-
Hey, @wondervictor thank you and your team for this wonderful model!
Could we have some clearance on how the demo in [HF](https://huggingface.co/spaces/stevengrove/YOLO-World) is different from you…
-
### Is your feature request related to a problem or challenge?
DataFusion optimizes queries like `... ORDER BY value LIMIT 10` by only keeping the top 10 ("limit") rows when sorting which is great!
…
alamb updated
4 months ago
-
我运行了READMD.md里面的命令python main.py --train --test --recur --get_rule --data family --topk 2400 --gpu 1,但是只获得了一个ours_top2400.txt的输出文件，请问关于MRR等评测数据是如何生成的呢？
-
Thank you for your excellent work! But I have a problem about the graph symmetry. The article says the dependency patterns between sensors need not be symmetric. However, in the case without prior inf…
-
For the topk function, when I used it as
, ind = tmpOutput:topk(1, true)
Where tmpOutput is a CudaTensor with size 1 by K, K is 1024.
The ind_ some times turn out to be some negative or float or out o…
-
In the deformable transformer block code in /Lite-DETR/models/dino/deformable_transformer.py, I had an issue while running batch inferences on CoCo eval data.
topk_proposals = torch.topk(enc_outpu…
-
Hi, @zpao @iassael @maraoz @achao @gfosco
I have met this error:
------------------------------------------------
./train.lua:137: bad argument #1 to 'topk' (k not in range for dimension
------…
-
The data is:
|class |name | score |
| - | - | - |
|1 | A | 90 |
|1 | B | 92 |
|1 | C | 92 |
|1 | D | 90 |
|1 | E | 99 |
|2 | F | 99 |
|2 | G| 100 |
|2 | H| 101 |
|2 | I| 110 |
|2 | J | 112…
-
**Is your feature request related to a problem? Please describe.**
Graph networks often use sort_pooling, which requires sorting all the elements in a segment vector. The segment sizes are variable…
-
作者您好，请问文章在使用topk生成mask进而计算注意力时，有考虑过此操作的可微性？
这种离散的操作我认为是不可微分的，那么神经网络在训练时是如何处理这个问题呢？换句话说，topk是如何奏效的？因为不可微分也许会干扰神经网络的训练，但是文章采用了topk效果非常好。
对这个问题我很困惑，还请作者在百忙之间解答一下我的疑惑，谢谢！