-
最近读了您的论文“SemiSANet: A Semi-Supervised High-Resolution Remote Sensing Image Change Detection Model Using Siamese Networks with Graph Attention”,受益匪浅。如果可能的话,我想向您寻求一份该代码的副本,以便更好地探索和理解您的研究成果。
如果您能够提供该代码或…
-
在LCQMC,亲测,直接用albert的fine-tuneing ,1 epoch,F1 score=0.84032。直接用你的模型,F1 score=0.80712
-
Hey,
First, thanks for the kind words in various places :). I came across your posts, which led me here.
I also spent quite some time working on similarity models. I think they're surprisingly d…
-
We need to convert keras.io examples to work with Keras 3.
This involves two stages:
## Stage 1: tf.keras backwards compatibility check
Keras 3 is intended as a drop-in replacement for tf.ker…
-
How are we not doing 10000 x 10000 operations in case of Sentence-BERT(for 10000 sentences)? It is understandable that we are using two different BERT architecture for a pair of sentences, but aren't …
-
[deeplearning.docx](https://github.com/ssfc/machine-learning/files/3323819/deeplearning.docx)
-
Deadline: 16/08
-
-
텍스트 리뷰 임베딩 데이터 저장하기
-> 2가지 방식
1. Spark NLP 모델 사용하기
2. 허깅페이스 모델 사용하기
리서치 영역
-> 데이터 마다 다르니깐, 쇼핑 리뷰에 가까운 모델을 찾으면 베스트.
-> sentence-bert
**1. SparkNLP**
https://sparknlp.org/2020/08/25/sent_s…
-
To train nets:
- we generate millions of selfplay games at a given visit number (1600 visits currently)
- we then only use ~10% of the positions generated to train candidate nets (the ~10% has to be…