skypilot
skypilot copied to clipboard
SkyPilot: Run LLMs, AI, and Batch jobs on any cloud. Get maximum savings, highest GPU availability, and managed execution—all with a simple interface.
Also, it would be nice to add some unit tests for the behavior of show-gpus. _Originally posted by @Michaelvll in https://github.com/skypilot-org/skypilot/pull/3492#pullrequestreview-2026918295_
On commit `7f30ce5e` Currently the `sky local up` will create a config yaml under `/tmp` and does not clean it up after local cluster creation. If another user on the...
#3377 changes the autostop for skyserve controller to autodown, which will teardown the controller when the sky serve controller job exited unexpectedly and remove any related replica information/logs. This PR...
I am experiencing an issue deploying the Mixtral-8x7B-Instruct-v0.1 LLM Model on RunPod via SkyPilot. The deployment process fails during resource allocation, specifically with the port configuration. Below are the details...
Based on the current version pinnings in requirements-dev.txt, format.sh is incompatible with Python versions >= 3.11. Is it possible to upgrade the format script to support versions >= 3.11? Also,...
A user job experience OOM will not be able to have the return code in `sky logs`. We should gracefully handle it.
Initial implementation of sustainability support. Includes cost plus carbon footprint optimization. To enable cost plus carbon footprint optimization launch skypilot with the command line option`--optimize-cost-target=carbon_footprint`. Tested (run the relevant ones):...
This is a prototype for client-server architecture. ## Architecture We now refactors the interface to 3 categories: 1. Restful API 2. SDK 3. CLI ![image](https://github.com/skypilot-org/skypilot/assets/6753189/e97dc483-f6ca-45c5-b7b7-4e74150c09e0) ## Features This is a...
@WoosukKwon raised this before, and I also just had a need to distinguish between several spot jobs -- it'd be great to show the entrypoint or command used for each...