Change8

b8188

📦 llama-cppView on GitHub →
2 features🔧 2 symbols

Summary

This release enhances ggml-webgpu support for binary operations by enabling handling of non-contiguous and overlapping source tensors. It also provides a wide array of updated pre-built binaries for diverse platforms.

✨ New Features

  • ggml-webgpu now supports non-contiguous src0 tensors in binary operations.
  • ggml-webgpu now supports overlapping src0 and src1 tensors in binary operations.

Affected Symbols