Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

`dvc.yaml`: pipeline-less params and metrics

See original GitHub issue

dvc.yaml now supports a top-level plots section. Similar sections should be supported for params and metrics:

plots:
  roc.json:
    ...
params:
  - data_path
  - train.lr
  - optimizer_config.json:
metrics:
  - metrics.json

Params and metrics should not be cached, but they should be collected during params/metrics/exp commands.

Along with #6746, this allows users to start tracking experiments without needing pipelines.

Issue Analytics

State:
Created a year ago
Comments:8 (6 by maintainers)

Top GitHub Comments

1reaction

daavoocommented, Dec 12, 2022

@skshetry Do you want to keep this open for now or close it?

Should we close this one and open a new one more specific about the internal follow-ups?

1reaction

skshetrycommented, Nov 5, 2022

In Stage-level params, the strings are param-keys to be tracked from params.yaml, but I don’t think we need to support that. The strings could be considered a file, and the granular support can be via key-value pair (if we intend to support).

params:
- completely_tracked_params1.yaml
- completely_tracked_params2.yaml:
- granularly_tracked_params.yaml:
  - key1
  - key2

I skipped granular support in https://github.com/iterative/dvc/pull/8529, we can add if we intend to in the future then, when we are sure (or someone asks for it).

Top Results From Across the Web

dvc.yaml Files | Data Version Control

Metrics are key/value pairs saved in structured files that map a metric name to a numeric value. See dvc metrics for more information...

Pipeline variables from params file #3633 - iterative/dvc - GitHub

With the introduction of the new multiple-stage pipeline, we will need to find a way of defining variables in the pipeline.

DVC - Pipeline Versioning - Laziness makes Great Engineer

Run git checkout out params.yaml dvc.lock to restore the previous state. Track metrics. We now know how to track parameters. Next, we'll see ......

The ultimate guide to building maintainable Machine Learning ...

The principle to build pipelines using DVC · Write a python script for each of these steps · Save the parameters each script...

shcheklein/example-get-started: Get started DVC project

Get started DVC project - shcheklein/example-get-started. ... DVC-metrics ... model.pkl # <-- Trained model file ├── params.yaml # <-- Parameters file ...