Skip to content

[auto] Update llama.cpp to latest version #21

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jun 6, 2025
Merged

[auto] Update llama.cpp to latest version #21

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
57d58f6
auto: update llama.cpp dependency
tattn Jun 6, 2025
e39a218
fix: remove unused linker flags and improve llama initialization logic
tattn Jun 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions Package.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@

import PackageDescription

let llamaVersion = "b5575"
let llamaVersion = "b5602"

// MARK: - Package Dependencies

Expand Down Expand Up @@ -94,7 +94,7 @@ packageTargets.append(contentsOf: [
name: "LocalLLMClientLlamaFramework",
url:
"https://github.com/ggml-org/llama.cpp/releases/download/\(llamaVersion)/llama-\(llamaVersion)-xcframework.zip",
checksum: "78ba1f1a8622e548f5e8a29b93473afe2e879b9b0781c3af0d31673c3310c7d6"
checksum: "88b75946b8fdedc1eb7eb97be1286d0f15c836f3676579305cf091cb977740ac"
),
.target(
name: "LocalLLMClientLlamaC",
Expand Down Expand Up @@ -159,7 +159,7 @@ packageTargets.append(contentsOf: [
],
linkerSettings: [
.unsafeFlags([
"-lggml-base", "-lggml-cpu", "-lggml-rpc", "-lggml", "-lllama", "-lmtmd"
"-lggml-base", "-lggml", "-lllama", "-lmtmd"
])
]
),
Expand Down
5 changes: 5 additions & 0 deletions Sources/LocalLLMClientLlama/Utility.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
#else
@preconcurrency import LocalLLMClientLlamaC
#endif
import Foundation
#if canImport(OSLog)
import OSLog
#endif
Expand All @@ -20,6 +21,10 @@ nonisolated(unsafe) private var llamaLogCallback: ((LlamaLogLevel, String) -> Vo
public func initializeLlama() {
guard !isLlamaInitialized else { return }
isLlamaInitialized = true
#if os(Linux)
ggml_backend_load_all_from_path(ProcessInfo.processInfo.environment["LD_LIBRARY_PATH"])
#endif

llama_backend_init()

if !isCustomLogEnabled {
Expand Down
2 changes: 1 addition & 1 deletion Sources/LocalLLMClientLlamaC/exclude/llama.cpp
Open in desktop
Submodule llama.cpp updated 81 files
+4 −4 .github/workflows/build.yml
+12 −5 .github/workflows/release.yml
+1 −1 .github/workflows/server.yml
+5 −0 CMakeLists.txt
+31 −11 README.md
+14 −1 ci/run.sh
+1 −0 common/arg.cpp
+8 −7 common/chat.cpp
+1 −1 common/chat.h
+2 −2 common/common.cpp
+2 −1 common/common.h
+6 −4 common/speculative.cpp
+1 −2 convert_hf_to_gguf.py
+4 −0 docs/build.md
+20 −16 docs/install.md
+1 −1 examples/batched.swift/Sources/main.swift
+18 −3 examples/embedding/embedding.cpp
+2 −2 examples/gritlm/gritlm.cpp
+4 −4 examples/llama.android/llama/src/main/cpp/llama-android.cpp
+4 −4 examples/llama.swiftui/llama.cpp.swift/LibLlama.swift
+8 −6 examples/lookahead/lookahead.cpp
+1 −1 examples/lookup/lookup.cpp
+7 −5 examples/parallel/parallel.cpp
+11 −9 examples/passkey/passkey.cpp
+1 −1 examples/retrieval/retrieval.cpp
+1 −1 examples/save-load-state/save-load-state.cpp
+2 −2 examples/simple-chat/simple-chat.cpp
+1 −1 examples/speculative-simple/speculative-simple.cpp
+14 −12 examples/speculative/speculative.cpp
+1 −1 ggml/CMakeLists.txt
+0 −1 ggml/src/CMakeLists.txt
+2 −2 ggml/src/ggml-cpu/ops.cpp
+4 −1 ggml/src/ggml-cuda/fattn-mma-f16.cuh
+5 −3 ggml/src/ggml-metal/ggml-metal.m
+52 −42 ggml/src/ggml-metal/ggml-metal.metal
+6 −0 ggml/src/ggml-opencl/CMakeLists.txt
+747 −3 ggml/src/ggml-opencl/ggml-opencl.cpp
+109 −0 ggml/src/ggml-opencl/kernels/concat.cl
+30 −0 ggml/src/ggml-opencl/kernels/pad.cl
+39 −0 ggml/src/ggml-opencl/kernels/repeat.cl
+63 −0 ggml/src/ggml-opencl/kernels/tanh.cl
+48 −0 ggml/src/ggml-opencl/kernels/tsembd.cl
+121 −0 ggml/src/ggml-opencl/kernels/upscale.cl
+162 −34 ggml/src/ggml-vulkan/ggml-vulkan.cpp
+98 −0 ggml/src/ggml-vulkan/vulkan-shaders/conv_transpose_1d.comp
+2 −0 ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
+3 −0 gguf-py/gguf/gguf_writer.py
+124 −29 include/llama.h
+0 −1 src/CMakeLists.txt
+240 −109 src/llama-context.cpp
+8 −6 src/llama-context.h
+2 −3 src/llama-graph.cpp
+1 −1 src/llama-graph.h
+16 −16 src/llama-kv-cache-recurrent.cpp
+13 −19 src/llama-kv-cache-recurrent.h
+34 −31 src/llama-kv-cache-unified-iswa.cpp
+22 −24 src/llama-kv-cache-unified-iswa.h
+141 −86 src/llama-kv-cache-unified.cpp
+66 −37 src/llama-kv-cache-unified.h
+0 −1 src/llama-kv-cache.cpp
+0 −44 src/llama-kv-cache.h
+41 −0 src/llama-memory.cpp
+65 −23 src/llama-memory.h
+1 −1 src/llama-mmap.cpp
+42 −17 src/llama-model-loader.cpp
+38 −10 src/llama-model.cpp
+3 −0 src/llama-model.h
+5 −1 src/llama-vocab.cpp
+14 −2 tests/test-backend-ops.cpp
+1 −1 tests/test-chat.cpp
+4 −2 tools/batched-bench/batched-bench.cpp
+1 −1 tools/cvector-generator/cvector-generator.cpp
+1 −1 tools/imatrix/imatrix.cpp
+2 −2 tools/llama-bench/llama-bench.cpp
+8 −6 tools/main/main.cpp
+1 −1 tools/mtmd/mtmd-cli.cpp
+6 −6 tools/perplexity/perplexity.cpp
+2 −2 tools/run/run.cpp
+17 −15 tools/server/server.cpp
+6 −7 tools/server/tests/unit/test_tool_call.py
+10 −1 tools/server/tests/utils.py