Add speculative decoding support to the server and command-line interfaces

This commit is contained in:
DeEMO
2025-06-30 09:35:35 +00:00
committed by DeEMO
parent 1ea2d61a97
commit 2e8e42a5ad
11 changed files with 591 additions and 31 deletions
+8
View File
@@ -963,6 +963,7 @@ OBJ_COMMON = \
common/console.o \
common/ngram-cache.o \
common/sampling.o \
common/speculative.o \
common/train.o \
common/build-info.o \
common/json-schema-to-grammar.o
@@ -1239,6 +1240,13 @@ common/json-schema-to-grammar.o: \
common/json-schema-to-grammar.h
$(CXX) $(CXXFLAGS) -c $< -o $@
# speculative
common/speculative.o: \
common/speculative.cpp \
common/speculative.h \
include/llama.h
$(CXX) $(CXXFLAGS) -c $< -o $@
common/train.o: \
common/train.cpp \
common/train.h