TAGC: Optimizing Gradient Communication in Distributed Transformer Training
This work addresses efficiency issues in distributed training for AI researchers and practitioners, but it is incremental as it builds on existing compression techniques.
The paper tackles the bottleneck of gradient synchronization in distributed training of large language models by introducing Transformer-Aware Gradient Compression (TAGC), which accelerates training by up to 15% compared to standard methods with minimal impact on model quality.
The increasing complexity of large language models (LLMs) necessitates efficient training strategies to mitigate the high computational costs associated with distributed training. A significant bottleneck in this process is gradient synchronization across multiple GPUs, particularly in the zero-redundancy parallelism mode. In this paper, we introduce Transformer-Aware Gradient Compression (TAGC), an optimized gradient compression algorithm designed specifically for transformer-based models. TAGC extends the lossless homomorphic compression method by adapting it for sharded models and incorporating transformer-specific optimizations, such as layer-selective compression and dynamic sparsification. Our experimental results demonstrate that TAGC accelerates training by up to 15% compared to the standard Fully Sharded Data Parallel (FSDP) approach, with minimal impact on model quality. We integrate TAGC into the PyTorch FSDP framework, the implementation is publicly available at https://github.com/ipolyakov/TAGC.