-
Notifications
You must be signed in to change notification settings - Fork 177
Pull requests: NVIDIA-NeMo/Automodel
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
feat(speculative): add EAGLE-3 sequence packing and reasoning-mode control
community-request
#2444
opened Jun 7, 2026 by
khazic
Contributor
Loading…
3 tasks done
feat(diffusion): improve qwen image finetuning configs
#2442
opened Jun 7, 2026 by
pthombre
Contributor
Loading…
3 tasks done
fix(moe): include MTP modules in FSDP sync traversal
#2441
opened Jun 6, 2026 by
HuiyingLi
Contributor
Loading…
refactor(datasets): unify reasoning_content coercion in agent chat
community-request
waiting-on-customer
Waiting on the original author to respond
#2440
opened Jun 6, 2026 by
khazic
Contributor
Loading…
fix(tokenizer): make NeMoAutoTokenizerWithBosEosEnforced picklable
community-request
#2439
opened Jun 6, 2026 by
khazic
Contributor
Loading…
chore(skills): refresh distributed training signature
docs-only
With great power comes great responsibility.
#2438
opened Jun 5, 2026 by
akoumpa
Contributor
Loading…
3 tasks done
feat(dllm): add I-DLM all-masked trainer (dllm.mode idlm)
community-request
#2437
opened Jun 5, 2026 by
kashif
Contributor
Loading…
[model][perf] feat: FFPA D=512 attention backend for Gemma4 (3× fwd / 6× bwd vs SDPA)
community-request
#2436
opened Jun 5, 2026 by
Butterfingrz
Loading…
refactor(speculative): reuse shared dflash mask and loss in the trainer
community-request
#2433
opened Jun 5, 2026 by
kashif
Contributor
Loading…
fix: load Qwen2.5-VL vision tower weights under FSDP2 - Do not merge
#2431
opened Jun 5, 2026 by
yuekaizhang
Contributor
Loading…
ci: Update transformers to latest version 5.10.2
#2430
opened Jun 5, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
feat(gemma4): add context parallel specifics
#2427
opened Jun 5, 2026 by
HuiyingLi
Contributor
Loading…
2 of 3 tasks
feat: Add query functionality of Model Capability Registry
#2423
opened Jun 4, 2026 by
athitten
Contributor
Loading…
3 tasks done
fix(precision): dtype contract bug fixes for FSDP2 mixed-dtype loads
#2419
opened Jun 4, 2026 by
yuhezhang-ai
Contributor
Loading…
8 tasks done
fix(checkpoint): add bounded retention window
#2416
opened Jun 4, 2026 by
oliverholworthy
Contributor
Loading…
3 tasks done
feat(speculative): add target_attn_implementation knob for EAGLE-3 target
community-request
waiting-on-customer
Waiting on the original author to respond
#2415
opened Jun 4, 2026 by
kashif
Contributor
Loading…
ci: Update transformers to latest version 5.10.1
#2410
opened Jun 4, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
feat(models): support fused linear cross-entropy across custom models
#2397
opened Jun 3, 2026 by
akoumpa
Contributor
Loading…
2 of 3 tasks
feat(moe): enable MXFP8 MoE training on GB200 (TransformerEngine + torchao)
#2394
opened Jun 3, 2026 by
hemildesai
Contributor
•
Draft
docs(fern): relocate Fern under docs/ and remove legacy Sphinx tree
#2391
opened Jun 2, 2026 by
lbliii
Contributor
Loading…
feat(distributed): add selective activation checkpointing for FSDP2
#2389
opened Jun 2, 2026 by
yuhezhang-ai
Contributor
Loading…
9 tasks done
Previous Next
ProTip!
Mix and match filters to narrow down what you’re looking for.