MPT: use upstream llama.cpp implementation (nomic-ai#1515)

activescott · Oct 19, 2023 · 4338e72 · 4338e72
1 parent 0fe2e19
commit 4338e72
Show file tree

Hide file tree

Showing 10 changed files with 24 additions and 1,038 deletions.
diff --git a/gpt4all-backend/CMakeLists.txt b/gpt4all-backend/CMakeLists.txt
@@ -102,10 +102,6 @@ foreach(BUILD_VARIANT IN LISTS BUILD_VARIANTS)
             gptj.cpp utils.h utils.cpp llmodel_shared.cpp llmodel_shared.h)
         prepare_target(gptj llama-mainline)
 
-        add_library(mpt-${BUILD_VARIANT} SHARED
-            mpt.cpp utils.h utils.cpp llmodel_shared.cpp llmodel_shared.h)
-        prepare_target(mpt llama-mainline)
-
         add_library(bert-${BUILD_VARIANT} SHARED
             bert.cpp utils.h utils.cpp llmodel_shared.cpp llmodel_shared.h)
         target_compile_definitions(bert-${BUILD_VARIANT} PRIVATE LLAMA_VERSIONS=>=3 LLAMA_DATE=999999)

diff --git a/gpt4all-backend/llama.cpp-mainline b/gpt4all-backend/llama.cpp-mainline
diff --git a/gpt4all-backend/llamamodel.cpp b/gpt4all-backend/llamamodel.cpp
@@ -397,7 +397,7 @@ DLL_EXPORT bool magic_match(const char * fname) {
 
     bool isValid = gguf_get_version(ctx_gguf) <= 2;
     auto arch = get_arch_name(ctx_gguf);
-    isValid = isValid && (arch == "llama" || arch == "starcoder" || arch == "falcon");
+    isValid = isValid && (arch == "llama" || arch == "starcoder" || arch == "falcon" || arch == "mpt");
 
     gguf_free(ctx_gguf);
     return isValid;
+1 −1		Makefile
+213 −0		convert-mpt-hf-to-gguf.py
+46 −2		ggml-cuda.cu
+7 −2		ggml.c
+2 −0		ggml.h
+465 −25		llama.cpp