make patches git am-able

raw diffs can be applied using `git apply` but not with `git am`. git
patches, e.g. through `git format-patch` are both apply-able and am-able

llm/generate/gen_common.sh

@@ -69,22 +69,10 @@ git_module_setup() {
 }
 
 apply_patches() {
-    # Wire up our CMakefile
+    # apply temporary patches until fix is upstream
-    if ! grep ollama ${LLAMACPP_DIR}/CMakeLists.txt; then
+    for patch in ../patches/*.patch; do
-        echo 'add_subdirectory(../ext_server ext_server) # ollama' >>${LLAMACPP_DIR}/CMakeLists.txt
+        git -c 'user.name=nobody' -c 'user.email=<>' -C ${LLAMACPP_DIR} am ${patch}
-    fi
+    done
-
-    if [ -n "$(ls -A ../patches/*.diff)" ]; then
-        # apply temporary patches until fix is upstream
-        for patch in ../patches/*.diff; do
-            for file in $(grep "^+++ " ${patch} | cut -f2 -d' ' | cut -f2- -d/); do
-                (cd ${LLAMACPP_DIR}; git checkout ${file})
-            done
-        done
-        for patch in ../patches/*.diff; do
-            (cd ${LLAMACPP_DIR} && git apply ${patch})
-        done
-    fi
 }
 
 build() {

llm/generate/gen_windows.ps1

@@ -83,29 +83,9 @@ function git_module_setup {
 }
 
 function apply_patches {
-    # Wire up our CMakefile
-    if (!(Select-String -Path "${script:llamacppDir}/CMakeLists.txt" -Pattern 'ollama')) {
-        Add-Content -Path "${script:llamacppDir}/CMakeLists.txt" -Value 'add_subdirectory(../ext_server ext_server) # ollama'
-    }
-
     # Apply temporary patches until fix is upstream
-    $patches = Get-ChildItem "../patches/*.diff"
+    foreach ($patch in $(Get-ChildItem "../patches/*.patch")) {
-    foreach ($patch in $patches) {
+        git -c 'user.name=nobody' -c 'user.email=<>' -C "${script:llamacppDir}" am $patch.FullName
-        # Extract file paths from the patch file
-        $filePaths = Get-Content $patch.FullName | Where-Object { $_ -match '^\+\+\+ ' } | ForEach-Object {
-            $parts = $_ -split ' '
-            ($parts[1] -split '/', 2)[1]
-        }
-
-        # Checkout each file
-        foreach ($file in $filePaths) {
-            git -C "${script:llamacppDir}" checkout $file
-        }
-    }
-
-    # Apply each patch
-    foreach ($patch in $patches) {
-        git -C "${script:llamacppDir}" apply $patch.FullName
     }
 }
 

llm/patches/0000-cmakelist.patch

@@ -0,0 +1,22 @@
+From 8b8d83ffca775840acc5dc700f3b3703e9f5cfe4 Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Fri, 23 Aug 2024 11:27:48 -0700
+Subject: [PATCH] patch cmakelist
+
+---
+ CMakeLists.txt | 2 ++
+ 1 file changed, 2 insertions(+)
+
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index a3132063..6a2a9912 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -199,3 +199,5 @@ if (LLAMA_BUILD_EXAMPLES)
+     add_subdirectory(examples)
+     add_subdirectory(pocs)
+ endif()
++
++add_subdirectory(../ext_server ext_server) # ollama
+-- 
+2.45.2
+

llm/patches/0001-load-progress.patch

@@ -1,8 +1,18 @@
+From 2cfaa0a04faa9c87ba8f1ac8527eb953e69c6cde Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:10 -0700
+Subject: [PATCH] 01-load-progress.diff
+
+---
+ common/common.cpp | 2 ++
+ common/common.h   | 7 +++++++
+ 2 files changed, 9 insertions(+)
+
 diff --git a/common/common.cpp b/common/common.cpp
-index 2c05a4d4..927f0e3d 100644
+index 9fa18472..48ff41e9 100644
 --- a/common/common.cpp
 +++ b/common/common.cpp
-@@ -2093,6 +2093,8 @@ struct llama_model_params llama_model_params_from_gpt_params(const gpt_params &
+@@ -2573,6 +2573,8 @@ struct llama_model_params llama_model_params_from_gpt_params(const gpt_params &
      mparams.use_mmap        = params.use_mmap;
      mparams.use_mlock       = params.use_mlock;
      mparams.check_tensors   = params.check_tensors;
@@ -12,10 +22,10 @@ index 2c05a4d4..927f0e3d 100644
          mparams.kv_overrides = NULL;
      } else {
 diff --git a/common/common.h b/common/common.h
-index 65c0ef81..ebca2c77 100644
+index cb5e7f6d..d8f043f7 100644
 --- a/common/common.h
 +++ b/common/common.h
-@@ -184,6 +184,13 @@ struct gpt_params {
+@@ -204,6 +204,13 @@ struct gpt_params {
      std::string mmproj = "";        // path to multimodal projector
      std::vector<std::string> image; // path to image file(s)
  
@@ -29,3 +39,6 @@ index 65c0ef81..ebca2c77 100644
      // embedding
      bool embedding         = false; // get only sentence embedding
      int32_t embd_normalize = 2;     // normalisation for embendings (-1=none, 0=max absolute int16, 1=taxicab, 2=euclidean, >2=p-norm)
+-- 
+2.46.0
+

llm/patches/0002-clip-log.patch

@@ -1,5 +1,14 @@
+From ba4bba80a744f76ac67b8234451c259a3c5da83b Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:11 -0700
+Subject: [PATCH] 02-clip-log.diff
+
+---
+ examples/llava/clip.cpp | 1 +
+ 1 file changed, 1 insertion(+)
+
 diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp
-index e431c7f7..f077e688 100644
+index 9b890571..cb51793d 100644
 --- a/examples/llava/clip.cpp
 +++ b/examples/llava/clip.cpp
 @@ -3,6 +3,7 @@
@@ -10,3 +19,6 @@ index e431c7f7..f077e688 100644
  #include "log.h"
  #include "ggml.h"
  #include "ggml-alloc.h"
+-- 
+2.46.0
+

llm/patches/0003-load_exception.patch

@@ -1,8 +1,17 @@
+From e43bfd3f607a6dfcaba2d490d35f412a52e55e30 Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:12 -0700
+Subject: [PATCH] 03-load_exception.diff
+
+---
+ src/llama.cpp | 25 ++++++++++++++++---------
+ 1 file changed, 16 insertions(+), 9 deletions(-)
+
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 73f52435..58a00fb1 100644
+index 88355971..926bb71a 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
-@@ -7241,7 +7241,7 @@ static int llama_model_load(const std::string & fname, llama_model & model, llam
+@@ -8635,7 +8635,7 @@ static int llama_model_load(const std::string & fname, llama_model & model, llam
          }
      } catch (const std::exception & err) {
          LLAMA_LOG_ERROR("%s: error loading model: %s\n", __func__, err.what());
@@ -11,7 +20,7 @@ index 73f52435..58a00fb1 100644
      }
  
      return 0;
-@@ -17564,16 +17564,23 @@ struct llama_model * llama_load_model_from_file(
+@@ -18022,16 +18022,23 @@ struct llama_model * llama_load_model_from_file(
          }
          model->rpc_servers.push_back(servers);
      }
@@ -43,3 +52,6 @@ index 73f52435..58a00fb1 100644
      }
  
      return model;
+-- 
+2.46.0
+

llm/patches/0004-metal.patch

@@ -1,8 +1,17 @@
+From 29411d9a9d2b6a0af6425ffe88498f17f71f7d5d Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:12 -0700
+Subject: [PATCH] 04-metal.diff
+
+---
+ ggml/src/ggml-metal.m | 30 +++++++++++++-----------------
+ 1 file changed, 13 insertions(+), 17 deletions(-)
+
 diff --git a/ggml/src/ggml-metal.m b/ggml/src/ggml-metal.m
-index 0207b787..b5e9884b 100644
+index 91b5e61b..9cfa72ac 100644
 --- a/ggml/src/ggml-metal.m
 +++ b/ggml/src/ggml-metal.m
-@@ -1396,27 +1396,23 @@ static enum ggml_status ggml_metal_graph_compute(
+@@ -1734,27 +1734,23 @@ static enum ggml_status ggml_metal_graph_compute(
                          // to the matrix-vector kernel
                          int ne11_mm_min = 1;
  
@@ -43,3 +52,6 @@ index 0207b787..b5e9884b 100644
  
                          // for now the matrix-matrix multiplication kernel only works on A14+/M1+ SoCs
                          // AMD GPU and older A-chips will reuse matrix-vector multiplication kernel
+-- 
+2.46.0
+

llm/patches/0005-default-pretokenizer.patch

@@ -1,5 +1,14 @@
+From b298ac8614d1e38da28f760eb1d2ae8af0fbbe62 Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:13 -0700
+Subject: [PATCH] 05-default-pretokenizer.diff
+
+---
+ src/llama.cpp | 14 +++-----------
+ 1 file changed, 3 insertions(+), 11 deletions(-)
+
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 88355971..dd7d41ed 100644
+index 926bb71a..d1e959fc 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
 @@ -6083,16 +6083,7 @@ static void llm_load_vocab(
@@ -30,3 +39,6 @@ index 88355971..dd7d41ed 100644
              }
          } else if (vocab.type == LLAMA_VOCAB_TYPE_SPM) {
              vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_DEFAULT;
+-- 
+2.46.0
+

llm/patches/0006-embeddings.patch

@@ -1,8 +1,17 @@
+From c9a6ca9fc039233dee746a4da9705762cd9e515d Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:14 -0700
+Subject: [PATCH] 06-embeddings.diff
+
+---
+ src/llama.cpp | 17 ++++++++++-------
+ 1 file changed, 10 insertions(+), 7 deletions(-)
+
 diff --git a/src/llama.cpp b/src/llama.cpp
-index 88355971..d7db689b 100644
+index d1e959fc..f79bd782 100644
 --- a/src/llama.cpp
 +++ b/src/llama.cpp
-@@ -15906,7 +15906,7 @@ static size_t llama_output_reserve(llama_context & lctx, size_t n_outputs) {
+@@ -15898,7 +15898,7 @@ static size_t llama_output_reserve(llama_context & lctx, size_t n_outputs) {
      const auto n_embd  = hparams.n_embd;
  
      // TODO: use a per-batch flag for logits presence instead
@@ -11,7 +20,7 @@ index 88355971..d7db689b 100644
      const bool has_embd   =  cparams.embeddings && (cparams.pooling_type == LLAMA_POOLING_TYPE_NONE);
  
      const size_t logits_size = has_logits ? n_vocab*n_outputs_max : 0;
-@@ -16175,20 +16175,23 @@ static int llama_decode_internal(
+@@ -16167,20 +16167,23 @@ static int llama_decode_internal(
              // no output
              res  = nullptr;
              embd = nullptr;
@@ -41,3 +50,6 @@ index 88355971..d7db689b 100644
          // LLAMA_LOG_INFO("graph build time: %.3f ms (%d nodes, %d leafs)\n", (ggml_time_us() - t_start_us)/1000.0, gf->n_nodes, gf->n_leafs);
  
          ggml_backend_sched_alloc_graph(lctx.sched, gf);
+-- 
+2.46.0
+

llm/patches/0007-clip-unicode.patch

@@ -1,8 +1,17 @@
+From ae2b188a679c83ce105aa1e823499441dfab3c57 Mon Sep 17 00:00:00 2001
+From: Michael Yang <mxyng@pm.me>
+Date: Mon, 16 Sep 2024 15:53:15 -0700
+Subject: [PATCH] 07-clip-unicode.diff
+
+---
+ examples/llava/clip.cpp | 23 +++++++++++++++++++++++
+ 1 file changed, 23 insertions(+)
+
 diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp
-index 95fbe3d0..5a02a6ec 100644
+index cb51793d..8716472b 100644
 --- a/examples/llava/clip.cpp
 +++ b/examples/llava/clip.cpp
-@@ -32,6 +33,14 @@
+@@ -41,6 +41,14 @@
  #include <cinttypes>
  #include <limits>
  
@@ -17,7 +26,7 @@ index 95fbe3d0..5a02a6ec 100644
  //#define CLIP_DEBUG_FUNCTIONS
  
  // RGB uint8 image
-@@ -1055,7 +1064,22 @@ struct clip_ctx * clip_model_load(const char * fname, const int verbosity = 1) {
+@@ -1223,7 +1231,22 @@ struct clip_ctx * clip_model_load(const char * fname, const int verbosity = 1) {
              return nullptr;
          }
  
@@ -40,3 +49,6 @@ index 95fbe3d0..5a02a6ec 100644
          if (!fin) {
              LOG_TEE("cannot open model file for loading tensors\n");
              clip_free(new_clip);
+-- 
+2.46.0
+