule to sagemaker, so we don't have to worry about the budget.

Status: Open.

#40 In slashml/magemaker;

· eff-kay opened on Jan 10, 2025

Document which version of TGI we are using as a base image for model deployment in gcp

Status: Open.

#35 In slashml/magemaker;

· eff-kay opened on Jan 9, 2025

we should not ask aws credentials, everytime we run <code>magemaker --cloud aws</code>.

Status: Open.

#33 In slashml/magemaker;

· eff-kay opened on Jan 9, 2025

Create a new flag for querying the model in one shot, assuming the relevant flag is provided and the relevant params are passed

Status: Open.

#31 In slashml/magemaker;

· eff-kay opened on Jan 9, 2025

remove the concept of a CONFIG_DIR, or .magemaker config, its a remnant of the old architecture.

Status: Open.

#30 In slashml/magemaker;

· eff-kay opened on Jan 9, 2025

docs should be in this repo

Status: Open.

#23 In slashml/magemaker;

· eff-kay opened on Jan 8, 2025

add Ci/CD for tests, and different python runtimes.

Status: Open.

#22 In slashml/magemaker;

· eff-kay opened on Jan 8, 2025