Gated Recurrent Unit (GRU)
وحدة البوابات التكرارية (GRU)
نوع من الوحدات التكرارية اقترحه Cho وزملاؤه (2014)، يستخدم بوابتين فقط — بوابة تحديث وبوابة إعادة تعيين — دون خلية ذاكرة منفصلة. يحقّق أداءً مكافئاً لـ LSTM بمعاملات أقل وتدريب أسرع في كثير من الحالات.
A type of recurrent unit proposed by Cho et al. (2014) that uses only two gates — update and reset — without a separate memory cell. It achieves comparable performance to LSTM with fewer parameters and often faster training.
Also translated asالوحدة التكرارية المبوَّبة
First appears in this corpus in: Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation (2014)
Appears in these papers
- Mastering Diverse Domains Through World Models2023in the sky ✦
- Neural Message Passing for Quantum Chemistry2017in the sky ✦
- RAFT: Recurrent All-Pairs Field Transforms for Optical Flow2020in the sky ✦
- Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation2014in the sky ✦
- Recurrent Neural Networks (RNNs): A Gentle Introduction and Overview2019in the sky ✦
- Recurrent Neural Networks (RNNs): A Gentle Introduction and Overview2019in the sky ✦