Nils Graef

Nils Graef, Andrew Wasielewski, Matthew Clapp

This paper presents FlashNorm, which is an exact but faster implementation of RMSNorm followed by linear layers. RMSNorm is used by many LLMs such as Llama, Mistral, and OpenELM. FlashNorm also speeds up Layer Normalization and its recently proposed replacement Dynamic Tanh (DyT) arXiv:2503.10622. FlashNorm also reduces the number of parameter tensors by simply merging the normalization weights with the weights of the next linear layer. See https://github.com/OpenMachine-ai/transformer-tricks for code and more transformer tricks.

Nils Graef

This micro-paper describes a trick to speed up inference of transformers with RoPE (such as LLaMA, Mistral, PaLM, and Gemma). For these models, a large portion of the first transformer layer can be precomputed, which results in slightly lower latency and lower cost-per-token. Because this trick optimizes only one layer, the relative savings depend on the total number of layers. For example, the maximum savings for a model with only 4 layers (such as Whisper tiny) is limited to 25%, while a 32-layer model is limited to 3% savings. See https://github.com/OpenMachine-ai/transformer-tricks for code and more transformer tricks.

Nils Graef, Andrew Wasielewski

Slim attention shrinks the context memory size by 2x for transformer models with MHA (multi-head attention), which can speed up inference by up to 2x for large context windows. Slim attention is an exact, mathematically identical implementation of the standard attention mechanism and therefore doesn't compromise model accuracy. In other words, slim attention losslessly compresses the context memory by a factor of 2. For encoder-decoder transformers, the context memory size can be reduced even further: For the Whisper models for example, slim attention reduces the context memory by 8x, which can speed up token generation by 5x for batch size 64 for example. And for the T5-11B model for example, the memory can be reduced by 32x because its MHA projection dimension is larger than the embedding dimension. See https://github.com/OpenMachine-ai/transformer-tricks for code and more transformer tricks, and https://www.youtube.com/watch?v=uVtk3B6YO4Y for this paper's YouTube video.

Nils Graef

He and Hofmann (arXiv:2311.01906) detailed a skipless transformer without the V and P (post-attention projection) linear layers, which reduces the total number of weights. However, this scheme is only applicable to MHA (multi-head attention), but not for MQA (multi-query attention) and GQA (grouped-query attention). The latter schemes are used by many popular LLMs such as Llama 2, Mistral, Mixtral, PaLM, and Gemma. Therefore, this micro-paper proposes mathematically equivalent versions that are suitable for MQA and GQA. For example, removing Q and P from a skipless version of Mistral-7B would remove 15% of its weights (and thus reduce its compute and memory complexity). Watch our explainer video https://youtu.be/Tx_lMpphd2g and see https://github.com/OpenMachine-ai/transformer-tricks for code and more transformer tricks.