Please register your pre-trained embeddings here by 23:59 on Sep 30th, 2019.
- Why do we need to register for using pre-trained embeddings?
-> To make sure the challenge is fair between participants, most challenges would also require registering pre-trained embeddings (e.g., see ). This process makes sure all pre-trained embeddings are publicly available, and also, to avoid the case of custom pre-trained embeddings using leaked data.
- Will registration for pre-trained embeddings affect your
-> Not really. Throughout the time of the challenge, all participants are able to learn from each other thanks to this registration process. When all participants register in advance, all others might also be able to try different embeddings, and hence, we have fair comparisons between different solutions, i.e., given all available resources, what will be the best architecture for the task?
- For other questions, please feel free to discuss in the discussion thread:
AIviVN 8 - VLSP-SHARED Task: Hate Speech Detection on Social Networks.