48 citations · 81 across the 9 of their papers we have counts for
9 papers
Skrull: Towards Efficient Long Context Fine-tuning through Dynamic Data Scheduling
Hongtao Xu, Wenting Shen, Yuanxin Wei +6
Long-context supervised fine-tuning (Long-SFT) plays a vital role in enhancing the performance of large language models (LLMs) on long-context tasks. To smoothly adapt LLMs to long…
Efficient Long Context Fine-tuning with Chunk Flow
Xiulong Yuan, Hongtao Xu, Wenting Shen +10
Long context fine-tuning of large language models(LLMs) involves training on datasets that are predominantly composed of short sequences and a small proportion of longer sequences.…
BladeDISC++: Memory Optimizations Based On Symbolic Shape
Xiulong Yuan, Xu Yan, Wenting Shen +5
Recent deep learning workloads exhibit dynamic characteristics, leading to the rising adoption of dynamic shape compilers. These compilers can generate efficient kernels for dynami…
ROAM: memory-efficient large DNN training via optimized operator ordering and memory layout
Huiyao Shu, Ang Wang, Ziji Shi +3
As deep learning models continue to increase in size, the memory requirements for training have surged. While high-level techniques like offloading, recomputation, and compression…
TAPAS: Fast and Automatic Derivation of Tensor Parallel Strategies for Large Neural Networks
Ziji Shi, Le Jiang, Ang Wang +6
Tensor parallelism is an essential technique for distributed training of large neural networks. However, automatically determining an optimal tensor parallel strategy is challengin…
Revisiting and Advancing Chinese Natural Language Understanding with Accelerated Heterogeneous Knowledge Pre-training
Taolin Zhang, Junwei Dong, Jianing Wang +6
Recently, knowledge-enhanced pre-trained language models (KEPLMs) improve context-aware representations via learning from structured relations in knowledge graphs, and/or linguisti…
M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining
Junyang Lin, An Yang, Jinze Bai +9
Recent expeditious developments in deep learning algorithms, distributed training, and even hardware design for large models have enabled training extreme-scale models, say GPT-3 a…
M6-T: Exploring Sparse Expert Models and Beyond
An Yang, Junyang Lin, Rui Men +12
Mixture-of-Experts (MoE) models can achieve promising results with outrageous large amount of parameters but constant computation cost, and thus it has become a trend in model scal…
M6: A Chinese Multimodal Pretrainer
Junyang Lin, Rui Men, An Yang +22
In this work, we construct the largest dataset for multimodal pretraining in Chinese, which consists of over 1.9TB images and 292GB texts that cover a wide range of domains. We pro…