Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-20 01:31:42 +02:00
Code Issues Packages Projects Releases Wiki Activity

added MPT support

Browse Source
This commit is contained in:
Concedo
2023-06-03 16:14:08 +08:00
parent 9839259b63
commit 6f82e17b7a
10 changed files with 983 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Makefile
+4 -2
View File
@@ -1,6 +1,6 @@
default: koboldcpp koboldcpp_failsafe koboldcpp_openblas koboldcpp_openblas_noavx2 koboldcpp_clblast
simple: koboldcpp koboldcpp_failsafe
tools: quantize_gpt2 quantize_gptj quantize_llama quantize_neox
tools: quantize_gpt2 quantize_gptj quantize_llama quantize_neox quantize_mpt
dev: koboldcpp_openblas
dev2: koboldcpp_clblast
@@ -281,7 +281,7 @@ gpttype_adapter_clblast.o: gpttype_adapter.cpp
$(CXX) $(CXXFLAGS) $(CLBLAST_FLAGS) -c $< -o $@
clean:
rm -vf *.o main quantize_llama quantize_gpt2 quantize_gptj quantize_neox quantize-stats perplexity embedding benchmark-matmult save-load-state main.exe quantize_llama.exe quantize_gptj.exe quantize_gpt2.exe quantize_neox.exe koboldcpp.dll koboldcpp_openblas.dll koboldcpp_failsafe.dll koboldcpp_openblas_noavx2.dll koboldcpp_clblast.dll koboldcpp.so koboldcpp_openblas.so koboldcpp_failsafe.so koboldcpp_openblas_noavx2.so koboldcpp_clblast.so gptj.exe gpt2.exe
rm -vf *.o main quantize_llama quantize_gpt2 quantize_gptj quantize_neox quantize_mpt quantize-stats perplexity embedding benchmark-matmult save-load-state main.exe quantize_llama.exe quantize_gptj.exe quantize_gpt2.exe quantize_neox.exe quantize_mpt.exe koboldcpp.dll koboldcpp_openblas.dll koboldcpp_failsafe.dll koboldcpp_openblas_noavx2.dll koboldcpp_clblast.dll koboldcpp.so koboldcpp_openblas.so koboldcpp_failsafe.so koboldcpp_openblas_noavx2.so koboldcpp_clblast.so
main: examples/main/main.cpp build-info.h ggml.o llama.o common.o $(OBJS)
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
@@ -308,6 +308,8 @@ quantize_gpt2: ggml.o llama.o otherarch/tools/gpt2_quantize.cpp otherarch/tools/
$(CXX) $(CXXFLAGS) $^ -o $@ $(LDFLAGS)
quantize_neox: ggml.o llama.o otherarch/tools/neox_quantize.cpp otherarch/tools/common-ggml.cpp
$(CXX) $(CXXFLAGS) $^ -o $@ $(LDFLAGS)
quantize_mpt: ggml.o llama.o otherarch/tools/mpt_quantize.cpp otherarch/tools/common-ggml.cpp
$(CXX) $(CXXFLAGS) $^ -o $@ $(LDFLAGS)
quantize-stats: examples/quantize-stats/quantize-stats.cpp ggml.o llama.o $(OBJS)
$(CXX) $(CXXFLAGS) $^ -o $@ $(LDFLAGS)
perplexity: examples/perplexity/perplexity.cpp ggml.o llama.o common.o $(OBJS)