THIS ONE AFTER THE LAST ONE: fix(#517): clear error when workload YAML missing; document supported matrix#563
Open
FileSystemGuy wants to merge 1 commit into
Open
Conversation
…t supported matrix
Before: passing an (model, accelerator-type) combination that has no
configs/dlio/workload/<model>_<accel>.yaml file produced a generic
"Configuration file not found: …/unet3d_mi355.yaml" error from
read_config_from_file. The user had no way to tell whether this was
a packaging bug, an install problem, or an unsupported combination.
After: DLIOBenchmark.process_dlio_params checks the resolved workload
path up front and raises ConfigurationError(CONFIG_FILE_NOT_FOUND)
with a specific "combination not supported" message that:
- names the (model, accelerator-type) pair
- points at the missing YAML so the failure is unambiguous
- lists the three v3.0 submittable combinations (unet3d/b200,
retinanet/b200, retinanet/mi355)
- notes that other pairs are available under `whatif` if a workload
file exists for them.
Docs (ManPage.md, training/README.md) gain the full (model x accelerator)
support matrix, marking each cell as v3.0-submittable, whatif-only, or
not supported.
Checkpointing uses a model-only filename so it gets a shorter variant
of the message via the same helper, branching on BENCHMARK_TYPE.
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
FileSystemGuy
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Resolves #517.
mlpstorage closed training unet3d datasize --accelerator-type mi355 …was producing a bare:…which is technically true but tells the user nothing about why the file is missing (because we don't support that combination in v3.0) or what they should do instead.
Code change
DLIOBenchmark.process_dlio_paramsnow does an explicit existence check on the resolved workload YAML and raisesConfigurationError(CONFIG_FILE_NOT_FOUND)with a specific "combination not supported" message before the lower-level loader can emit its generic one. The message:(model, accelerator-type)pair.unet3d/b200,retinanet/b200,retinanet/mi355.The helper branches on `BENCHMARK_TYPE` so checkpointing (model-only filename) gets the same treatment with a shorter message.
Example output
Docs
ManPage.md(near `--accelerator-type`) and `training/README.md` (top of §"Training Models") gain the full support matrix:Test plan