eval-callback: renamed from ggml-debug

.gitignore

@@ -48,7 +48,7 @@ models-mnt
 /convert-llama2c-to-ggml
 /embd-input-test
 /embedding
-/ggml-debug
+/eval-callback
 /gguf
 /gguf-llama-simple
 /gguf-split

Makefile

@@ -1,7 +1,7 @@
 # Define the default target now so that it is always the first target
 BUILD_TARGETS = \
 	main quantize quantize-stats perplexity imatrix embedding vdot q8dot train-text-from-scratch convert-llama2c-to-ggml \
-	simple batched batched-bench save-load-state server gguf gguf-split ggml-debug llama-bench libllava.a llava-cli baby-llama beam-search  \
+	simple batched batched-bench save-load-state server gguf gguf-split eval-callback llama-bench libllava.a llava-cli baby-llama beam-search  \
 	retrieval speculative infill tokenize benchmark-matmult parallel finetune export-lora lookahead lookup passkey gritlm tests/test-c.o
 
 # Binaries only useful for tests
@@ -800,7 +800,7 @@ gguf-split: examples/gguf-split/gguf-split.cpp ggml.o llama.o $(COMMON_DEPS) $(O
 	$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h $<,$^) $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS)
 
-ggml-debug: examples/ggml-debug/ggml-debug.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
+eval-callback: examples/eval-callback ggml.o llama.o $(COMMON_DEPS) $(OBJS)
 	$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h $<,$^) $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS)
 

docs/HOWTO-add-model.md

@@ -100,7 +100,7 @@ Have a look to existing implementation like `build_llama`, `build_dbrx` or `buil
 
 When implementing a new graph, please note that the underlying `ggml` backends might not support them all, support of missing backend operations can be added in another PR.
 
-Note: to debug the inference graph: you can use [ggml-debug](../examples/ggml-debug).
+Note: to debug the inference graph: you can use [eval-callback](../examples/eval-callback).
 
 ## GGUF specification
 

examples/CMakeLists.txt

@@ -19,10 +19,10 @@ else()
     add_subdirectory(benchmark)
     add_subdirectory(convert-llama2c-to-ggml)
     add_subdirectory(embedding)
+    add_subdirectory(eval-callback)
     add_subdirectory(finetune)
     add_subdirectory(gritlm)
     add_subdirectory(gguf-split)
-    add_subdirectory(ggml-debug)
     add_subdirectory(infill)
     add_subdirectory(llama-bench)
     add_subdirectory(llava)

examples/eval-callback/CMakeLists.txt

@@ -0,0 +1,9 @@
+set(TARGET eval-callback)
+add_executable(${TARGET} eval-callback.cpp)
+install(TARGETS ${TARGET} RUNTIME)
+target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
+target_compile_features(${TARGET} PRIVATE cxx_std_11)
+
+set(TEST_TARGET test-eval-callback)
+add_test(NAME ${TEST_TARGET} COMMAND eval-callback --hf-repo ggml-org/models --hf-file tinyllamas/stories260K.gguf --model stories260K.gguf --prompt hello --seed 42)
+set_property(TEST ${TEST_TARGET} PROPERTY LABELS eval-callback curl)

examples/eval-callback/README.md

@@ -1,4 +1,4 @@
-# llama.cpp/examples/ggml-debug
+# llama.cpp/examples/eval-callback
 
 A simple example which demonstrates how to use callback during the inference.
 It simply prints to the console all operations and tensor data.
@@ -6,7 +6,7 @@ It simply prints to the console all operations and tensor data.
 Usage:
 
 ```shell
-ggml-debug \
+eval-callback \
   --hf-repo ggml-org/models \
   --hf-file phi-2/ggml-model-q4_0.gguf \
   --model phi-2-q4_0.gguf \

examples/ggml-debug/CMakeLists.txt

@@ -1,9 +0,0 @@
-set(TARGET ggml-debug)
-add_executable(${TARGET} ggml-debug.cpp)
-install(TARGETS ${TARGET} RUNTIME)
-target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
-
-set(TEST_TARGET test-ggml-debug)
-add_test(NAME ${TEST_TARGET} COMMAND ggml-debug --hf-repo ggml-org/models --hf-file tinyllamas/stories260K.gguf --model stories260K.gguf --prompt hello --seed 42)
-set_property(TEST ${TEST_TARGET} PROPERTY LABELS ggml-debug curl)