Search results

ASR batch transcription crash with torch.cuda.empty_cache()

bugSomething isn't workingSomething isn't working

Status: Open.

#14727 In NVIDIA-NeMo/NeMo;

· SystemPanic opened on Sep 13, 2025

Activation recomputation doc does not match implementation

bugSomething isn't workingSomething isn't working

Status: Open.

#14726 In NVIDIA-NeMo/NeMo;

· zinccat opened on Sep 12, 2025

NeMo CPU conversion

Status: Open.

#14725 In NVIDIA-NeMo/NeMo;

· mgilmore-relace opened on Sep 12, 2025

where is nemo/collections/diffusion/train.py?

Status: Open.

#14723 In NVIDIA-NeMo/NeMo;

· parryppp opened on Sep 12, 2025

Wrong computation of <code>max_train_samples</code> in <code>FineTuningDataModule</code>

bugSomething isn't workingSomething isn't working

Status: Open.

#14720 In NVIDIA-NeMo/NeMo;

· albertoveneri opened on Sep 11, 2025

Resume training causes ckpt saving problem

bugSomething isn't workingSomething isn't working

Status: Open.

#14718 In NVIDIA-NeMo/NeMo;

· yspMing opened on Sep 11, 2025

<code>Missing key in checkpoint</code> error while converting a legacy Nemotron checkpoint

Status: Open.

#14713 In NVIDIA-NeMo/NeMo;

· albertoveneri opened on Sep 10, 2025

Training GPT-OSS-20B caused OOM

bugSomething isn't workingSomething isn't working

Status: Open.

#14712 In NVIDIA-NeMo/NeMo;

· kltlj1 opened on Sep 10, 2025

parakeet-tdt_ctc loss scale

Status: Open.

#14710 In NVIDIA-NeMo/NeMo;

· ehgusehgus opened on Sep 10, 2025

Is <code>MegatronOptimizerModule</code>'s <code>adam</code> optimizer actually <code>adamw</code>?

Status: Open.

#14695 In NVIDIA-NeMo/NeMo;

· pilot7747 opened on Sep 9, 2025

Error while using aggregate_moe_loss_stats

bugSomething isn't workingSomething isn't working

Status: Open.

#14692 In NVIDIA-NeMo/NeMo;

· prathamk-tw opened on Sep 9, 2025

GPT-OSS has the wrong value of yarn_original_max_position_embeddings

bugSomething isn't workingSomething isn't working

Status: Open.

#14686 In NVIDIA-NeMo/NeMo;

· kzhou92 opened on Sep 9, 2025