-
-
Notifications
You must be signed in to change notification settings - Fork 4.1k
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.
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
[Core] Fix tracking of model forward time to the span traces in case of PP>1 #7440
Conversation
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge). To run full CI, you can do one of these:
🚀 |
cc @rkooo567 |
91f253a
to
d43fe90
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Leave to @rkooo567
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
QQ: is it possible to add unit tests for this?
orig_model_forward_time = 0.0 | ||
if intermediate_tensors is not None: | ||
orig_model_forward_time = intermediate_tensors.tensors.get( | ||
"model_forward_time", torch.tensor(0.0)).item() |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why do we store this to tensor? any way to just use cpu data?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
as far as I can tell, the only thing passed from the pipeline workers is the IntermediateTensors in serialized form. Hence added it to that. Is there a wrapper object of some form that holds these ?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
can you try a regular python object here to see if it works?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done. It looks like the worker serializes a Dict[Str, Any], so it can serialize floats too.
Let me look into that. There is one just the reporting of these metrics and two the PP>1 case. Let me look into see how doable these are. I will circle back later in the day. |
840e0b1
to
ac3db8d
Compare
I could not figure out how to get a unittest for this part of the worker. I instead added a test in the overall tracing test to test for these detailed trace data. It does not test for the pp>1 case though. please take a look. |
orig_model_forward_time = 0.0 | ||
if intermediate_tensors is not None: | ||
orig_model_forward_time = intermediate_tensors.tensors.get( | ||
"model_forward_time", torch.tensor(0.0)).item() |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
can you try a regular python object here to see if it works?
assert metrics.model_execute_time is None | ||
|
||
|
||
def test_traces_with_detailed_steps(trace_service): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Add the same test with pp= 2?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It was a bit more involved, but done.
@rkooo567 Are you comfortable with this PR on the whole ? |
/ready |
Head branch was pushed to by a user without write access
bec7bc6
to
125cd07
Compare
…ct#7440) [Core] Fix tracking of model forward time to the span traces in case of PP>1 (vllm-project#7440)
…ct#7440) [Core] Fix tracking of model forward time to the span traces in case of PP>1 (vllm-project#7440)
…ct#7440) [Core] Fix tracking of model forward time to the span traces in case of PP>1 (vllm-project#7440)
…ct#7440) [Core] Fix tracking of model forward time to the span traces in case of PP>1 (vllm-project#7440)
This is a quick follow up to #7089. In that PR we left the PP>1 unsupported for the model forward time. Fixing that here.
PR Checklist (Click to Expand)
Thank you for your contribution to vLLM! Before submitting the pull request, please ensure the PR meets the following criteria. This helps vLLM maintain the code quality and improve the efficiency of the review process.
PR Title and Classification
Only specific types of PRs will be reviewed. The PR title is prefixed appropriately to indicate the type of change. Please use one of the following:
[Bugfix]
for bug fixes.[CI/Build]
for build or continuous integration improvements.[Doc]
for documentation fixes and improvements.[Model]
for adding a new model or improving an existing model. Model name should appear in the title.[Frontend]
For changes on the vLLM frontend (e.g., OpenAI API server,LLM
class, etc.)[Kernel]
for changes affecting CUDA kernels or other compute kernels.[Core]
for changes in the core vLLM logic (e.g.,LLMEngine
,AsyncLLMEngine
,Scheduler
, etc.)[Hardware][Vendor]
for hardware-specific changes. Vendor name should appear in the prefix (e.g.,[Hardware][AMD]
).[Misc]
for PRs that do not fit the above categories. Please use this sparingly.Note: If the PR spans more than one category, please include all relevant prefixes.
Code Quality
The PR need to meet the following code quality standards:
format.sh
to format your code.docs/source/
if the PR modifies the user-facing behaviors of vLLM. It helps vLLM user understand and utilize the new features or changes.Notes for Large Changes
Please keep the changes as concise as possible. For major architectural changes (>500 LOC excluding kernel/data/config/test), we would expect a GitHub issue (RFC) discussing the technical design and justification. Otherwise, we will tag it with
rfc-required
and might not go through the PR.What to Expect for the Reviews
The goal of the vLLM team is to be a transparent reviewing machine. We would like to make the review process transparent and efficient and make sure no contributor feel confused or frustrated. However, the vLLM team is small, so we need to prioritize some PRs over others. Here is what you can expect from the review process:
action-required
label on the PR if there are changes required. The contributor should address the comments and ping the reviewer to re-review the PR.Thank You
Finally, thank you for taking the time to read these guidelines and for your interest in contributing to vLLM. Your contributions make vLLM a great tool for everyone!