Skip to content

Actions: verl-project/verl

Actions

pre-commit

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
pre-commit.yml
8,855 workflow runs
8,855 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[doc] chore: fix npu docs pre-commit #12169: Pull request #5428 synchronize by wucong25
36s wucong25:wucong/add_docker
wucong25:wucong/add_docker
36s
[doc] chore: fix npu docs pre-commit #12168: Pull request #5428 opened by wucong25
31s wucong25:wucong/add_docker
wucong25:wucong/add_docker
31s
[doc] add Dr. MAS to awesome work pre-commit #12167: Pull request #5427 synchronize by langfengQ
39s langfengQ:main
langfengQ:main
39s
[BREAKING][rollout,cfg] refactor: get rid of actor_rollout_ref config from rollout pre-commit #12165: Pull request #5418 synchronize by wuxibin89
36s wuxibin89:wuxibin/refactor_config
wuxibin89:wuxibin/refactor_config
36s
[tool] refactor: remove tool schema plumbing from SingleTurnAgentLoop… pre-commit #12164: Commit 48b367b pushed by wuxibin89
36s main
main
36s
[megatron, vllm] feat: NVFP4 (W4A16) QAT training support via ModelOpt pre-commit #12163: Pull request #5254 synchronize by jQizhang
Action required jQizhang:nvfp4_cleanup
jQizhang:nvfp4_cleanup
Action required
[perf, trtllm] feat: Add Nsight support for rollout server mode (trtllm) pre-commit #12162: Pull request #5391 synchronize by davidmlw
34s joyang-nv:liweim/nsys
joyang-nv:liweim/nsys
34s
[trainer] feat: Add Nemo-Automodel as alternative training engine pre-commit #12161: Pull request #5407 synchronize by HuiyingLi
32s HuiyingLi:add_automodel_sft_backend
HuiyingLi:add_automodel_sft_backend
32s
[tool, rollout] refactor: cache tool list initialization with MCP timeout pre-commit #12160: Pull request #5426 opened by denismegerle
Action required denismegerle:public/tool-list-caching
denismegerle:public/tool-list-caching
Action required
[trainer] feat: Add Nemo-Automodel as alternative training engine pre-commit #12159: Pull request #5407 synchronize by HuiyingLi
Action required HuiyingLi:add_automodel_sft_backend
HuiyingLi:add_automodel_sft_backend
Action required
[trainer] feat: Add Nemo-Automodel as alternative training engine pre-commit #12158: Pull request #5407 synchronize by HuiyingLi
Action required HuiyingLi:add_automodel_sft_backend
HuiyingLi:add_automodel_sft_backend
Action required
[WIP][algo] Migrate and implement the GDPO algorithm into the existing framework. pre-commit #12157: Pull request #5422 synchronize by Rhetee
Action required Rhetee:GDPO
Rhetee:GDPO
Action required
[tool] refactor: remove tool schema plumbing from SingleTurnAgentLoop pre-commit #12156: Pull request #5425 opened by denismegerle
36s denismegerle:public/single-turn-disable-tools
denismegerle:public/single-turn-disable-tools
36s
[trainer] feat: add support for the GDPO algorithm pre-commit #12155: Pull request #5409 synchronize by yue-zeng-yue
Action required yue-zeng-yue:feat-gdpo
yue-zeng-yue:feat-gdpo
Action required
[rollout, data] fix: honor train_max_samples/val_max_samples in fully… pre-commit #12154: Commit c3e3970 pushed by ArronHZG
29s main
main
29s
[tool] fix: handle empty image inputs in ToolAgentLoop (#5420) pre-commit #12153: Commit 9dd447e pushed by wuxibin89
34s main
main
34s
[trainer] feat: add support for the GDPO algorithm pre-commit #12152: Pull request #5409 synchronize by yue-zeng-yue
Action required yue-zeng-yue:feat-gdpo
yue-zeng-yue:feat-gdpo
Action required
[BREAKING][rollout,cfg] refactor: get rid of actor_rollout_ref config from rollout pre-commit #12151: Pull request #5418 synchronize by wuxibin89
38s wuxibin89:wuxibin/refactor_config
wuxibin89:wuxibin/refactor_config
38s
[misc] feat: Add code for data grouping in no-padding scenario pre-commit #12150: Pull request #5424 synchronize by Kite0011
33s Kite0011:add_rm_support
Kite0011:add_rm_support
33s
[rollout] feat: Fix partial load problem, Add vlm support for trtllm rollout pre-commit #12147: Pull request #5149 synchronize by SchumiDing
Action required SchumiDing:vlm_trtllm_support
SchumiDing:vlm_trtllm_support
Action required
why are there multiple settings for actor_rollout_ref.model.enable_gradient_checkpointing? Is this a deliberate design choice? pre-commit #12146: Pull request #4263 synchronize by khazic
Action required khazic:main
khazic:main
Action required
[rollout] feat: Fix partial load problem, Add vlm support for trtllm rollout pre-commit #12145: Pull request #5149 synchronize by SchumiDing
Action required SchumiDing:vlm_trtllm_support
SchumiDing:vlm_trtllm_support
Action required
[fsdp] fix: make GDN work correctly when using ulysses and varlen sequence pre-commit #12144: Pull request #5346 synchronize by nono-Sang
Action required nono-Sang:gdn_fix
nono-Sang:gdn_fix
Action required
[ci] feat: add profiling tests to vLLM ci pre-commit #12143: Pull request #5215 synchronize by Gary-cjy
33s Gary-cjy:main
Gary-cjy:main
33s
[BREAKING][megatron] feat: support Megatron-FSDP as a new training backend pre-commit #12142: Pull request #5423 opened by conver334
Action required conver334:megatron-fsdp
conver334:megatron-fsdp
Action required