scripts : working curl pipe

* ggml : move FP16 <-> FP32 stuff to ggml-impl.h

ggml-ci

* tests : fix ARM build

* ggml : explicitly initialize deprecated type traits

* ggml : add math.h to ggml-impl.h

* ggml : remove duplicate static assert macros

* ggml : prefix lookup tables with ggml_

ggml-ci

* ggml-impl : move extern "C" to start of file

.github/ISSUE_TEMPLATE/bug.md

@@ -1,8 +1,7 @@
 ---
-name: Issue and enhancement template
-about: Used to report issues and request enhancements for llama.cpp
-title: "[User] Insert summary of your issue or enhancement.."
-labels: ''
+name: Bug template
+about: Used to report bugs in llama.cpp
+labels: ["bug-unconfirmed"]
 assignees: ''
 
 ---
@@ -46,7 +45,7 @@ $ g++ --version
 
 # Failure Information (for bugs)
 
-Please help provide information about the failure if this is a bug. If it is not a bug, please remove the rest of this template.
+Please help provide information about the failure / bug.
 
 # Steps to Reproduce
 

.github/ISSUE_TEMPLATE/enhancement.md

@@ -0,0 +1,28 @@
+---
+name: Enhancement template
+about: Used to request enhancements for llama.cpp
+labels: ["enhancement"]
+assignees: ''
+
+---
+
+# Prerequisites
+
+Please answer the following questions for yourself before submitting an issue.
+
+- [ ] I am running the latest code. Development is very rapid so there are no tagged versions as of now.
+- [ ] I carefully followed the [README.md](https://github.com/ggerganov/llama.cpp/blob/master/README.md).
+- [ ] I [searched using keywords relevant to my issue](https://docs.github.com/en/issues/tracking-your-work-with-issues/filtering-and-searching-issues-and-pull-requests) to make sure that I am creating a new issue that is not already open (or closed).
+- [ ] I reviewed the [Discussions](https://github.com/ggerganov/llama.cpp/discussions), and have a new bug or useful enhancement to share.
+
+# Feature Description
+
+Please provide a detailed written description of what you were trying to do, and what you expected `llama.cpp` to do as an enhancement.
+
+# Motivation
+
+Please provide a detailed written description of reasons why this feature is necessary and how it is useful to `llama.cpp` users.
+
+# Possible Implementation
+
+If you have an idea as to how it can be implemented, please write a detailed description. Feel free to give links to external sources or share visuals that might be helpful to understand the details better.

CMakeLists.txt

@@ -82,6 +82,7 @@ set(LLAMA_BLAS_VENDOR "Generic" CACHE STRING "llama: BLAS library vendor")
 option(LLAMA_CUBLAS                          "llama: use CUDA"                                  OFF)
 #option(LLAMA_CUDA_CUBLAS                     "llama: use cuBLAS for prompt processing"          OFF)
 option(LLAMA_CUDA_FORCE_DMMV                 "llama: use dmmv instead of mmvq CUDA kernels"     OFF)
+option(LLAMA_CUDA_FORCE_MMQ                  "llama: use mmq kernels instead of cuBLAS"         OFF)
 set(LLAMA_CUDA_DMMV_X      "32" CACHE STRING "llama: x stride for dmmv CUDA kernels")
 set(LLAMA_CUDA_MMV_Y        "1" CACHE STRING "llama: y block size for mmv CUDA kernels")
 option(LLAMA_CUDA_F16                        "llama: use 16 bit floats for some calculations"   OFF)
@@ -93,7 +94,6 @@ option(LLAMA_CLBLAST                         "llama: use CLBlast"
 option(LLAMA_METAL                           "llama: use Metal"                                 ${LLAMA_METAL_DEFAULT})
 option(LLAMA_METAL_NDEBUG                    "llama: disable Metal debugging"                   OFF)
 option(LLAMA_MPI                             "llama: use MPI"                                   OFF)
-option(LLAMA_K_QUANTS                        "llama: use k-quants"                              ON)
 option(LLAMA_QKK_64                          "llama: use super-block size of 64 for k-quants"   OFF)
 
 option(LLAMA_BUILD_TESTS                "llama: build tests"    ${LLAMA_STANDALONE})
@@ -277,13 +277,8 @@ if (LLAMA_BLAS)
     endif()
 endif()
 
-if (LLAMA_K_QUANTS)
-    set(GGML_HEADERS_EXTRA k_quants.h)
-    set(GGML_SOURCES_EXTRA k_quants.c)
-    add_compile_definitions(GGML_USE_K_QUANTS)
-    if (LLAMA_QKK_64)
-        add_compile_definitions(GGML_QKK_64)
-    endif()
+if (LLAMA_QKK_64)
+    add_compile_definitions(GGML_QKK_64)
 endif()
 
 if (LLAMA_CUBLAS)
@@ -305,6 +300,9 @@ if (LLAMA_CUBLAS)
         if (LLAMA_CUDA_FORCE_DMMV)
             add_compile_definitions(GGML_CUDA_FORCE_DMMV)
         endif()
+        if (LLAMA_CUDA_FORCE_MMQ)
+            add_compile_definitions(GGML_CUDA_FORCE_MMQ)
+        endif()
         add_compile_definitions(GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X})
         add_compile_definitions(GGML_CUDA_MMV_Y=${LLAMA_CUDA_MMV_Y})
         if (DEFINED LLAMA_CUDA_DMMV_Y)
@@ -331,6 +329,7 @@ if (LLAMA_CUBLAS)
             set(CMAKE_CUDA_ARCHITECTURES "60;61;70") # needed for f16 CUDA intrinsics
         else()
             set(CMAKE_CUDA_ARCHITECTURES "52;61;70") # lowest CUDA 12 standard + lowest for integer intrinsics
+            #set(CMAKE_CUDA_ARCHITECTURES "") # use this to compile much faster, but only F16 models work
         endif()
     endif()
     message(STATUS "Using CUDA architectures: ${CMAKE_CUDA_ARCHITECTURES}")
@@ -404,6 +403,9 @@ if (LLAMA_HIPBLAS)
         if (LLAMA_CUDA_FORCE_DMMV)
             target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_FORCE_DMMV)
         endif()
+        if (LLAMA_CUDA_FORCE_MMQ)
+            target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_FORCE_MMQ)
+        endif()
         target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X})
         target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_MMV_Y=${LLAMA_CUDA_MMV_Y})
         target_compile_definitions(ggml-rocm PRIVATE K_QUANTS_PER_ITERATION=${LLAMA_CUDA_KQUANTS_ITER})
@@ -665,6 +667,8 @@ add_library(ggml OBJECT
             ggml-alloc.h
             ggml-backend.c
             ggml-backend.h
+            ggml-quants.c
+            ggml-quants.h
             ${GGML_SOURCES_CUDA} ${GGML_HEADERS_CUDA}
             ${GGML_SOURCES_OPENCL} ${GGML_HEADERS_OPENCL}
             ${GGML_SOURCES_METAL} ${GGML_HEADERS_METAL}

Makefile

@@ -342,13 +342,9 @@ else
 	MK_CXXFLAGS += -march=rv64gcv -mabi=lp64d
 endif
 
-ifndef LLAMA_NO_K_QUANTS
-	MK_CPPFLAGS += -DGGML_USE_K_QUANTS
-	OBJS     += k_quants.o
 ifdef LLAMA_QKK_64
 	MK_CPPFLAGS += -DGGML_QKK_64
 endif
-endif
 
 ifndef LLAMA_NO_ACCELERATE
 	# Mac OS - include Accelerate framework.
@@ -365,7 +361,7 @@ ifdef LLAMA_MPI
 	MK_CPPFLAGS += -DGGML_USE_MPI
 	MK_CFLAGS   += -Wno-cast-qual
 	MK_CXXFLAGS += -Wno-cast-qual
-	OBJS     += ggml-mpi.o
+	OBJS        += ggml-mpi.o
 endif # LLAMA_MPI
 
 ifdef LLAMA_OPENBLAS
@@ -382,7 +378,7 @@ endif # LLAMA_BLIS
 ifdef LLAMA_CUBLAS
 	MK_CPPFLAGS  += -DGGML_USE_CUBLAS -I/usr/local/cuda/include -I/opt/cuda/include -I$(CUDA_PATH)/targets/x86_64-linux/include
 	MK_LDFLAGS   += -lcublas -lculibos -lcudart -lcublasLt -lpthread -ldl -lrt -L/usr/local/cuda/lib64 -L/opt/cuda/lib64 -L$(CUDA_PATH)/targets/x86_64-linux/lib
-	OBJS      += ggml-cuda.o
+	OBJS         += ggml-cuda.o
 	NVCCFLAGS = --forward-unknown-to-host-compiler -use_fast_math
 ifdef LLAMA_CUDA_NVCC
 	NVCC = $(LLAMA_CUDA_NVCC)
@@ -397,6 +393,9 @@ endif # CUDA_DOCKER_ARCH
 ifdef LLAMA_CUDA_FORCE_DMMV
 	NVCCFLAGS += -DGGML_CUDA_FORCE_DMMV
 endif # LLAMA_CUDA_FORCE_DMMV
+ifdef LLAMA_CUDA_FORCE_MMQ
+	NVCCFLAGS += -DGGML_CUDA_FORCE_MMQ
+endif # LLAMA_CUDA_FORCE_MMQ
 ifdef LLAMA_CUDA_DMMV_X
 	NVCCFLAGS += -DGGML_CUDA_DMMV_X=$(LLAMA_CUDA_DMMV_X)
 else
@@ -494,11 +493,6 @@ ggml-mpi.o: ggml-mpi.c ggml-mpi.h
 	$(CC) $(CFLAGS) -c $< -o $@
 endif # LLAMA_MPI
 
-ifndef LLAMA_NO_K_QUANTS
-k_quants.o: k_quants.c k_quants.h
-	$(CC) $(CFLAGS) -c $< -o $@
-endif # LLAMA_NO_K_QUANTS
-
 # combine build flags with cmdline overrides
 override CFLAGS        := $(MK_CPPFLAGS) $(CPPFLAGS) $(MK_CFLAGS) $(CFLAGS)
 override CXXFLAGS      := $(MK_CPPFLAGS) $(CPPFLAGS) $(MK_CXXFLAGS) $(CXXFLAGS)
@@ -539,15 +533,18 @@ ggml-alloc.o: ggml-alloc.c ggml.h ggml-alloc.h
 ggml-backend.o: ggml-backend.c ggml.h ggml-backend.h
 	$(CC)  $(CFLAGS)   -c $< -o $@
 
-OBJS += ggml-alloc.o ggml-backend.o
+ggml-quants.o: ggml-quants.c ggml.h ggml-quants.h
+	$(CC) $(CFLAGS)    -c $< -o $@
+
+OBJS += ggml-alloc.o ggml-backend.o ggml-quants.o
 
 llama.o: llama.cpp ggml.h ggml-alloc.h ggml-backend.h ggml-cuda.h ggml-metal.h llama.h
 	$(CXX) $(CXXFLAGS) -c $< -o $@
 
-COMMON_H_DEPS = common/common.h common/sampling.h build-info.h common/log.h
-COMMON_DEPS   = $(COMMON_H_DEPS) common.o sampling.o grammar-parser.o
+COMMON_H_DEPS = common/common.h common/sampling.h common/log.h
+COMMON_DEPS   = common.o sampling.o grammar-parser.o
 
-common.o: common/common.cpp $(COMMON_H_DEPS)
+common.o: common/common.cpp build-info.h $(COMMON_H_DEPS)
 	$(CXX) $(CXXFLAGS) -c $< -o $@
 
 sampling.o: common/sampling.cpp $(COMMON_H_DEPS)

Package.swift

@@ -42,13 +42,12 @@ let package = Package(
                 "llama.cpp",
                 "ggml-alloc.c",
                 "ggml-backend.c",
-                "k_quants.c",
+                "ggml-quants.c",
             ] + additionalSources,
             resources: resources,
             publicHeadersPath: "spm-headers",
             cSettings: [
                 .unsafeFlags(["-Wno-shorten-64-to-32", "-O3", "-DNDEBUG"]),
-                .define("GGML_USE_K_QUANTS"),
                 .define("GGML_USE_ACCELERATE")
                 // NOTE: NEW_LAPACK will required iOS version 16.4+
                 // We should consider add this in the future when we drop support for iOS 14

build.zig

@@ -116,15 +116,10 @@ pub fn build(b: *std.build.Builder) !void {
     var make = try Maker.init(b);
     make.enable_lto = b.option(bool, "lto", "Enable LTO optimization, (default: false)") orelse false;
 
-    if (b.option(bool, "k-quants", "Enable K-quants, (default: true)") orelse true) {
-        try make.addFlag("-DGGML_USE_K_QUANTS");
-        const k_quants = make.obj("k_quants", "k_quants.c");
-        try make.objs.append(k_quants);
-    }
-
     const ggml = make.obj("ggml", "ggml.c");
     const ggml_alloc = make.obj("ggml-alloc", "ggml-alloc.c");
     const ggml_backend = make.obj("ggml-backend", "ggml-backend.c");
+    const ggml_quants = make.obj("ggml-quants", "ggml-quants.c");
     const llama = make.obj("llama", "llama.cpp");
     const common = make.obj("common", "common/common.cpp");
     const console = make.obj("console", "common/console.cpp");
@@ -133,14 +128,14 @@ pub fn build(b: *std.build.Builder) !void {
     const train = make.obj("train", "common/train.cpp");
     const clip = make.obj("clip", "examples/llava/clip.cpp");
 
-    _ = make.exe("main", "examples/main/main.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common, sampling, console, grammar_parser });
-    _ = make.exe("quantize", "examples/quantize/quantize.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common });
-    _ = make.exe("perplexity", "examples/perplexity/perplexity.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common });
-    _ = make.exe("embedding", "examples/embedding/embedding.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common });
-    _ = make.exe("finetune", "examples/finetune/finetune.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common, train });
-    _ = make.exe("train-text-from-scratch", "examples/train-text-from-scratch/train-text-from-scratch.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common, train });
+    _ = make.exe("main", "examples/main/main.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common, sampling, console, grammar_parser });
+    _ = make.exe("quantize", "examples/quantize/quantize.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common });
+    _ = make.exe("perplexity", "examples/perplexity/perplexity.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common });
+    _ = make.exe("embedding", "examples/embedding/embedding.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common });
+    _ = make.exe("finetune", "examples/finetune/finetune.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common, train });
+    _ = make.exe("train-text-from-scratch", "examples/train-text-from-scratch/train-text-from-scratch.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common, train });
 
-    const server = make.exe("server", "examples/server/server.cpp", &.{ ggml, ggml_alloc, ggml_backend, llama, common, sampling, grammar_parser, clip });
+    const server = make.exe("server", "examples/server/server.cpp", &.{ ggml, ggml_alloc, ggml_backend, ggml_quants, llama, common, sampling, grammar_parser, clip });
     if (server.target.isWindows()) {
         server.linkSystemLibrary("ws2_32");
     }

common/common.cpp

@@ -224,6 +224,7 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
                 break;
             }
             sparams.temp = std::stof(argv[i]);
+            sparams.temp = std::max(sparams.temp, 0.0f);
         } else if (arg == "--tfs") {
             if (++i >= argc) {
                 invalid_param = true;
@@ -743,7 +744,7 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
 #endif // GGML_USE_CUBLAS
 #endif
     printf("  --verbose-prompt      print prompt before generation\n");
-    fprintf(stderr, "  --simple-io           use basic IO for better compatibility in subprocesses and limited consoles\n");
+    printf("  --simple-io           use basic IO for better compatibility in subprocesses and limited consoles\n");
     printf("  --lora FNAME          apply LoRA adapter (implies --no-mmap)\n");
     printf("  --lora-scaled FNAME S apply LoRA adapter with user defined scaling S (implies --no-mmap)\n");
     printf("  --lora-base FNAME     optional model to use as a base for the layers modified by the LoRA adapter\n");
@@ -880,15 +881,15 @@ std::tuple<struct llama_model *, struct llama_context *> llama_init_from_gpt_par
     }
 
     if (params.ignore_eos) {
-        params.sparams.logit_bias[llama_token_eos(lctx)] = -INFINITY;
+        params.sparams.logit_bias[llama_token_eos(model)] = -INFINITY;
     }
 
     {
         LOG("warming up the model with an empty run\n");
 
-        std::vector<llama_token> tmp = { llama_token_bos(lctx), llama_token_eos(lctx), };
+        std::vector<llama_token> tmp = { llama_token_bos(model), llama_token_eos(model), };
         llama_decode(lctx, llama_batch_get_one(tmp.data(), std::min(tmp.size(), (size_t) params.n_batch), 0, 0));
-        llama_kv_cache_tokens_rm(lctx, -1, -1);
+        llama_kv_cache_clear(lctx);
         llama_reset_timings(lctx);
     }
 
@@ -941,7 +942,7 @@ std::string llama_token_to_piece(const struct llama_context * ctx, llama_token t
 }
 
 std::string llama_detokenize_spm(llama_context * ctx, const std::vector<llama_token> & tokens) {
-    const llama_token bos_id = llama_token_bos(ctx);
+    const llama_token bos_id = llama_token_bos(llama_get_model(ctx));
 
     std::string piece;
     std::string result;
@@ -1186,7 +1187,7 @@ void dump_non_result_info_yaml(FILE * stream, const gpt_params & params, const l
     fprintf(stream, "hellaswag: %s # default: false\n", params.hellaswag ? "true" : "false");
     fprintf(stream, "hellaswag_tasks: %zu # default: 400\n", params.hellaswag_tasks);
 
-    const auto logit_bias_eos = sparams.logit_bias.find(llama_token_eos(lctx));
+    const auto logit_bias_eos = sparams.logit_bias.find(llama_token_eos(llama_get_model(lctx)));
     const bool ignore_eos = logit_bias_eos != sparams.logit_bias.end() && logit_bias_eos->second == -INFINITY;
     fprintf(stream, "ignore_eos: %s # default: false\n", ignore_eos ? "true" : "false");
 

common/log.h

@@ -97,22 +97,23 @@
     #define LOG_TEE_TARGET stderr
 #endif
 
+// NOTE: currently disabled as it produces too many log files
 // Utility to obtain "pid" like unique process id and use it when creating log files.
-inline std::string log_get_pid()
-{
-    static std::string pid;
-    if (pid.empty())
-    {
-        // std::this_thread::get_id() is the most portable way of obtaining a "process id"
-        //  it's not the same as "pid" but is unique enough to solve multiple instances
-        //  trying to write to the same log.
-        std::stringstream ss;
-        ss << std::this_thread::get_id();
-        pid = ss.str();
-    }
-
-    return pid;
-}
+//inline std::string log_get_pid()
+//{
+//    static std::string pid;
+//    if (pid.empty())
+//    {
+//        // std::this_thread::get_id() is the most portable way of obtaining a "process id"
+//        //  it's not the same as "pid" but is unique enough to solve multiple instances
+//        //  trying to write to the same log.
+//        std::stringstream ss;
+//        ss << std::this_thread::get_id();
+//        pid = ss.str();
+//    }
+//
+//    return pid;
+//}
 
 // Utility function for generating log file names with unique id based on thread id.
 //  invocation with log_filename_generator( "llama", "log" ) creates a string "llama.<number>.log"
@@ -126,8 +127,8 @@ inline std::string log_filename_generator_impl(const std::string & log_file_base
     std::stringstream buf;
 
     buf << log_file_basename;
-    buf << ".";
-    buf << log_get_pid();
+    //buf << ".";
+    //buf << log_get_pid();
     buf << ".";
     buf << log_file_extension;
 

common/sampling.cpp

@@ -147,7 +147,7 @@ llama_token llama_sampling_sample(
 
     // apply penalties
     if (!prev.empty()) {
-        const float nl_logit = logits[llama_token_nl(ctx_main)];
+        const float nl_logit = logits[llama_token_nl(llama_get_model(ctx_main))];
 
         llama_sample_repetition_penalties(ctx_main, &cur_p,
                 prev.data() + prev.size() - penalty_last_n,
@@ -155,7 +155,7 @@ llama_token llama_sampling_sample(
 
         if (!penalize_nl) {
             for (size_t idx = 0; idx < cur_p.size; idx++) {
-                if (cur_p.data[idx].id == llama_token_nl(ctx_main)) {
+                if (cur_p.data[idx].id == llama_token_nl(llama_get_model(ctx_main))) {
                     cur_p.data[idx].logit = nl_logit;
                     break;
                 }
@@ -167,8 +167,12 @@ llama_token llama_sampling_sample(
         llama_sample_grammar(ctx_main, &cur_p, ctx_sampling->grammar);
     }
 
-    if (temp <= 0) {
-        // greedy sampling
+    if (temp < 0.0) {
+        // greedy sampling, with probs
+        llama_sample_softmax(ctx_main, &cur_p);
+        id = cur_p.data[0].id;
+    } else if (temp == 0.0) {
+        // greedy sampling, no probs
         id = llama_sample_token_greedy(ctx_main, &cur_p);
     } else {
         if (mirostat == 1) {

common/train.cpp

@@ -236,8 +236,8 @@ int64_t get_example_targets_batch(
     int64_t used_samples = 0;
 
     ggml_set_f32(target_probs, 0.0f);
-    llama_token bos = llama_token_bos(lctx);
-    llama_token eos = llama_token_eos(lctx);
+    llama_token bos = llama_token_bos(llama_get_model(lctx));
+    llama_token eos = llama_token_eos(llama_get_model(lctx));
     // printf("%s: example_id=%d n_batch=%d n_train_samples=%zu\n", __func__, example_id, n_batch, n_train_samples);
     for (int k=0; k<n_batch; ++k) {
         // printf("%s: batch %d\n", __func__, k);
@@ -924,7 +924,7 @@ size_t tokenize_file(
         for (llama_token token=0; token < n_vocab; ++token) {
             max_token_text_size = std::max(
                 max_token_text_size,
-                strlen(llama_token_get_text(lctx, token)));
+                strlen(llama_token_get_text(llama_get_model(lctx), token)));
         }
 
         // upper bound of context byte length.

convert-baichuan-hf-to-gguf.py

@@ -110,7 +110,7 @@ print("gguf: loading model "+dir_model.name)
 with open(dir_model / "config.json", "r", encoding="utf-8") as f:
     hparams = json.load(f)
 print("hello print: ",hparams["architectures"][0])
-if hparams["architectures"][0] != "BaichuanForCausalLM":
+if hparams["architectures"][0] != "BaichuanForCausalLM" and hparams["architectures"][0] != "BaiChuanForCausalLM":
     print("Model architecture not supported: " + hparams["architectures"][0])
 
     sys.exit()

convert-bloom-hf-to-gguf.py

@@ -118,15 +118,24 @@ tokenizer = AutoTokenizer.from_pretrained(dir_model)
 vocab_size = hparams.get("vocab_size", len(tokenizer.vocab))
 assert max(tokenizer.vocab.values()) < vocab_size
 
+added_vocab = tokenizer.get_added_vocab()
 reverse_vocab = {id: encoded_tok for encoded_tok, id in tokenizer.vocab.items()}
 
 for i in range(vocab_size):
-    tokens.append(reverse_vocab[i] if i in reverse_vocab else f"[PAD{i}]")
-    scores.append(0.0)  # dummy
-    toktypes.append(gguf.TokenType.NORMAL)
+    if i not in reverse_vocab:
+        tokens.append(f"[PAD{i}]")
+        toktypes.append(gguf.TokenType.USER_DEFINED)
+    elif reverse_vocab[i] in added_vocab:
+        tokens.append(reverse_vocab[i])
+        if tokenizer.added_tokens_decoder[i].special:
+            toktypes.append(gguf.TokenType.CONTROL)
+        else:
+            toktypes.append(gguf.TokenType.USER_DEFINED)
+    else:
+        tokens.append(reverse_vocab[i])
+        toktypes.append(gguf.TokenType.NORMAL)
 
 gguf_writer.add_token_list(tokens)
-gguf_writer.add_token_scores(scores)
 gguf_writer.add_token_types(toktypes)
 
 special_vocab = gguf.SpecialVocab(dir_model, load_merges=True, n_vocab = len(tokens))

convert-gptneox-hf-to-gguf.py

@@ -123,15 +123,24 @@ tokenizer = AutoTokenizer.from_pretrained(dir_model)
 vocab_size = hparams.get("vocab_size", len(tokenizer.vocab))
 assert max(tokenizer.vocab.values()) < vocab_size
 
+added_vocab = tokenizer.get_added_vocab()
 reverse_vocab = {id: encoded_tok for encoded_tok, id in tokenizer.vocab.items()}
 
 for i in range(vocab_size):
-    tokens.append(reverse_vocab[i] if i in reverse_vocab else f"[PAD{i}]")
-    scores.append(0.0) # dummy
-    toktypes.append(gguf.TokenType.NORMAL)
+    if i not in reverse_vocab:
+        tokens.append(f"[PAD{i}]")
+        toktypes.append(gguf.TokenType.USER_DEFINED)
+    elif reverse_vocab[i] in added_vocab:
+        tokens.append(reverse_vocab[i])
+        if tokenizer.added_tokens_decoder[i].special:
+            toktypes.append(gguf.TokenType.CONTROL)
+        else:
+            toktypes.append(gguf.TokenType.USER_DEFINED)
+    else:
+        tokens.append(reverse_vocab[i])
+        toktypes.append(gguf.TokenType.NORMAL)
 
 gguf_writer.add_token_list(tokens)
-gguf_writer.add_token_scores(scores)
 gguf_writer.add_token_types(toktypes)
 
 special_vocab = gguf.SpecialVocab(dir_model, load_merges = True, n_vocab = len(tokens))

convert-mpt-hf-to-gguf.py

@@ -136,9 +136,11 @@ for i in range(vocab_size):
         tokens.append(f"[PAD{i}]")
         toktypes.append(gguf.TokenType.USER_DEFINED)
     elif reverse_vocab[i] in added_vocab:
-        # NOTE: wouldn't we like to distinguish CONTROL tokens here?
         tokens.append(reverse_vocab[i])
-        toktypes.append(gguf.TokenType.USER_DEFINED)
+        if tokenizer.added_tokens_decoder[i].special:
+            toktypes.append(gguf.TokenType.CONTROL)
+        else:
+            toktypes.append(gguf.TokenType.USER_DEFINED)
     else:
         tokens.append(reverse_vocab[i])
         toktypes.append(gguf.TokenType.NORMAL)

convert-refact-hf-to-gguf.py

@@ -139,15 +139,24 @@ tokenizer = AutoTokenizer.from_pretrained(dir_model)
 vocab_size = hparams.get("vocab_size", len(tokenizer.vocab))
 assert max(tokenizer.vocab.values()) < vocab_size
 
+added_vocab = tokenizer.get_added_vocab()
 reverse_vocab = {id: encoded_tok for encoded_tok, id in tokenizer.vocab.items()}
 
 for i in range(vocab_size):
-    tokens.append(reverse_vocab[i] if i in reverse_vocab else f"[PAD{i}]")
-    scores.append(0.0) # dummy
-    toktypes.append(gguf.TokenType.NORMAL)
+    if i not in reverse_vocab:
+        tokens.append(f"[PAD{i}]")
+        toktypes.append(gguf.TokenType.USER_DEFINED)
+    elif reverse_vocab[i] in added_vocab:
+        tokens.append(reverse_vocab[i])
+        if tokenizer.added_tokens_decoder[i].special:
+            toktypes.append(gguf.TokenType.CONTROL)
+        else:
+            toktypes.append(gguf.TokenType.USER_DEFINED)
+    else:
+        tokens.append(reverse_vocab[i])
+        toktypes.append(gguf.TokenType.NORMAL)
 
 gguf_writer.add_token_list(tokens)
-gguf_writer.add_token_scores(scores)
 gguf_writer.add_token_types(toktypes)
 
 special_vocab = gguf.SpecialVocab(dir_model, load_merges=True, n_vocab = len(tokens))

convert-starcoder-hf-to-gguf.py

@@ -111,17 +111,25 @@ tokenizer = AutoTokenizer.from_pretrained(dir_model)
 vocab_size = hparams.get("vocab_size", len(tokenizer.vocab))
 assert max(tokenizer.vocab.values()) < vocab_size
 
+added_vocab = tokenizer.get_added_vocab()
 reverse_vocab = {id: encoded_tok for encoded_tok, id in tokenizer.vocab.items()}
 
 for i in range(vocab_size):
-    tokens.append(reverse_vocab[i] if i in reverse_vocab else f"[PAD{i}]")
-    scores.append(0.0) # dummy
-    toktypes.append(gguf.TokenType.NORMAL)
+    if i not in reverse_vocab:
+        tokens.append(f"[PAD{i}]")
+        toktypes.append(gguf.TokenType.USER_DEFINED)
+    elif reverse_vocab[i] in added_vocab:
+        tokens.append(reverse_vocab[i])
+        if tokenizer.added_tokens_decoder[i].special:
+            toktypes.append(gguf.TokenType.CONTROL)
+        else:
+            toktypes.append(gguf.TokenType.USER_DEFINED)
+    else:
+        tokens.append(reverse_vocab[i])
+        toktypes.append(gguf.TokenType.NORMAL)
 
 gguf_writer.add_token_list(tokens)
-gguf_writer.add_token_scores(scores)
 gguf_writer.add_token_types(toktypes)
-
 special_vocab = gguf.SpecialVocab(dir_model, load_merges = True, n_vocab = len(tokens))
 special_vocab.add_to_gguf(gguf_writer)
 

convert.py

@@ -366,16 +366,19 @@ class SentencePieceVocab:
             added_tokens = {}
 
         vocab_size: int = self.sentencepiece_tokenizer.vocab_size()
-        expected_ids = list(range(vocab_size, vocab_size + len(added_tokens)))
-        actual_ids   = sorted(added_tokens.values())
-        if expected_ids != actual_ids:
-            raise Exception(f"Expected added token IDs to be sequential and start at {vocab_size}; got {actual_ids}")
 
-        items = sorted(added_tokens.items(), key=lambda text_idx: text_idx[1])
-        self.added_tokens_list = [text for (text, idx) in items]
-        self.vocab_size_base: int = vocab_size
-        self.vocab_size: int = self.vocab_size_base + len(self.added_tokens_list)
-        self.fname_tokenizer = fname_tokenizer
+        new_tokens       = {id: piece for piece, id in added_tokens.items() if id >= vocab_size}
+        expected_new_ids = list(range(vocab_size, vocab_size + len(new_tokens)))
+        actual_new_ids   = sorted(new_tokens.keys())
+
+        if expected_new_ids != actual_new_ids:
+            raise ValueError(f"Expected new token IDs {expected_new_ids} to be sequential; got {actual_new_ids}")
+
+        # Token pieces that were added to the base vocabulary.
+        self.added_tokens_list  = [new_tokens[id] for id in actual_new_ids]
+        self.vocab_size_base    = vocab_size
+        self.vocab_size         = self.vocab_size_base + len(self.added_tokens_list)
+        self.fname_tokenizer    = fname_tokenizer
         self.fname_added_tokens = fname_added_tokens
 
     def sentencepiece_tokens(self) -> Iterable[tuple[bytes, float, gguf.TokenType]]:

examples/batched-bench/batched-bench.cpp

@@ -154,6 +154,10 @@ int main(int argc, char ** argv) {
         }
     }
 
+    LOG_TEE("\n");
+    LOG_TEE("%s: n_kv_max = %d, is_pp_shared = %d, n_gpu_layers = %d, mmq = %d\n", __func__, n_kv_max, is_pp_shared, n_gpu_layers, mmq);
+    LOG_TEE("\n");
+
     LOG_TEE("|%6s | %6s | %4s | %6s | %8s | %8s | %8s | %8s | %8s | %8s |\n", "PP",     "TG",     "B",    "N_KV",     "T_PP s",   "S_PP t/s", "T_TG s",   "S_TG t/s", "T s",      "S t/s");
     LOG_TEE("|%6s-|-%6s-|-%4s-|-%6s-|-%8s-|-%8s-|-%8s-|-%8s-|-%8s-|-%8s-|\n", "------", "------", "----", "------", "--------", "--------", "--------", "--------", "--------", "--------");
 
@@ -181,7 +185,7 @@ int main(int argc, char ** argv) {
 
                 const auto t_pp_start = ggml_time_us();
 
-                llama_kv_cache_tokens_rm(ctx, -1, -1);
+                llama_kv_cache_clear(ctx);
 
                 if (!decode_helper(ctx, batch, ctx_params.n_batch)) {
                     LOG_TEE("%s: llama_decode() failed\n", __func__);

examples/batched/batched.cpp

@@ -11,7 +11,7 @@ int main(int argc, char ** argv) {
     gpt_params params;
 
     if (argc == 1 || argv[1][0] == '-') {
-        printf("usage: %s MODEL_PATH [PROMPT] [PARALLEL] [LEN]\n" , argv[0]);
+        printf("usage: %s MODEL_PATH [PROMPT] [PARALLEL] [LEN] [NGL]\n" , argv[0]);
         return 1 ;
     }
 
@@ -21,6 +21,9 @@ int main(int argc, char ** argv) {
     // total length of the sequences including the prompt
     int n_len = 32;
 
+    // number of layers to offload to the GPU
+    int n_gpu_layers = 0;
+
     if (argc >= 2) {
         params.model = argv[1];
     }
@@ -37,6 +40,10 @@ int main(int argc, char ** argv) {
         n_len = std::atoi(argv[4]);
     }
 
+    if (argc >= 6) {
+        n_gpu_layers = std::atoi(argv[5]);
+    }
+
     if (params.prompt.empty()) {
         params.prompt = "Hello my name is";
     }
@@ -49,7 +56,7 @@ int main(int argc, char ** argv) {
 
     llama_model_params model_params = llama_model_default_params();
 
-    // model_params.n_gpu_layers = 99; // offload all layers to the GPU
+    model_params.n_gpu_layers = n_gpu_layers;
 
     llama_model * model = llama_load_model_from_file(params.model.c_str(), model_params);
 
@@ -180,7 +187,7 @@ int main(int argc, char ** argv) {
             //const llama_token new_token_id = llama_sample_token_greedy(ctx, &candidates_p);
 
             // is it an end of stream? -> mark the stream as finished
-            if (new_token_id == llama_token_eos(ctx) || n_cur == n_len) {
+            if (new_token_id == llama_token_eos(model) || n_cur == n_len) {
                 i_batch[i] = -1;
                 LOG_TEE("\n");
                 if (n_parallel > 1) {

examples/beam-search/beam-search.cpp

@@ -47,7 +47,7 @@ struct beam_search_callback_data {
 // In this case, end-of-beam (eob) is equivalent to end-of-sentence (eos) but this need not always be the same.
 // For example, eob can be flagged due to maximum token length, stop words, etc.
 static bool is_at_eob(const beam_search_callback_data & callback_data, const llama_token * tokens, size_t n_tokens) {
-    return n_tokens && tokens[n_tokens-1] == llama_token_eos(callback_data.ctx);
+    return n_tokens && tokens[n_tokens-1] == llama_token_eos(llama_get_model(callback_data.ctx));
 }
 
 // Function matching type llama_beam_search_callback_fn_t.

examples/infill/infill.cpp

@@ -246,14 +246,14 @@ int main(int argc, char ** argv) {
     if (suff_rm_leading_spc && inp_sfx[0] == space_token) {
         inp_sfx.erase(inp_sfx.begin());
     }
-    inp_pfx.insert(inp_pfx.begin(), llama_token_prefix(ctx));
+    inp_pfx.insert(inp_pfx.begin(), llama_token_prefix(model));
     if (add_bos) {
-        inp_pfx.insert(inp_pfx.begin(), llama_token_bos(ctx));
+        inp_pfx.insert(inp_pfx.begin(), llama_token_bos(model));
     }
-    inp_sfx.insert(inp_sfx.begin(), llama_token_suffix(ctx));
+    inp_sfx.insert(inp_sfx.begin(), llama_token_suffix(model));
     embd_inp = inp_pfx;
     embd_inp.insert(embd_inp.end(), inp_sfx.begin(), inp_sfx.end());
-    embd_inp.push_back(llama_token_middle(ctx));
+    embd_inp.push_back(llama_token_middle(model));
 
     LOG("prefix: \"%s\"\n", log_tostr(params.input_prefix));
     LOG("suffix: \"%s\"\n", log_tostr(params.input_suffix));
@@ -261,7 +261,7 @@ int main(int argc, char ** argv) {
 
     // Should not run without any tokens
     if (embd_inp.empty()) {
-        embd_inp.push_back(llama_token_bos(ctx));
+        embd_inp.push_back(llama_token_bos(model));
         LOG("embd_inp was considered empty and bos was added: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, embd_inp).c_str());
     }
 
@@ -577,10 +577,10 @@ int main(int argc, char ** argv) {
         if ((int) embd_inp.size() <= n_consumed) {
 
             // deal with eot token in infill mode
-            if ((llama_sampling_last(ctx_sampling) == llama_token_eot(ctx) || is_interacting) && params.interactive){
+            if ((llama_sampling_last(ctx_sampling) == llama_token_eot(model) || is_interacting) && params.interactive){
                 if(is_interacting && !params.interactive_first) {
                     // print an eot token
-                    printf("%s", llama_token_to_piece(ctx, llama_token_eot(ctx)).c_str());
+                    printf("%s", llama_token_to_piece(ctx, llama_token_eot(model)).c_str());
                 }
                 fflush(stdout);
                 printf("\n");
@@ -627,14 +627,14 @@ int main(int argc, char ** argv) {
                 if (suff_rm_leading_spc && inp_sfx[0] == space_token) {
                     inp_sfx.erase(inp_sfx.begin());
                 }
-                inp_pfx.insert(inp_pfx.begin(), llama_token_prefix(ctx));
+                inp_pfx.insert(inp_pfx.begin(), llama_token_prefix(model));
                 if (add_bos) {
-                    inp_pfx.insert(inp_pfx.begin(), llama_token_bos(ctx));
+                    inp_pfx.insert(inp_pfx.begin(), llama_token_bos(model));
                 }
-                inp_sfx.insert(inp_sfx.begin(), llama_token_suffix(ctx));
+                inp_sfx.insert(inp_sfx.begin(), llama_token_suffix(model));
                 embd_inp = inp_pfx;
                 embd_inp.insert(embd_inp.end(), inp_sfx.begin(), inp_sfx.end());
-                embd_inp.push_back(llama_token_middle(ctx));
+                embd_inp.push_back(llama_token_middle(model));
                 embd.clear();
                 embd_guidance.clear();
                 n_remain = params.n_predict;
@@ -644,7 +644,7 @@ int main(int argc, char ** argv) {
                 is_interacting = false;
             }
             // deal with end of text token in interactive mode
-            else if (llama_sampling_last(ctx_sampling) == llama_token_eos(ctx)) {
+            else if (llama_sampling_last(ctx_sampling) == llama_token_eos(model)) {
                 LOG("found EOS token\n");
 
                 if (params.interactive) {
@@ -661,7 +661,7 @@ int main(int argc, char ** argv) {
 
                 if (params.input_prefix_bos) {
                     LOG("adding input prefix BOS token\n");
-                    embd_inp.push_back(llama_token_bos(ctx));
+                    embd_inp.push_back(llama_token_bos(model));
                 }
 
                 std::string buffer;
@@ -724,7 +724,7 @@ int main(int argc, char ** argv) {
         }
 
         // end of text token
-        if (!embd.empty() && embd.back() == llama_token_eos(ctx) && !params.interactive) {
+        if (!embd.empty() && embd.back() == llama_token_eos(model) && !params.interactive) {
             break;
         }
 
@@ -736,7 +736,7 @@ int main(int argc, char ** argv) {
         }
     }
     if (!params.interactive && n_remain <= 0) {
-        printf("%s", llama_token_to_piece(ctx, llama_token_eot(ctx)).c_str());
+        printf("%s", llama_token_to_piece(ctx, llama_token_eot(model)).c_str());
         fflush(stdout);
     }
 

examples/llama-bench/llama-bench.cpp

@@ -933,7 +933,7 @@ struct sql_printer : public printer {
 };
 
 static void test_prompt(llama_context * ctx, int n_prompt, int n_past, int n_batch, int n_threads) {
-    std::vector<llama_token> tokens(n_batch, llama_token_bos(ctx));
+    std::vector<llama_token> tokens(n_batch, llama_token_bos(llama_get_model(ctx)));
     int n_processed = 0;
 
     llama_set_n_threads(ctx, n_threads, n_threads);
@@ -946,7 +946,7 @@ static void test_prompt(llama_context * ctx, int n_prompt, int n_past, int n_bat
 }
 
 static void test_gen(llama_context * ctx, int n_gen, int n_past, int n_threads) {
-    llama_token token = llama_token_bos(ctx);
+    llama_token token = llama_token_bos(llama_get_model(ctx));
 
     llama_set_n_threads(ctx, n_threads, n_threads);
 
@@ -1037,7 +1037,7 @@ int main(int argc, char ** argv) {
 
         test t(inst, lmodel, ctx);
 
-        llama_kv_cache_tokens_rm(ctx, -1, -1);
+        llama_kv_cache_clear(ctx);
 
         // warmup run
         if (t.n_prompt > 0) {
@@ -1048,7 +1048,7 @@ int main(int argc, char ** argv) {
         }
 
         for (int i = 0; i < params.reps; i++) {
-            llama_kv_cache_tokens_rm(ctx, -1, -1);
+            llama_kv_cache_clear(ctx);
 
             uint64_t t_start = get_time_ns();
             if (t.n_prompt > 0) {

examples/llava/llava-utils.h

@@ -137,7 +137,7 @@ inline llama_token sample_id(llama_context * ctx_llama, gpt_params & params) {
 inline const char * sample(struct llama_context * ctx_llama, gpt_params & params, int * n_past) {
     int id = sample_id(ctx_llama, params);
     static std::string ret;
-    if (id == llama_token_eos(ctx_llama)) {
+    if (id == llama_token_eos(llama_get_model(ctx_llama))) {
         ret = "</s>";
     } else {
         ret = llama_token_to_piece(ctx_llama, id);

examples/main-cmake-pkg/CMakeLists.txt

@@ -16,6 +16,8 @@ add_library(common OBJECT
     ${_common_path}/console.cpp
     ${_common_path}/grammar-parser.h
     ${_common_path}/grammar-parser.cpp
+    ${_common_path}/sampling.h
+    ${_common_path}/sampling.cpp
     )
 
 # WARNING: because build-info.h is auto-generated, it will only

examples/main/main.cpp

@@ -248,7 +248,7 @@ int main(int argc, char ** argv) {
 
     // Should not run without any tokens
     if (embd_inp.empty()) {
-        embd_inp.push_back(llama_token_bos(ctx));
+        embd_inp.push_back(llama_token_bos(model));
         LOG("embd_inp was considered empty and bos was added: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, embd_inp).c_str());
     }
 
@@ -298,7 +298,7 @@ int main(int argc, char ** argv) {
         }
 
         // remove any "future" tokens that we might have inherited from the previous session
-        llama_kv_cache_tokens_rm(ctx, n_matching_session_tokens, -1);
+        llama_kv_cache_seq_rm(ctx, -1, n_matching_session_tokens, -1);
     }
 
     LOGLN(
@@ -693,7 +693,7 @@ int main(int argc, char ** argv) {
             }
 
             // deal with end of text token in interactive mode
-            if (llama_sampling_last(ctx_sampling) == llama_token_eos(ctx)) {
+            if (llama_sampling_last(ctx_sampling) == llama_token_eos(model)) {
                 LOG("found EOS token\n");
 
                 if (params.interactive) {
@@ -720,7 +720,7 @@ int main(int argc, char ** argv) {
 
                 if (params.input_prefix_bos) {
                     LOG("adding input prefix BOS token\n");
-                    embd_inp.push_back(llama_token_bos(ctx));
+                    embd_inp.push_back(llama_token_bos(model));
                 }
 
                 std::string buffer;
@@ -804,7 +804,7 @@ int main(int argc, char ** argv) {
         }
 
         // end of text token
-        if (!embd.empty() && embd.back() == llama_token_eos(ctx) && !(params.instruct || params.interactive)) {
+        if (!embd.empty() && embd.back() == llama_token_eos(model) && !(params.instruct || params.interactive)) {
             LOG_TEE(" [end of text]\n");
             break;
         }

examples/parallel/parallel.cpp

@@ -347,7 +347,7 @@ int main(int argc, char ** argv) {
                 //        client.id, client.seq_id, id, client.n_decoded, client.i_batch, token_str.c_str());
 
                 if (client.n_decoded > 2 &&
-                        (id == llama_token_eos(ctx) ||
+                        (id == llama_token_eos(model) ||
                          (params.n_predict > 0 && client.n_decoded + client.n_prompt >= params.n_predict) ||
                          client.response.find("User:") != std::string::npos ||
                          client.response.find('\n') != std::string::npos)) {

examples/perplexity/perplexity.cpp

@@ -210,7 +210,7 @@ static results_perplexity perplexity_v2(llama_context * ctx, const gpt_params &
         const auto t_start = std::chrono::high_resolution_clock::now();
 
         // clear the KV cache
-        llama_kv_cache_tokens_rm(ctx, -1, -1);
+        llama_kv_cache_clear(ctx);
 
         for (int j = 0; j < num_batches; ++j) {
             const int batch_start = start + j * n_batch;
@@ -227,7 +227,7 @@ static results_perplexity perplexity_v2(llama_context * ctx, const gpt_params &
 
             // add BOS token for the first batch of each chunk
             if (add_bos && j == 0) {
-                tokens[batch_start] = llama_token_bos(ctx);
+                tokens[batch_start] = llama_token_bos(llama_get_model(ctx));
             }
 
             const auto batch_logits = llama_get_logits(ctx);
@@ -339,7 +339,7 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
         const auto t_start = std::chrono::high_resolution_clock::now();
 
         // clear the KV cache
-        llama_kv_cache_tokens_rm(ctx, -1, -1);
+        llama_kv_cache_clear(ctx);
 
         for (int j = 0; j < num_batches; ++j) {
             const int batch_start = start + j * n_batch;
@@ -350,7 +350,7 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
 
             // add BOS token for the first batch of each chunk
             if (add_bos && j == 0) {
-                tokens[batch_start] = llama_token_bos(ctx);
+                tokens[batch_start] = llama_token_bos(llama_get_model(ctx));
             }
 
             if (llama_decode(ctx, llama_batch_get_one(tokens.data() + batch_start, batch_size, j * n_batch, 0))) {
@@ -573,7 +573,7 @@ static void hellaswag_score(llama_context * ctx, const gpt_params & params) {
         }
 
         // clear the KV cache
-        llama_kv_cache_tokens_rm(ctx, -1, -1);
+        llama_kv_cache_clear(ctx);
 
         auto logits = hellaswag_evaluate_tokens(ctx, query_embd, 0, params.n_batch, n_vocab);
         if (logits.empty()) {

examples/quantize/quantize.cpp

@@ -18,7 +18,6 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
     { "Q4_1",   LLAMA_FTYPE_MOSTLY_Q4_1,   " 3.90G, +0.1585 ppl @ LLaMA-v1-7B", },
     { "Q5_0",   LLAMA_FTYPE_MOSTLY_Q5_0,   " 4.33G, +0.0683 ppl @ LLaMA-v1-7B", },
     { "Q5_1",   LLAMA_FTYPE_MOSTLY_Q5_1,   " 4.70G, +0.0349 ppl @ LLaMA-v1-7B", },
-#ifdef GGML_USE_K_QUANTS
     { "Q2_K",   LLAMA_FTYPE_MOSTLY_Q2_K,   " 2.63G, +0.6717 ppl @ LLaMA-v1-7B", },
     { "Q3_K",   LLAMA_FTYPE_MOSTLY_Q3_K_M, "alias for Q3_K_M" },
     { "Q3_K_S", LLAMA_FTYPE_MOSTLY_Q3_K_S, " 2.75G, +0.5551 ppl @ LLaMA-v1-7B", },
@@ -31,7 +30,6 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
     { "Q5_K_S", LLAMA_FTYPE_MOSTLY_Q5_K_S, " 4.33G, +0.0400 ppl @ LLaMA-v1-7B", },
     { "Q5_K_M", LLAMA_FTYPE_MOSTLY_Q5_K_M, " 4.45G, +0.0122 ppl @ LLaMA-v1-7B", },
     { "Q6_K",   LLAMA_FTYPE_MOSTLY_Q6_K,   " 5.15G, -0.0008 ppl @ LLaMA-v1-7B", },
-#endif
     { "Q8_0",   LLAMA_FTYPE_MOSTLY_Q8_0,   " 6.70G, +0.0004 ppl @ LLaMA-v1-7B", },
     { "F16",    LLAMA_FTYPE_MOSTLY_F16,    "13.00G              @ 7B", },
     { "F32",    LLAMA_FTYPE_ALL_F32,       "26.00G              @ 7B", },
@@ -70,13 +68,14 @@ static bool try_parse_ftype(const std::string & ftype_str_in, llama_ftype & ftyp
 }
 
 // usage:
-//  ./quantize [--allow-requantize] [--leave-output-tensor] models/llama/ggml-model.gguf [models/llama/ggml-model-quant.gguf] type [nthreads]
+//  ./quantize [--allow-requantize] [--leave-output-tensor] [--pure] models/llama/ggml-model.gguf [models/llama/ggml-model-quant.gguf] type [nthreads]
 //
 [[noreturn]]
 static void usage(const char * executable) {
-    printf("usage: %s [--help] [--allow-requantize] [--leave-output-tensor] model-f32.gguf [model-quant.gguf] type [nthreads]\n\n", executable);
+    printf("usage: %s [--help] [--allow-requantize] [--leave-output-tensor] [--pure] model-f32.gguf [model-quant.gguf] type [nthreads]\n\n", executable);
     printf("  --allow-requantize: Allows requantizing tensors that have already been quantized. Warning: This can severely reduce quality compared to quantizing from 16bit or 32bit\n");
     printf("  --leave-output-tensor: Will leave output.weight un(re)quantized. Increases model size but may also increase quality, especially when requantizing\n");
+    printf("  --pure: Disable k-quant mixtures and quantize all tensors to the same type\n");
     printf("\nAllowed quantization types:\n");
     for (auto & it : QUANT_OPTIONS) {
         if (it.name != "COPY") {
@@ -103,6 +102,8 @@ int main(int argc, char ** argv) {
             params.quantize_output_tensor = false;
         } else if (strcmp(argv[arg_idx], "--allow-requantize") == 0) {
             params.allow_requantize = true;
+        } else if (strcmp(argv[arg_idx], "--pure") == 0) {
+            params.pure = true;
         } else {
             usage(argv[0]);
         }

examples/server/server.cpp

@@ -454,7 +454,7 @@ struct llama_client_slot
     }
 
     void release() {
-        if (state == PROCESSING)
+        if (state == IDLE || state == PROCESSING)
         {
             t_token_generation = (ggml_time_us() - t_start_genereration) / 1e3;
             command = RELEASE;
@@ -726,7 +726,7 @@ struct llama_server_context
 
         if (json_value(data, "ignore_eos", false))
         {
-            slot->sparams.logit_bias[llama_token_eos(ctx)] = -INFINITY;
+            slot->sparams.logit_bias[llama_token_eos(model)] = -INFINITY;
         }
 
         const auto &logit_bias = data.find("logit_bias");
@@ -754,6 +754,7 @@ struct llama_server_context
         }
 
         slot->params.antiprompt.clear();
+
         const auto &stop = data.find("stop");
         if (stop != data.end() && stop->is_array())
         {
@@ -856,7 +857,7 @@ struct llama_server_context
 
     void kv_cache_clear() {
         // clear the entire KV cache
-        llama_kv_cache_tokens_rm(ctx, -1, -1);
+        llama_kv_cache_clear(ctx);
         clean_kv_cache = false;
     }
 
@@ -867,7 +868,7 @@ struct llama_server_context
 
         kv_cache_clear();
 
-        for (int32_t i = 0; i < batch.n_tokens; ++i)
+        for (int i = 0; i < (int) system_tokens.size(); ++i)
         {
             llama_batch_add(batch, system_tokens[i], i, { 0 }, false);
         }
@@ -894,16 +895,8 @@ struct llama_server_context
         {
             slot.release();
         }
-        wait_all_are_idle();
-        all_slots_are_idle = true;
 
-        // wait until system prompt load
         system_need_update = true;
-        while (system_need_update)
-        {
-            std::this_thread::sleep_for(std::chrono::milliseconds(5));
-        }
-        // system prompt loaded, continue
     }
 
     void process_system_prompt_data(const json &sys_props) {
@@ -915,26 +908,6 @@ struct llama_server_context
         {
             notify_system_prompt_changed();
         }
-        else
-        {
-            system_need_update = true;
-        }
-    }
-
-    void wait_all_are_idle() {
-        bool wait = true;
-        while (wait)
-        {
-            wait = false;
-            for (auto &slot : slots)
-            {
-                if (!slot.available())
-                {
-                    wait = true;
-                    break;
-                }
-            }
-        }
     }
 
     static size_t find_stopping_strings(const std::string &text, const size_t last_token_size,
@@ -965,7 +938,6 @@ struct llama_server_context
                     slot.has_next_token = false;
                 }
                 stop_pos = pos;
-
             }
         }
 
@@ -1056,7 +1028,7 @@ struct llama_server_context
             slot.has_next_token = false;
         }
 
-        if (!slot.cache_tokens.empty() && result.tok == llama_token_eos(ctx))
+        if (!slot.cache_tokens.empty() && result.tok == llama_token_eos(model))
         {
             slot.stopped_eos = true;
             slot.has_next_token = false;
@@ -1130,7 +1102,7 @@ struct llama_server_context
 
     json get_formated_generation(llama_client_slot &slot)
     {
-        const auto eos_bias = slot.sparams.logit_bias.find(llama_token_eos(ctx));
+        const auto eos_bias = slot.sparams.logit_bias.find(llama_token_eos(model));
         const bool ignore_eos = eos_bias != slot.sparams.logit_bias.end() &&
                                 eos_bias->second < 0.0f && std::isinf(eos_bias->second);
         return json {
@@ -1444,7 +1416,7 @@ struct llama_server_context
         process_tasks();
 
         // update the system prompt wait until all slots are idle state
-        if (system_need_update)
+        if (system_need_update && all_slots_are_idle)
         {
             LOG_TEE("updating system prompt\n");
             update_system_prompt();
@@ -1498,7 +1470,7 @@ struct llama_server_context
         for (auto & slot : slots)
         {
             // release the slot
-            if (slot.state == PROCESSING && slot.command == RELEASE)
+            if (slot.command == RELEASE)
             {
                 slot.state = IDLE;
                 slot.command = NONE;
@@ -1509,7 +1481,7 @@ struct llama_server_context
                 continue;
             }
 
-            if (slot.state == IDLE || slot.command == RELEASE)
+            if (slot.state == IDLE)
             {
                 continue;
             }
@@ -1530,6 +1502,17 @@ struct llama_server_context
         {
             for (auto & slot : slots)
             {
+                const bool has_prompt = slot.prompt.is_array() || (slot.prompt.is_string() && !slot.prompt.get<std::string>().empty()) || !slot.images.empty();
+
+                // empty prompt passed -> release the slot and send empty response
+                if (slot.state == IDLE && slot.command == LOAD_PROMPT && !has_prompt)
+                {
+                    slot.release();
+                    slot.print_timings();
+                    send_final_response(slot);
+                    continue;
+                }
+
                 // need process the prompt
                 if (slot.state == IDLE && slot.command == LOAD_PROMPT)
                 {
@@ -1555,11 +1538,11 @@ struct llama_server_context
                             suffix_tokens.erase(suffix_tokens.begin());
                         }
 
-                        prefix_tokens.insert(prefix_tokens.begin(), llama_token_prefix(ctx));
-                        prefix_tokens.insert(prefix_tokens.begin(), llama_token_bos(ctx)); // always add BOS
-                        prefix_tokens.insert(prefix_tokens.end(), llama_token_suffix(ctx));
+                        prefix_tokens.insert(prefix_tokens.begin(), llama_token_prefix(model));
+                        prefix_tokens.insert(prefix_tokens.begin(), llama_token_bos(model)); // always add BOS
+                        prefix_tokens.insert(prefix_tokens.end(), llama_token_suffix(model));
                         prefix_tokens.insert(prefix_tokens.end(), suffix_tokens.begin(), suffix_tokens.end());
-                        prefix_tokens.push_back(llama_token_middle(ctx));
+                        prefix_tokens.push_back(llama_token_middle(model));
                         prompt_tokens = prefix_tokens;
                     }
                     else
@@ -1749,8 +1732,8 @@ struct llama_server_context
                 if (!process_token(result, slot))
                 {
                     slot.release();
-                    send_final_response(slot);
                     slot.print_timings();
+                    send_final_response(slot);
                 }
 
                 slot.i_batch = -1;
@@ -1766,15 +1749,16 @@ static void server_print_usage(const char *argv0, const gpt_params &params,
     printf("usage: %s [options]\n", argv0);
     printf("\n");
     printf("options:\n");
-    printf("  -h, --help            show this help message and exit\n");
-    printf("  -v, --verbose         verbose output (default: %s)\n", server_verbose ? "enabled" : "disabled");
-    printf("  -t N, --threads N     number of threads to use during computation (default: %d)\n", params.n_threads);
-    printf("  -c N, --ctx-size N    size of the prompt context (default: %d)\n", params.n_ctx);
-    printf("  --rope-freq-base N    RoPE base frequency (default: loaded from model)\n");
-    printf("  --rope-freq-scale N   RoPE frequency scaling factor (default: loaded from model)\n");
-    printf("  -b N, --batch-size N  batch size for prompt processing (default: %d)\n", params.n_batch);
-    printf("  --memory-f32          use f32 instead of f16 for memory key+value (default: disabled)\n");
-    printf("                        not recommended: doubles context memory required and no measurable increase in quality\n");
+    printf("  -h, --help                show this help message and exit\n");
+    printf("  -v, --verbose             verbose output (default: %s)\n", server_verbose ? "enabled" : "disabled");
+    printf("  -t N,  --threads N        number of threads to use during computation (default: %d)\n", params.n_threads);
+    printf("  -tb N, --threads-batch N  number of threads to use during batch and prompt processing (default: same as --threads)\n");
+    printf("  -c N,  --ctx-size N       size of the prompt context (default: %d)\n", params.n_ctx);
+    printf("  --rope-freq-base N        RoPE base frequency (default: loaded from model)\n");
+    printf("  --rope-freq-scale N       RoPE frequency scaling factor (default: loaded from model)\n");
+    printf("  -b N,  --batch-size N     batch size for prompt processing (default: %d)\n", params.n_batch);
+    printf("  --memory-f32              use f32 instead of f16 for memory key+value (default: disabled)\n");
+    printf("                            not recommended: doubles context memory required and no measurable increase in quality\n");
     if (llama_mlock_supported())
     {
         printf("  --mlock               force system to keep model in RAM rather than swapping or compressing\n");
@@ -1924,6 +1908,15 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
             }
             params.n_threads = std::stoi(argv[i]);
         }
+        else if (arg == "--threads-batch" || arg == "-tb")
+        {
+            if (++i >= argc)
+            {
+                invalid_param = true;
+                break;
+            }
+            params.n_threads_batch = std::stoi(argv[i]);
+        }
         else if (arg == "-b" || arg == "--batch-size")
         {
             if (++i >= argc)
@@ -2285,7 +2278,7 @@ int main(int argc, char **argv)
                 if (!json_value(data, "stream", false)) {
                     std::string completion_text;
                     task_result result = llama.next_result(task_id);
-                    if(!result.error && result.stop) {
+                    if (!result.error && result.stop) {
                         res.set_content(result.result_json.dump(-1, ' ', false, json::error_handler_t::replace), "application/json");
                     }
                     else
@@ -2312,7 +2305,7 @@ int main(int argc, char **argv)
                                 {
                                     return false;
                                 }
-                                if(result.stop) {
+                                if (result.stop) {
                                     break;
                                 }
                             } else {

examples/simple/simple.cpp

@@ -95,13 +95,8 @@ int main(int argc, char ** argv) {
     llama_batch batch = llama_batch_init(512, 0, 1);
 
     // evaluate the initial prompt
-    batch.n_tokens = tokens_list.size();
-
-    for (int32_t i = 0; i < batch.n_tokens; i++) {
-        batch.token[i]  = tokens_list[i];
-        batch.pos[i]    = i;
-        batch.seq_id[i] = 0;
-        batch.logits[i] = false;
+    for (size_t i = 0; i < tokens_list.size(); i++) {
+        llama_batch_add(batch, tokens_list[i], i, { 0 }, false);
     }
 
     // llama_decode will output logits only for the last token of the prompt
@@ -138,7 +133,7 @@ int main(int argc, char ** argv) {
             const llama_token new_token_id = llama_sample_token_greedy(ctx, &candidates_p);
 
             // is it an end of stream?
-            if (new_token_id == llama_token_eos(ctx) || n_cur == n_len) {
+            if (new_token_id == llama_token_eos(model) || n_cur == n_len) {
                 LOG_TEE("\n");
 
                 break;
@@ -148,15 +143,10 @@ int main(int argc, char ** argv) {
             fflush(stdout);
 
             // prepare the next batch
-            batch.n_tokens = 0;
+            llama_batch_clear(batch);
 
             // push this new token for next evaluation
-            batch.token [batch.n_tokens] = new_token_id;
-            batch.pos   [batch.n_tokens] = n_cur;
-            batch.seq_id[batch.n_tokens] = 0;
-            batch.logits[batch.n_tokens] = true;
-
-            batch.n_tokens += 1;
+            llama_batch_add(batch, new_token_id, n_cur, { 0 }, true);
 
             n_decode += 1;
         }

examples/speculative/speculative.cpp

@@ -8,6 +8,9 @@
 #include <string>
 #include <vector>
 
+#define SPEC_VOCAB_MAX_SIZE_DIFFERENCE  100
+#define SPEC_VOCAB_CHECK_START_TOKEN_ID 5
+
 struct seq_draft {
     bool active   = false;
     bool drafting = false;
@@ -64,6 +67,33 @@ int main(int argc, char ** argv) {
     params.n_gpu_layers = params.n_gpu_layers_draft;
     std::tie(model_dft, ctx_dft) = llama_init_from_gpt_params(params);
 
+    {
+        const int n_vocab_tgt = llama_n_vocab(model_tgt);
+        const int n_vocab_dft = llama_n_vocab(model_dft);
+        const int vocab_diff  = n_vocab_tgt > n_vocab_dft
+            ? n_vocab_tgt - n_vocab_dft
+            : n_vocab_dft - n_vocab_tgt;
+
+        if (vocab_diff > SPEC_VOCAB_MAX_SIZE_DIFFERENCE) {
+            fprintf(stderr, "%s: error: draft model vocab must closely match target model to use speculation but ", __func__);
+            fprintf(stderr, "target vocab size %d does not match draft vocab size %d - difference %d, max allowed %d\n",
+                    n_vocab_tgt, llama_n_vocab(model_dft), vocab_diff, SPEC_VOCAB_MAX_SIZE_DIFFERENCE);
+            return 1;
+        }
+
+        for (int i = SPEC_VOCAB_CHECK_START_TOKEN_ID; i < std::min(n_vocab_tgt, n_vocab_dft); ++i) {
+            const char * token_text_tgt = llama_token_get_text(model_tgt, i);
+            const char * token_text_dft = llama_token_get_text(model_dft, i);
+            if (std::strcmp(token_text_tgt, token_text_dft) != 0) {
+                fprintf(stderr, "%s: error: draft model vocab must match target model to use speculation but ", __func__);
+                fprintf(stderr, "token %d content differs - target '%s', draft '%s'\n", i,
+                        llama_token_to_piece(ctx_tgt, i).c_str(),
+                        llama_token_to_piece(ctx_dft, i).c_str());
+                return 1;
+            }
+        }
+    }
+
     // tokenize the prompt
     std::vector<llama_token> inp;
     inp = ::llama_tokenize(ctx_tgt, params.prompt, true);
@@ -118,7 +148,7 @@ int main(int argc, char ** argv) {
     std::vector<seq_draft> drafts(n_seq_dft);
 
     params.sparams.grammar.clear(); // the draft samplers will copy the target sampler's grammar
-    params.sparams.temp = std::max(0.01f, params.sparams.temp);
+    params.sparams.temp = -1.0f;    // force greedy sampling with probs for the draft model
 
     for (int s = 0; s < n_seq_dft; ++s) {
         drafts[s].ctx_sampling = llama_sampling_init(params.sparams);
@@ -163,7 +193,7 @@ int main(int argc, char ** argv) {
             printf("%s", token_str.c_str());
             fflush(stdout);
 
-            if (id == llama_token_eos(ctx_tgt)) {
+            if (id == llama_token_eos(model_tgt)) {
                 has_eos = true;
             }
 
@@ -227,6 +257,7 @@ int main(int argc, char ** argv) {
             llama_batch_add  (batch_dft, id, n_past_dft, { 0 }, true);
 
             llama_kv_cache_seq_rm(ctx_dft, 0, n_past_dft, -1);
+            // LOG("dft batch: %s\n", LOG_BATCH_TOSTR_PRETTY(ctx_dft, batch_dft).c_str());
             llama_decode         (ctx_dft, batch_dft);
 
             ++n_past_dft;
@@ -370,7 +401,7 @@ int main(int argc, char ** argv) {
                 llama_kv_cache_seq_cp(ctx_tgt, 0, s, -1, -1);
             }
 
-            //LOG("target batch: %s\n", LOG_BATCH_TOSTR_PRETTY(ctx_tgt, batch_tgt));
+            // LOG("target batch: %s\n", LOG_BATCH_TOSTR_PRETTY(ctx_tgt, batch_tgt).c_str());
             llama_decode(ctx_tgt, batch_tgt);
             ++n_past_tgt;
         }

flake.lock

@@ -20,11 +20,11 @@
     },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1692913444,
-        "narHash": "sha256-1SvMQm2DwofNxXVtNWWtIcTh7GctEVrS/Xel/mdc6iY=",
+        "lastModified": 1698134075,
+        "narHash": "sha256-foCD+nuKzfh49bIoiCBur4+Fx1nozo+4C/6k8BYk4sg=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "18324978d632ffc55ef1d928e81630c620f4f447",
+        "rev": "8efd5d1e283604f75a808a20e6cde0ef313d07d4",
         "type": "github"
       },
       "original": {

flake.nix

@@ -51,6 +51,9 @@
         };
         llama-python =
           pkgs.python3.withPackages (ps: with ps; [ numpy sentencepiece ]);
+        # TODO(Green-Sky): find a better way to opt-into the heavy ml python runtime
+        llama-python-extra =
+          pkgs.python3.withPackages (ps: with ps; [ numpy sentencepiece torchWithoutCuda transformers ]);
         postPatch = ''
           substituteInPlace ./ggml-metal.m \
             --replace '[bundle pathForResource:@"ggml-metal" ofType:@"metal"];' "@\"$out/bin/ggml-metal.metal\";"
@@ -126,5 +129,9 @@
           buildInputs = [ llama-python ];
           packages = nativeBuildInputs ++ osSpecific;
         };
+        devShells.extra = pkgs.mkShell {
+          buildInputs = [ llama-python-extra ];
+          packages = nativeBuildInputs ++ osSpecific;
+        };
       });
 }

ggml-cuda.cu

@@ -29,6 +29,8 @@
 #define __shfl_xor_sync(mask, var, laneMask, width) __shfl_xor(var, laneMask, width)
 #define cublasCreate hipblasCreate
 #define cublasGemmEx hipblasGemmEx
+#define cublasGemmBatchedEx hipblasGemmBatchedEx
+#define cublasGemmStridedBatchedEx hipblasGemmStridedBatchedEx
 #define cublasHandle_t hipblasHandle_t
 #define cublasSetMathMode(handle, mode) CUBLAS_STATUS_SUCCESS
 #define cublasSetStream hipblasSetStream
@@ -85,6 +87,24 @@
 #define CC_OFFSET_AMD 1000000
 #define CC_RDNA2      (CC_OFFSET_AMD + 1030)
 
+// define this if you want to always fallback to MMQ kernels and not use cuBLAS for matrix multiplication
+// on modern hardware, using cuBLAS is recommended as it utilizes F16 tensor cores which are very performant
+// for large computational tasks. the drawback is that this requires some extra amount of VRAM:
+// -  7B quantum model: +100-200 MB
+// - 13B quantum model: +200-400 MB
+//
+//#define GGML_CUDA_FORCE_MMQ
+
+// TODO: improve this to be correct for more hardware
+//       for example, currently fails for GeForce GTX 1660 which is TURING arch (> VOLTA) but does not have tensor cores
+//       probably other such cases, and not sure what happens on AMD hardware
+#if !defined(GGML_CUDA_FORCE_MMQ)
+#define CUDA_USE_TENSOR_CORES
+#endif
+
+// max batch size to use MMQ kernels when tensor cores are available
+#define MMQ_MAX_BATCH_SIZE 32
+
 #if defined(GGML_USE_HIPBLAS)
 #define __CUDA_ARCH__ 1300
 
@@ -468,7 +488,6 @@ static int g_device_count = -1;
 static int g_main_device = 0;
 static int g_compute_capabilities[GGML_CUDA_MAX_DEVICES];
 static float g_tensor_split[GGML_CUDA_MAX_DEVICES] = {0};
-static bool g_mul_mat_q = true;
 
 static void * g_scratch_buffer = nullptr;
 static size_t g_scratch_size = 0; // disabled by default
@@ -3552,9 +3571,15 @@ static __device__ __forceinline__ void mul_mat_q(
 #define  MMQ_X_Q4_0_RDNA1  64
 #define  MMQ_Y_Q4_0_RDNA1  64
 #define NWARPS_Q4_0_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q4_0_AMPERE 4
+#define  MMQ_Y_Q4_0_AMPERE 32
+#define NWARPS_Q4_0_AMPERE 4
+#else
 #define  MMQ_X_Q4_0_AMPERE 64
 #define  MMQ_Y_Q4_0_AMPERE 128
 #define NWARPS_Q4_0_AMPERE 4
+#endif
 #define  MMQ_X_Q4_0_PASCAL 64
 #define  MMQ_Y_Q4_0_PASCAL 64
 #define NWARPS_Q4_0_PASCAL 8
@@ -3613,9 +3638,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q4_1_RDNA1  64
 #define  MMQ_Y_Q4_1_RDNA1  64
 #define NWARPS_Q4_1_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q4_1_AMPERE 4
+#define  MMQ_Y_Q4_1_AMPERE 32
+#define NWARPS_Q4_1_AMPERE 4
+#else
 #define  MMQ_X_Q4_1_AMPERE 64
 #define  MMQ_Y_Q4_1_AMPERE 128
 #define NWARPS_Q4_1_AMPERE 4
+#endif
 #define  MMQ_X_Q4_1_PASCAL 64
 #define  MMQ_Y_Q4_1_PASCAL 64
 #define NWARPS_Q4_1_PASCAL 8
@@ -3676,9 +3707,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q5_0_RDNA1  64
 #define  MMQ_Y_Q5_0_RDNA1  64
 #define NWARPS_Q5_0_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q5_0_AMPERE 4
+#define  MMQ_Y_Q5_0_AMPERE 32
+#define NWARPS_Q5_0_AMPERE 4
+#else
 #define  MMQ_X_Q5_0_AMPERE 128
 #define  MMQ_Y_Q5_0_AMPERE 64
 #define NWARPS_Q5_0_AMPERE 4
+#endif
 #define  MMQ_X_Q5_0_PASCAL 64
 #define  MMQ_Y_Q5_0_PASCAL 64
 #define NWARPS_Q5_0_PASCAL 8
@@ -3737,9 +3774,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q5_1_RDNA1  64
 #define  MMQ_Y_Q5_1_RDNA1  64
 #define NWARPS_Q5_1_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q5_1_AMPERE 4
+#define  MMQ_Y_Q5_1_AMPERE 32
+#define NWARPS_Q5_1_AMPERE 4
+#else
 #define  MMQ_X_Q5_1_AMPERE 128
 #define  MMQ_Y_Q5_1_AMPERE 64
 #define NWARPS_Q5_1_AMPERE 4
+#endif
 #define  MMQ_X_Q5_1_PASCAL 64
 #define  MMQ_Y_Q5_1_PASCAL 64
 #define NWARPS_Q5_1_PASCAL 8
@@ -3798,9 +3841,15 @@ mul_mat_q5_1(
 #define  MMQ_X_Q8_0_RDNA1  64
 #define  MMQ_Y_Q8_0_RDNA1  64
 #define NWARPS_Q8_0_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q8_0_AMPERE 4
+#define  MMQ_Y_Q8_0_AMPERE 32
+#define NWARPS_Q8_0_AMPERE 4
+#else
 #define  MMQ_X_Q8_0_AMPERE 128
 #define  MMQ_Y_Q8_0_AMPERE 64
 #define NWARPS_Q8_0_AMPERE 4
+#endif
 #define  MMQ_X_Q8_0_PASCAL 64
 #define  MMQ_Y_Q8_0_PASCAL 64
 #define NWARPS_Q8_0_PASCAL 8
@@ -3859,9 +3908,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q2_K_RDNA1  128
 #define  MMQ_Y_Q2_K_RDNA1  32
 #define NWARPS_Q2_K_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q2_K_AMPERE 4
+#define  MMQ_Y_Q2_K_AMPERE 32
+#define NWARPS_Q2_K_AMPERE 4
+#else
 #define  MMQ_X_Q2_K_AMPERE 64
 #define  MMQ_Y_Q2_K_AMPERE 128
 #define NWARPS_Q2_K_AMPERE 4
+#endif
 #define  MMQ_X_Q2_K_PASCAL 64
 #define  MMQ_Y_Q2_K_PASCAL 64
 #define NWARPS_Q2_K_PASCAL 8
@@ -3920,9 +3975,15 @@ mul_mat_q2_K(
 #define  MMQ_X_Q3_K_RDNA1  32
 #define  MMQ_Y_Q3_K_RDNA1  128
 #define NWARPS_Q3_K_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q3_K_AMPERE 4
+#define  MMQ_Y_Q3_K_AMPERE 32
+#define NWARPS_Q3_K_AMPERE 4
+#else
 #define  MMQ_X_Q3_K_AMPERE 128
 #define  MMQ_Y_Q3_K_AMPERE 128
 #define NWARPS_Q3_K_AMPERE 4
+#endif
 #define  MMQ_X_Q3_K_PASCAL 64
 #define  MMQ_Y_Q3_K_PASCAL 64
 #define NWARPS_Q3_K_PASCAL 8
@@ -3983,9 +4044,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q4_K_RDNA1  32
 #define  MMQ_Y_Q4_K_RDNA1  64
 #define NWARPS_Q4_K_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q4_K_AMPERE 4
+#define  MMQ_Y_Q4_K_AMPERE 32
+#define NWARPS_Q4_K_AMPERE 4
+#else
 #define  MMQ_X_Q4_K_AMPERE 64
 #define  MMQ_Y_Q4_K_AMPERE 128
 #define NWARPS_Q4_K_AMPERE 4
+#endif
 #define  MMQ_X_Q4_K_PASCAL 64
 #define  MMQ_Y_Q4_K_PASCAL 64
 #define NWARPS_Q4_K_PASCAL 8
@@ -4046,9 +4113,15 @@ template <bool need_check> static __global__ void
 #define  MMQ_X_Q5_K_RDNA1  32
 #define  MMQ_Y_Q5_K_RDNA1  64
 #define NWARPS_Q5_K_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q5_K_AMPERE 4
+#define  MMQ_Y_Q5_K_AMPERE 32
+#define NWARPS_Q5_K_AMPERE 4
+#else
 #define  MMQ_X_Q5_K_AMPERE 64
 #define  MMQ_Y_Q5_K_AMPERE 128
 #define NWARPS_Q5_K_AMPERE 4
+#endif
 #define  MMQ_X_Q5_K_PASCAL 64
 #define  MMQ_Y_Q5_K_PASCAL 64
 #define NWARPS_Q5_K_PASCAL 8
@@ -4107,9 +4180,15 @@ mul_mat_q5_K(
 #define  MMQ_X_Q6_K_RDNA1  32
 #define  MMQ_Y_Q6_K_RDNA1  64
 #define NWARPS_Q6_K_RDNA1  8
+#if defined(CUDA_USE_TENSOR_CORES)
+#define  MMQ_X_Q6_K_AMPERE 4
+#define  MMQ_Y_Q6_K_AMPERE 32
+#define NWARPS_Q6_K_AMPERE 4
+#else
 #define  MMQ_X_Q6_K_AMPERE 64
 #define  MMQ_Y_Q6_K_AMPERE 64
 #define NWARPS_Q6_K_AMPERE 4
+#endif
 #define  MMQ_X_Q6_K_PASCAL 64
 #define  MMQ_Y_Q6_K_PASCAL 64
 #define NWARPS_Q6_K_PASCAL 8
@@ -4326,13 +4405,13 @@ static __global__ void mul_mat_vec_nc_f16_f32( // nc == non-contiguous
 
     const half * x = (const half *) vx;
 
-    const int row_x = blockDim.y*blockIdx.y + threadIdx.y;
-    const int channel = blockDim.z*blockIdx.z + threadIdx.z;
+    const int row_x     = blockDim.y*blockIdx.y + threadIdx.y;
+    const int channel   = blockDim.z*blockIdx.z + threadIdx.z;
     const int channel_x = channel / channel_x_divisor;
 
-    const int nrows_y = ncols_x;
+    const int nrows_y   = ncols_x;
     const int nrows_dst = nrows_x;
-    const int row_dst = row_x;
+    const int row_dst   = row_x;
 
     const int idst = channel*nrows_dst + row_dst;
 
@@ -4345,13 +4424,13 @@ static __global__ void mul_mat_vec_nc_f16_f32( // nc == non-contiguous
             break;
         }
 
-        const int ix = channel_x*channel_stride_x + row_x*row_stride_x + col_x;
-        const float xi = __half2float(x[ix]);
-
         const int row_y = col_x;
 
+        const int ix = channel_x*channel_stride_x + row_x*row_stride_x + col_x;
         const int iy = channel*nrows_y + row_y;
 
+        const float xi = __half2float(x[ix]);
+
         tmp += xi * y[iy];
     }
 
@@ -5661,11 +5740,21 @@ void ggml_init_cublas() {
         CUDA_CHECK(cudaGetDeviceCount(&g_device_count));
         GGML_ASSERT(g_device_count <= GGML_CUDA_MAX_DEVICES);
         int64_t total_vram = 0;
+#if defined(GGML_CUDA_FORCE_MMQ)
+        fprintf(stderr, "%s: GGML_CUDA_FORCE_MMQ:   yes\n", __func__);
+#else
+        fprintf(stderr, "%s: GGML_CUDA_FORCE_MMQ:   no\n", __func__);
+#endif
+#if defined(CUDA_USE_TENSOR_CORES)
+        fprintf(stderr, "%s: CUDA_USE_TENSOR_CORES: yes\n", __func__);
+#else
+        fprintf(stderr, "%s: CUDA_USE_TENSOR_CORES: no\n", __func__);
+#endif
         fprintf(stderr, "%s: found %d " GGML_CUDA_NAME " devices:\n", __func__, g_device_count);
-        for (int64_t id = 0; id < g_device_count; ++id) {
+        for (int id = 0; id < g_device_count; ++id) {
             cudaDeviceProp prop;
             CUDA_CHECK(cudaGetDeviceProperties(&prop, id));
-            fprintf(stderr, "  Device %ld: %s, compute capability %d.%d\n", id, prop.name, prop.major, prop.minor);
+            fprintf(stderr, "  Device %d: %s, compute capability %d.%d\n", id, prop.name, prop.major, prop.minor);
 
             g_tensor_split[id] = total_vram;
             total_vram += prop.totalGlobalMem;
@@ -5675,15 +5764,15 @@ void ggml_init_cublas() {
             g_compute_capabilities[id] = 100*prop.major + 10*prop.minor;
 #endif // defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)
         }
-        for (int64_t id = 0; id < g_device_count; ++id) {
+        for (int id = 0; id < g_device_count; ++id) {
             g_tensor_split[id] /= total_vram;
         }
 
-        for (int64_t id = 0; id < g_device_count; ++id) {
+        for (int id = 0; id < g_device_count; ++id) {
             CUDA_CHECK(ggml_cuda_set_device(id));
 
             // create cuda streams
-            for (int64_t is = 0; is < MAX_STREAMS; ++is) {
+            for (int is = 0; is < MAX_STREAMS; ++is) {
                 CUDA_CHECK(cudaStreamCreateWithFlags(&g_cudaStreams[id][is], cudaStreamNonBlocking));
             }
 
@@ -6252,16 +6341,15 @@ inline void ggml_cuda_op_mul_mat_cublas(
     const char * src1_ddq_i, float * dst_dd_i, const int64_t row_low, const int64_t row_high, const int64_t src1_ncols,
     const int64_t src1_padded_row_size, const cudaStream_t & stream) {
 
-    GGML_ASSERT(src0_dd_i != nullptr);
+    GGML_ASSERT(src0_dd_i  != nullptr);
     GGML_ASSERT(src1_ddf_i != nullptr);
-    GGML_ASSERT(dst_dd_i != nullptr);
-
+    GGML_ASSERT(dst_dd_i   != nullptr);
 
     const int64_t ne00 = src0->ne[0];
-
     const int64_t ne10 = src1->ne[0];
 
     const int64_t ne0 = dst->ne[0];
+
     const int64_t row_diff = row_high - row_low;
 
     int id;
@@ -6346,7 +6434,7 @@ inline void ggml_cuda_op_mul_mat_cublas(
             cublasSgemm(g_cublas_handles[id], CUBLAS_OP_T, CUBLAS_OP_N,
                     row_diff, src1_ncols, ne10,
                     &alpha, src0_ddf_i, ne00,
-                            src1_ddf_i,  ne10,
+                            src1_ddf_i, ne10,
                     &beta,  dst_dd_i,   ldc));
 
         if (src0_as != 0) {
@@ -7013,7 +7101,8 @@ static void ggml_cuda_mul_mat_vec_p021(const ggml_tensor * src0, const ggml_tens
 }
 
 static void ggml_cuda_mul_mat_vec_nc(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst){
-    GGML_ASSERT(!ggml_is_contiguous(src0) && ggml_is_contiguous(src1));
+    GGML_ASSERT(!ggml_is_transposed(src0));
+    GGML_ASSERT(!ggml_is_transposed(src1));
     GGML_ASSERT(!ggml_is_permuted(src0));
     GGML_ASSERT(src0->backend != GGML_BACKEND_GPU_SPLIT);
     GGML_ASSERT(src0->type == GGML_TYPE_F16);
@@ -7023,11 +7112,11 @@ static void ggml_cuda_mul_mat_vec_nc(const ggml_tensor * src0, const ggml_tensor
     const int64_t ne01 = src0->ne[1];
     const int64_t ne02 = src0->ne[2];
 
-    const int64_t ne12 = src1->ne[2];
-
     const int64_t nb01 = src0->nb[1];
     const int64_t nb02 = src0->nb[2];
 
+    const int64_t ne12 = src1->ne[2];
+
     CUDA_CHECK(ggml_cuda_set_device(g_main_device));
     cudaStream_t main_stream = g_cudaStreams[g_main_device][0];
 
@@ -7046,27 +7135,200 @@ static void ggml_cuda_mul_mat_vec_nc(const ggml_tensor * src0, const ggml_tensor
     ggml_mul_mat_vec_nc_f16_f32_cuda(src0_ddq, src1_ddf, dst_ddf, ne00, ne01, row_stride_x, ne02, ne12, channel_stride_x, main_stream);
 }
 
+static void ggml_cuda_mul_mat_mat_batched_cublas(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+    GGML_ASSERT(!ggml_is_transposed(src0));
+    GGML_ASSERT(!ggml_is_transposed(src1));
+
+    GGML_ASSERT(src0->backend != GGML_BACKEND_GPU_SPLIT);
+    GGML_ASSERT(src0->type == GGML_TYPE_F16);
+    GGML_ASSERT(src1->type == GGML_TYPE_F32);
+
+    const int64_t ne00 = src0->ne[0]; GGML_UNUSED(ne00);
+    const int64_t ne01 = src0->ne[1];
+    const int64_t ne02 = src0->ne[2];
+    const int64_t ne03 = src0->ne[3];
+
+    const int64_t nb01 = src0->nb[1];
+    const int64_t nb02 = src0->nb[2]; GGML_UNUSED(nb02);
+    const int64_t nb03 = src0->nb[3]; GGML_UNUSED(nb03);
+
+    const int64_t ne10 = src1->ne[0];
+    const int64_t ne11 = src1->ne[1];
+    const int64_t ne12 = src1->ne[2];
+    const int64_t ne13 = src1->ne[3];
+
+    const int64_t nb11 = src1->nb[1];
+    const int64_t nb12 = src1->nb[2]; GGML_UNUSED(nb12);
+    const int64_t nb13 = src1->nb[3]; GGML_UNUSED(nb13);
+
+    const int64_t ne1 = ggml_nelements(src1);
+    const int64_t ne  = ggml_nelements(dst);
+
+    CUDA_CHECK(ggml_cuda_set_device(g_main_device));
+    cudaStream_t main_stream = g_cudaStreams[g_main_device][0];
+
+    int id;
+    CUDA_CHECK(cudaGetDevice(&id));
+    CUBLAS_CHECK(cublasSetStream(g_cublas_handles[id], main_stream));
+
+    ggml_tensor_extra_gpu * src0_extra = (ggml_tensor_extra_gpu *) src0->extra;
+    void * src0_ddq = src0_extra->data_device[g_main_device];
+    half * src0_as_f16 = (half *) src0_ddq;
+
+    ggml_tensor_extra_gpu * src1_extra = (ggml_tensor_extra_gpu *) src1->extra;
+    float * src1_ddf = (float *) src1_extra->data_device[g_main_device];
+
+    ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra;
+    float * dst_ddf = (float *) dst_extra->data_device[g_main_device];
+
+    // convert src1 to fp16
+    const to_fp16_cuda_t to_fp16_cuda = ggml_get_to_fp16_cuda(src1->type);
+    GGML_ASSERT(to_fp16_cuda != nullptr);
+
+    size_t src1_as = 0;
+    half * src1_as_f16 = (half *) ggml_cuda_pool_malloc(ne1 * sizeof(half), &src1_as);
+    to_fp16_cuda(src1_ddf, src1_as_f16, ne1, main_stream);
+
+    size_t dst_as = 0;
+    half * dst_f16 = (half *) ggml_cuda_pool_malloc(ne * sizeof(half), &dst_as);
+
+    GGML_ASSERT(ne12 % ne02 == 0);
+    GGML_ASSERT(ne13 % ne03 == 0);
+
+    // broadcast factors
+    const int64_t r2 = ne12/ne02;
+    const int64_t r3 = ne13/ne03;
+
+    const half alpha_f16 = 1.0f;
+    const half beta_f16  = 0.0f;
+
+#if 0
+    // use cublasGemmEx
+    {
+        for (int i13 = 0; i13 < ne13; ++i13) {
+            for (int i12 = 0; i12 < ne12; ++i12) {
+                int i03 = i13 / r3;
+                int i02 = i12 / r2;
+
+                CUBLAS_CHECK(
+                        cublasGemmEx(g_cublas_handles[id], CUBLAS_OP_T, CUBLAS_OP_N,
+                            ne01, ne11, ne10,
+                            &alpha_f16, (const char *) src0_as_f16 + i02*src0->nb[2]   + i03*src0->nb[3]  , CUDA_R_16F, nb01/sizeof(half),
+                                        (const char *) src1_as_f16 + i12*src1->nb[2]/2 + i13*src1->nb[3]/2, CUDA_R_16F, nb11/sizeof(float),
+                            &beta_f16,  (      char *)     dst_f16 + i12* dst->nb[2]/2 + i13* dst->nb[3]/2, CUDA_R_16F, ne01,
+                            CUBLAS_COMPUTE_16F,
+                            CUBLAS_GEMM_DEFAULT_TENSOR_OP));
+            }
+        }
+    }
+#else
+    if (r2 == 1 && r3 == 1 && src0->nb[2]*src0->ne[2] == src0->nb[3] && src1->nb[2]*src1->ne[2] == src1->nb[3]) {
+        // there is no broadcast and src0, src1 are contiguous across dims 2, 3
+        // use cublasGemmStridedBatchedEx
+        CUBLAS_CHECK(
+        cublasGemmStridedBatchedEx(g_cublas_handles[id], CUBLAS_OP_T, CUBLAS_OP_N,
+                ne01, ne11, ne10,
+                &alpha_f16, (const char *) src0_as_f16, CUDA_R_16F, nb01/sizeof(half),  src0->nb[2]/sizeof(half),  // strideA
+                            (const char *) src1_as_f16, CUDA_R_16F, nb11/sizeof(float), src1->nb[2]/sizeof(float), // strideB
+                &beta_f16,  (      char *)     dst_f16, CUDA_R_16F, ne01,                dst->nb[2]/sizeof(float), // strideC
+                ne12*ne13,
+                CUBLAS_COMPUTE_16F,
+                CUBLAS_GEMM_DEFAULT_TENSOR_OP));
+    } else {
+        // use cublasGemmBatchedEx
+        // TODO: https://github.com/ggerganov/llama.cpp/pull/3749#discussion_r1369997000
+        const int ne23 = ne12*ne13;
+
+        // TODO: avoid this alloc
+        void ** ptrs = (void **) malloc(3*ne23*sizeof(void *));
+
+        for (int i13 = 0; i13 < ne13; ++i13) {
+            for (int i12 = 0; i12 < ne12; ++i12) {
+                int i03 = i13 / r3;
+                int i02 = i12 / r2;
+
+                ptrs[0*ne23 + i12 + i13*ne12] = (char *) src0_as_f16 + i02*src0->nb[2]   + i03*src0->nb[3];
+                ptrs[1*ne23 + i12 + i13*ne12] = (char *) src1_as_f16 + i12*src1->nb[2]/2 + i13*src1->nb[3]/2;
+                ptrs[2*ne23 + i12 + i13*ne12] = (char *)     dst_f16 + i12* dst->nb[2]/2 + i13* dst->nb[3]/2;
+            }
+        }
+
+        // allocate device memory for pointers
+        void ** ptrs_as = nullptr;
+        CUDA_CHECK(cudaMalloc(&ptrs_as, 3*ne23*sizeof(void *)));
+
+        // TODO: this does not work for some reason -- not sure why?
+        //size_t ptrs_s = 0;
+        //ptrs_as = (void **) ggml_cuda_pool_malloc(3*ne23*sizeof(void *), &ptrs_s);
+
+        // copy pointers to device
+        CUDA_CHECK(cudaMemcpy(ptrs_as, ptrs, 3*ne23*sizeof(void *), cudaMemcpyHostToDevice));
+
+        free(ptrs);
+
+        CUBLAS_CHECK(
+        cublasGemmBatchedEx(g_cublas_handles[id], CUBLAS_OP_T, CUBLAS_OP_N,
+                ne01, ne11, ne10,
+                &alpha_f16, (const void **) (ptrs_as + 0*ne23), CUDA_R_16F, nb01/sizeof(half),
+                            (const void **) (ptrs_as + 1*ne23), CUDA_R_16F, nb11/sizeof(float),
+                &beta_f16,  (      void **) (ptrs_as + 2*ne23), CUDA_R_16F, ne01,
+                ne23,
+                CUBLAS_COMPUTE_16F,
+                CUBLAS_GEMM_DEFAULT_TENSOR_OP));
+
+        // free device memory for pointers
+        CUDA_CHECK(cudaFree(ptrs_as));
+        //ggml_cuda_pool_free(ptrs_as, ptrs_s);
+    }
+#endif
+
+    const to_fp32_cuda_t to_fp32_cuda = ggml_get_to_fp32_cuda(GGML_TYPE_F16);
+    to_fp32_cuda(dst_f16, dst_ddf, ne, main_stream);
+
+    ggml_cuda_pool_free(src1_as_f16, src1_as);
+    ggml_cuda_pool_free(dst_f16, dst_as);
+}
+
 static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    bool all_on_device = (src0->backend == GGML_BACKEND_GPU || src0->backend == GGML_BACKEND_GPU_SPLIT) &&
-        src1->backend == GGML_BACKEND_GPU && dst->backend == GGML_BACKEND_GPU;
+    const bool all_on_device =
+        (src0->backend == GGML_BACKEND_GPU) &&
+        (src1->backend == GGML_BACKEND_GPU) &&
+        ( dst->backend == GGML_BACKEND_GPU);
 
     int64_t min_compute_capability = INT_MAX;
     for (int64_t id = 0; id < g_device_count; ++id) {
-        if (min_compute_capability > g_compute_capabilities[id]
-                && g_tensor_split[id] < (id + 1 < g_device_count ? g_tensor_split[id + 1] : 1.0f)) {
+        if (min_compute_capability > g_compute_capabilities[id] && g_tensor_split[id] < (id + 1 < g_device_count ? g_tensor_split[id + 1] : 1.0f)) {
             min_compute_capability = g_compute_capabilities[id];
         }
     }
 
-    if (all_on_device && src0->type == GGML_TYPE_F16 && ggml_is_permuted(src0) && ggml_is_permuted(src1) && src1->ne[1] == 1) {
+#ifdef CUDA_USE_TENSOR_CORES
+    const bool use_tensor_cores = true;
+#else
+    const bool use_tensor_cores = false;
+#endif
+
+    // debug helpers
+    //printf("src0: %8d %8d %8d %8d\n", src0->ne[0], src0->ne[1], src0->ne[2], src0->ne[3]);
+    //printf("      %8d %8d %8d %8d\n", src0->nb[0], src0->nb[1], src0->nb[2], src0->nb[3]);
+    //printf("src1: %8d %8d %8d %8d\n", src1->ne[0], src1->ne[1], src1->ne[2], src1->ne[3]);
+    //printf("      %8d %8d %8d %8d\n", src1->nb[0], src1->nb[1], src1->nb[2], src1->nb[3]);
+    //printf("src0 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src0), ggml_is_transposed(src0), ggml_type_name(src0->type), src0->name);
+    //printf("src1 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src1), ggml_is_transposed(src1), ggml_type_name(src1->type), src1->name);
+
+    if (all_on_device && !use_tensor_cores && src0->type == GGML_TYPE_F16 && ggml_is_permuted(src0) && ggml_is_permuted(src1) && src1->ne[1] == 1) {
+        // KQ single-batch
         ggml_cuda_mul_mat_vec_p021(src0, src1, dst);
-    } else if (all_on_device && !ggml_is_contiguous(src0) && ggml_is_contiguous(src1) && src1->ne[1] == 1) {
+    } else if (all_on_device && !use_tensor_cores && src0->type == GGML_TYPE_F16 && !ggml_is_contiguous(src0) && !ggml_is_transposed(src1) && src1->ne[1] == 1) {
+        // KQV single-batch
         ggml_cuda_mul_mat_vec_nc(src0, src1, dst);
+    } else if (all_on_device && src0->type == GGML_TYPE_F16 && src1->type == GGML_TYPE_F32 && !ggml_is_transposed(src0) && !ggml_is_transposed(src1)) {
+        // KQ + KQV multi-batch
+        ggml_cuda_mul_mat_mat_batched_cublas(src0, src1, dst);
     } else if (src0->type == GGML_TYPE_F32) {
         ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, false);
     } else if (ggml_is_quantized(src0->type) || src0->type == GGML_TYPE_F16) {
         if (src1->ne[1] == 1 && src0->ne[0] % GGML_CUDA_DMMV_X == 0) {
-
 #ifdef GGML_CUDA_FORCE_DMMV
             const bool use_mul_mat_vec_q = false;
 #else
@@ -7079,7 +7341,15 @@ static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
                 ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_dequantize_mul_mat_vec, false);
             }
         } else {
-            if (g_mul_mat_q && ggml_is_quantized(src0->type) && min_compute_capability >= MIN_CC_DP4A) {
+            bool use_mul_mat_q = min_compute_capability >= MIN_CC_DP4A && ggml_is_quantized(src0->type);
+
+            // when tensor cores are available, use them for large batch size
+            // ref: https://github.com/ggerganov/llama.cpp/pull/3776
+            if (use_tensor_cores && min_compute_capability >= CC_VOLTA && src1->ne[1] > MMQ_MAX_BATCH_SIZE) {
+                use_mul_mat_q = false;
+            }
+
+            if (use_mul_mat_q) {
                 ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_mul_mat_q, true);
             } else {
                 ggml_cuda_op_mul_mat(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, false);
@@ -7433,10 +7703,6 @@ void ggml_cuda_set_main_device(const int main_device) {
     }
 }
 
-void ggml_cuda_set_mul_mat_q(const bool mul_mat_q) {
-    g_mul_mat_q = mul_mat_q;
-}
-
 void ggml_cuda_set_scratch_size(const size_t scratch_size) {
     // this is a hack to not completely break llama.cpp when using multiple models or contexts simultaneously
     // it still won't always work as expected, but it's better than nothing

ggml-impl.h

@@ -0,0 +1,237 @@
+#pragma once
+
+#include "ggml.h"
+
+// GGML internal header
+
+#include <assert.h>
+#include <stddef.h>
+#include <stdbool.h>
+#include <string.h> // memcpy
+#include <math.h>   // fabsf
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+// static_assert should be a #define, but if it's not,
+// fall back to the _Static_assert C11 keyword.
+// if C99 - static_assert is noop
+// ref: https://stackoverflow.com/a/53923785/4039976
+#ifndef static_assert
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201100L)
+#define static_assert(cond, msg) _Static_assert(cond, msg)
+#else
+#define static_assert(cond, msg) struct global_scope_noop_trick
+#endif
+#endif
+
+// __FMA__ and __F16C__ are not defined in MSVC, however they are implied with AVX2/AVX512
+#if defined(_MSC_VER) && (defined(__AVX2__) || defined(__AVX512F__))
+#ifndef __FMA__
+#define __FMA__
+#endif
+#ifndef __F16C__
+#define __F16C__
+#endif
+#ifndef __SSE3__
+#define __SSE3__
+#endif
+#endif
+
+#undef MIN
+#undef MAX
+
+#define MIN(a, b) ((a) < (b) ? (a) : (b))
+#define MAX(a, b) ((a) > (b) ? (a) : (b))
+
+// 16-bit float
+// on Arm, we use __fp16
+// on x86, we use uint16_t
+#if defined(__ARM_NEON) && !defined(_MSC_VER)
+
+// if YCM cannot find <arm_neon.h>, make a symbolic link to it, for example:
+//
+//   $ ln -sfn /Library/Developer/CommandLineTools/usr/lib/clang/13.1.6/include/arm_neon.h ./src/
+//
+#include <arm_neon.h>
+
+#define GGML_COMPUTE_FP16_TO_FP32(x) ((float) (x))
+#define GGML_COMPUTE_FP32_TO_FP16(x) (x)
+
+#define GGML_FP16_TO_FP32(x) ((float) (x))
+#define GGML_FP32_TO_FP16(x) (x)
+
+#else
+
+#ifdef __wasm_simd128__
+#include <wasm_simd128.h>
+#else
+#ifdef __POWER9_VECTOR__
+#include <altivec.h>
+#undef bool
+#define bool _Bool
+#else
+#if defined(_MSC_VER) || defined(__MINGW32__)
+#include <intrin.h>
+#else
+#if defined(__AVX__) || defined(__AVX2__) || defined(__AVX512F__) || defined(__SSSE3__) || defined(__SSE3__)
+#if !defined(__riscv)
+#include <immintrin.h>
+#endif
+#endif
+#endif
+#endif
+#endif
+
+#ifdef __riscv_v_intrinsic
+#include <riscv_vector.h>
+#endif
+
+#ifdef __F16C__
+
+#ifdef _MSC_VER
+#define GGML_COMPUTE_FP16_TO_FP32(x) _mm_cvtss_f32(_mm_cvtph_ps(_mm_cvtsi32_si128(x)))
+#define GGML_COMPUTE_FP32_TO_FP16(x) _mm_extract_epi16(_mm_cvtps_ph(_mm_set_ss(x), 0), 0)
+#else
+#define GGML_COMPUTE_FP16_TO_FP32(x) _cvtsh_ss(x)
+#define GGML_COMPUTE_FP32_TO_FP16(x) _cvtss_sh(x, 0)
+#endif
+
+#elif defined(__POWER9_VECTOR__)
+
+#define GGML_COMPUTE_FP16_TO_FP32(x) ggml_compute_fp16_to_fp32(x)
+#define GGML_COMPUTE_FP32_TO_FP16(x) ggml_compute_fp32_to_fp16(x)
+/* the inline asm below is about 12% faster than the lookup method */
+#define GGML_FP16_TO_FP32(x) GGML_COMPUTE_FP16_TO_FP32(x)
+#define GGML_FP32_TO_FP16(x) GGML_COMPUTE_FP32_TO_FP16(x)
+
+static inline float ggml_compute_fp16_to_fp32(ggml_fp16_t h) {
+    register float f;
+    register double d;
+    __asm__(
+        "mtfprd %0,%2\n"
+        "xscvhpdp %0,%0\n"
+        "frsp %1,%0\n" :
+        /* temp */ "=d"(d),
+        /* out */  "=f"(f):
+        /* in */   "r"(h));
+    return f;
+}
+
+static inline ggml_fp16_t ggml_compute_fp32_to_fp16(float f) {
+    register double d;
+    register ggml_fp16_t r;
+    __asm__( /* xscvdphp can work on double or single precision */
+        "xscvdphp %0,%2\n"
+        "mffprd %1,%0\n" :
+        /* temp */ "=d"(d),
+        /* out */  "=r"(r):
+        /* in */   "f"(f));
+    return r;
+}
+
+#else
+
+// FP16 <-> FP32
+// ref: https://github.com/Maratyszcza/FP16
+
+static inline float fp32_from_bits(uint32_t w) {
+    union {
+        uint32_t as_bits;
+        float as_value;
+    } fp32;
+    fp32.as_bits = w;
+    return fp32.as_value;
+}
+
+static inline uint32_t fp32_to_bits(float f) {
+    union {
+        float as_value;
+        uint32_t as_bits;
+    } fp32;
+    fp32.as_value = f;
+    return fp32.as_bits;
+}
+
+static inline float ggml_compute_fp16_to_fp32(ggml_fp16_t h) {
+    const uint32_t w = (uint32_t) h << 16;
+    const uint32_t sign = w & UINT32_C(0x80000000);
+    const uint32_t two_w = w + w;
+
+    const uint32_t exp_offset = UINT32_C(0xE0) << 23;
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) || defined(__GNUC__) && !defined(__STRICT_ANSI__)
+    const float exp_scale = 0x1.0p-112f;
+#else
+    const float exp_scale = fp32_from_bits(UINT32_C(0x7800000));
+#endif
+    const float normalized_value = fp32_from_bits((two_w >> 4) + exp_offset) * exp_scale;
+
+    const uint32_t magic_mask = UINT32_C(126) << 23;
+    const float magic_bias = 0.5f;
+    const float denormalized_value = fp32_from_bits((two_w >> 17) | magic_mask) - magic_bias;
+
+    const uint32_t denormalized_cutoff = UINT32_C(1) << 27;
+    const uint32_t result = sign |
+        (two_w < denormalized_cutoff ? fp32_to_bits(denormalized_value) : fp32_to_bits(normalized_value));
+    return fp32_from_bits(result);
+}
+
+static inline ggml_fp16_t ggml_compute_fp32_to_fp16(float f) {
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) || defined(__GNUC__) && !defined(__STRICT_ANSI__)
+    const float scale_to_inf = 0x1.0p+112f;
+    const float scale_to_zero = 0x1.0p-110f;
+#else
+    const float scale_to_inf = fp32_from_bits(UINT32_C(0x77800000));
+    const float scale_to_zero = fp32_from_bits(UINT32_C(0x08800000));
+#endif
+    float base = (fabsf(f) * scale_to_inf) * scale_to_zero;
+
+    const uint32_t w = fp32_to_bits(f);
+    const uint32_t shl1_w = w + w;
+    const uint32_t sign = w & UINT32_C(0x80000000);
+    uint32_t bias = shl1_w & UINT32_C(0xFF000000);
+    if (bias < UINT32_C(0x71000000)) {
+        bias = UINT32_C(0x71000000);
+    }
+
+    base = fp32_from_bits((bias >> 1) + UINT32_C(0x07800000)) + base;
+    const uint32_t bits = fp32_to_bits(base);
+    const uint32_t exp_bits = (bits >> 13) & UINT32_C(0x00007C00);
+    const uint32_t mantissa_bits = bits & UINT32_C(0x00000FFF);
+    const uint32_t nonsign = exp_bits + mantissa_bits;
+    return (sign >> 16) | (shl1_w > UINT32_C(0xFF000000) ? UINT16_C(0x7E00) : nonsign);
+}
+
+#define GGML_COMPUTE_FP16_TO_FP32(x) ggml_compute_fp16_to_fp32(x)
+#define GGML_COMPUTE_FP32_TO_FP16(x) ggml_compute_fp32_to_fp16(x)
+
+#endif // __F16C__
+
+#endif // __ARM_NEON
+
+// precomputed f32 table for f16 (256 KB)
+// defined in ggml.c, initialized in ggml_init()
+extern float ggml_table_f32_f16[1 << 16];
+
+// On ARM NEON, it's quicker to directly convert x -> x instead of calling into ggml_lookup_fp16_to_fp32,
+// so we define GGML_FP16_TO_FP32 and GGML_FP32_TO_FP16 elsewhere for NEON.
+// This is also true for POWER9.
+#if !defined(GGML_FP16_TO_FP32) || !defined(GGML_FP32_TO_FP16)
+
+inline static float ggml_lookup_fp16_to_fp32(ggml_fp16_t f) {
+    uint16_t s;
+    memcpy(&s, &f, sizeof(uint16_t));
+    return ggml_table_f32_f16[s];
+}
+
+#define GGML_FP16_TO_FP32(x) ggml_lookup_fp16_to_fp32(x)
+#define GGML_FP32_TO_FP16(x) GGML_COMPUTE_FP32_TO_FP16(x)
+
+#endif
+
+    // TODO: backend v2 PR
+
+#ifdef __cplusplus
+}
+#endif

ggml-metal.m

@@ -62,6 +62,7 @@ struct ggml_metal_context {
     GGML_METAL_DECL_KERNEL(mul);
     GGML_METAL_DECL_KERNEL(mul_row); // TODO: avoid this extra kernel, instead extend the "mul" kernel to support broadcast
     GGML_METAL_DECL_KERNEL(scale);
+    GGML_METAL_DECL_KERNEL(scale_4);
     GGML_METAL_DECL_KERNEL(silu);
     GGML_METAL_DECL_KERNEL(relu);
     GGML_METAL_DECL_KERNEL(gelu);
@@ -209,6 +210,10 @@ struct ggml_metal_context * ggml_metal_init(int n_cb) {
             GGML_METAL_LOG_INFO("%s: default.metallib not found, loading from source\n", __func__);
 
             NSString * sourcePath = [bundle pathForResource:@"ggml-metal" ofType:@"metal"];
+            if (sourcePath == nil) {
+                GGML_METAL_LOG_WARN("%s: error: could not use bundle path to find ggml-metal.metal, falling back to trying cwd\n", __func__);
+                sourcePath = @"ggml-metal.metal";
+            }
             GGML_METAL_LOG_INFO("%s: loading '%s'\n", __func__, [sourcePath UTF8String]);
             NSString * src = [NSString stringWithContentsOfFile:sourcePath encoding:NSUTF8StringEncoding error:&error];
             if (error) {
@@ -249,6 +254,7 @@ struct ggml_metal_context * ggml_metal_init(int n_cb) {
         GGML_METAL_ADD_KERNEL(mul);
         GGML_METAL_ADD_KERNEL(mul_row);
         GGML_METAL_ADD_KERNEL(scale);
+        GGML_METAL_ADD_KERNEL(scale_4);
         GGML_METAL_ADD_KERNEL(silu);
         GGML_METAL_ADD_KERNEL(relu);
         GGML_METAL_ADD_KERNEL(gelu);
@@ -347,6 +353,7 @@ void ggml_metal_free(struct ggml_metal_context * ctx) {
     GGML_METAL_DEL_KERNEL(mul);
     GGML_METAL_DEL_KERNEL(mul_row);
     GGML_METAL_DEL_KERNEL(scale);
+    GGML_METAL_DEL_KERNEL(scale_4);
     GGML_METAL_DEL_KERNEL(silu);
     GGML_METAL_DEL_KERNEL(relu);
     GGML_METAL_DEL_KERNEL(gelu);
@@ -923,15 +930,20 @@ void ggml_metal_graph_compute(
 
                             const float scale = *(const float *) src1->data;
 
-                            [encoder setComputePipelineState:ctx->pipeline_scale];
+                            int64_t n = ggml_nelements(dst);
+
+                            if (n % 4 == 0) {
+                                n /= 4;
+                                [encoder setComputePipelineState:ctx->pipeline_scale_4];
+                            } else {
+                                [encoder setComputePipelineState:ctx->pipeline_scale];
+                            }
+
                             [encoder setBuffer:id_src0 offset:offs_src0 atIndex:0];
                             [encoder setBuffer:id_dst  offset:offs_dst  atIndex:1];
                             [encoder setBytes:&scale length:sizeof(scale) atIndex:2];
 
-                            const int64_t n = ggml_nelements(dst);
-                            GGML_ASSERT(n % 4 == 0);
-
-                            [encoder dispatchThreadgroups:MTLSizeMake(n/4, 1, 1) threadsPerThreadgroup:MTLSizeMake(1, 1, 1)];
+                            [encoder dispatchThreadgroups:MTLSizeMake(n, 1, 1) threadsPerThreadgroup:MTLSizeMake(1, 1, 1)];
                         } break;
                     case GGML_OP_UNARY:
                         switch (ggml_get_unary_op(gf->nodes[i])) {

ggml-metal.metal

@@ -125,9 +125,17 @@ kernel void kernel_mul_row(
 }
 
 kernel void kernel_scale(
+        device const float * src0,
+        device       float * dst,
+        constant     float & scale,
+        uint tpig[[thread_position_in_grid]]) {
+    dst[tpig] = src0[tpig] * scale;
+}
+
+kernel void kernel_scale_4(
         device const float4 * src0,
         device       float4 * dst,
-        constant     float & scale,
+        constant     float  & scale,
         uint tpig[[thread_position_in_grid]]) {
     dst[tpig] = src0[tpig] * scale;
 }

ggml-quants.h

@@ -1,11 +1,63 @@
 #pragma once
 
-#include "ggml.h"
+#include "ggml-impl.h"
+
+// GGML internal header
 
 #include <stdint.h>
-#include <assert.h>
 #include <stddef.h>
 
+#define QK4_0 32
+typedef struct {
+    ggml_fp16_t d;          // delta
+    uint8_t qs[QK4_0 / 2];  // nibbles / quants
+} block_q4_0;
+static_assert(sizeof(block_q4_0) == sizeof(ggml_fp16_t) + QK4_0 / 2, "wrong q4_0 block size/padding");
+
+#define QK4_1 32
+typedef struct {
+    ggml_fp16_t d;          // delta
+    ggml_fp16_t m;          // min
+    uint8_t qs[QK4_1 / 2];  // nibbles / quants
+} block_q4_1;
+static_assert(sizeof(block_q4_1) == 2 * sizeof(ggml_fp16_t) + QK4_1 / 2, "wrong q4_1 block size/padding");
+
+#define QK5_0 32
+typedef struct {
+    ggml_fp16_t d;         // delta
+    uint8_t qh[4];         // 5-th bit of quants
+    uint8_t qs[QK5_0 / 2]; // nibbles / quants
+} block_q5_0;
+static_assert(sizeof(block_q5_0) == sizeof(ggml_fp16_t) + sizeof(uint32_t) + QK5_0 / 2, "wrong q5_0 block size/padding");
+
+#define QK5_1 32
+typedef struct {
+    ggml_fp16_t d;         // delta
+    ggml_fp16_t m;         // min
+    uint8_t qh[4];         // 5-th bit of quants
+    uint8_t qs[QK5_1 / 2]; // nibbles / quants
+} block_q5_1;
+static_assert(sizeof(block_q5_1) == 2 * sizeof(ggml_fp16_t) + sizeof(uint32_t) + QK5_1 / 2, "wrong q5_1 block size/padding");
+
+#define QK8_0 32
+typedef struct {
+    ggml_fp16_t d;         // delta
+    int8_t  qs[QK8_0];     // quants
+} block_q8_0;
+static_assert(sizeof(block_q8_0) == sizeof(ggml_fp16_t) + QK8_0, "wrong q8_0 block size/padding");
+
+#define QK8_1 32
+typedef struct {
+    float d;               // delta
+    float s;               // d * sum(qs[i])
+    int8_t  qs[QK8_1];     // quants
+} block_q8_1;
+static_assert(sizeof(block_q8_1) == 2*sizeof(float) + QK8_1, "wrong q8_1 block size/padding");
+
+//
+// Super-block quantization structures
+//
+
 // Super-block size
 #ifdef GGML_QKK_64
 #define QK_K 64
@@ -15,18 +67,6 @@
 #define K_SCALE_SIZE 12
 #endif
 
-#ifndef static_assert
-#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201100L)
-#define static_assert(cond, msg) _Static_assert(cond, msg)
-#else
-#define static_assert(cond, msg) struct global_scope_noop_trick
-#endif
-#endif
-
-//
-// Super-block quantization structures
-//
-
 // 2-bit quantization
 // weight is represented as x = a * q + b
 // 16 blocks of 16 elements each
@@ -127,6 +167,13 @@ static_assert(sizeof(block_q8_K) == sizeof(float) + QK_K + QK_K/16*sizeof(int16_
 
 
 // Quantization
+void quantize_row_q4_0_reference(const float * restrict x, block_q4_0 * restrict y, int k);
+void quantize_row_q4_1_reference(const float * restrict x, block_q4_1 * restrict y, int k);
+void quantize_row_q5_0_reference(const float * restrict x, block_q5_0 * restrict y, int k);
+void quantize_row_q5_1_reference(const float * restrict x, block_q5_1 * restrict y, int k);
+void quantize_row_q8_0_reference(const float * restrict x, block_q8_0 * restrict y, int k);
+void quantize_row_q8_1_reference(const float * restrict x, block_q8_1 * restrict y, int k);
+
 void quantize_row_q2_K_reference(const float * restrict x, block_q2_K * restrict y, int k);
 void quantize_row_q3_K_reference(const float * restrict x, block_q3_K * restrict y, int k);
 void quantize_row_q4_K_reference(const float * restrict x, block_q4_K * restrict y, int k);
@@ -134,6 +181,13 @@ void quantize_row_q5_K_reference(const float * restrict x, block_q5_K * restrict
 void quantize_row_q6_K_reference(const float * restrict x, block_q6_K * restrict y, int k);
 void quantize_row_q8_K_reference(const float * restrict x, block_q8_K * restrict y, int k);
 
+void quantize_row_q4_0(const float * restrict x, void * restrict y, int k);
+void quantize_row_q4_1(const float * restrict x, void * restrict y, int k);
+void quantize_row_q5_0(const float * restrict x, void * restrict y, int k);
+void quantize_row_q5_1(const float * restrict x, void * restrict y, int k);
+void quantize_row_q8_0(const float * restrict x, void * restrict y, int k);
+void quantize_row_q8_1(const float * restrict x, void * restrict y, int k);
+
 void quantize_row_q2_K(const float * restrict x, void * restrict y, int k);
 void quantize_row_q3_K(const float * restrict x, void * restrict y, int k);
 void quantize_row_q4_K(const float * restrict x, void * restrict y, int k);
@@ -142,6 +196,13 @@ void quantize_row_q6_K(const float * restrict x, void * restrict y, int k);
 void quantize_row_q8_K(const float * restrict x, void * restrict y, int k);
 
 // Dequantization
+void dequantize_row_q4_0(const block_q4_0 * restrict x, float * restrict y, int k);
+void dequantize_row_q4_1(const block_q4_1 * restrict x, float * restrict y, int k);
+void dequantize_row_q5_0(const block_q5_0 * restrict x, float * restrict y, int k);
+void dequantize_row_q5_1(const block_q5_1 * restrict x, float * restrict y, int k);
+void dequantize_row_q8_0(const block_q8_0 * restrict x, float * restrict y, int k);
+//void dequantize_row_q8_1(const block_q8_1 * restrict x, float * restrict y, int k);
+
 void dequantize_row_q2_K(const block_q2_K * restrict x, float * restrict y, int k);
 void dequantize_row_q3_K(const block_q3_K * restrict x, float * restrict y, int k);
 void dequantize_row_q4_K(const block_q4_K * restrict x, float * restrict y, int k);
@@ -150,16 +211,14 @@ void dequantize_row_q6_K(const block_q6_K * restrict x, float * restrict y, int
 void dequantize_row_q8_K(const block_q8_K * restrict x, float * restrict y, int k);
 
 // Dot product
+void ggml_vec_dot_q4_0_q8_0(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q4_1_q8_1(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q5_0_q8_0(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q5_1_q8_1(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q8_0_q8_0(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+
 void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
 void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
 void ggml_vec_dot_q4_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
 void ggml_vec_dot_q5_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
 void ggml_vec_dot_q6_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
-
-// Quantization with histogram collection
-size_t ggml_quantize_q2_K(const float * src, void * dst, int n, int k, int64_t * hist);
-size_t ggml_quantize_q3_K(const float * src, void * dst, int n, int k, int64_t * hist);
-size_t ggml_quantize_q4_K(const float * src, void * dst, int n, int k, int64_t * hist);
-size_t ggml_quantize_q5_K(const float * src, void * dst, int n, int k, int64_t * hist);
-size_t ggml_quantize_q6_K(const float * src, void * dst, int n, int k, int64_t * hist);
-

ggml.h

@@ -401,15 +401,16 @@ extern "C" {
         GGML_OP_ALIBI,
         GGML_OP_CLAMP,
         GGML_OP_CONV_1D,
-        GGML_OP_CONV_2D,
+        GGML_OP_CONV_1D_STAGE_0,  // internal
+        GGML_OP_CONV_1D_STAGE_1,  // internal
         GGML_OP_CONV_TRANSPOSE_1D,
+        GGML_OP_CONV_2D,
+        GGML_OP_CONV_2D_STAGE_0, // internal
+        GGML_OP_CONV_2D_STAGE_1, // internal
         GGML_OP_CONV_TRANSPOSE_2D,
         GGML_OP_POOL_1D,
         GGML_OP_POOL_2D,
 
-        GGML_OP_CONV_1D_STAGE_0,  // internal
-        GGML_OP_CONV_1D_STAGE_1,  // internal
-
         GGML_OP_UPSCALE, // nearest interpolate
 
         GGML_OP_FLASH_ATTN,
@@ -1020,9 +1021,9 @@ extern "C" {
             struct ggml_tensor  * b,
             float                 eps);
 
-    // A: n columns, m rows
-    // B: n columns, p rows  (i.e. we transpose it internally)
-    // result is m columns, p rows
+    // A: k columns, n rows => [ne03, ne02, n, k]
+    // B: k columns, m rows  (i.e. we transpose it internally) => [ne03 * x, ne02 * y, m, k]
+    // result is n columns, m rows => [ne03 * x, ne02 * y, m, n]
     GGML_API struct ggml_tensor * ggml_mul_mat(
             struct ggml_context * ctx,
             struct ggml_tensor  * a,
@@ -1929,12 +1930,19 @@ extern "C" {
     // quantization
     //
 
+    // TODO: these would probably get removed in favor of the more general ggml_quantize_chunk
     GGML_API size_t ggml_quantize_q4_0(const float * src, void * dst, int n, int k, int64_t * hist);
     GGML_API size_t ggml_quantize_q4_1(const float * src, void * dst, int n, int k, int64_t * hist);
     GGML_API size_t ggml_quantize_q5_0(const float * src, void * dst, int n, int k, int64_t * hist);
     GGML_API size_t ggml_quantize_q5_1(const float * src, void * dst, int n, int k, int64_t * hist);
     GGML_API size_t ggml_quantize_q8_0(const float * src, void * dst, int n, int k, int64_t * hist);
 
+    GGML_API size_t ggml_quantize_q2_K(const float * src, void * dst, int n, int k, int64_t * hist);
+    GGML_API size_t ggml_quantize_q3_K(const float * src, void * dst, int n, int k, int64_t * hist);
+    GGML_API size_t ggml_quantize_q4_K(const float * src, void * dst, int n, int k, int64_t * hist);
+    GGML_API size_t ggml_quantize_q5_K(const float * src, void * dst, int n, int k, int64_t * hist);
+    GGML_API size_t ggml_quantize_q6_K(const float * src, void * dst, int n, int k, int64_t * hist);
+
     GGML_API size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst, int start, int n, int64_t * hist);
 
     //

llama.cpp

@@ -19,13 +19,11 @@
 #ifdef GGML_USE_MPI
 #  include "ggml-mpi.h"
 #endif
-#ifdef GGML_USE_K_QUANTS
-#  ifndef QK_K
-#    ifdef GGML_QKK_64
-#      define QK_K 64
-#    else
-#      define QK_K 256
-#    endif
+#ifndef QK_K
+#  ifdef GGML_QKK_64
+#    define QK_K 64
+#  else
+#    define QK_K 256
 #  endif
 #endif
 
@@ -1468,17 +1466,12 @@ static int32_t llama_kv_cache_cell_max(const struct llama_kv_cache & cache) {
     return 0;
 }
 
-static void llama_kv_cache_tokens_rm(struct llama_kv_cache & cache, int32_t c0, int32_t c1) {
-    if (c0 < 0) c0 = 0;
-    if (c1 < 0) c1 = cache.size;
-
-    for (int32_t i = c0; i < c1; ++i) {
+static void llama_kv_cache_clear(struct llama_kv_cache & cache) {
+    for (int32_t i = 0; i < (int32_t) cache.size; ++i) {
         cache.cells[i].pos = -1;
         cache.cells[i].seq_id.clear();
     }
-
-    // Searching for a free slot can start here since we know it will be empty.
-    cache.head = uint32_t(c0);
+    cache.head = 0;
 }
 
 static void llama_kv_cache_seq_rm(
@@ -1492,8 +1485,14 @@ static void llama_kv_cache_seq_rm(
     if (p1 < 0) p1 = std::numeric_limits<llama_pos>::max();
 
     for (uint32_t i = 0; i < cache.size; ++i) {
-        if (cache.cells[i].has_seq_id(seq_id) && cache.cells[i].pos >= p0 && cache.cells[i].pos < p1) {
-            cache.cells[i].seq_id.erase(seq_id);
+        if (cache.cells[i].pos >= p0 && cache.cells[i].pos < p1) {
+            if (seq_id < 0) {
+                cache.cells[i].seq_id.clear();
+            } else if (cache.cells[i].has_seq_id(seq_id)) {
+                cache.cells[i].seq_id.erase(seq_id);
+            } else {
+                continue;
+            }
             if (cache.cells[i].seq_id.empty()) {
                 cache.cells[i].pos = -1;
                 if (new_head == cache.size) new_head = i;
@@ -1554,14 +1553,14 @@ static void llama_kv_cache_seq_shift(
 
     for (uint32_t i = 0; i < cache.size; ++i) {
         if (cache.cells[i].has_seq_id(seq_id) && cache.cells[i].pos >= p0 && cache.cells[i].pos < p1) {
-            cache.cells[i].pos += delta;
+            cache.has_shift = true;
+            cache.cells[i].pos   += delta;
+            cache.cells[i].delta += delta;
+
             if (cache.cells[i].pos < 0) {
                 cache.cells[i].pos = -1;
                 cache.cells[i].seq_id.clear();
                 if (new_head == cache.size) new_head = i;
-            } else {
-                cache.has_shift = true;
-                cache.cells[i].delta = delta;
             }
         }
     }
@@ -1578,12 +1577,14 @@ static void llama_kv_cache_seq_shift(
 enum llama_fver {
     GGUF_FILE_VERSION_V1 = 1,
     GGUF_FILE_VERSION_V2 = 2,
+    GGUF_FILE_VERSION_V3 = 3,
 };
 
 static const char * llama_file_version_name(llama_fver version) {
     switch (version) {
         case GGUF_FILE_VERSION_V1: return "GGUF V1 (support until nov 2023)";
-        case GGUF_FILE_VERSION_V2: return "GGUF V2 (latest)";
+        case GGUF_FILE_VERSION_V2: return "GGUF V2";
+        case GGUF_FILE_VERSION_V3: return "GGUF V3 (latest)";
     }
 
     return "unknown";
@@ -2693,8 +2694,8 @@ static void llm_load_tensors(
                 } break;
             case LLM_ARCH_STARCODER:
                 {
-                    model.tok_embeddings = ml.create_tensor(ctx, tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, GGML_BACKEND_CPU);
-                    model.pos_embeddings = ml.create_tensor(ctx, tn(LLM_TENSOR_POS_EMBD, "weight"), {n_embd, hparams.n_ctx_train}, GGML_BACKEND_CPU);
+                    model.tok_embeddings = ml.create_tensor(ctx, tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab},             GGML_BACKEND_CPU);
+                    model.pos_embeddings = ml.create_tensor(ctx, tn(LLM_TENSOR_POS_EMBD, "weight"),   {n_embd, hparams.n_ctx_train}, GGML_BACKEND_CPU);
 
                     // output
                     {
@@ -2745,19 +2746,19 @@ static void llm_load_tensors(
                         layer.attn_norm_b = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_NORM,   "bias", i),   {n_embd}, backend);
 
                         layer.wqkv = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_QKV, "weight", i), {n_embd, n_embd + 2*n_embd_gqa}, backend_split);
-                        layer.bqkv = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_QKV, "bias", i),   {n_embd + 2*n_embd_gqa},         backend_split);
+                        layer.bqkv = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_QKV, "bias", i),   {n_embd + 2*n_embd_gqa},         backend);
 
                         layer.wo   = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "weight", i), {n_embd, n_embd},   backend_split);
-                        layer.bo   = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "bias", i),   {n_embd},           backend_split);
+                        layer.bo   = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "bias", i),   {n_embd},           backend);
 
                         layer.ffn_norm   = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_NORM, "weight", i), {n_embd}, backend);
                         layer.ffn_norm_b = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_NORM, "bias", i),   {n_embd}, backend);
 
                         layer.w2 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_DOWN, "weight", i), {n_ff, n_embd}, backend_split);
-                        layer.b2 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_DOWN, "bias", i),   {n_embd},       backend_split);
+                        layer.b2 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_DOWN, "bias", i),   {n_embd},       backend);
 
                         layer.w3 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_UP,   "weight", i), {n_embd,   n_ff}, backend_split);
-                        layer.b3 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_UP,   "bias", i),   {n_ff},           backend_split);
+                        layer.b3 = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_UP,   "bias", i),   {n_ff},           backend);
 
                         if (backend == GGML_BACKEND_GPU) {
                             vram_weights +=
@@ -4614,6 +4615,8 @@ static struct ggml_cgraph * llm_build_starcoder(
 
     const float norm_eps = hparams.f_norm_eps;
 
+    const int n_gpu_layers = model.n_gpu_layers;
+
     const int32_t n_tokens = batch.n_tokens;
     const int32_t n_kv     = ggml_allocr_is_measure(lctx.alloc) ? n_ctx            : kv_self.n;
     const int32_t kv_head  = ggml_allocr_is_measure(lctx.alloc) ? n_ctx - n_tokens : kv_self.head;
@@ -4658,6 +4661,27 @@ static struct ggml_cgraph * llm_build_starcoder(
         }
     }
 
+    const int i_gpu_start = n_layer - n_gpu_layers;
+    (void) i_gpu_start;
+
+    // offload functions set the tensor output backend to GPU
+    // tensors are GPU-accelerated if any input or the output has been offloaded
+    offload_func_t offload_func_nr = llama_nop; // nr = non-repeating
+    offload_func_t offload_func_kq = llama_nop;
+    offload_func_t offload_func_v  = llama_nop;
+
+#ifdef GGML_USE_CUBLAS
+    if (n_gpu_layers > n_layer) {
+        offload_func_nr = ggml_cuda_assign_buffers_no_alloc;
+    }
+    if (n_gpu_layers > n_layer + 1) {
+        offload_func_v  = ggml_cuda_assign_buffers_no_alloc;
+    }
+    if (n_gpu_layers > n_layer + 2) {
+        offload_func_kq = ggml_cuda_assign_buffers_no_alloc;
+    }
+#endif // GGML_USE_CUBLAS
+
     {
         // Compute position embeddings.
         struct ggml_tensor * inp_positions = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, n_tokens);
@@ -4683,6 +4707,7 @@ static struct ggml_cgraph * llm_build_starcoder(
     // KQ_mask (mask for 1 head, it will be broadcasted to all heads)
     struct ggml_tensor * KQ_mask = ggml_new_tensor_3d(ctx0, GGML_TYPE_F32, n_kv, n_tokens, 1);
     ggml_set_name(KQ_mask, "KQ_mask");
+    offload_func_kq(KQ_mask);
     ggml_allocr_alloc(lctx.alloc, KQ_mask);
     if (!ggml_allocr_is_measure(lctx.alloc)) {
         float * data = (float *) KQ_mask->data;
@@ -4706,44 +4731,67 @@ static struct ggml_cgraph * llm_build_starcoder(
     ggml_set_name(inpL, "inpL");
 
     for (int il = 0; il < n_layer; ++il) {
+        offload_func_t offload_func = llama_nop;
+
+#ifdef GGML_USE_CUBLAS
+        if (il >= i_gpu_start) {
+            offload_func = ggml_cuda_assign_buffers_no_alloc;
+        }
+#endif // GGML_USE_CUBLAS
+
         {
             // Norm
             cur = ggml_norm(ctx0, inpL, norm_eps);
+            offload_func(cur);
+
             cur = ggml_add(ctx0, ggml_mul(ctx0, cur, model.layers[il].attn_norm), model.layers[il].attn_norm_b);
+            offload_func(cur);
         }
 
         {
             // Self Attention
-            cur = ggml_add(ctx0, ggml_mul_mat(ctx0, model.layers[il].wqkv, cur), model.layers[il].bqkv);
+            cur = ggml_mul_mat(ctx0, model.layers[il].wqkv, cur);
+            offload_func_kq(cur);
 
-            struct ggml_tensor * tmpq = ggml_view_2d(ctx0, cur, n_embd, n_tokens, cur->nb[1], 0*sizeof(float)*n_embd);
-            struct ggml_tensor * tmpk = ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], sizeof(float)*n_embd);
-            struct ggml_tensor * tmpv = ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], sizeof(float)*(n_embd + n_embd_gqa));
+            cur = ggml_add(ctx0, cur, model.layers[il].bqkv);
+            offload_func_kq(cur);
 
-            struct ggml_tensor * Qcur = tmpq;
+            struct ggml_tensor * tmpq = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd,     n_tokens, cur->nb[1], 0*sizeof(float)*(n_embd)));
+            struct ggml_tensor * tmpk = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], 1*sizeof(float)*(n_embd)));
+            struct ggml_tensor * tmpv = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], 1*sizeof(float)*(n_embd + n_embd_gqa)));
+
+            ggml_set_name(tmpq, "tmpq");
+            ggml_set_name(tmpk, "tmpk");
+            ggml_set_name(tmpv, "tmpv");
+
+            offload_func_kq(tmpq);
+            offload_func_kq(tmpk);
+            offload_func_v (tmpv);
+
+            struct ggml_tensor * Qcur = ggml_reshape_3d(ctx0, tmpq, n_embd_head, n_head, n_tokens);
             struct ggml_tensor * Kcur = tmpk;
 
             {
-                struct ggml_tensor * Vcur = ggml_transpose(ctx0, ggml_reshape_2d(ctx0, ggml_cont(ctx0, tmpv), n_embd_gqa, n_tokens));
+                struct ggml_tensor * Vcur = ggml_transpose(ctx0, tmpv);
+                offload_func_v(Vcur);
                 ggml_set_name(Vcur, "Vcur");
 
                 struct ggml_tensor * k = ggml_view_1d(ctx0, kv_self.k, n_tokens*n_embd_gqa, (ggml_element_size(kv_self.k)*n_embd_gqa)*(il*n_ctx + kv_head));
+                offload_func_kq(k);
                 ggml_set_name(k, "k");
 
                 struct ggml_tensor * v = ggml_view_2d(ctx0, kv_self.v, n_tokens, n_embd_gqa,
                         (   n_ctx)*ggml_element_size(kv_self.v),
                         (il*n_ctx)*ggml_element_size(kv_self.v)*n_embd_gqa + kv_head*ggml_element_size(kv_self.v));
+                offload_func_v(v);
+                ggml_set_name(v, "v");
 
                 ggml_build_forward_expand(gf, ggml_cpy(ctx0, Kcur, k));
                 ggml_build_forward_expand(gf, ggml_cpy(ctx0, Vcur, v));
             }
 
-            struct ggml_tensor * Q =
-                ggml_permute(ctx0,
-                        ggml_cpy(ctx0,
-                            Qcur,
-                            ggml_new_tensor_3d(ctx0, GGML_TYPE_F32, n_embd_head, n_head, n_tokens)),
-                        0, 2, 1, 3);
+            struct ggml_tensor * Q = ggml_permute(ctx0, Qcur, 0, 2, 1, 3);
+            offload_func_kq(Q);
             ggml_set_name(Q, "Q");
 
             struct ggml_tensor * K =
@@ -4752,23 +4800,28 @@ static struct ggml_cgraph * llm_build_starcoder(
                         ggml_element_size(kv_self.k)*n_embd_gqa,
                         ggml_element_size(kv_self.k)*n_embd_head,
                         ggml_element_size(kv_self.k)*n_embd_gqa*n_ctx*il);
+            offload_func_kq(K);
             ggml_set_name(K, "K");
 
             // K * Q
             struct ggml_tensor * KQ = ggml_mul_mat(ctx0, K, Q);
+            offload_func_kq(KQ);
             ggml_set_name(KQ, "KQ");
 
             // KQ_scaled = KQ / sqrt(n_embd_head)
             // KQ_scaled shape [n_past + n_tokens, n_tokens, n_head, 1]
             struct ggml_tensor * KQ_scaled = ggml_scale_inplace(ctx0, KQ, KQ_scale);
+            offload_func_kq(KQ_scaled);
             ggml_set_name(KQ_scaled, "KQ_scaled");
 
             // KQ_masked = mask_past(KQ_scaled)
             struct ggml_tensor * KQ_masked = ggml_add(ctx0, KQ_scaled, KQ_mask);
+            offload_func_kq(KQ_masked);
             ggml_set_name(KQ_masked, "KQ_masked");
 
             // KQ = soft_max(KQ_masked)
             struct ggml_tensor * KQ_soft_max = ggml_soft_max_inplace(ctx0, KQ_masked);
+            offload_func_v(KQ_soft_max);
             ggml_set_name(KQ_soft_max, "KQ_soft_max");
 
             // split cached V into n_head heads
@@ -4781,22 +4834,25 @@ static struct ggml_cgraph * llm_build_starcoder(
             ggml_set_name(V, "V");
 
             struct ggml_tensor * KQV = ggml_mul_mat(ctx0, V, KQ_soft_max);
+            offload_func_v(KQV);
             ggml_set_name(KQV, "KQV");
 
-            // KQV_merged = KQV.permute(0, 2, 1, 3)
             struct ggml_tensor * KQV_merged = ggml_permute(ctx0, KQV, 0, 2, 1, 3);
+            offload_func_v(KQV_merged);
             ggml_set_name(KQV_merged, "KQV_merged");
 
-            // cur = KQV_merged.contiguous().view(n_embd, n_tokens)
             cur = ggml_cont_2d(ctx0, KQV_merged, n_embd, n_tokens);
+            offload_func_v(cur);
             ggml_set_name(cur, "KQV_merged_contiguous");
         }
 
         // Projection
         cur = ggml_add(ctx0, ggml_mul_mat(ctx0, model.layers[il].wo, cur), model.layers[il].bo);
+        offload_func(cur);
 
         // Add the input
         cur = ggml_add(ctx0, cur, inpL);
+        offload_func(cur);
 
         struct ggml_tensor * inpFF = cur;
 
@@ -4805,27 +4861,36 @@ static struct ggml_cgraph * llm_build_starcoder(
             // Norm
             {
                 cur = ggml_norm(ctx0, inpFF, norm_eps);
+                offload_func_nr(cur);
+
                 cur = ggml_add(ctx0, ggml_mul(ctx0, cur, model.layers[il].ffn_norm), model.layers[il].ffn_norm_b);
+                offload_func_nr(cur);
             }
 
             cur = ggml_add(ctx0, ggml_mul_mat(ctx0, model.layers[il].w3, cur), model.layers[il].b3);
+            offload_func(cur);
 
             // GELU activation
             cur = ggml_gelu(ctx0, cur);
+            offload_func(cur);
 
             // Projection
             cur = ggml_add(ctx0, ggml_mul_mat(ctx0, model.layers[il].w2, cur), model.layers[il].b2);
+            offload_func(cur);
         }
 
         inpL = ggml_add(ctx0, cur, inpFF);
+
     }
 
     // Output Norm
     {
         cur = ggml_norm(ctx0, inpL, norm_eps);
+        offload_func_nr(cur);
+
         cur = ggml_add(ctx0, ggml_mul(ctx0, cur, model.output_norm), model.output_norm_b);
+        ggml_set_name(cur, "result_norm");
     }
-    ggml_set_name(cur, "result_norm");
 
     cur = ggml_mul_mat(ctx0, model.output, cur);
     ggml_set_name(cur, "result_output");
@@ -5959,8 +6024,6 @@ static int llama_decode_internal(
         }
     }
 
-    ggml_cuda_set_mul_mat_q(cparams.mul_mat_q);
-
     // HACK: ggml-alloc may change the tensor backend when reusing a parent, so force output to be on the CPU here if needed
     if (!lctx.embedding.empty()) {
         embeddings->backend = GGML_BACKEND_CPU;
@@ -6011,11 +6074,20 @@ static int llama_decode_internal(
 #endif
 
     // update the kv ring buffer
-    lctx.kv_self.has_shift  = false;
-    lctx.kv_self.head      += n_tokens;
-    // Ensure kv cache head points to a valid index.
-    if (lctx.kv_self.head >= lctx.kv_self.size) {
-        lctx.kv_self.head = 0;
+    {
+        if (kv_self.has_shift) {
+            kv_self.has_shift = false;
+            for (uint32_t i = 0; i < kv_self.size; ++i) {
+                kv_self.cells[i].delta = 0;
+            }
+        }
+
+        kv_self.head += n_tokens;
+
+        // Ensure kv cache head points to a valid index.
+        if (kv_self.head >= kv_self.size) {
+            kv_self.head = 0;
+        }
     }
 
 #ifdef GGML_PERF
@@ -7493,7 +7565,7 @@ void llama_sample_grammar(struct llama_context * ctx, llama_token_data_array * c
         }
     }
 
-    const llama_token eos = llama_token_eos(ctx);
+    const llama_token eos = llama_token_eos(&ctx->model);
 
     std::vector<std::pair<std::vector<uint32_t>, llama_partial_utf8>> candidates_decoded;
     std::vector<llama_grammar_candidate>                              candidates_grammar;
@@ -7703,7 +7775,7 @@ llama_token llama_sample_token(struct llama_context * ctx, llama_token_data_arra
 void llama_grammar_accept_token(struct llama_context * ctx, struct llama_grammar * grammar, llama_token token) {
     const int64_t t_start_sample_us = ggml_time_us();
 
-    if (token == llama_token_eos(ctx)) {
+    if (token == llama_token_eos(&ctx->model)) {
         for (const auto & stack : grammar->stacks) {
             if (stack.empty()) {
                 return;
@@ -7985,6 +8057,24 @@ struct no_init {
     no_init() { /* do nothing */ }
 };
 
+struct quantize_state_internal {
+    const llama_model                 & model;
+    const llama_model_quantize_params * params;
+
+    int n_attention_wv    = 0;
+    int n_feed_forward_w2 = 0;
+    int i_attention_wv    = 0;
+    int i_feed_forward_w2 = 0;
+
+    int n_k_quantized     = 0;
+    int n_fallback        = 0;
+
+    quantize_state_internal(const llama_model & model, const llama_model_quantize_params * params)
+        : model(model)
+        , params(params)
+        {}
+};
+
 static void llama_convert_tensor_internal(
     struct ggml_tensor * tensor, std::vector<no_init<float>> & output, std::vector<std::thread> & workers,
     const size_t nelements, const int nthread
@@ -8043,14 +8133,14 @@ static void llama_convert_tensor_internal(
     workers.clear();
 }
 
-#ifdef GGML_USE_K_QUANTS
 static ggml_type get_k_quant_type(
-    ggml_type new_type, const ggml_tensor * tensor, const llama_model & model, llama_ftype ftype, int * i_attention_wv,
-    int n_attention_wv, int * i_feed_forward_w2, int n_feed_forward_w2
+    quantize_state_internal & qs,
+    ggml_type new_type, const ggml_tensor * tensor, llama_ftype ftype
 ) {
     const std::string name = ggml_get_name(tensor);
     // TODO: avoid hardcoded tensor names - use the TN_* constants
-    const auto tn = LLM_TN(model.arch);
+    const llm_arch arch = qs.model.arch;
+    const auto       tn = LLM_TN(arch);
 
     auto use_more_bits = [](int i_layer, int num_layers) -> bool {
         return i_layer < num_layers/8 || i_layer >= 7*num_layers/8 || (i_layer - num_layers/8)%3 == 2;
@@ -8058,7 +8148,7 @@ static ggml_type get_k_quant_type(
 
     if (name == tn(LLM_TENSOR_OUTPUT, "weight")) {
         int nx = tensor->ne[0];
-        if (model.arch == LLM_ARCH_FALCON || nx % QK_K != 0) {
+        if (arch == LLM_ARCH_FALCON || nx % QK_K != 0) {
             new_type = GGML_TYPE_Q8_0;
         }
         else if (new_type != GGML_TYPE_Q8_0) {
@@ -8067,46 +8157,46 @@ static ggml_type get_k_quant_type(
     } else if (name.find("attn_v.weight") != std::string::npos) {
         if      (ftype == LLAMA_FTYPE_MOSTLY_Q2_K) new_type = GGML_TYPE_Q3_K;
         else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_M) {
-            new_type = *i_attention_wv < 2 ? GGML_TYPE_Q5_K : GGML_TYPE_Q4_K;
+            new_type = qs.i_attention_wv < 2 ? GGML_TYPE_Q5_K : GGML_TYPE_Q4_K;
         }
         else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_L) new_type = GGML_TYPE_Q5_K;
         else if ((ftype == LLAMA_FTYPE_MOSTLY_Q4_K_M || ftype == LLAMA_FTYPE_MOSTLY_Q5_K_M) &&
-                use_more_bits(*i_attention_wv, n_attention_wv)) new_type = GGML_TYPE_Q6_K;
-        else if (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_S && *i_attention_wv < 4) new_type = GGML_TYPE_Q5_K;
+                use_more_bits(qs.i_attention_wv, qs.n_attention_wv)) new_type = GGML_TYPE_Q6_K;
+        else if (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_S && qs.i_attention_wv < 4) new_type = GGML_TYPE_Q5_K;
         else if (QK_K == 64 && (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_S || ftype == LLAMA_FTYPE_MOSTLY_Q3_K_S) &&
-                (*i_attention_wv < n_attention_wv/8 || *i_attention_wv >= 7*n_attention_wv/8)) new_type = GGML_TYPE_Q6_K;
-        if (model.type == MODEL_70B) {
+                (qs.i_attention_wv < qs.n_attention_wv/8 || qs.i_attention_wv >= 7*qs.n_attention_wv/8)) new_type = GGML_TYPE_Q6_K;
+        if (qs.model.type == MODEL_70B) {
             // In the 70B model we have 8 heads sharing the same attn_v weights. As a result, the attn_v.weight tensor is
             // 8x smaller compared to attn_q.weight. Hence, we can get a nice boost in quantization accuracy with
             // nearly negligible increase in model size by quantizing this tensor with more bits:
             if (new_type == GGML_TYPE_Q3_K || new_type == GGML_TYPE_Q4_K) new_type = GGML_TYPE_Q5_K;
         }
-        ++*i_attention_wv;
+        ++qs.i_attention_wv;
     } else if (name.find("ffn_down.weight") != std::string::npos) {
         if      (ftype == LLAMA_FTYPE_MOSTLY_Q2_K) new_type = GGML_TYPE_Q3_K;
         else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_M) {
-            new_type = *i_feed_forward_w2 < 2 ? GGML_TYPE_Q5_K
-                     : model.arch != LLM_ARCH_FALCON || use_more_bits(*i_feed_forward_w2, n_feed_forward_w2) ? GGML_TYPE_Q4_K
+            new_type = qs.i_feed_forward_w2 < 2 ? GGML_TYPE_Q5_K
+                     : arch != LLM_ARCH_FALCON || use_more_bits(qs.i_feed_forward_w2, qs.n_feed_forward_w2) ? GGML_TYPE_Q4_K
                      : GGML_TYPE_Q3_K;
         }
         else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_L) {
-            new_type = model.arch == LLM_ARCH_FALCON ? GGML_TYPE_Q4_K : GGML_TYPE_Q5_K;
+            new_type = arch == LLM_ARCH_FALCON ? GGML_TYPE_Q4_K : GGML_TYPE_Q5_K;
         }
         else if (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_M) {
-            if (model.arch == LLM_ARCH_FALCON) {
-                new_type = *i_feed_forward_w2 < 2 ? GGML_TYPE_Q6_K :
-                           use_more_bits(*i_feed_forward_w2, n_feed_forward_w2) ? GGML_TYPE_Q5_K : GGML_TYPE_Q4_K;
+            if (arch == LLM_ARCH_FALCON) {
+                new_type = qs.i_feed_forward_w2 < 2 ? GGML_TYPE_Q6_K :
+                           use_more_bits(qs.i_feed_forward_w2, qs.n_feed_forward_w2) ? GGML_TYPE_Q5_K : GGML_TYPE_Q4_K;
             } else {
-                if (use_more_bits(*i_feed_forward_w2, n_feed_forward_w2)) new_type = GGML_TYPE_Q6_K;
+                if (use_more_bits(qs.i_feed_forward_w2, qs.n_feed_forward_w2)) new_type = GGML_TYPE_Q6_K;
             }
         }
-        else if (ftype == LLAMA_FTYPE_MOSTLY_Q5_K_M && use_more_bits(*i_feed_forward_w2, n_feed_forward_w2)) new_type = GGML_TYPE_Q6_K;
-        else if (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_S && model.arch != LLM_ARCH_FALCON && *i_feed_forward_w2 < 4) {
+        else if (ftype == LLAMA_FTYPE_MOSTLY_Q5_K_M && use_more_bits(qs.i_feed_forward_w2, qs.n_feed_forward_w2)) new_type = GGML_TYPE_Q6_K;
+        else if (ftype == LLAMA_FTYPE_MOSTLY_Q4_K_S && arch != LLM_ARCH_FALCON && qs.i_feed_forward_w2 < 4) {
             new_type = GGML_TYPE_Q5_K;
         }
-        ++*i_feed_forward_w2;
+        ++qs.i_feed_forward_w2;
     } else if (name.find("attn_output.weight") != std::string::npos) {
-        if (model.arch != LLM_ARCH_FALCON) {
+        if (arch != LLM_ARCH_FALCON) {
             if      (ftype == LLAMA_FTYPE_MOSTLY_Q2_K  ) new_type = GGML_TYPE_Q3_K;
             else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_M) new_type = GGML_TYPE_Q4_K;
             else if (ftype == LLAMA_FTYPE_MOSTLY_Q3_K_L) new_type = GGML_TYPE_Q5_K;
@@ -8133,25 +8223,27 @@ static ggml_type get_k_quant_type(
         int nx = tensor->ne[0];
         int ny = tensor->ne[1];
         if (nx % QK_K != 0) {
-            LLAMA_LOG_WARN("\n\n%s : tensor cols %d x %d are not divisible by %d, required for k-quants\n", __func__, nx, ny, QK_K);
+            LLAMA_LOG_WARN("\n\n%s : tensor cols %d x %d are not divisible by %d, required for %s", __func__, nx, ny, QK_K, ggml_type_name(new_type));
             convert_incompatible_tensor = true;
+        } else {
+            ++qs.n_k_quantized;
         }
     }
     if (convert_incompatible_tensor) {
-        if (name == tn(LLM_TENSOR_OUTPUT, "weight")) {
-            new_type = GGML_TYPE_F16; //fall back to F16 instead of just failing.
-            LLAMA_LOG_WARN("F16 will be used for this tensor instead.\n");
-        } else if (name == tn(LLM_TENSOR_TOKEN_EMBD, "weight")) {
-            new_type = GGML_TYPE_Q4_0; //fall back to Q4_0 instead of just failing.
-            LLAMA_LOG_WARN("Q4_0 will be used for this tensor instead.\n");
-        } else {
-            throw std::runtime_error("Unsupported tensor size encountered\n");
+        switch (new_type) {
+            case GGML_TYPE_Q2_K: new_type = GGML_TYPE_Q4_0; break;
+            case GGML_TYPE_Q3_K: new_type = GGML_TYPE_Q4_1; break;
+            case GGML_TYPE_Q4_K: new_type = GGML_TYPE_Q5_0; break;
+            case GGML_TYPE_Q5_K: new_type = GGML_TYPE_Q5_1; break;
+            case GGML_TYPE_Q6_K: new_type = GGML_TYPE_Q8_0; break;
+            default: throw std::runtime_error("\nUnsupported tensor size encountered\n");
         }
+        LLAMA_LOG_WARN(" - using fallback quantization %s\n", ggml_type_name(new_type));
+        ++qs.n_fallback;
     }
 
     return new_type;
 }
-#endif
 
 static void llama_model_quantize_internal(const std::string & fname_inp, const std::string & fname_out, const llama_model_quantize_params * params) {
     ggml_type quantized_type;
@@ -8166,7 +8258,6 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
         case LLAMA_FTYPE_MOSTLY_F16:  quantized_type = GGML_TYPE_F16;  break;
         case LLAMA_FTYPE_ALL_F32:     quantized_type = GGML_TYPE_F32;  break;
 
-#ifdef GGML_USE_K_QUANTS
         // K-quants
         case LLAMA_FTYPE_MOSTLY_Q2_K:   quantized_type = GGML_TYPE_Q2_K; break;
         case LLAMA_FTYPE_MOSTLY_Q3_K_S:
@@ -8177,7 +8268,7 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
         case LLAMA_FTYPE_MOSTLY_Q5_K_S:
         case LLAMA_FTYPE_MOSTLY_Q5_K_M: quantized_type = GGML_TYPE_Q5_K; break;
         case LLAMA_FTYPE_MOSTLY_Q6_K:   quantized_type = GGML_TYPE_Q6_K; break;
-#endif
+
         default: throw std::runtime_error(format("invalid output file type %d\n", ftype));
     }
 
@@ -8204,6 +8295,8 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
     llm_load_arch(ml, model);
     llm_load_hparams(ml, model);
 
+    struct quantize_state_internal qs(model, params);
+
     if (params->only_copy) {
         ftype = model.ftype;
     }
@@ -8216,10 +8309,6 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
     gguf_set_val_u32(ctx_out, "general.quantization_version", GGML_QNT_VERSION);
     gguf_set_val_u32(ctx_out, "general.file_type", ftype);
 
-#ifdef GGML_USE_K_QUANTS
-    int n_attention_wv    = 0;
-    int n_feed_forward_w2 = 0;
-
     for (int i = 0; i < ml.n_tensors; ++i) {
         struct ggml_tensor * meta = ml.get_tensor_meta(i);
 
@@ -8227,21 +8316,17 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
 
         // TODO: avoid hardcoded tensor names - use the TN_* constants
         if (name.find("attn_v.weight") != std::string::npos || name.find("attn_qkv.weight") != std::string::npos) {
-            ++n_attention_wv;
+            ++qs.n_attention_wv;
         }
         else if (name.find("ffn_down.weight") != std::string::npos) {
-            ++n_feed_forward_w2;
+            ++qs.n_feed_forward_w2;
         }
     }
-    if (n_attention_wv != n_feed_forward_w2 || (uint32_t)n_attention_wv != model.hparams.n_layer) {
+    if (qs.n_attention_wv != qs.n_feed_forward_w2 || (uint32_t)qs.n_attention_wv != model.hparams.n_layer) {
         LLAMA_LOG_WARN("%s ============ Strange model: n_attention_wv = %d, n_feed_forward_w2 = %d, hparams.n_layer = %d\n",
-                __func__, n_attention_wv, n_feed_forward_w2, model.hparams.n_layer);
+                __func__, qs.n_attention_wv, qs.n_feed_forward_w2, model.hparams.n_layer);
     }
 
-    int i_attention_wv = 0;
-    int i_feed_forward_w2 = 0;
-#endif
-
     size_t total_size_org = 0;
     size_t total_size_new = 0;
     std::vector<int64_t> hist_all(1 << 4, 0);
@@ -8305,11 +8390,10 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
 
         if (quantize) {
             new_type = quantized_type;
-#ifdef GGML_USE_K_QUANTS
-            new_type = get_k_quant_type(
-                new_type, tensor, model, ftype, &i_attention_wv, n_attention_wv, &i_feed_forward_w2, n_feed_forward_w2
-            );
-#endif
+            if (!params->pure) {
+                new_type = get_k_quant_type(qs, new_type, tensor, ftype);
+            }
+
             // If we've decided to quantize to the same type the tensor is already
             // in then there's nothing to do.
             quantize = tensor->type != new_type;
@@ -8434,6 +8518,11 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
             LLAMA_LOG_INFO("\n");
         }
     }
+
+    if (qs.n_fallback > 0) {
+        LLAMA_LOG_WARN("%s: WARNING: %d of %d tensor(s) incompatible with k-quants and required fallback quantization\n",
+                __func__, qs.n_fallback, qs.n_k_quantized + qs.n_fallback);
+    }
 }
 
 static int llama_apply_lora_from_file_internal(
@@ -8758,6 +8847,7 @@ struct llama_model_quantize_params llama_model_quantize_default_params() {
         /*.allow_requantize            =*/ false,
         /*.quantize_output_tensor      =*/ true,
         /*.only_copy                   =*/ false,
+        /*.pure                        =*/ false,
     };
 
     return result;
@@ -8912,7 +9002,7 @@ struct llama_context * llama_new_context_with_model(
             // build worst-case graph
             int n_tokens = (int)std::min(cparams.n_ctx, cparams.n_batch);
             int n_past = cparams.n_ctx - n_tokens;
-            llama_token token = llama_token_bos(ctx); // not actually used by llama_build_graph, but required to choose between token and embedding inputs graph
+            llama_token token = llama_token_bos(&ctx->model); // not actually used by llama_build_graph, but required to choose between token and embedding inputs graph
             ggml_cgraph * gf = llama_build_graph(*ctx, llama_batch_get_one(&token, n_tokens, n_past, 0));
 
 #ifdef GGML_USE_METAL
@@ -9118,8 +9208,8 @@ int llama_get_kv_cache_token_count(const struct llama_context * ctx) {
     return ctx->kv_self.head;
 }
 
-void llama_kv_cache_tokens_rm(struct llama_context * ctx, int32_t c0, int32_t c1) {
-    llama_kv_cache_tokens_rm(ctx->kv_self, c0, c1);
+void llama_kv_cache_clear(struct llama_context * ctx) {
+    llama_kv_cache_clear(ctx->kv_self);
 }
 
 void llama_kv_cache_seq_rm(struct llama_context * ctx, llama_seq_id seq_id, llama_pos p0, llama_pos p1) {
@@ -9565,7 +9655,7 @@ int llama_eval(
                  llama_token * tokens,
                      int32_t   n_tokens,
                          int   n_past) {
-    llama_kv_cache_tokens_rm(ctx->kv_self, n_past, -1);
+    llama_kv_cache_seq_rm(ctx->kv_self, -1, n_past, -1);
 
     const int ret = llama_decode_internal(*ctx, llama_batch_get_one(tokens, n_tokens, n_past, 0));
     if (ret < 0) {
@@ -9580,7 +9670,7 @@ int llama_eval_embd(
                            float * embd,
                          int32_t   n_tokens,
                              int   n_past) {
-    llama_kv_cache_tokens_rm(ctx->kv_self, n_past, -1);
+    llama_kv_cache_seq_rm(ctx->kv_self, -1, n_past, -1);
 
     llama_batch batch = { n_tokens, nullptr, embd, nullptr, nullptr, nullptr, nullptr, n_past, 1, 0, };
 
@@ -9673,43 +9763,44 @@ float * llama_get_embeddings(struct llama_context * ctx) {
     return ctx->embedding.data();
 }
 
-const char * llama_token_get_text(const struct llama_context * ctx, llama_token token) {
-    return ctx->model.vocab.id_to_token[token].text.c_str();
+const char * llama_token_get_text(const struct llama_model * model, llama_token token) {
+    return model->vocab.id_to_token[token].text.c_str();
 }
 
-float llama_token_get_score(const struct llama_context * ctx, llama_token token) {
-    return ctx->model.vocab.id_to_token[token].score;
+float llama_token_get_score(const struct llama_model * model, llama_token token) {
+    return model->vocab.id_to_token[token].score;
 }
 
-llama_token_type llama_token_get_type(const struct llama_context * ctx, llama_token token) {
-    return ctx->model.vocab.id_to_token[token].type;
+llama_token_type llama_token_get_type(const struct llama_model * model, llama_token token) {
+    return model->vocab.id_to_token[token].type;
 }
 
-llama_token llama_token_bos(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_bos_id;
+llama_token llama_token_bos(const struct llama_model * model) {
+    return model->vocab.special_bos_id;
 }
 
-llama_token llama_token_eos(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_eos_id;
+llama_token llama_token_eos(const struct llama_model * model) {
+    return model->vocab.special_eos_id;
 }
 
-llama_token llama_token_nl(const struct llama_context * ctx) {
-    return ctx->model.vocab.linefeed_id;
-}
-llama_token llama_token_prefix(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_prefix_id;
+llama_token llama_token_nl(const struct llama_model * model) {
+    return model->vocab.linefeed_id;
 }
 
-llama_token llama_token_middle(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_middle_id;
+llama_token llama_token_prefix(const struct llama_model * model) {
+    return model->vocab.special_prefix_id;
 }
 
-llama_token llama_token_suffix(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_suffix_id;
+llama_token llama_token_middle(const struct llama_model * model) {
+    return model->vocab.special_middle_id;
 }
 
-llama_token llama_token_eot(const struct llama_context * ctx) {
-    return ctx->model.vocab.special_eot_id;
+llama_token llama_token_suffix(const struct llama_model * model) {
+    return model->vocab.special_suffix_id;
+}
+
+llama_token llama_token_eot(const struct llama_model * model) {
+    return model->vocab.special_eot_id;
 }
 
 int llama_tokenize(

llama.h

@@ -178,7 +178,7 @@ extern "C" {
         float rope_freq_scale; // RoPE frequency scaling factor, 0 = from model
 
         // Keep the booleans together to avoid misalignment during copy-by-value.
-        bool mul_mat_q;  // if true, use experimental mul_mat_q kernels
+        bool mul_mat_q;  // if true, use experimental mul_mat_q kernels (DEPRECATED - always true)
         bool f16_kv;     // use fp16 for KV cache, fp32 otherwise
         bool logits_all; // the llama_eval() call computes all logits, not just the last one
         bool embedding;  // embedding mode only
@@ -191,6 +191,7 @@ extern "C" {
         bool allow_requantize;       // allow quantizing non-f32/f16 tensors
         bool quantize_output_tensor; // quantize output.weight
         bool only_copy;              // only copy tensors - ftype, allow_requantize and quantize_output_tensor are ignored
+        bool pure;                   // disable k-quant mixtures and quantize all tensors to the same type
     } llama_model_quantize_params;
 
     // grammar types
@@ -333,17 +334,14 @@ extern "C" {
     LLAMA_API DEPRECATED(int llama_get_kv_cache_token_count(const struct llama_context * ctx),
             "avoid using this, it will be removed in the future, instead - count the tokens in user code");
 
-    // Remove all tokens data of cells in [c0, c1)
-    // c0 < 0 : [0,  c1]
-    // c1 < 0 : [c0, inf)
-    LLAMA_API void llama_kv_cache_tokens_rm(
-            struct llama_context * ctx,
-                         int32_t   c0,
-                         int32_t   c1);
+    // Clear the KV cache
+    LLAMA_API void llama_kv_cache_clear(
+            struct llama_context * ctx);
 
     // Removes all tokens that belong to the specified sequence and have positions in [p0, p1)
-    // p0 < 0 : [0,  p1]
-    // p1 < 0 : [p0, inf)
+    // seq_id < 0 : match any sequence
+    // p0 < 0     : [0,  p1]
+    // p1 < 0     : [p0, inf)
     LLAMA_API void llama_kv_cache_seq_rm(
             struct llama_context * ctx,
                     llama_seq_id   seq_id,
@@ -494,21 +492,22 @@ extern "C" {
     // Vocab
     //
 
-    LLAMA_API const char * llama_token_get_text(const struct llama_context * ctx, llama_token token);
+    LLAMA_API const char * llama_token_get_text(const struct llama_model * model, llama_token token);
 
-    LLAMA_API float llama_token_get_score(const struct llama_context * ctx, llama_token token);
+    LLAMA_API float llama_token_get_score(const struct llama_model * model, llama_token token);
 
-    LLAMA_API enum llama_token_type llama_token_get_type(const struct llama_context * ctx, llama_token token);
+    LLAMA_API enum llama_token_type llama_token_get_type(const struct llama_model * model, llama_token token);
 
     // Special tokens
-    LLAMA_API llama_token llama_token_bos(const struct llama_context * ctx);  // beginning-of-sentence
-    LLAMA_API llama_token llama_token_eos(const struct llama_context * ctx);  // end-of-sentence
-    LLAMA_API llama_token llama_token_nl (const struct llama_context * ctx);  // next-line
+    LLAMA_API llama_token llama_token_bos(const struct llama_model * model); // beginning-of-sentence
+    LLAMA_API llama_token llama_token_eos(const struct llama_model * model); // end-of-sentence
+    LLAMA_API llama_token llama_token_nl (const struct llama_model * model); // next-line
+
     // codellama infill tokens
-    LLAMA_API llama_token llama_token_prefix(const struct llama_context * ctx); // Beginning of infill prefix
-    LLAMA_API llama_token llama_token_middle(const struct llama_context * ctx); // Beginning of infill middle
-    LLAMA_API llama_token llama_token_suffix(const struct llama_context * ctx); // Beginning of infill suffix
-    LLAMA_API llama_token llama_token_eot   (const struct llama_context * ctx); // End of infill middle
+    LLAMA_API llama_token llama_token_prefix(const struct llama_model * model); // Beginning of infill prefix
+    LLAMA_API llama_token llama_token_middle(const struct llama_model * model); // Beginning of infill middle
+    LLAMA_API llama_token llama_token_suffix(const struct llama_model * model); // Beginning of infill suffix
+    LLAMA_API llama_token llama_token_eot   (const struct llama_model * model); // End of infill middle
 
     //
     // Tokenization
@@ -657,6 +656,7 @@ extern "C" {
                            float * mu);
 
     /// @details Selects the token with the highest probability.
+    ///          Does not compute the token probabilities. Use llama_sample_softmax() instead.
     LLAMA_API llama_token llama_sample_token_greedy(
             struct llama_context * ctx,
           llama_token_data_array * candidates);

scripts/server-llm.sh

@@ -0,0 +1,391 @@
+#!/bin/bash
+#
+# Helper script for deploying llama.cpp server with a single Bash command
+#
+# - Works on Linux and macOS
+# - Supports: CPU, CUDA, Metal, OpenCL
+# - Can run all GGUF models from HuggingFace
+# - Can serve requests in parallel
+# - Always builds latest llama.cpp from GitHub
+#
+# Limitations
+#
+# - Chat templates are poorly supported (base models recommended)
+# - Might be unstable!
+#
+# Usage:
+#   ./server-llm.sh [--port] [--repo] [--wtype] [--backend] [--gpu-id] [--n-parallel] [--n-kv] [--verbose]
+#
+#   --port:       port number, default is 8888
+#   --repo:       path to a repo containing GGUF model files
+#   --wtype:      weights type (f16, q8_0, q4_0, q4_1), default is user-input
+#   --backend:    cpu, cuda, metal, opencl, depends on the OS
+#   --gpu-id:     gpu id, default is 0
+#   --n-parallel: number of parallel requests, default is 8
+#   --n-kv:       KV cache size, default is 4096
+#   --verbose:    verbose output
+#
+# Example:
+#
+#   bash -c "$(curl -s https://ggml.ai/server-llm.sh)"
+#
+
+set -e
+
+# required utils: curl, git, make
+if ! command -v curl &> /dev/null; then
+    printf "[-] curl not found\n"
+    exit 1
+fi
+if ! command -v git &> /dev/null; then
+    printf "[-] git not found\n"
+    exit 1
+fi
+if ! command -v make &> /dev/null; then
+    printf "[-] make not found\n"
+    exit 1
+fi
+
+# parse arguments
+port=8888
+repo=""
+wtype=""
+backend="cpu"
+
+# if macOS, use metal backend by default
+if [[ "$OSTYPE" == "darwin"* ]]; then
+    backend="metal"
+elif command -v nvcc &> /dev/null; then
+    backend="cuda"
+fi
+
+gpu_id=0
+n_parallel=8
+n_kv=4096
+verbose=0
+
+function print_usage {
+    printf "Usage:\n"
+    printf "  ./server-llm.sh [--port] [--repo] [--wtype] [--backend] [--gpu-id] [--n-parallel] [--n-kv] [--verbose]\n\n"
+    printf "  --port:       port number, default is 8888\n"
+    printf "  --repo:       path to a repo containing GGUF model files\n"
+    printf "  --wtype:      weights type (f16, q8_0, q4_0, q4_1), default is user-input\n"
+    printf "  --backend:    cpu, cuda, metal, opencl, depends on the OS\n"
+    printf "  --gpu-id:     gpu id, default is 0\n"
+    printf "  --n-parallel: number of parallel requests, default is 8\n"
+    printf "  --n-kv:       KV cache size, default is 4096\n"
+    printf "  --verbose:    verbose output\n\n"
+    printf "Example:\n\n"
+    printf '  bash -c "$(curl -s https://ggml.ai/server-llm.sh)"\n\n'
+}
+
+while [[ $# -gt 0 ]]; do
+    key="$1"
+    case $key in
+        --port)
+            port="$2"
+            shift
+            shift
+            ;;
+        --repo)
+            repo="$2"
+            shift
+            shift
+            ;;
+        --wtype)
+            wtype="$2"
+            shift
+            shift
+            ;;
+        --backend)
+            backend="$2"
+            shift
+            shift
+            ;;
+        --gpu-id)
+            gpu_id="$2"
+            shift
+            shift
+            ;;
+        --n-parallel)
+            n_parallel="$2"
+            shift
+            shift
+            ;;
+        --n-kv)
+            n_kv="$2"
+            shift
+            shift
+            ;;
+        --verbose)
+            verbose=1
+            shift
+            ;;
+        --help)
+            print_usage
+            exit 0
+            ;;
+        *)
+            echo "Unknown argument: $key"
+            print_usage
+            exit 1
+            ;;
+    esac
+done
+
+# available weights types
+wtypes=("F16" "Q8_0" "Q4_0" "Q4_1" "Q5_0" "Q5_1" "Q6_K" "Q5_K_M" "Q5_K_S" "Q4_K_M" "Q4_K_S" "Q3_K_L" "Q3_K_M" "Q3_K_S" "Q2_K")
+
+wfiles=()
+for wt in "${wtypes[@]}"; do
+    wfiles+=("")
+done
+
+# sample repos
+repos=(
+    "https://huggingface.co/TheBloke/Llama-2-7B-GGUF"
+    "https://huggingface.co/TheBloke/Llama-2-13B-GGUF"
+    "https://huggingface.co/TheBloke/Llama-2-70B-GGUF"
+    "https://huggingface.co/TheBloke/CodeLlama-7B-GGUF"
+    "https://huggingface.co/TheBloke/CodeLlama-13B-GGUF"
+    "https://huggingface.co/TheBloke/CodeLlama-34B-GGUF"
+    "https://huggingface.co/TheBloke/Mistral-7B-v0.1-GGUF"
+    "https://huggingface.co/TheBloke/zephyr-7B-beta-GGUF"
+    "https://huggingface.co/TheBloke/OpenHermes-2-Mistral-7B-GGUF"
+    "https://huggingface.co/TheBloke/CausalLM-7B-GGUF"
+)
+
+printf "\n"
+printf "[I] This is a helper script for deploying llama.cpp's server on this machine.\n\n"
+printf "    Based on the options that follow, the script might download a model file\n"
+printf "    from the internet, which can be a few GBs in size. The script will also\n"
+printf "    build the latest llama.cpp source code from GitHub, which can be unstable.\n"
+printf "\n"
+printf "    Upon success, an HTTP server will be started and it will serve the selected\n"
+printf "    model using llama.cpp for demonstration purposes.\n"
+printf "\n"
+printf "    Please note:\n"
+printf "\n"
+printf "    - All new data will be stored in the current folder\n"
+printf "    - The server will be listening on all network interfaces\n"
+printf "    - The server will run with default settings which are not always optimal\n"
+printf "    - Do not judge the quality of a model based on the results from this script\n"
+printf "    - Do not use this script to benchmark llama.cpp\n"
+printf "    - Do not use this script in production\n"
+printf "    - This script is only for demonstration purposes\n"
+printf "\n"
+printf "    If you don't know what you are doing, please press Ctrl-C to abort now\n"
+printf "\n"
+printf "    Press Enter to continue ...\n\n"
+
+read
+
+if [[ -z "$repo" ]]; then
+    printf "[+] No repo provided from the command line\n"
+    printf "    Please select a number from the list below or enter an URL:\n\n"
+
+    is=0
+    for r in "${repos[@]}"; do
+        printf "    %2d) %s\n" $is "$r"
+        is=$((is+1))
+    done
+
+    # ask for repo until index of sample repo is provided or an URL
+    while [[ -z "$repo" ]]; do
+        printf "\n    Or choose one from: https://huggingface.co/models?sort=trending&search=gguf\n\n"
+        read -p "[+] Select repo: " repo
+
+        # check if the input is a number
+        if [[ "$repo" =~ ^[0-9]+$ ]]; then
+            if [[ "$repo" -ge 0 && "$repo" -lt ${#repos[@]} ]]; then
+                repo="${repos[$repo]}"
+            else
+                printf "[-] Invalid repo index: %s\n" "$repo"
+                repo=""
+            fi
+        elif [[ "$repo" =~ ^https?:// ]]; then
+            repo="$repo"
+        else
+            printf "[-] Invalid repo URL: %s\n" "$repo"
+            repo=""
+        fi
+    done
+fi
+
+# remove suffix
+repo=$(echo "$repo" | sed -E 's/\/tree\/main$//g')
+
+printf "[+] Checking for GGUF model files in %s\n" "$repo"
+
+# find GGUF files in the source
+# TODO: better logic
+model_tree="${repo%/}/tree/main"
+model_files=$(curl -s "$model_tree" | grep -i "\\.gguf</span>" | sed -E 's/.*<span class="truncate group-hover:underline">(.*)<\/span><\/a>/\1/g')
+
+# list all files in the provided git repo
+printf "[+] Model files:\n\n"
+for file in $model_files; do
+    # determine iw by grepping the filename with wtypes
+    iw=-1
+    is=0
+    for wt in "${wtypes[@]}"; do
+        # uppercase
+        ufile=$(echo "$file" | tr '[:lower:]' '[:upper:]')
+        if [[ "$ufile" =~ "$wt" ]]; then
+            iw=$is
+            break
+        fi
+        is=$((is+1))
+    done
+
+    if [[ $iw -eq -1 ]]; then
+        continue
+    fi
+
+    wfiles[$iw]="$file"
+
+    have=" "
+    if [[ -f "$file" ]]; then
+        have="*"
+    fi
+
+    printf "    %2d) %s %s\n" $iw "$have" "$file"
+done
+
+# ask for weights type until provided and available
+while [[ -z "$wtype" ]]; do
+    printf "\n"
+    read -p "[+] Select weight type: " wtype
+    wfile="${wfiles[$wtype]}"
+
+    if [[ -z "$wfile" ]]; then
+        printf "[-] Invalid weight type: %s\n" "$wtype"
+        wtype=""
+    fi
+done
+
+printf "[+] Selected weight type: %s (%s)\n" "$wtype" "$wfile"
+
+url="${repo%/}/resolve/main/$wfile"
+
+# check file if the model has been downloaded before
+chk="$wfile.chk"
+
+# check if we should download the file
+# - if $wfile does not exist
+# - if $wfile exists but $chk does not exist
+# - if $wfile exists and $chk exists but $wfile is newer than $chk
+# TODO: better logic using git lfs info
+
+do_download=0
+
+if [[ ! -f "$wfile" ]]; then
+    do_download=1
+elif [[ ! -f "$chk" ]]; then
+    do_download=1
+elif [[ "$wfile" -nt "$chk" ]]; then
+    do_download=1
+fi
+
+if [[ $do_download -eq 1 ]]; then
+    printf "[+] Downloading weights from %s\n" "$url"
+
+    # download the weights file
+    curl -o "$wfile" -# -L "$url"
+
+    # create a check file if successful
+    if [[ $? -eq 0 ]]; then
+        printf "[+] Creating check file %s\n" "$chk"
+        touch "$chk"
+    fi
+else
+    printf "[+] Using cached weights %s\n" "$wfile"
+fi
+
+# get latest llama.cpp and build
+
+printf "[+] Downloading latest llama.cpp\n"
+
+llama_cpp_dir="__llama_cpp_port_${port}__"
+
+if [[ -d "$llama_cpp_dir" && ! -f "$llama_cpp_dir/__ggml_script__" ]]; then
+    # if the dir exists and there isn't a file "__ggml_script__" in it, abort
+    printf "[-] Directory %s already exists\n" "$llama_cpp_dir"
+    printf "[-] Please remove it and try again\n"
+    exit 1
+elif [[ -d "$llama_cpp_dir" ]]; then
+    printf "[+] Directory %s already exists\n" "$llama_cpp_dir"
+    printf "[+] Using cached llama.cpp\n"
+
+    cd "$llama_cpp_dir"
+    git reset --hard
+    git fetch
+    git checkout origin/master
+
+    cd ..
+else
+    printf "[+] Cloning llama.cpp\n"
+
+    git clone https://github.com/ggerganov/llama.cpp "$llama_cpp_dir"
+fi
+
+# mark that that the directory is made by this script
+touch "$llama_cpp_dir/__ggml_script__"
+
+if [[ $verbose -eq 1 ]]; then
+    set -x
+fi
+
+# build
+cd "$llama_cpp_dir"
+
+make clean
+
+log="--silent"
+if [[ $verbose -eq 1 ]]; then
+    log=""
+fi
+
+if [[ "$backend" == "cuda" ]]; then
+    printf "[+] Building with CUDA backend\n"
+    LLAMA_CUBLAS=1 make -j server $log
+elif [[ "$backend" == "cpu" ]]; then
+    printf "[+] Building with CPU backend\n"
+    make -j server $log
+elif [[ "$backend" == "metal" ]]; then
+    printf "[+] Building with Metal backend\n"
+    make -j server $log
+elif [[ "$backend" == "opencl" ]]; then
+    printf "[+] Building with OpenCL backend\n"
+    LLAMA_CLBLAST=1 make -j server $log
+else
+    printf "[-] Unknown backend: %s\n" "$backend"
+    exit 1
+fi
+
+# run the server
+
+printf "[+] Running server\n"
+
+args=""
+if [[ "$backend" == "cuda" ]]; then
+    export CUDA_VISIBLE_DEVICES=$gpu_id
+    args="-ngl 999"
+elif [[ "$backend" == "cpu" ]]; then
+    args="-ngl 0"
+elif [[ "$backend" == "metal" ]]; then
+    args="-ngl 999"
+elif [[ "$backend" == "opencl" ]]; then
+    args="-ngl 999"
+else
+    printf "[-] Unknown backend: %s\n" "$backend"
+    exit 1
+fi
+
+if [[ $verbose -eq 1 ]]; then
+    args="$args --verbose"
+fi
+
+./server -m "../$wfile" --host 0.0.0.0 --port "$port" -c $n_kv -np "$n_parallel" $args
+
+exit 0

tests/CMakeLists.txt

@@ -28,10 +28,14 @@ llama_build_executable(test-tokenizer-0-falcon.cpp)
 llama_test_executable (test-tokenizer-0-falcon test-tokenizer-0-falcon.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-falcon.gguf)
 llama_build_executable(test-tokenizer-1-llama.cpp)
 llama_test_executable (test-tokenizer-1-llama test-tokenizer-1-llama.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-llama.gguf)
+llama_test_executable(test-tokenizer-1-baichuan test-tokenizer-1-llama.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-baichuan.gguf)
 llama_build_executable(test-tokenizer-1-bpe.cpp)
 llama_test_executable (test-tokenizer-1-falcon test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-falcon.gguf)
 llama_test_executable(test-tokenizer-1-aquila test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-aquila.gguf)
 llama_test_executable(test-tokenizer-1-mpt test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-mpt.gguf)
+llama_test_executable(test-tokenizer-1-gpt-neox test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-gpt-neox.gguf)
+llama_test_executable(test-tokenizer-1-refact test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-refact.gguf)
+llama_test_executable(test-tokenizer-1-starcoder test-tokenizer-1-bpe.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-starcoder.gguf)
 llama_build_and_test_executable(test-grammar-parser.cpp)
 llama_build_and_test_executable(test-llama-grammar.cpp)
 llama_build_and_test_executable(test-grad0.cpp) # SLOW

tests/test-double-float.cpp

@@ -4,7 +4,7 @@
 
 #undef NDEBUG
 #include <cassert>
-#if !defined(__riscv) && !defined(__s390__)
+#if !defined(__riscv) && !defined(__s390__) && !defined(__ARM_NEON)
 #include <immintrin.h>
 #endif
 #include <cmath>

tests/test-quantize-fns.cpp

@@ -129,6 +129,13 @@ int main(int argc, char * argv[]) {
         ggml_type type = (ggml_type) i;
         ggml_type_traits_t qfns = ggml_internal_get_type_traits(type);
 
+        // deprecated - skip
+        if (qfns.blck_size == 0) {
+            continue;
+        }
+
+        printf("Testing %s\n", ggml_type_name((ggml_type) i));
+
         if (qfns.from_float && qfns.to_float) {
             const float total_error = total_quantization_error(qfns, test_size, test_data.data());
             const float max_quantization_error =

tests/test-tokenizer-1-bpe.cpp

@@ -91,9 +91,19 @@ int main(int argc, char **argv) {
             }
         }
     }
-    // TODO: why doesn't this work for the full range of Unicodes?
+    // Restrict to assigned unicode planes
     // for (uint32_t cp = 0x10000; cp < 0x0010ffff; ++cp) {
-    for (uint32_t cp = 0x10000; cp < 0x00080000; ++cp) {
+    for (uint32_t cp = 0x10000; cp < 0x00040000; ++cp) {
+        std::string str = codepoint_to_utf8(cp);
+        std::vector<llama_token> tokens = llama_tokenize(ctx, str, false);
+        std::string check = llama_detokenize_bpe(ctx, tokens);
+        if (str != check) {
+            fprintf(stderr, "%s : error: codepoint %x detokenizes to '%s'(%zu) instead of '%s'(%zu)\n",
+                __func__, cp, check.c_str(), check.length(), str.c_str(), str.length());
+            return 4;
+        }
+    }
+    for (uint32_t cp = 0x000e0000; cp < 0x0010ffff; ++cp) {
         std::string str = codepoint_to_utf8(cp);
         std::vector<llama_token> tokens = llama_tokenize(ctx, str, false);
         std::string check = llama_detokenize_bpe(ctx, tokens);
@@ -103,7 +113,6 @@ int main(int argc, char **argv) {
             return 4;
         }
     }
-
     llama_free_model(model);
     llama_free(ctx);