mlcommons / mlcommons/inference
Increase expected QPS so the loadgen pre-generates a larger (coalesced) query
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.6k
- Forks
- 650
- Avg merge
- 1d 22h
- Merged PRs (30d)
- 6
Description
I used the Makefile to run the BERT example. Where can a user Increase expected QPS ? Thanks.
make run_pytorch_performance
Loading BERT configs...
Loading PyTorch model...
Constructing SUT...
Finished constructing SUT.
Constructing QSL...
No cached features at 'eval_features.pickle'... converting from examples...
Creating tokenizer...
Reading examples...
Converting examples to features...
Caching features at 'eval_features.pickle'...
Finished constructing QSL.
Running LoadGen test...
================================================
MLPerf Results Summary
================================================
SUT name : PySUT
Scenario : Offline
Mode : PerformanceOnly
Samples per second: 113.38
Result is : INVALID
Min duration satisfied : NO
Min queries satisfied : Yes
Early stopping satisfied: Yes
Recommendations:
* Increase expected QPS so the loadgen pre-generates a larger (coalesced) query.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the Makefile target run_pytorch_performance and trace the BERT example's LoadGen configuration for the expected-QPS value. Document where that setting is controlled and how a user can change it, including how to verify the resulting run is valid.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- documentation, machine-learning, performance
- Issue type
- Documentation
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100