Search results

Integrate FlashQLA

enhancementNew feature or requestNew feature or request

Status: Open.

#2356 In NVIDIA-NeMo/Automodel;

· hemildesai opened on May 29, 2026

Refactor gemma4 drafter specific from recipes/vlm/finetune.py

Status: Open.

#2355 In NVIDIA-NeMo/Automodel;

· athitten opened on May 29, 2026

Integrate Transformer Engine EP

enhancementNew feature or requestNew feature or request

Status: Open.

#2342 In NVIDIA-NeMo/Automodel;

· hemildesai opened on May 28, 2026

[26.06] AutoModel Roadmap

Status: Open.

#2340 In NVIDIA-NeMo/Automodel;

· akoumpa opened on May 28, 2026

·  

26.06

Support MiniCPM5-1B

community-request

enhancementNew feature or requestNew feature or request

good first issueGood for newcomersGood for newcomers

Status: Open.

#2337 In NVIDIA-NeMo/Automodel;

· HuiyingLi opened on May 28, 2026

[NeMo AutoModel] 26.06 Roadmap

enhancementNew feature or requestNew feature or request

Status: Open.

#2329 In NVIDIA-NeMo/Automodel;

· anwithk opened on May 28, 2026

Fix full-training parameter dtype defaults (fp32) and torch_dtype enforcement

bugSomething isn't workingSomething isn't working

Status: Open.

#2328 In NVIDIA-NeMo/Automodel;

· yuhezhang-ai opened on May 27, 2026

Bug: deepseekv3_pretrain.yaml fails on first forward with DeepEP dispatch timeout

bugSomething isn't workingSomething isn't working

Status: Open.

#2327 In NVIDIA-NeMo/Automodel;

· yuhezhang-ai opened on May 27, 2026

Support Hy MT2

enhancementNew feature or requestNew feature or request

Feature

good first issueGood for newcomersGood for newcomers

Status: Open.

#2318 In NVIDIA-NeMo/Automodel;

· HuiyingLi opened on May 26, 2026

Feature Request: Make save_consolidated parallelism resilient to non-standard upstream shard filenames (Qwen3.5 family) — currently wastes hundreds of GPU-hours on hangs

community-request

enhancementNew feature or requestNew feature or request

waiting-on-customerWaiting on the original author to respondWaiting on the original author to respond

Status: Open.

#2279 In NVIDIA-NeMo/Automodel;

· linmuchuiyang opened on May 19, 2026

Support configurable HSDP in both expert and non-expert groups for MoEs

enhancementNew feature or requestNew feature or request

Status: Open.

#2267 In NVIDIA-NeMo/Automodel;

· hemildesai opened on May 18, 2026

Support Falcon H1

enhancementNew feature or requestNew feature or request

good first issueGood for newcomersGood for newcomers

Status: Open.

Feature

#2241 In NVIDIA-NeMo/Automodel;

· HuiyingLi opened on May 15, 2026