aws / aws/amazon-sagemaker-feedback

Floating point arithmetic display problem with metrics in SageMaker Studio Models

Open
#192 0 comments 0 reactions 0 assignees View on GitHub
bug
Dominant language
No language data
Stars
10
Forks
3
PR merge metrics
No merged PRs in 30d

Description

### Product Version

- [ ] Amazon SageMaker Studio Classic
- [x] Amazon SageMaker Studio
- [ ] Issue is not related to SageMaker Studio

### Issue Description

I have a training job that emits the following log line:
```
2025-07-02 13:44:32,875 - signals - INFO - Lift (candidate) over random: 1.2623;
```
I capture this metric in a training job using the following metrics definition: `Lift \(candidate\) over random: (.*?);`, which correctly records the cloud watch metric and displays it in SageMaker Studio:

Image

However, once a model is registered, the digits are expanded and I see this:

Image

This is visually disturbing.

### Expected Behavior

See as many digits after the decimal points in the registered model metrics as in the captured metrics from the training job.

### Observed Behavior

_No response_

### Product Category

Models

### Feedback Category

User Interface

### Other Details

I submitted a Support Case with AWS Support: 175148780700690

Contributor guide

Open the contributing guide

Research direction

The issue identifies SageMaker Studio Models and registered model metrics, but names no repository files, tests, or implementation entry point. Start by locating the registered-model metrics display and compare its formatting with the captured training-job metric; done means preserving the captured decimal precision.

Written by the indexing model from the issue text.

Assessment

Tech stack
aws
Domain
cloud, frontend
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.