Track current best and test result records
This commit is contained in:
@@ -34,4 +34,13 @@ These are the values used to balance speed and quality:
|
||||
|
||||
The workflow files retain the tested baseline values. Each workflow README documents its model, inputs, dimensions, and measured runtime.
|
||||
|
||||
## Test result convention
|
||||
|
||||
When a test is run, store its output and a matching metadata record beside the workflow:
|
||||
|
||||
- `results/<test-id>.png` — output image labelled with the exact table ID.
|
||||
- `results/<test-id>.json` — exact prompt, seed, input-image names or hashes, resolution, CFG, steps, sampler, scheduler, runtime, and verdict.
|
||||
|
||||
Update the table's **Tested?** cell to `Yes` after the run. If the result is judged **Better**, update the workflow's optimization settings and mark that row's **Current Best** cell `Yes`; all other rows must remain `No`. Keep the static prompt, seed, and base images unchanged unless the test plan explicitly changes them.
|
||||
|
||||
The previous prompt-specific and specialized workflows are retained under [`archive/`](archive/) for reference.
|
||||
|
||||
Reference in New Issue
Block a user