chore: bump llama.cpp

Signed-off-by: thxCode <[email protected]>
gpustack · Dec 18, 2024 · 529c2ab · 529c2ab
1 parent 358b44f
commit 529c2ab
Show file tree

Hide file tree

Showing 8 changed files with 8 additions and 8 deletions.
diff --git a/llama-box/patches/llama.cpp/embedding.patch b/llama-box/patches/llama.cpp/embedding.patch
@@ -11,7 +11,7 @@ index efbb27d2..6b944ef2 100644
      LLAMA_API int32_t llama_n_vocab    (const struct llama_model * model);
      LLAMA_API int32_t llama_n_ctx_train(const struct llama_model * model);
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..4ddde98a 100644
+index b7b04a41..98f00cd4 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -20025,9 +20025,17 @@ struct llama_context * llama_new_context_with_model(

diff --git a/llama-box/patches/llama.cpp/model.patch b/llama-box/patches/llama.cpp/model.patch
@@ -1,5 +1,5 @@
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..030019d7 100644
+index b7b04a41..2d4423b1 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -158,6 +158,7 @@ enum llm_arch {

diff --git a/llama-box/patches/llama.cpp/mrope.patch b/llama-box/patches/llama.cpp/mrope.patch
@@ -11,7 +11,7 @@ index efbb27d2..73bdbf78 100644
      LLAMA_API const struct llama_model * llama_get_model(const struct llama_context * ctx);
 
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..e7f7dafb 100644
+index b7b04a41..96daaced 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -20361,6 +20361,10 @@ int32_t llama_n_head(const struct llama_model * model) {

diff --git a/llama-box/patches/llama.cpp/template.patch b/llama-box/patches/llama.cpp/template.patch
@@ -1,5 +1,5 @@
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..f3d623ed 100644
+index b7b04a41..9c923330 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -1630,6 +1630,9 @@ enum llm_chat_template {

diff --git a/llama-box/patches/llama.cpp/token.patch b/llama-box/patches/llama.cpp/token.patch
@@ -38,7 +38,7 @@ index 4bb16d2e..528f0291 100644
  llama_token llama_token_prefix_impl(const struct llama_vocab & vocab);
  llama_token llama_token_middle_impl(const struct llama_vocab & vocab);
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..22d2f51f 100644
+index b7b04a41..eb01c6b6 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -22089,6 +22089,10 @@ llama_token llama_token_pad(const struct llama_model * model) {

diff --git a/llama-box/patches/llama.cpp/tokenizer.patch b/llama-box/patches/llama.cpp/tokenizer.patch
@@ -1,5 +1,5 @@
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..5ee0d984 100644
+index b7b04a41..e0d88735 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -6559,7 +6559,8 @@ static void llm_load_vocab(

diff --git a/llama-box/patches/llama.cpp/tool_calling.patch b/llama-box/patches/llama.cpp/tool_calling.patch
@@ -158,7 +158,7 @@ index efbb27d2..19031e22 100644
                                    char * buf,
                                 int32_t   length);
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 8b799e0e..ade86b46 100644
+index b7b04a41..a742db1b 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -22274,12 +22274,52 @@ static llm_chat_template llama_chat_detect_template(const std::string & tmpl) {

diff --git a/llama.cpp b/llama.cpp
+2 −2		README.md
+6 −2		examples/server/README.md
+ −		examples/server/public/index.html.gz
+54 −20		examples/server/server.cpp
+12 −4		examples/server/tests/unit/test_completion.py
+59 −0		examples/server/tests/unit/test_embedding.py
+23 −0		examples/server/tests/unit/test_rerank.py
+13 −6		examples/server/utils.hpp
+7 −0		examples/server/webui/package-lock.json
+1 −0		examples/server/webui/package.json
+5 −2		examples/server/webui/src/main.js
+0 −1		ggml/src/ggml-alloc.c
+1 −1		ggml/src/ggml-cpu/CMakeLists.txt
+3 −0		ggml/src/ggml-cpu/ggml-cpu.cpp
+16 −0		ggml/src/ggml-impl.h
+1 −1		ggml/src/ggml-vulkan/vulkan-shaders/soft_max.comp
+26 −43		ggml/src/ggml.c
+1 −1		scripts/sync-ggml.last
+8 −8		src/llama.cpp
+1 −0		tests/CMakeLists.txt
+2 −2		tests/test-backend-ops.cpp
+1,303 −0		tests/test-gguf.cpp