Inference v6.0
Pre-release
Pre-release
What's Changed
- [VLM] Offline scenario, performance-only mode of the reference implementation by @wangshangsam in #2381
- update docs with submission instruction + hyperlinks for heading by @anandhu-eng in #2390
- [VLM] Server Scenario, performance and accuracy by @johncalesp in #2388
- Submission dir update by @pgmpablo157321 in #2366
- Add note on LoadGen PyPI auto-build workflow by @anandhu-eng in #2398
- Revert changes used for testing new submission directory change by @anandhu-eng in #2401
- [VLM] Accuracy Evaluation by @johncalesp in #2393
- [VLM] Add brand field to accuracy evaluation by @johncalesp in #2404
- [VLM] Update the notebook to reflect the latest version of the dataset (that we are going to freeze) by @wangshangsam in #2406
- Update loadgen to 6.0; Bulk update 6.0 checker bits by @nvzhihanj in #2415
- feat: add MTP to ds-r1 ref. impl by @viraatc in #2403
- Add tests to detect failures in igbh dataset download by @anandhu-eng in #2359
- Text to Video Reference Implementation by @hvagadia in #2413
- [VLM] Fixing request timeout error, and enable VllmDeployer to fail fast if the underying
vllm serveprocess already failed by @wangshangsam in #2409 - Initial draft for Inference submission guide by @anandhu-eng in #2378
New Contributors
- @johncalesp made their first contribution in #2388
Full Changelog: v5.1.1...v6.0.0pre