-
Notifications
You must be signed in to change notification settings - Fork 319
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
Loss Divergence after 250 Steps during Gemma 4 4B Joint Drafter Fine-Tuning
bugSomething isn't workingSomething isn't workingStatus: Open.#4022聽In NVIDIA-NeMo/Automodel;[RFC] Adopt PyTorch per-parameter mixed precision in AutoModel FSDP2
enhancementNew feature or requestNew feature or requestStatus: Open.#4021聽In NVIDIA-NeMo/Automodel;uv sync --extra moerebuildstransformer-engine-torchand fails on bare metalbugSomething isn't workingSomething isn't workingStatus: Open.#4020聽In NVIDIA-NeMo/Automodel;Question about pretraining model support
enhancementNew feature or requestNew feature or requestwaiting-on-customerWaiting on the original author to respondWaiting on the original author to respondStatus: Open.#4012聽In NVIDIA-NeMo/Automodel;Questions about reproducing the Qwen3.8-Flash-Next example
bugSomething isn't workingSomething isn't workingwaiting-on-customerWaiting on the original author to respondWaiting on the original author to respondStatus: Open.#4009聽In NVIDIA-NeMo/Automodel;Config loading intermittently fails when ranks share a Hugging Face cache
bugSomething isn't workingSomething isn't workingStatus: Open.#3975聽In NVIDIA-NeMo/Automodel;- Status: Open.#3971聽In NVIDIA-NeMo/Automodel;
DeepEP internode dispatch times out with NVSHMEM 3.6.5; backport DeepEP #564 into the Automodel image
bugSomething isn't workingSomething isn't workingStatus: Open.#3967聽In NVIDIA-NeMo/Automodel;[bug/guidance] DeepSeek-V4.1 Engram ID all-to-all receives byte-shifted IDs on H100 + InfiniBand unless
NCCL_PXN_DISABLE=1waiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#3959聽In NVIDIA-NeMo/Automodel;- Status: Open.#3957聽In NVIDIA-NeMo/Automodel;
[Tracking] DSV4.1 perf
enhancementNew feature or requestNew feature or requestStatus: Open.#3956聽In NVIDIA-NeMo/Automodel;GLM-5.3-Flash: which Transformers version produced the published HF logit parity?
waiting-on-customerWaiting on the original author to respondWaiting on the original author to respondStatus: Open.#3940聽In NVIDIA-NeMo/Automodel;