Skip to content

Latest commit

 

History

History
53 lines (47 loc) · 3.13 KB

.CurrentChangelog.md

File metadata and controls

53 lines (47 loc) · 3.13 KB

0.13.5 (2023-11-03)

Features:

  • [Component] share loading entry for component and module (#2945)
    • Initial support for the component model proposal.
    • This PR allows WasmEdge to recognize the component and module format.
  • [WASI-NN] ggml backend:
    • Provide options for enabling OpenBLAS, Metal, and cuBLAS.
    • Bump llama.cpp to b1383
    • Build thirdparty/ggml only when the ggml backend is enabled.
    • Enable the ggml plugin on the macOS platform.
    • Introduce AUTO detection. Wasm application will no longer need to specify the hardware spec (e.g., CPU or GPU). It will auto-detect by the runtime.
    • Unified the preload options with case-insensitive matching
    • Introduce metadata for setting the ggml options.
      • The following options are supported:
        • enable-log: true to enable logging. (default: false)
        • stream-stdout: true to print the inferred tokens in the streaming mode to standard output. (default: false)
        • ctx-size: Set the context size the same as the --ctx-size parameter in llama.cpp. (default: 512)
        • n-predict: Set the number of tokens to predict, the same as the --n-predict parameter in llama.cpp. (default: 512)
        • n-gpu-layers: Set the number of layers to store in VRAM, the same as the --n-gpu-layers parameter in llama.cpp. (default: 0)
        • reverse-prompt: Set the token pattern at which you want to halt the generation. Similar to the --reverse-prompt parameter in llama.cpp. (default: "")
        • batch-size: Set the number of batch sizes for prompt processing, the same as the --batch-size parameter in llama.cpp. (default: 512)
    • Notice: Because of the limitation of the WASI-NN proposal, there is no way to set the metadata during the loading process. The current workaround will re-load the model when n_gpu_layers is set to a non-zero value.
    • Installer: Support WASI-NN ggml plugin on both macOS Intel model (CPU only) and macOS Apple Silicon model. (#2882)
  • [Java Bindings] provide platform-specific jni and jar for Java bindings (#2980)
  • [C API]:
    • Provide getData API for FunctionInstance (#2937)
    • Add the API to set WASI-NN preloads. (#2827)
  • [Plugin]:
    • [zlib]:
      • initial support of the zlib plugin (#2562)
      • With a simple building guide and basic working examples
  • [MSVC] Support MSVC for building WasmEdge
  • [AOT] Support LLVM 17

Fixed issues:

  • [Installer]: Double quote the strings to prevent splitting in env file (#2994)
  • [AOT]:
    • Validate AOT section header fields
    • Add invariant attribute for memory and global pointer
  • [C API]:
    • Fix the wrong logic of getting types from exports.
  • [Example] Fix get-string with the latest C++ internal getSpan API. Fixes #2887 (#2929)
  • [CI] install llvm@16 to fix macOS build (#2878)

Misc:

  • [Example] Update wit-bindgen version from 0.7.0 to 0.11.0 (#2770)

Thank all the contributors who made this release possible!

dm4, hydai, Lîm Tsú-thuàn, Meenu Yadav, michael1017, proohit, Saikat Dey, Shen-Ta Hsieh, Shreyas Atre, Wang Jikai, Wck-iipi, YiYing He

If you want to build from source, please use WasmEdge-0.13.5-src.tar.gz instead of the zip or tarball provided by GitHub directly.