Skip to content

Some fixes #248

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Sep 1, 2023
Merged

Some fixes #248

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2228883
Some fixes
yunfeng-scale Sep 1, 2023
6efb9a8
Merge branch 'main' into yunfeng-fixes
yunfeng-scale Sep 1, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion charts/model-engine/templates/balloon_cpu_deployment.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ spec:
resources:
limits:
memory: 28Gi
cpu: 8
Copy link
Contributor Author

@yunfeng-scale yunfeng-scale Sep 1, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

m5.2xlarge has only 8 CPUs and about 1 is used for system cores

cpu: 6
command:
- /bin/bash
- -c
Expand Down
6 changes: 3 additions & 3 deletions clients/python/llmengine/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,10 +41,10 @@ def create(
quantize: Optional[Quantization] = None,
checkpoint_path: Optional[str] = None,
# General endpoint fields
cpus: int = 32,
memory: str = "192Gi",
cpus: int = 8,
memory: str = "40Gi",
Copy link
Contributor Author

@yunfeng-scale yunfeng-scale Sep 1, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@gargutsav @auag92 i'm updating these default values, so if you're still going with 4 GPUs, please make sure values are set same as how you're using them right now

storage: str = "96Gi",
gpus: int = 4,
gpus: int = 1,
min_workers: int = 0,
max_workers: int = 1,
per_worker: int = 10,
Expand Down
10 changes: 8 additions & 2 deletions model-engine/model_engine_server/infra/services/image_cache_service.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -128,7 +128,9 @@ async def execute(self, endpoint_infra_states: Dict[str, Tuple[bool, ModelEndpoi
if state.resource_state.gpus == 0 and (
(
state.image not in images_to_cache_priority["cpu"]
or last_updated_at
or last_updated_at.replace(
tzinfo=images_to_cache_priority["cpu"][state.image].last_updated_at.tzinfo
)
> images_to_cache_priority["cpu"][state.image].last_updated_at
)
and self.docker_repository.image_exists(image_tag, repository_name)
Expand All @@ -143,7 +145,11 @@ async def execute(self, endpoint_infra_states: Dict[str, Tuple[bool, ModelEndpoi
if state.resource_state.gpu_type == gpu_type and (
(
state.image not in images_to_cache_priority[key]
or last_updated_at
or last_updated_at.replace(
tzinfo=images_to_cache_priority[key][
state.image
].last_updated_at.tzinfo
)
> images_to_cache_priority[key][state.image].last_updated_at
)
and self.docker_repository.image_exists(image_tag, repository_name)
Expand Down