Search results

Bump cudnn to 9.14 to support packed sequence fused attention

enhancementNew feature or requestNew feature or request

Status: Open.

#1819 In NVIDIA-NeMo/RL;

· guyueh1 opened on Jan 23, 2026

Online DPO Support

community-request

enhancementNew feature or requestNew feature or request

Status: Open.

#1816 In NVIDIA-NeMo/RL;

· dhineshkumar-r opened on Jan 23, 2026

Async data collector is single-epoch

bugSomething isn't workingSomething isn't working

Status: Open.

#1814 In NVIDIA-NeMo/RL;

· smahdavi4 opened on Jan 23, 2026

config option to disable monotonicity checks and on policy token id correction

enhancementNew feature or requestNew feature or request

Status: Open.

#1812 In NVIDIA-NeMo/RL;

· cmunley1 opened on Jan 22, 2026

GPT-OSS on-policy distillation fails during checkpoint load (Missing key: model.layers.0.mlp.experts.down_proj)

bugSomething isn't workingSomething isn't working

community-request

t-gptoss

Status: Open.

#1810 In NVIDIA-NeMo/RL;

· uygarmv opened on Jan 22, 2026

token_mult_prob_error continuously increases during GRPO training with DeepSeek-R1-0528-Qwen3-8B using Megatron backend, leading to training crash

bugSomething isn't workingSomething isn't working

Status: Open.

#1801 In NVIDIA-NeMo/RL;

· binhu-nv opened on Jan 21, 2026

Async validation takes too long in some envs

bugSomething isn't workingSomething isn't working

Status: Open.

#1798 In NVIDIA-NeMo/RL;

· smahdavi4 opened on Jan 20, 2026

DSV3 on GB200 + seq-packing not supported

bugSomething isn't workingSomething isn't working

Status: Open.

#1775 In NVIDIA-NeMo/RL;

· terrykong opened on Jan 14, 2026

GPT-OSS has high logprob errors on GB200

bugSomething isn't workingSomething isn't working

Status: Open.

#1774 In NVIDIA-NeMo/RL;

· ashors1 opened on Jan 14, 2026

Remove Gym dependency contract to smooth out Gym-side dependency changes

enhancementNew feature or requestNew feature or request

Status: Open.

#1769 In NVIDIA-NeMo/RL;

· bxyu-nvidia opened on Jan 13, 2026

Polish lambda ray.sub and upstream

enhancementNew feature or requestNew feature or request

Status: Open.

#1765 In NVIDIA-NeMo/RL;

· terrykong opened on Jan 13, 2026

Support QAT for low precision RL

enhancementNew feature or requestNew feature or request

Status: Open.

#1750 In NVIDIA-NeMo/RL;

· sharonyu-115 opened on Jan 9, 2026