Finetune SAM3 with LoRA — optimized for images. A simple setup for training SAM3 on image datasets. Video finetuning is not yet supported but planned for future releases.
Python · active 2025-12-05 → 2026-07-05 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 213 days from 2025-12-05 to 2026-07-05. Pushes: 24 total, peak 4 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 8 total, peak 2 in a day. Comments: 28 total, peak 5 in a day. Stars: 43 total, peak 4 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| Sompote | 39 | 24 | 0 | 15 |
| ALingll | 4 | 0 | 0 | 1 |
| mythosx6 | 4 | 0 | 0 | 4 |
| blackdyc | 3 | 0 | 0 | 3 |
| shenxhui | 3 | 0 | 0 | 2 |
| Ayoob7 | 1 | 0 | 0 | 0 |
| kerongYan | 1 | 0 | 0 | 0 |
| atharvadeore999 | 1 | 0 | 0 | 1 |
| DaryAvelagrom | 1 | 0 | 0 | 1 |
| clearisoft | 1 | 0 | 0 | 0 |
| Wangduo428 | 1 | 0 | 0 | 0 |
| mathanprasannakumar | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#28Sompote2026-07-05 00:02apply_to_mask_decoder doesn’t work？
- Issue comment#30Sompote2026-07-04 23:57How to merge the Lora weights to the SAM3
- Issue comment#33Sompote2026-07-04 23:51有相关论文的发布或投稿吗？如有是否可以提供预印版本
- Issue#37Wangduo4282026-06-28 07:57About the validation code
- Issue comment#29mathanprasannakumar2026-04-04 05:40VRAM requirements for full lora finetuning ?
- Issue comment#15ALingll2026-03-30 01:34FileNotFoundError(f"LoRA weights not found: {weights_path}")
- Issue comment#24Sompote2026-03-07 05:13Does this codebase support performing inference with visual prompts (e.g., bounding boxes or point prompts)?
- Issue comment#22Sompote2026-02-22 02:20Preserving semantic alignment and class discriminability after fine-tuning
- Issue#21clearisoft2026-02-04 11:35when support video？
- Issue comment#20shenxhui2026-02-04 10:23Query only gets single mode category for multi-class images with multiple masks
- Issue comment#20DaryAvelagrom2026-02-04 09:22Query only gets single mode category for multi-class images with multiple masks
- Issue#19shenxhui2026-02-02 03:04predicted bounding boxes deviate significantly from the ground-truth annotations
- Issue comment#19shenxhui2026-02-02 03:04predicted bounding boxes deviate significantly from the ground-truth annotations
- Issue comment#14Sompote2026-01-27 03:59Point prompts
- Issue comment#16Sompote2026-01-27 03:57Grounding格式数据训练
- Issue#12ALingll2026-01-12 09:55Missing sam3_lora_wrapper.py
- Issue#11ALingll2026-01-12 09:51Multiple GPUs Training Issue
- Issue#15Ayoob72026-01-10 18:47FileNotFoundError(f"LoRA weights not found: {weights_path}")
- Issue comment#6mythosx62026-01-06 01:29Model Training Performance Analysis Request
- Issue comment#10atharvadeore9992026-01-05 08:00CheckpointError
- Issue comment#10Sompote2026-01-05 03:35CheckpointError
- Issue comment#11Sompote2026-01-05 03:33Multiple GPUs Training Issue
- Issue comment#11Sompote2026-01-05 03:32Multiple GPUs Training Issue
- Issue comment#6mythosx62026-01-05 01:15Model Training Performance Analysis Request
- Issue comment#6mythosx62026-01-04 02:21Model Training Performance Analysis Request
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 43 stars here means stars gained during the window, not the repo's star count.