Issues · huggingface/pytorch-image-models · GitHub

Labels Milestones New issue

[BUG] torch.compile applied at task level has no effect during validation and causes multi-node stalls

#2693

· janbernloehr opened

on Apr 10, 2026

[BUG] The behaviour of `param_groups_layer_decay()` towards `requires_grad==False` parameters

#2691

· sjiang95 opened

on Apr 9, 2026

[FEATURE] EUPE pretrained weights

#2690

· ankandrew opened

on Apr 4, 2026

[FEATURE] 3D model support for volumetric / medical imaging in timm

#2628

· CowboyH opened

on Dec 10, 2025

[FEATURE] Add DeepSpeed / ZeRO support in training script

#2623

· Tianyi-Franklin-Wang opened

on Dec 5, 2025

[FEATURE] [RFC] Support for interchangable attention backends

#2607

· fffffgggg54 opened

on Nov 8, 2025

[FEATURE] Support for Sapiens

#2597

· chenzhekl opened

on Oct 17, 2025

[FEATURE] Support for DPT decoder

#2551

· dqj5182 opened

on Jul 16, 2025

[BUG] naflexvit_so400m_patch16_siglip has undocumented different default pos_embed_interp_mode of "bicubic" instead of "bilinear"

#2542

· redhottensors opened

on Jul 9, 2025

[FEATURE] Unified `embed` Interfaces for Vision Transformer Models for MIM Pretrain Research

#2540

· ryan-minato opened

on Jul 9, 2025

[FEATURE] Adding Agglomerative Vision Foundation Model such as RADIO

#2505

· RaptorMai opened

on May 30, 2025

[FEATURE] timm.models.adapt_input_conv: beyond RGB weights

#2445

· adamjstewart opened

on Feb 25, 2025