feat: Update llama.cpp · abetlen/llama-cpp-python@9497bcd

Commit

feat: Update llama.cpp

Browse files

abetlen committed Sep 5, 2024

1 parent c3fc80a commit 9497bcd

+1 −1		.devops/full-cuda.Dockerfile
+46 −7		.devops/nix/devshells.nix
+8 −10		.devops/nix/nixpkgs-instances.nix
+36 −0		.devops/nix/package-gguf-py.nix
+153 −229		.devops/nix/package.nix
+66 −0		.devops/nix/python-scripts.nix
+31 −9		.devops/nix/scope.nix
+3 −21		README.md
+4 −0		common/common.cpp
+1 −1		docs/docker.md
+43 −22		examples/llama-bench/README.md
+85 −56		examples/llama-bench/llama-bench.cpp
+2 −2		examples/main/main.cpp
+313 −446		examples/server/server.cpp
+19 −19		examples/server/tests/features/steps/steps.py
+3 −0		examples/server/tests/features/wrong_usages.feature
+33 −0		examples/server/utils.hpp
+6 −6		flake.lock
+4 −1		flake.nix
+612 −0		ggml/src/ggml-aarch64.c
+2 −2		ggml/src/ggml-sycl/dmmv.cpp
+3 −0		ggml/src/ggml.c
+1 −0		gguf-py/pyproject.toml
+1 −1		grammars/README.md
+1 −1		pyproject.toml
+47 −40		src/llama.cpp