-
Notifications
You must be signed in to change notification settings - Fork 491
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
Distillation loss path upcasts the full [B, S, V] logits to fp32 when only K columns are read
bugSomething isn't workingSomething isn't workingStatus: Open.#3495 In NVIDIA-NeMo/RL;NeMo RL v0.8.0 Roadmap
DocumentationImprovements or additions to documentationImprovements or additions to documentationStatus: Open.#3493 In NVIDIA-NeMo/RL;Add official NeMo RL container support and validation for GB300
DocumentationImprovements or additions to documentationImprovements or additions to documentationStatus: Open.#3492 In NVIDIA-NeMo/RL;async grpo: trainer waits on an empty buffer forever after the trajectory collector dies
bugSomething isn't workingSomething isn't workingStatus: Open.#3488 In NVIDIA-NeMo/RL;Support hetgroups for ray head
enhancementNew feature or requestNew feature or requestStatus: Open.#3474 In NVIDIA-NeMo/RL;- Status: Open.#3464 In NVIDIA-NeMo/RL;
- Status: Open.#3462 In NVIDIA-NeMo/RL;
- Status: Open.#3459 In NVIDIA-NeMo/RL;
Avoid repeated weight fake-quantization during QAT logprob rescoring
bugSomething isn't workingSomething isn't workingwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#3440 In NVIDIA-NeMo/RL;- Status: Open.#3432 In NVIDIA-NeMo/RL;
- Status: Open.#3431 In NVIDIA-NeMo/RL;
- Status: Open.#3425 In NVIDIA-NeMo/RL;
