Skip to content

Actions: NVIDIA-NeMo/RL

Actions

Create PR to main with cherry-pick from release

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
cherry-pick-release-commit.yml
688 workflow runs
688 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

feat: dynamic batching for training and log prob stages (#274) Create PR to main with cherry-pick from release #163: Commit 7d8ce74 pushed by github-merge-queue bot
19s main
main
19s
feat: add deepscaler guide (#391) Create PR to main with cherry-pick from release #162: Commit f4e95ab pushed by github-merge-queue bot
20s main
main
20s
feat: add aime24 validation set (#388) Create PR to main with cherry-pick from release #161: Commit 281eacb pushed by github-merge-queue bot
23s main
main
23s
feat: SFT on OpenMathInstruct-2 (#360) Create PR to main with cherry-pick from release #160: Commit f4526ee pushed by github-merge-queue bot
15s main
main
15s
feat: Fixed metric calculation and made all grpo metrics token-level … Create PR to main with cherry-pick from release #159: Commit b960794 pushed by github-merge-queue bot
22s main
main
22s
feat: Handle Gemma3 special cases in code (#379) Create PR to main with cherry-pick from release #158: Commit fdb565c pushed by github-merge-queue bot
19s main
main
19s
ci: Add release build stage (#361) Create PR to main with cherry-pick from release #157: Commit 91c95c1 pushed by github-merge-queue bot
18s main
main
18s
fix: Save last checkpoint (#368) Create PR to main with cherry-pick from release #156: Commit 2873851 pushed by github-merge-queue bot
16s main
main
16s
fix: add missing multi-turn, container information in README (#369) Create PR to main with cherry-pick from release #155: Commit de18808 pushed by github-merge-queue bot
18s main
main
18s
ci: Migrate to use cuda image as base for container (#312) Create PR to main with cherry-pick from release #154: Commit edfd362 pushed by github-merge-queue bot
17s main
main
17s
test: make dpo functional test threshold higher until flakiness resol… Create PR to main with cherry-pick from release #153: Commit 3fbdfb7 pushed by github-merge-queue bot
21s main
main
21s
fix: recipes missing args (#365) Create PR to main with cherry-pick from release #152: Commit 0d299fb pushed by github-merge-queue bot
17s main
main
17s
docs: remove license that was erroneously copy-pasted (#357) Create PR to main with cherry-pick from release #151: Commit 69c3939 pushed by github-merge-queue bot
16s main
main
16s
fix: fix issues preventing running grpo on volta (#294) Create PR to main with cherry-pick from release #150: Commit 0a674bb pushed by github-merge-queue bot
17s main
main
17s
feat: add and log a very rough entropy approximation (#342) Create PR to main with cherry-pick from release #149: Commit b7302ad pushed by github-merge-queue bot
21s main
main
21s
fix: update the comment about why we init in fp32 (#354) Create PR to main with cherry-pick from release #148: Commit 75dfe10 pushed by github-merge-queue bot
22s main
main
22s
docs: update readme.md features (#348) Create PR to main with cherry-pick from release #147: Commit cd16f99 pushed by github-merge-queue bot
17s main
main
17s
fix: fix accumulation of loss across microbatches (#266) Create PR to main with cherry-pick from release #146: Commit af9f6e8 pushed by github-merge-queue bot
14s main
main
14s
feat: pin our python to 3.12 since python 3.13 can break ray (#343) Create PR to main with cherry-pick from release #145: Commit bb75f66 pushed by github-merge-queue bot
13s main
main
13s
docs: add docs for local concurrent clusters and fix paths (#346) Create PR to main with cherry-pick from release #144: Commit fa066ba pushed by github-merge-queue bot
18s main
main
18s
fix: sliding_window_overwrite (#331) Create PR to main with cherry-pick from release #143: Commit 35a0e09 pushed by github-merge-queue bot
15s main
main
15s
feat: improve eval (#325) Create PR to main with cherry-pick from release #142: Commit 790888f pushed by github-merge-queue bot
16s main
main
16s
feat: dual-clip in grpo loss (#311) Create PR to main with cherry-pick from release #141: Commit bc8cb65 pushed by github-merge-queue bot
14s main
main
14s
fix: reinitialize ray cluster if required (#341) Create PR to main with cherry-pick from release #140: Commit 3021098 pushed by github-merge-queue bot
18s main
main
18s
feat: Add deepscaler dataset (#335) Create PR to main with cherry-pick from release #139: Commit 9d8c1ef pushed by github-merge-queue bot
15s main
main
15s
ProTip! You can narrow down the results and go further in time using created:<2025-05-09 or the other filters available.