skypilot

skypilot copied to clipboard

We may want to expose our python APIs, make them easier to use, and add them to our document. It could be related to #871.

Michaelvll

Some inference tasks can be run on multiple different resources, e.g. V100:8, V100:4, or V100:1. Based on the availability, the user wants the sky to failover from 8 to 1....

Michaelvll

Add Imagenet Benchmark

2

To test out how well goofys performs for a real deep learning workload, I made this benchmark. ## ImageNet Dataset Information ### Stats * 1M training images, and 50K val...

Michaelvll

Fail to launch GCP instance when the account does not have IAM permission

2

A GCP user with no IAM setting permission (the permission in the following figure) will not be able to get setIAMPolicy for the ray-autoscaler service IAM and cause the following...

Michaelvll

Do we need specific versions for Keras and other frameworks to support callback function? _Originally posted by @infwinston in https://github.com/sky-proj/sky/pull/832#discussion_r917123658_ To quickly merge #832, I will resolve this after the...

WoosukKwon

Sky Storage Download Feature

1

Add a Sky Storage Download Feature. Also fixes #932. To download a folder from a bucket using Sky Storage, run: ``` from sky.data import storage, StoreType storage_obj = storage.Storage(name='imagenet-bucket') storage_obj.add_store(StoreType.S3)...

michaelzhiluo

Bash Error in all `*-ray.yml.j2` files

2

In `head_start_ray_commands` for every Cloud jinja file, there is the command: ``` (ps aux | grep "-m sky.skylet.skylet" | grep -q python3) || nohup python3 -m sky.skylet.skylet >> ~/.sky/skylet.log 2>&1...

michaelzhiluo

This came up during a user conversation. The user has been using auto-stopping, and he'd also like an auto-down equivalent. To design this, we need to think about - Do...

concretevitamin

Force `ray==1.10.0` in setup.py?

2

I am working on a docker image to run the sky worker (remote VM in cloud parlance) in a container. I am trying to reuse the sky client docker image...

romilbhardwaj

`$(nproc)` in setup and run sections

2

My task on a m6i.2xlarge instance ```yaml name: nproc resources: cloud: aws setup: | echo "setup $(nproc)" run: | echo "run $(nproc)" ``` returns ``` setup 1 run 1 ```...

WoosukKwon