v0.8.rc1

What's Changed

• Reorg conf directory by @srinivas212 in #174

Full Changelog: v0.8.rc0...v0.8.rc1

v0.8.rc0

What's Changed

• update container version by @jeffnvidia in #122
• Update copyright format by @amaslenn in #155
• Enhance job status error message for clarity and user guidance by @TaekyungHeo in #156
• Upgrade pyright, fix warning by @amaslenn in #157
• Allow extra srun args set in system config by @amaslenn in #160
• Automatically define version from the git tags by @amaslenn in #162
• add support for NeMo Launcher in the reservation by @jeffnvidia in #161
• Add job_status_check flag to disable checks for specific tests by @TaekyungHeo in #146
• Defer docker image URL accessibility check to srun when not caching locally by @TaekyungHeo in #164
• Check TOMLs formatting with taplo by @amaslenn in #163
• Hierarchical Test template for support Grok/GPT via PAXML by @srivatsankrishnan in #141

Full Changelog: v0.7.14...v0.8.rc0

v0.7.14

What's Changed

• Make sure to dry-run all tests in a test scenario by @TaekyungHeo in #149
• Add warning for insufficient epochs in JaxToolbox report generation by @TaekyungHeo in #148
• Remove subtest name check in NcclTestSlurmCommandGenStrategy by @TaekyungHeo in #147
• Handle disk quota exceeded error in cache_docker_image method by @TaekyungHeo in #150
• Remove unused properties from TestTemplateStrategy by @amaslenn in #151
• Enhance error messages to provide guidance for missing schemas by @TaekyungHeo in #152
• Bump version to v0.7.14 by @TaekyungHeo in #153

Full Changelog: v0.7.13...v0.7.14

v0.7.13

What's Changed

• Pass all env vars to final command in NeMo launcher test template by @TaekyungHeo in #134
• Added JaxToolbox (Grok) troubleshooting steps by @TaekyungHeo in #142
• Improve tokenizer path handling in NeMo Launcher Slurm strategy by @TaekyungHeo in #136
• Remove identical if-else branches by @amaslenn in #143
• Move parts of srun CLI generation into base class by @amaslenn in #140

Full Changelog: v0.7.12...v0.7.13

v0.7.12

What's Changed

• Add how to download NeMo launcher tokenizer in the USER_GUIDE by @jeffnvidia in #115
• Update logging to use dynamic log file name from args.log_file in run mode by @TaekyungHeo in #131
• Update README: add mandatory args test-templates-dir and tests-dir by @TaekyungHeo in #132
• Update NeMo launcher troubleshooting guide for clarity and conciseness by @TaekyungHeo in #135
• Add support for Sleep on Slurm systems by @TaekyungHeo in #121
• Raise descriptive exceptions when strategies are missing by @TaekyungHeo in #137
• Add L40s test configs by @srinivas212 in #138
• Bump version to v0.7.12 by @TaekyungHeo in #139

Full Changelog: v0.7.11...v0.7.12

v0.7.11

What's Changed

• Enhance Quick Start guide for Docker repo access and API key by @TaekyungHeo in #99
• Add copyright headers for TOML, update its format by @amaslenn in #117
• Add pyxis mktemp error handling and test cases in JaxToolbox strategy by @TaekyungHeo in #118
• Add section on downloading NeMo datasets to USER_GUIDE.md by @TaekyungHeo in #116
• Enhance USER_GUIDE.md with system schema description and troubleshooting steps by @TaekyungHeo in #120
• Fix bug in generating NeMo launcher command by @TaekyungHeo in #124
• Add section on describing a test scenario to USER_GUIDE.md by @TaekyungHeo in #123
• Add cache_docker_images_locally field to system schema in USER_GUIDE.md by @TaekyungHeo in #125
• Allow local docker image caching for JaxToolbox by @TaekyungHeo in #126
• Update BaseRunner to include scenario name in output directory by @TaekyungHeo in #119
• Add mpi field to SlurmSystem and allow different MPI options in schema by @TaekyungHeo in #127
• Bump version to v0.7.11 by @TaekyungHeo in #128

Full Changelog: v0.7.10...v0.7.11

v0.7.10

What's Changed

• Reduce logs for stdout by @amaslenn in #108
• Update USER_GUIDE by @amaslenn in #109
• Fix incorrect argument name in README by @TaekyungHeo in #110
• Remove bisection tests from NCCL-test template by @TaekyungHeo in #113
• Fix bug in NeMo launcher report generation by @TaekyungHeo in #112
• Add pip install requirements.txt step to NeMoLauncherSlurmInstallStrategy by @TaekyungHeo in #111
• Bump version to v0.7.10 by @TaekyungHeo in #114

Full Changelog: v0.7.9...v0.7.10

v0.7.9

What's Changed

• Fix _check_docker_image_accessibility condition and add detailed logging by @TaekyungHeo in #107
• Refactor handle_install_and_uninstall to identify unique test templates by @TaekyungHeo in #105
• Bump version to v0.7.9 by @TaekyungHeo in #106

Full Changelog: v0.7.8...v0.7.9

v0.7.8

What's Changed

• Handle 401 Unauthorized error with detailed instructions for Docker image access by @TaekyungHeo in #103
• Bump version to v0.7.8 by @TaekyungHeo in #104

Full Changelog: v0.7.7...v0.7.8

v0.7.7

What's Changed

• Refactor _check_docker_image_accessibility to remove srun usage by @TaekyungHeo in #98
• Update NcclTestJobStatusRetrievalStrategy to improve error messages by @TaekyungHeo in #100
• Update JaxToolboxJobStatusRetrievalStrategy to improve error messages by @TaekyungHeo in #101
• Bump version to v0.7.7 by @TaekyungHeo in #102

Full Changelog: v0.7.6...v0.7.7