Ubuntu
ggml package

ggml 0.9.6-3 source package in Ubuntu

Changelog

ggml (0.9.6-3) unstable; urgency=medium

  * Add another workaround for ROCm on trixie

 -- Christian Kastner <email address hidden>  Sun, 08 Feb 2026 10:10:41 +0100

Upload details

Uploaded by:
Debian Deep Learning Team
Uploaded to:
Sid
Original maintainer:
Debian Deep Learning Team
Architectures:
amd64 arm64 loong64 ppc64 ppc64el riscv64 s390x
Section:
misc
Urgency:
Medium Urgency

See full publishing history Publishing

Series Pocket Published Component Section

Downloads

File Size SHA-256 Checksum
ggml_0.9.6-3.dsc 2.7 KiB 52af53f9ef931f28c530e06aa97828f7d76c3cecbb4714aaf257b3f70810d953
ggml_0.9.6.orig.tar.gz 2.4 MiB c5a7c9d737c25bb080630f1285924ac006e1d909ebc3b2df946e72c6ae0bc222
ggml_0.9.6-3.debian.tar.xz 16.1 KiB 203497b7d4ef9767835a1b7ff95779b665f48861bc998126965e227d64064df5

Available diffs

No changes file available.

Binary packages built by this source

libggml-dev: Tensor library for machine learning - development files

 ggml is a tensor library for machine learning with the following features:
 .
  * Low-level cross-platform implementation
  * Integer quantization support
  * Broad hardware support
  * Automatic differentiation
  * ADAM and L-BFGS optimizers
  * No third-party dependencies
  * Zero memory allocations during runtime
 .
 On Debian, ggml is shipped as a package that depends on one of the CPU or
 GPU backends.
 .
 This package provides the development files.

libggml0: Tensor library for machine learning - libraries

 ggml is a tensor library for machine learning with the following features:
 .
  * Low-level cross-platform implementation
  * Integer quantization support
  * Broad hardware support
  * Automatic differentiation
  * ADAM and L-BFGS optimizers
  * No third-party dependencies
  * Zero memory allocations during runtime
 .
 On Debian, ggml is shipped as a package that depends on one of the CPU or
 GPU backends.
 .
 This package provides the main libraries, and the CPU backend.

libggml0-backend-blas: Tensor library for machine learning - BLAS backend

 ggml is a tensor library for machine learning with the following features:
 .
  * Low-level cross-platform implementation
  * Integer quantization support
  * Broad hardware support
  * Automatic differentiation
  * ADAM and L-BFGS optimizers
  * No third-party dependencies
  * Zero memory allocations during runtime
 .
 On Debian, ggml is shipped as a package that depends on one of the CPU or
 GPU backends.
 .
 This package provides the BLAS backend.

libggml0-backend-blas-dbgsym: debug symbols for libggml0-backend-blas
libggml0-backend-vulkan: Tensor library for machine learning - Vulkan backend

 ggml is a tensor library for machine learning with the following features:
 .
  * Low-level cross-platform implementation
  * Integer quantization support
  * Broad hardware support
  * Automatic differentiation
  * ADAM and L-BFGS optimizers
  * No third-party dependencies
  * Zero memory allocations during runtime
 .
 On Debian, ggml is shipped as a package that depends on one of the CPU or
 GPU backends.
 .
 This package provides the Vulkan backend.

libggml0-backend-vulkan-dbgsym: debug symbols for libggml0-backend-vulkan
libggml0-dbgsym: debug symbols for libggml0