Issues

• ImageNet-1K DINOv2 checkpoints

  facebookresearch/dinov2#600 · Filter by author tomburgert opened on Oct 2, 2026

• cv技术交流

  facebookresearch/dinov2#597 · Filter by author zhaoshuo1223 opened on Jun 1, 2026

• Trouble accessing DINO-BoC weights

  facebookresearch/dinov2#596 · Filter by author pressm opened on May 18, 2026

• 📋 Documentation Enhancement Suggestion

  facebookresearch/dinov2#587 · Filter by author croviatrust opened on Mar 4, 2026

• Training DINO-BoC on CHAMMI Train Split (Table 4) and ViT-S Stability

  facebookresearch/dinov2#581 · Filter by author Bulat27 opened on Feb 3, 2026
  1

• Has anyone tried implementing dinov2 as backbone to any of the config files from MMdetection?

  facebookresearch/dinov2#579 · Filter by author shvardhan1994 opened on Jan 21, 2026

• torch.jit.script

  facebookresearch/dinov2#574 · Filter by author mxjtufeigmail opened on Dec 3, 2025
  1

• Application of Remote Sensing Dataset

  facebookresearch/dinov2#571 · Filter by author Jay2002-whu opened on Nov 24, 2025

• LVD Dataset containes retinal images?

  facebookresearch/dinov2#568 · Filter by author pnderitu opened on Nov 10, 2025
  3

  

• Can DINOv2 be used to train my own unlabeled dataset?

  facebookresearch/dinov2#567 · Filter by author linglizhang-seu opened on Nov 7, 2025
  1

• Feature request: Semantic audit module for visual embeddings

  facebookresearch/dinov2#566 · Filter by author elly99-AI opened on Oct 29, 2025

• real cases tested about your segmentation function

  facebookresearch/dinov2#565 · Filter by author shukami1 opened on Oct 10, 2025

• Question regarding input resolution during distillation to the smaller model

  facebookresearch/dinov2#560 · Filter by author ajamal-isrg opened on Aug 21, 2025

• Fine-tuning DINOv2 ViT-L/14 for Jewelry Image Retrieval Results Worse Than Pretrained

  facebookresearch/dinov2#559 · Filter by author Sushant200210 opened on Aug 21, 2025

• High GPU Memory Usage When Fine-tuning DINOv2 Compared to Similar ViTs

  facebookresearch/dinov2#553 · Filter by author WeiQijie opened on Aug 11, 2025

• randon result for multiple trials with the same input image

  facebookresearch/dinov2#551 · Filter by author qyjyork opened on Aug 8, 2025
  1

• Clarification on the definition of N in Section 3.1 (Scaling Channel-Adaptive SSL)

  facebookresearch/dinov2#550 · Filter by author LIYangggggg opened on Aug 6, 2025

• layerwise decay different value

  facebookresearch/dinov2#545 · Filter by author chagmgang opened on Jul 22, 2025

• Checkpoint state_dict contains keys that are not used by the model when resuming pre-training from intermediate checkpoint

  facebookresearch/dinov2#544 · Filter by author ajamal-isrg opened on Jul 17, 2025
  6

• The example provided in dinotxt.

  facebookresearch/dinov2#543 · Filter by author llcshappy opened on Jul 14, 2025

• Fine Tuning using Self Supervised Learning to OOD Data

  facebookresearch/dinov2#541 · Filter by author alexisdrakopoulos opened on Jul 5, 2025

• How to train from 224 to 518?

  facebookresearch/dinov2#540 · Filter by author Jokerjokerspy opened on Jul 3, 2025

• Notebook examples don't run on master (outdated conda envs)

  facebookresearch/dinov2#536 · Filter by author janblumenkamp opened on Jun 26, 2025

• Error in student forward?

  facebookresearch/dinov2#535 · Filter by author Jeffkang-94 opened on Jun 24, 2025
  4

• DINOv2 meets text - Zero-shot segmentation reproductibility

  facebookresearch/dinov2#530 · Filter by author alexcbb opened on Jun 16, 2025
  18