Skip to content

Issues: volcengine/verl

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

94 Open 85 Closed
94 Open 85 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Question] How to call reward model in rule based reward function?
#433 opened Mar 1, 2025 by zsychina
1
[Bug?] Attention Mask Not Rearranged in Critic Model When use_dynamic_bsz is Enabled
#431 opened Mar 1, 2025 by Viol2000
Maybe a memory check tool is need.
#430 opened Mar 1, 2025 by Jinyi6
Ray OOM causes the process to be killed
#429 opened Mar 1, 2025 by PKU-Fgx
请问训练完的actor model 里边的参数文件怎么加载测试呀
#422 opened Feb 28, 2025 by aaa559
运行你们的demo遇到这个报错是咋回事啊
#421 opened Feb 28, 2025 by xzagit
1
[Feature Request] Save checkpoint with huggingface format.
#420 opened Feb 28, 2025 by wenxueru
1
optim.warmup_style do not take effect
#418 opened Feb 28, 2025 by wenxueru
Does it support the update of the ref model and how to control its update frequency?
#415 opened Feb 28, 2025 by nomadlx
Load weight issue when world_size==1 and vllm>=0.7.0
#413 opened Feb 28, 2025 by wenxueru
SFT with Megatron backend
#412 opened Feb 28, 2025 by kechunFIVE
SPMD mode fail when VLLM_USE_V1=1 and VLLM_ENABLE_V1_MULTIPROCESSING=0 with vllm 0.7.4.dev nightly
#410 opened Feb 27, 2025 by SwordFaith
RuntimeError: CUDA error: uncorrectable ECC error encountered！！！
#407 opened Feb 27, 2025 by asirgogogo
2
[Feature Request]: Support Parallel Reward Calculation for Time-consuming Methods
#406 opened Feb 27, 2025 by AIBionics
How many H20 (96GB) GPUs are needed to train Qwen7B with the GRPO algorithm?
#401 opened Feb 27, 2025 by Tuziking
2
TP rollout + FSDP / TP actor call for contribution enhancement New feature or request
#400 opened Feb 27, 2025 by jeromeku
4
Support multi-turn rollout
#398 opened Feb 26, 2025 by casper-hansen
reward_model.micro_batch_size_per_gpu not work
#395 opened Feb 26, 2025 by yuki-666
vllm 0.6.3 error
#393 opened Feb 26, 2025 by CurryxIaoHu
1
[Question] How to use customized chat template
#390 opened Feb 26, 2025 by zpqiu
Using Megatron backend, OOM occurs when running the PPO of qwen25-32b model on 4-node H800
#388 opened Feb 26, 2025 by echo-valor
1
PPO Training Hangs at Step 0 when use_remove_padding
#387 opened Feb 26, 2025 by maksimstw
1
Support for mutliturn online RL training
#385 opened Feb 25, 2025 by UbeCc
13
Startup times are slow
#384 opened Feb 25, 2025 by casper-hansen
3
ray.exceptions.ActorDiedError: The actor died unexpectedly before finishing this task.
#383 opened Feb 25, 2025 by fengyang95
5
ProTip! Add no:assignee to see everything that’s not assigned.