-
Thank you very much for your contribution to the multimodal recommendation system!
I would like to ask if the train_mat in the dataset you provided was edited and generated by you, because I download…
-
-
Thank you for the incredible work and your outstanding contributions to the field of blade defect detection.
Could you please provide the estimated release date for the dataset and accompanying cod…
-
Hello. Thanks for your contribution.
I want to reproduce your work using dataset PRID. However, the code said it couldn't find the file "splits_prid2011.json", which I actually didn't find it the dow…
-
### OpenLineage integration/client
Spark integration
### OpenLineage version
openlineage-spark_2.12-1.23.0
### Technology and package versions
_No response_
### Environment configuration
_No re…
-
# **Description**
The **Picker component** in the [Hyperswitch SDK](https://docs.hyperswitch.io/hyperswitch-cloud/integration-guide) currently struggles with performance lags when handling large data…
-
First, I would like to express my appreciation for this incredible work. The initiative to unify and standardize open datasets is a significant contribution to the community, as it not only simplifies…
-
### Feature request
I am not sure if this is a new feature, but I wanted to post this problem here, and hear if others have ways of optimizing and speeding up this process.
Let's say I have a real…
-
Hello Authors, Fascinated by the simplicity but an efficient solution. Currently if i am not wrong, the test data has 100 samples, which are used to test the LLM's. So is it possible for the opensourc…
-
Dear author,
Thank you for your work, it's an outstanding contribution, but I have a problem when running your code.
![111](https://github.com/user-attachments/assets/be4edef5-6ebc-4c90-806a-e6c8a7a…