-
Notifications
You must be signed in to change notification settings - Fork 51
chore: initial telemetry impl #118
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
Conversation
|
All contributors have signed the DCO ✍️ ✅ |
|
I have read the DCO document and I hereby sign the DCO. |
src/data_designer/engine/dataset_builders/column_wise_builder.py
Outdated
Show resolved
Hide resolved
for this, I'd recommend excluding |
src/data_designer/engine/dataset_builders/column_wise_builder.py
Outdated
Show resolved
Hide resolved
nabinchha
left a comment
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.
Nice, thanks @johntmyers!
src/data_designer/engine/dataset_builders/column_wise_builder.py
Outdated
Show resolved
Hide resolved
johnnygreco
left a comment
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.
thanks @johntmyers 🙌
Co-authored-by: Nabin Mulepati <[email protected]>
f6678c5
…ataDesigner into johntmyers/chore/telemetry
This PR adds anonymous model usage telemetry collection. Notable changes:
We generate an
InferenceEventwhich tracks model and token counts after each DD batch is created. This provides higher fidelity during longer running jobs. In order to do this we need to snapshot the current token usage from the cumulative counts in the model registry.The telemetry module is currently being vendored in. It's fully tested from its source which we can discuss offline if need be. It's a best-effort approach to reduce risk of interfering with any actual library usage.