wingvortex

How to speed up the fid?

2

Hi, thanks for sharing your work. In my case, I have a reference folder (fdir1), and about 50 target folders (fdir2), each containing hundreds of images. It takes a long...

Does knowledge distillation support multi-gpu?

2

Hi, thanks for your sharing. When I tried to use multi-gpu to train Knowledge Distillation: `python3 -m torch.distributed.run --nproc_per_node $N_GPU distillation.py ...` I got the error: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: distillation.py FAILED Failures:

NameError: name '_update_fn' is not defined when using the SAM

13

Hi, I installed the ml-backend according to the official steps: ` git clone https://github.com/HumanSignal/label-studio-ml-backend.git` ` cd label-studio-ml-backend/` ` pip install -e .` Then I installed SAM and started label-studio-ml: `pip...

ModuleNotFoundError: No module named 'primitives_fitting.py_primitives_fitting'

5

Hi, when I run `import primitives_fitting as pf,` I got the this error: ModuleNotFoundError: No module named 'primitives_fitting.py_primitives_fitting' Since I only need the primitives fitting, so it is better to...

Hi, thank you for sharing the code. I’m currently evaluating the use of PointDSC for point cloud registration in my project. However, I couldn’t find a license file. Could you...

Hi, I tried to load diffusers dreambooth flux lora，but failed with no error. It turns out that, the match results returned 'None': ` match_results = lora.match(model, state_dict)` Any ideas how...

Discrepancy in ROUGE BLEU Scores During Training vs. Final Results

2

### Reminder - [x] I have read the above rules and searched the existing issues. ### System Info Ubuntu ### Reproduction 你好。我发现训练中保存的ROUGE BLEU Scores（trainer_state.json）比最终的分数（eval_results.json）低很多。 配置如下： ### model model_name_or_path: Qwen/Qwen2.5-VL-7B-Instruct image_max_pixels:...