TorchSparse++ is a high-performance computing library for efficient 3D sparse convolution. It offers significant performance improvement over TorchSparse++ by overlapping computation with memory access. It also searches for the best execution strategy for sparse workloads within a large design space through auto-tuning.
This project introduce PockEngine: a tiny, sparse and efficient engine to enable fine-tuning on various edge devices. PockEngine supports sparse backpropagation: it prunes the backward graph and sparsely updates the model with measured memory saving and latency reduction while maintaining the model quality.
LongLoRA takes advantage of shifted sparse attention to greatly reduce the finetuning cost of long context LLMs.
We propose a efficient fine-tuning method that does not need full model weights and data sharing.
TorchSparse is a high-performance computing library for efficient 3D sparse convolution. This library aims at accelerating sparse computation in 3D, in particular the Sparse Convolution operation.
In MCUNetV3, we enable on-device training under 256KB SRAM and 1MB Flash, using less than 1/1000 memory of PyTorch while matching the accuracy on the visual wake words application. It enables the model to adapt to newly collected sensor data and users can enjoy customized services without uploading the data to the cloud thus protecting privacy.