-
按照默认代码跑了三次,把earlystop都调大了,也没跑出作者的声称的效果
Best Iter = recall:0.13969, pres:0.02932, hr:0.37911, ndcg:0.17657
Best Iter = recall:0.13646, pres:0.02885, hr:0.37792, ndcg:0.17680
Best Iter = recall:0.132…
-
Multi-vector MaxSim is increasingly important and we have optimizations for float cell precision, but I think we should also consider optimize for int8 with hamming as it approximates the dotproduct f…
-
Hello,
I am following the tutorial here https://www.tensorflow.org/recommenders/examples/dcn to build a dcn model, but I want to have ndcg as a metric. Then I did
`model.compile(optimizer='sgd', …
-
Hello!
I am trying to reproduce quality of monoT5 on [BEIR](https://github.com/beir-cellar/beir) benchmark from the recent [article](https://arxiv.org/pdf/2206.02873.pdf). But after running script …
-
Hi,
Thanks for the repo! One thing that is unclear to me is how you evaluated your models on the Amazon-C4 dataset. In your repo, you mentioned that _"sampled_item_metadata_1M.jsonl contains ~1M it…
-
When training, the test loss does not decrease (as it should), but the Hit Ratio keeps increasing. Doesn't this suggest that the loss function being used is incorrect?
For example here when training …
-
I tried to evaluate both unicamp-dl/mt5-base-en-msmarco and unicamp-dl/mt5-13b-mmarco-100k, but the performance of 13b is lower than base model. Here is a simple comparison of reranking results of BM2…
-
## Why
Machine Learning 輪講は最新の技術や論文を追うことで、エンジニアが「技術で解決できること」のレベルをあげていくことを目的にした会です。
prev. #256
## What
話したいことがある人はここにコメントしましょう!
面白いものを見つけた時点でとりあえず話すという宣言だけでもしましょう!
-
I did an experiment that delete the propagation on graph (comment line:194-221), which means that only do dot product. It is amazing that it can get comparible results in gollaw and amazon dataset.
…
-
@massquantity 网络损失从第三轮左右就变得非常大,ndcg一直很小,你的运行结果也是这样吗,不知道哪里的问题
![image](https://user-images.githubusercontent.com/96457748/229278761-f623c220-6433-4c28-962e-2c8cf20f5a06.png)