4 citations · 6 across the 6 of their papers we have counts for
8 papers
Virtual Head Attention
Xibo Ding, Guoxia Wang, Jinle Zeng +4
The design of attention mechanisms in long-context large language models faces an ``impossible triangle'' among KV cache size, training and prefill FLOPs, and model quality. GQA ac…
FlashMask-3: Efficient and Expressive Mask-Aware Distributed Attention
Guoxia Wang, Qianyue He, Siming Wu +6
Long-context training of large language and multimodal models couples structured attention masks (document, prefix-LM, share-question, sliding-window, block-sparse) with multi-node…
SimReg: Achieving Higher Performance in the Pretraining via Embedding Similarity Regularization
Yan Sun, Guoxia Wang, Jinle Zeng +6
Pretraining large language models (LLMs) with next-token prediction has led to remarkable advances, yet the context-dependent nature of token embeddings in such models results in h…
RRAttention: Dynamic Block Sparse Attention via Per-Head Round-Robin Shifts for Long-Context Inference
Siran Liu, Guoxia Wang, Sa Wang +7
The quadratic complexity of attention mechanisms poses a critical bottleneck for large language models processing long contexts. While dynamic sparse attention methods offer input-…
ERNIE 5.0 Technical Report
Haifeng Wang, Hua Wu, Tian Wu +432
In this report, we introduce ERNIE 5.0, a natively autoregressive foundation model desinged for unified multimodal understanding and generation across text, image, video, and audio…
AdaGC: Enhancing LLM Pretraining Stability via Adaptive Gradient Clipping
Guoxia Wang, Shuai Li, Congliang Chen +5
Loss spikes remain a persistent obstacle in large-scale language model pretraining. While previous research has attempted to identify the root cause of loss spikes by investigating…
FlashMask: Efficient and Rich Mask Extension of FlashAttention
Guoxia Wang, Jinle Zeng, Xiyuan Xiao +7
The computational and memory demands of vanilla attention scale quadratically with the sequence length , posing significant challenges for processing long sequences in Transform…
Boosting Distributed Training Performance of the Unpadded BERT Model
Jinle Zeng, Min Li, Zhihua Wu +4
Pre-training models are an important tool in Natural Language Processing (NLP), while the BERT model is a classic pre-training model whose structure has been widely adopted by foll…