Skip to content

Actions: verl-project/verl

Actions

e2e_one_step_off_policy

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
e2e_one_step_off_policy.yml
4,318 workflow runs
4,318 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[rollout] feat: support auto resume on abort in FullyAsyncLLMServerManager e2e_one_step_off_policy #6020: Pull request #5430 synchronize by wuxibin89
12m 56s wuxibin89:wuxibin/resume_client
wuxibin89:wuxibin/resume_client
12m 56s
[rollout] feat: support auto resume on abort in FullyAsyncLLMServerManager e2e_one_step_off_policy #6019: Pull request #5430 synchronize by wuxibin89
26m 17s wuxibin89:wuxibin/resume_client
wuxibin89:wuxibin/resume_client
26m 17s
[model] feat: add qwen3.5 fsdp grpo training support with npu optimization e2e_one_step_off_policy #6018: Pull request #5434 opened by Kirrito-k423
Action required Kirrito-k423:qwen3.5_fsdp_npu
Kirrito-k423:qwen3.5_fsdp_npu
Action required
[BREAKING][rollout,cfg] refactor: get rid of actor_rollout_ref config… e2e_one_step_off_policy #6017: Commit de6a1bc pushed by ArronHZG
25m 10s main
main
25m 10s
[megatron, vllm] feat: NVFP4 (W4A16) QAT training support via ModelOpt e2e_one_step_off_policy #6016: Pull request #5254 synchronize by jQizhang
Action required jQizhang:nvfp4_cleanup
jQizhang:nvfp4_cleanup
Action required
[WIP][algo] Migrate and implement the GDPO algorithm into the existing framework. e2e_one_step_off_policy #6015: Pull request #5422 synchronize by Rhetee
Action required Rhetee:GDPO
Rhetee:GDPO
Action required
[WIP][algo] Migrate and implement the GDPO algorithm into the existing framework. e2e_one_step_off_policy #6014: Pull request #5422 synchronize by Rhetee
Action required Rhetee:GDPO
Rhetee:GDPO
Action required
[WIP][algo] Migrate and implement the GDPO algorithm into the existing framework. e2e_one_step_off_policy #6013: Pull request #5422 synchronize by Rhetee
Action required Rhetee:GDPO
Rhetee:GDPO
Action required
[rollout] refactor: bucketed transfer utils e2e_one_step_off_policy #6012: Pull request #5309 synchronize by pengwu22
6m 29s pengwu22:pw/test-wt-utils
pengwu22:pw/test-wt-utils
6m 29s
[rollout] refactor: bucketed transfer utils e2e_one_step_off_policy #6011: Pull request #5309 synchronize by pengwu22
12m 56s pengwu22:pw/test-wt-utils
pengwu22:pw/test-wt-utils
12m 56s
[rollout] feat: support auto resume on abort in FullyAsyncLLMServerManager e2e_one_step_off_policy #6010: Pull request #5430 opened by wuxibin89
8m 57s wuxibin89:wuxibin/resume_client
wuxibin89:wuxibin/resume_client
8m 57s
[misc] feat: Add code for data grouping in no-padding scenario (#5424) e2e_one_step_off_policy #6009: Commit c3fc222 pushed by vermouth1992
19m 38s main
main
19m 38s
[ci] feat: add profiling tests to vLLM ci e2e_one_step_off_policy #6008: Pull request #5215 synchronize by Gary-cjy
Action required Gary-cjy:main
Gary-cjy:main
Action required
[BREAKING][rollout,cfg] refactor: get rid of actor_rollout_ref config from rollout e2e_one_step_off_policy #6007: Pull request #5418 synchronize by wuxibin89
51m 57s wuxibin89:wuxibin/refactor_config
wuxibin89:wuxibin/refactor_config
51m 57s
[tool] refactor: remove tool schema plumbing from SingleTurnAgentLoop… e2e_one_step_off_policy #6006: Commit 48b367b pushed by wuxibin89
57m 16s main
main
57m 16s
[megatron, vllm] feat: NVFP4 (W4A16) QAT training support via ModelOpt e2e_one_step_off_policy #6005: 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) e2e_one_step_off_policy #6004: Pull request #5391 synchronize by davidmlw
51m 33s joyang-nv:liweim/nsys
joyang-nv:liweim/nsys
51m 33s
[trainer] feat: Add Nemo-Automodel as alternative training engine e2e_one_step_off_policy #6003: Pull request #5407 synchronize by HuiyingLi
Queued HuiyingLi:add_automodel_sft_backend
HuiyingLi:add_automodel_sft_backend
Queued
[tool, rollout] refactor: cache tool list initialization with MCP timeout e2e_one_step_off_policy #6002: 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 e2e_one_step_off_policy #6001: 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 e2e_one_step_off_policy #6000: 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. e2e_one_step_off_policy #5999: Pull request #5422 synchronize by Rhetee
Action required Rhetee:GDPO
Rhetee:GDPO
Action required
[tool] refactor: remove tool schema plumbing from SingleTurnAgentLoop e2e_one_step_off_policy #5998: Pull request #5425 opened by denismegerle
4m 55s denismegerle:public/single-turn-disable-tools
denismegerle:public/single-turn-disable-tools
4m 55s
[trainer] feat: add support for the GDPO algorithm e2e_one_step_off_policy #5997: 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… e2e_one_step_off_policy #5996: Commit c3e3970 pushed by ArronHZG
2h 14m 17s main
main
2h 14m 17s