xiaomao19970819

Maybe there's a bug here

2

YOLO v7  YOLOv5  Hi, this code differs from yolov5. I think **m.anchors** should not be divided by the **m.stride** before running **check_anchor_order**, and that sorting should be done...

Hello,i find you use "tri_mat = np.triu(rerank_dist, 1)" in selftraining.py

1

# DBSCAN cluster tri_mat = np.triu(rerank_dist, 1) # tri_mat.dim=2 tri_mat = tri_mat[np.nonzero(tri_mat)] # tri_mat.dim=1 tri_mat = np.sort(tri_mat,axis=None) top_num = np.round(args.rho*tri_mat.size).astype(int) eps = tri_mat[:top_num].mean() print('eps in cluster: {:.3f}'.format(eps)) cluster = DBSCAN(eps=eps,min_samples=4,metric='precomputed',...

Hi, does this have a pre-training model weights?

1

Thank you for your work. Does this have a pre-training model weights? if i use my own dataset，is it accurate to use the Inceptionv3 model parameters pretrained on Imagenet to...

嗨，大佬，我们团队想拿 kuairand 数据集做一下实验，看到数据集主页写到user_id 和 video_id 都re-indexed，想问一下这个 原始id 有保存吗，我是快手商业化团队的，kim上给你留言了哈 ~

希望得到你的回复，谢谢

你好，我有一个问题想请教一下

1

DELF特征的提取匹配的过程，是否可以用于实时的图像检索系统，我实验中发现使用ransac这一步会耗费很长的时间，希望得到你的回复

你好，我在自己的数据集上对qwen2-7b进行sft后，想在相同数据结构上进行评测，遇到了问题

1

我的eval脚本配置如下 torchrun $DISTRIBUTED_ARGS src/train.py \ --deepspeed $DS_CONFIG_PATH \ --stage sft \ --do_eval \ --use_fast_tokenizer \ --flash_attn auto \ --model_name_or_path $MODEL_PATH \ --eval_dataset chg_sft_demo_test_usertask \ --template qwen \ --finetuning_type full \...