Change8

b9565

📦 llama-cppView on GitHub →
2 features🐛 1 fixes🔧 4 symbols

Summary

This release focuses on improving WebGPU support by fixing buffer overlap issues in the concat operator and setting up dedicated CI for WebGPU.

✨ New Features

  • Added WebGPU CI workflow.
  • Enabled handling of buffer overlap/aliasing for the concat operator in ggml-webgpu.

🐛 Bug Fixes

  • Fixed buffer overlap/aliasing issue for the concat operator in ggml-webgpu.

Affected Symbols