root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-27 20:04:35 +00:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

Merge branch 'master' into sycl_q3s_q1s

Browse Source
This commit is contained in:
Abhilash Majumder 2024-03-11 08:41:35 +05:30 committed by GitHub
commit 989e15b3c1
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
68 changed files with 49403 additions and 53237 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
.github/workflows/server.yml vendored
View File

@ -47,6 +47,8 @@ jobs:
- name: Clone
id: checkout
uses: actions/checkout@v3
with:
fetch-depth: 0
- name: Dependencies
id: depends
@ -58,7 +60,7 @@ jobs:
cmake \
python3-pip \
wget \
psmisc
language-pack-en
- name: Build
id: cmake_build
@ -89,3 +91,46 @@ jobs:
run: |
cd examples/server/tests
PORT=8888 ./tests.sh --stop --no-skipped --no-capture --tags slow
server-windows:
runs-on: windows-latest
steps:
- name: Clone
id: checkout
uses: actions/checkout@v3
with:
fetch-depth: 0
- name: Build
id: cmake_build
run: |
mkdir build
cd build
cmake .. -DLLAMA_BUILD_SERVER=ON -DCMAKE_BUILD_TYPE=Release ;
cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS} --target server
- name: Python setup
id: setup_python
uses: actions/setup-python@v5
with:
python-version: '3.11'
- name: Tests dependencies
id: test_dependencies
run: |
pip install -r examples/server/tests/requirements.txt
- name: Tests
id: server_integration_tests
run: |
cd examples/server/tests
behave.exe --summary --stop --no-capture --exclude 'issues|wrong_usages|passkey' --tags llama.cpp
- name: Slow tests
id: server_integration_tests_slow
if: ${{ github.event.schedule != '' || github.event.inputs.slow_tests == 'true' }}
run: |
cd examples/server/tests
behave.exe --stop --no-skipped --no-capture --tags slow

1
.gitignore vendored
View File

@ -45,6 +45,7 @@ models-mnt
/embedding
/gguf
/gguf-llama-simple
/gritlm
/imatrix
/infill
/libllama.so

15
CMakeLists.txt
View File

@ -116,6 +116,7 @@ option(LLAMA_MPI "llama: use MPI"
option(LLAMA_QKK_64 "llama: use super-block size of 64 for k-quants" OFF)
option(LLAMA_SYCL "llama: use SYCL" OFF)
option(LLAMA_SYCL_F16 "llama: use 16 bit floats for sycl calculations" OFF)
set(LLAMA_SYCL_TARGET "INTEL" CACHE STRING "llama: sycl target device")
option(LLAMA_CPU_HBM "llama: use memkind for CPU HBM" OFF)
option(LLAMA_BUILD_TESTS "llama: build tests" ${LLAMA_STANDALONE})
@ -199,7 +200,8 @@ if (LLAMA_METAL)
# get full path to the file
#add_compile_definitions(GGML_METAL_DIR_KERNELS="${CMAKE_CURRENT_SOURCE_DIR}/")
# copy ggml-metal.metal to bin directory
# copy ggml-common.h and ggml-metal.metal to bin directory
configure_file(ggml-common.h ${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/ggml-common.h COPYONLY)
configure_file(ggml-metal.metal ${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/ggml-metal.metal COPYONLY)
if (LLAMA_METAL_EMBED_LIBRARY)
@ -533,6 +535,10 @@ if (LLAMA_HIPBLAS)
endif()
if (LLAMA_SYCL)
if (NOT LLAMA_SYCL_TARGET MATCHES "^(INTEL|NVIDIA)$")
message(FATAL_ERROR "Invalid backend chosen, supported options are INTEL or NVIDIA")
endif()
if ( NOT DEFINED ENV{ONEAPI_ROOT})
message(FATAL_ERROR "Not detect ENV {ONEAPI_ROOT}, please install oneAPI & source it, like: source /opt/intel/oneapi/setvars.sh")
endif()
@ -554,6 +560,9 @@ if (LLAMA_SYCL)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-narrowing")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O3")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsycl -L${MKLROOT}/lib")
if (LLAMA_SYCL_TARGET STREQUAL "NVIDIA")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsycl-targets=nvptx64-nvidia-cuda")
endif()
set(GGML_HEADERS_SYCL ggml-sycl.h)
set(GGML_SOURCES_SYCL ggml-sycl.cpp)
@ -561,7 +570,11 @@ if (LLAMA_SYCL)
if (WIN32)
set(LLAMA_EXTRA_LIBS ${LLAMA_EXTRA_LIBS} -fsycl sycl7 OpenCL mkl_sycl_blas_dll.lib mkl_intel_ilp64_dll.lib mkl_sequential_dll.lib mkl_core_dll.lib)
else()
if (LLAMA_SYCL_TARGET STREQUAL "INTEL")
set(LLAMA_EXTRA_LIBS ${LLAMA_EXTRA_LIBS} -fsycl OpenCL mkl_core pthread m dl mkl_sycl_blas mkl_intel_ilp64 mkl_tbb_thread)
elseif (LLAMA_SYCL_TARGET STREQUAL "NVIDIA")
set(LLAMA_EXTRA_LIBS ${LLAMA_EXTRA_LIBS} -fsycl pthread m dl onemkl)
endif()
endif()
endif()

19
Makefile
View File

@ -2,7 +2,7 @@
BUILD_TARGETS = \
main quantize quantize-stats perplexity imatrix embedding vdot q8dot train-text-from-scratch convert-llama2c-to-ggml \
simple batched batched-bench save-load-state server gguf llama-bench libllava.a llava-cli baby-llama beam-search \
speculative infill tokenize benchmark-matmult parallel finetune export-lora lookahead lookup passkey tests/test-c.o
speculative infill tokenize benchmark-matmult parallel finetune export-lora lookahead lookup passkey gritlm tests/test-c.o
# Binaries only useful for tests
TEST_TARGETS = \
@ -201,6 +201,10 @@ ifdef LLAMA_SERVER_VERBOSE
MK_CPPFLAGS += -DSERVER_VERBOSE=$(LLAMA_SERVER_VERBOSE)
endif
ifdef LLAMA_SERVER_SSL
MK_CPPFLAGS += -DCPPHTTPLIB_OPENSSL_SUPPORT
MK_LDFLAGS += -lssl -lcrypto
endif
ifdef LLAMA_CODE_COVERAGE
MK_CXXFLAGS += -fprofile-arcs -ftest-coverage -dumpbase ''
@ -449,7 +453,7 @@ endif # LLAMA_CUDA_PEER_MAX_BATCH_SIZE
ifdef LLAMA_CUDA_CCBIN
MK_NVCCFLAGS += -ccbin $(LLAMA_CUDA_CCBIN)
endif
ggml-cuda.o: ggml-cuda.cu ggml-cuda.h
ggml-cuda.o: ggml-cuda.cu ggml-cuda.h ggml-common.h
ifdef JETSON_EOL_MODULE_DETECT
$(NVCC) -I. -Icommon -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -DNDEBUG -DGGML_USE_CUBLAS -I/usr/local/cuda/include -I/opt/cuda/include -I/usr/local/cuda/targets/aarch64-linux/include -std=c++11 -O3 $(NVCCFLAGS) $(CPPFLAGS) -Xcompiler "$(CUDA_CXXFLAGS)" -c $< -o $@
else
@ -626,7 +630,7 @@ 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 $@
ggml-quants.o: ggml-quants.c ggml.h ggml-quants.h
ggml-quants.o: ggml-quants.c ggml.h ggml-quants.h ggml-common.h
$(CC) $(CFLAGS) -c $< -o $@
OBJS += ggml-alloc.o ggml-backend.o ggml-quants.o
@ -720,14 +724,17 @@ embedding: examples/embedding/embedding.cpp ggml.o llama.o $(C
$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
$(CXX) $(CXXFLAGS) $(filter-out %.h $<,$^) $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS)
gritlm: examples/gritlm/gritlm.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
$(CXX) $(CXXFLAGS) $(filter-out %.h $<,$^) $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS)
save-load-state: examples/save-load-state/save-load-state.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
$(CXX) $(CXXFLAGS) $(filter-out %.h $<,$^) $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS)
server: examples/server/server.cpp examples/server/oai.hpp examples/server/utils.hpp examples/server/httplib.h examples/server/json.hpp examples/server/index.html.hpp examples/server/index.js.hpp examples/server/completion.js.hpp examples/llava/clip.cpp examples/llava/clip.h examples/llava/llava.h examples/llava/llava.cpp common/stb_image.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
server: examples/server/server.cpp examples/server/utils.hpp examples/server/httplib.h examples/server/json.hpp examples/server/index.html.hpp examples/server/index.js.hpp examples/server/completion.js.hpp common/stb_image.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)
$(CXX) $(CXXFLAGS) -c examples/llava/clip.cpp -o $(call GET_OBJ_FILE, examples/llava/clip.cpp) -Wno-cast-qual
$(CXX) $(CXXFLAGS) -Iexamples/server $(filter-out %.h %.hpp $< examples/llava/clip.cpp,$^) $(call GET_OBJ_FILE, $<) $(call GET_OBJ_FILE, examples/llava/clip.cpp) -o $@ $(LDFLAGS) $(LWINSOCK2)
$(CXX) $(CXXFLAGS) $(filter-out %.h %.hpp $<,$^) -Iexamples/server $(call GET_OBJ_FILE, $<) -o $@ $(LDFLAGS) $(LWINSOCK2)
gguf: examples/gguf/gguf.cpp ggml.o $(OBJS)
$(CXX) $(CXXFLAGS) -c $< -o $(call GET_OBJ_FILE, $<)

26
README-sycl.md
View File

@ -73,6 +73,29 @@ For iGPU, please make sure the shared memory from host memory is enough. For lla
For dGPU, please make sure the device memory is enough. For llama-2-7b.Q4_0, recommend the device memory is 4GB+.
## Nvidia GPU
### Verified
|Intel GPU| Status | Verified Model|
|-|-|-|
|Ampere Series| Support| A100|
### oneMKL
The current oneMKL release does not contain the oneMKL cuBlas backend.
As a result for Nvidia GPU's oneMKL must be built from source.
```
git clone https://github.com/oneapi-src/oneMKL
cd oneMKL
mkdir build
cd build
cmake -G Ninja .. -DCMAKE_CXX_COMPILER=icpx -DCMAKE_C_COMPILER=icx -DENABLE_MKLGPU_BACKEND=OFF -DENABLE_MKLCPU_BACKEND=OFF -DENABLE_CUBLAS_BACKEND=ON
ninja
// Add paths as necessary
```
## Docker
Note:
@ -186,6 +209,9 @@ source /opt/intel/oneapi/setvars.sh
# Or, for FP32:
cmake .. -DLLAMA_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx
# For Nvidia GPUs
cmake .. -DLLAMA_SYCL=ON -DLLAMA_SYCL_TARGET=NVIDIA -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx
# Build example/main only
#cmake --build . --config Release --target main

10
README.md
View File

@ -10,16 +10,15 @@ Inference of Meta's [LLaMA](https://arxiv.org/abs/2302.13971) model (and others)
### Recent API changes
- [2024 Mar 8] `llama_kv_cache_seq_rm()` returns a `bool` instead of `void`, and new `llama_n_max_seq()` returns the upper limit of acceptable `seq_id` in batches (relevant when dealing with multiple sequences) https://github.com/ggerganov/llama.cpp/pull/5328
- [2024 Mar 4] Embeddings API updated https://github.com/ggerganov/llama.cpp/pull/5796
- [2024 Mar 3] `struct llama_context_params` https://github.com/ggerganov/llama.cpp/pull/5849
### Hot topics
- The `api_like_OAI.py` script has been removed - use `server` instead ([#5766](https://github.com/ggerganov/llama.cpp/issues/5766#issuecomment-1969037761))
- Support for chat templates: [Wiki (contributions welcome)](https://github.com/ggerganov/llama.cpp/wiki/Templates-supported-by-llama_chat_apply_template)
- Support for Gemma models: https://github.com/ggerganov/llama.cpp/pull/5631
- Non-linear quantization IQ4_NL: https://github.com/ggerganov/llama.cpp/pull/5590
- Looking for contributions to improve and maintain the `server` example: https://github.com/ggerganov/llama.cpp/issues/4216
- Looking for contributions to add Deepseek support: https://github.com/ggerganov/llama.cpp/issues/5981
- Quantization blind testing: https://github.com/ggerganov/llama.cpp/discussions/5962
- Initial Mamba support has been added: https://github.com/ggerganov/llama.cpp/pull/5328
----
@ -110,6 +109,7 @@ Typically finetunes of the base models below are supported as well.
- [x] [InternLM2](https://huggingface.co/models?search=internlm2)
- [x] [CodeShell](https://github.com/WisdomShell/codeshell)
- [x] [Gemma](https://ai.google.dev/gemma)
- [x] [Mamba](https://github.com/state-spaces/mamba)
**Multimodal models:**

3
ci/run.sh
View File

@ -45,7 +45,8 @@ fi
if [ ! -z ${GG_BUILD_SYCL} ]; then
if [ -z ${ONEAPI_ROOT} ]; then
echo "Not detected ONEAPI_ROOT, please install oneAPI base toolkit and enable it by:\n source /opt/intel/oneapi/setvars.sh"
echo "Not detected ONEAPI_ROOT, please install oneAPI base toolkit and enable it by:"
echo "source /opt/intel/oneapi/setvars.sh"
exit 1
fi

16
common/common.cpp
View File

@ -1288,6 +1288,7 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
cparams.n_ctx = params.n_ctx;
cparams.n_batch = params.n_batch;
cparams.n_parallel = params.n_parallel;
cparams.n_threads = params.n_threads;
cparams.n_threads_batch = params.n_threads_batch == -1 ? params.n_threads : params.n_threads_batch;
cparams.seed = params.seed;
@ -1851,3 +1852,18 @@ void dump_kv_cache_view_seqs(const llama_kv_cache_view & view, int row_size) {
printf("\n=== Done dumping\n");
}
void llama_embd_normalize(const float * inp, float * out, int n) {
double sum = 0.0;
for (int i = 0; i < n; i++) {
sum += inp[i] * inp[i];
}
sum = sqrt(sum);
const float norm = sum > 0.0 ? 1.0f / sum : 0.0f;
for (int i = 0; i < n; i++) {
out[i] = inp[i] * norm;
}
}

7
common/common.h
View File

@ -260,3 +260,10 @@ void dump_kv_cache_view(const llama_kv_cache_view & view, int row_size = 80);
// Dump the KV cache view showing individual sequences in each cell (long output).
void dump_kv_cache_view_seqs(const llama_kv_cache_view & view, int row_size = 40);
//
// Embedding utils
//
void llama_embd_normalize(const float * inp, float * out, int n);

16
common/grammar-parser.cpp
View File

@ -278,6 +278,22 @@ namespace grammar_parser {
while (*pos) {
pos = parse_rule(state, pos);
}
// Validate the state to ensure that all rules are defined
for (const auto & rule : state.rules) {
for (const auto & elem : rule) {
if (elem.type == LLAMA_GRETYPE_RULE_REF) {
// Ensure that the rule at that location exists
if (elem.value >= state.rules.size() || state.rules[elem.value].empty()) {
// Get the name of the rule that is missing
for (const auto & kv : state.symbol_ids) {
if (kv.second == elem.value) {
throw std::runtime_error("Undefined rule identifier '" + kv.first + "'");
}
}
}
}
}
}
return state;
} catch (const std::exception & err) {
fprintf(stderr, "%s: error parsing grammar: %s\n", __func__, err.what());

8
common/log.h
View File

@ -297,7 +297,7 @@ inline std::string log_filename_generator_impl(LogTriState multilog, const std::
#ifndef _MSC_VER
#define LOG(...) LOG_IMPL(__VA_ARGS__, "")
#else
#define LOG(str, ...) LOG_IMPL("%s" str, "", __VA_ARGS__, "")
#define LOG(str, ...) LOG_IMPL("%s" str, "", ##__VA_ARGS__, "")
#endif
// Main TEE macro.
@ -311,7 +311,7 @@ inline std::string log_filename_generator_impl(LogTriState multilog, const std::
#ifndef _MSC_VER
#define LOG_TEE(...) LOG_TEE_IMPL(__VA_ARGS__, "")
#else
#define LOG_TEE(str, ...) LOG_TEE_IMPL("%s" str, "", __VA_ARGS__, "")
#define LOG_TEE(str, ...) LOG_TEE_IMPL("%s" str, "", ##__VA_ARGS__, "")
#endif
// LOG macro variants with auto endline.
@ -319,8 +319,8 @@ inline std::string log_filename_generator_impl(LogTriState multilog, const std::
#define LOGLN(...) LOG_IMPL(__VA_ARGS__, "\n")
#define LOG_TEELN(...) LOG_TEE_IMPL(__VA_ARGS__, "\n")
#else
#define LOGLN(str, ...) LOG_IMPL("%s" str, "", __VA_ARGS__, "\n")
#define LOG_TEELN(str, ...) LOG_TEE_IMPL("%s" str, "", __VA_ARGS__, "\n")
#define LOGLN(str, ...) LOG_IMPL("%s" str, "", ##__VA_ARGS__, "\n")
#define LOG_TEELN(str, ...) LOG_TEE_IMPL("%s" str, "", ##__VA_ARGS__, "\n")
#endif
// INTERNAL, DO NOT USE

118
convert-hf-to-gguf.py
View File

@ -1847,6 +1847,124 @@ class StarCoder2Model(Model):
model_arch = gguf.MODEL_ARCH.STARCODER2
@Model.register("MambaForCausalLM", "MambaLMHeadModel")
class MambaModel(Model):
model_arch = gguf.MODEL_ARCH.MAMBA
def set_vocab(self):
vocab_size = self.hparams["vocab_size"]
# Round vocab size to next multiple of 8
pad_vocab = self.hparams.get("pad_vocab_size_multiple", 8)
# pad using ceiling division
# ref: https://stackoverflow.com/a/17511341/22827863
vocab_size = -(vocab_size // -pad_vocab) * pad_vocab
self.hparams["vocab_size"] = vocab_size
if (self.dir_model / "tokenizer.json").is_file():
self._set_vocab_gpt2()
else:
# Use the GPT-NeoX tokenizer when no tokenizer files are present
tokenizer_path = Path(sys.path[0]) / "models" / "ggml-vocab-gpt-neox.gguf"
print(f"Using tokenizer from '{os.path.relpath(tokenizer_path, os.getcwd())}'")
neox_reader = gguf.GGUFReader(tokenizer_path, "r")
field = neox_reader.get_field(gguf.Keys.Tokenizer.MODEL)
self.gguf_writer.add_tokenizer_model(bytes(field.parts[-1]))
field = neox_reader.get_field(gguf.Keys.Tokenizer.LIST)
self.gguf_writer.add_token_list([bytes(field.parts[i]) for i in field.data][:vocab_size])
field = neox_reader.get_field(gguf.Keys.Tokenizer.TOKEN_TYPE)
self.gguf_writer.add_token_types([field.parts[i].tolist()[0] for i in field.data][:vocab_size])
field = neox_reader.get_field(gguf.Keys.Tokenizer.MERGES)
self.gguf_writer.add_token_merges([bytes(field.parts[i]) for i in field.data])
field = neox_reader.get_field(gguf.Keys.Tokenizer.BOS_ID)
self.gguf_writer.add_bos_token_id(field.parts[-1].tolist()[0])
field = neox_reader.get_field(gguf.Keys.Tokenizer.EOS_ID)
self.gguf_writer.add_eos_token_id(field.parts[-1].tolist()[0])
field = neox_reader.get_field(gguf.Keys.Tokenizer.UNK_ID)
self.gguf_writer.add_unk_token_id(field.parts[-1].tolist()[0])
def set_gguf_parameters(self):
d_model = self.find_hparam(["hidden_size", "d_model"])
d_conv = self.find_hparam(["conv_kernel", "d_conv"], optional=True) or 4
d_inner = self.find_hparam(["intermediate_size", "d_inner"], optional=True) or 2 * d_model
d_state = self.find_hparam(["state_size", "d_state"], optional=True) or 16
# ceiling division
# ref: https://stackoverflow.com/a/17511341/22827863
# ref: https://github.com/state-spaces/mamba/blob/ce59daea3a090d011d6476c6e5b97f6d58ddad8b/mamba_ssm/modules/mamba_simple.py#L58
dt_rank = self.find_hparam(["time_step_rank", "dt_rank"], optional=True) or -(d_model // -16)
rms_norm_eps = self.find_hparam(["layer_norm_epsilon", "rms_norm_eps"], optional=True) or 1e-5
# Fail early for models which don't have a block expansion factor of 2
assert d_inner == 2 * d_model
self.gguf_writer.add_name(self.dir_model.name)
self.gguf_writer.add_context_length(2**20) # arbitrary value; for those who use the default
self.gguf_writer.add_embedding_length(d_model)
self.gguf_writer.add_feed_forward_length(0) # unused, but seemingly required when loading
self.gguf_writer.add_head_count(0) # unused, but seemingly required when loading
self.gguf_writer.add_block_count(self.hparams["n_layer"])
self.gguf_writer.add_ssm_conv_kernel(d_conv)
self.gguf_writer.add_ssm_inner_size(d_inner)
self.gguf_writer.add_ssm_state_size(d_state)
self.gguf_writer.add_ssm_time_step_rank(dt_rank)
self.gguf_writer.add_layer_norm_rms_eps(rms_norm_eps)
self.gguf_writer.add_file_type(self.ftype)
def write_tensors(self):
block_count = self.hparams["n_layer"]
tensor_map = gguf.get_tensor_name_map(self.model_arch, block_count)
tok_embd = None
tok_embd_name = gguf.TENSOR_NAMES[gguf.MODEL_TENSOR.TOKEN_EMBD] + ".weight"
output_name = gguf.TENSOR_NAMES[gguf.MODEL_TENSOR.OUTPUT] + ".weight"
for name, data_torch in self.get_tensors():
old_dtype = data_torch.dtype
# convert any unsupported data types to float32
if data_torch.dtype not in (torch.float16, torch.float32):
data_torch = data_torch.to(torch.float32)
# map tensor names
new_name = tensor_map.get_name(name, try_suffixes=(".weight", ".bias"))
if new_name is None:
print(f"Can not map tensor {name!r}")
sys.exit()
if name.endswith(".A_log"):
print("A_log --> A ==> " + new_name)
data_torch = -torch.exp(data_torch)
# assuming token_embd.weight is seen before output.weight
if tok_embd is not None and new_name == output_name:
if torch.equal(tok_embd, data_torch):
print(f"{output_name} is equivalent to {tok_embd_name}, omitting")
continue
if new_name == tok_embd_name:
tok_embd = data_torch
data = data_torch.squeeze().numpy()
n_dims = len(data.shape)
data_dtype = data.dtype
# if f32 desired, convert any float16 to float32
if self.ftype == 0 and data_dtype == np.float16:
data = data.astype(np.float32)
# TODO: Why cant we use these float16 as-is? There should be not reason to store float16 as float32
if self.ftype == 1 and data_dtype == np.float16 and n_dims == 1:
data = data.astype(np.float32)
# if f16 desired, convert big float32 2-dim weight tensors to float16
if self.ftype == 1 and data_dtype == np.float32 and new_name.removesuffix(".weight").endswith((".ssm_in", ".ssm_out", "token_embd", "output")) and n_dims == 2:
data = data.astype(np.float16)
print(f"{new_name}, n_dims = {n_dims}, {old_dtype} --> {data.dtype}")
self.gguf_writer.add_tensor(new_name, data)
###### CONVERSION LOGIC ######

13
convert.py
View File

@ -1377,7 +1377,6 @@ def main(args_in: list[str] | None = None) -> None:
# We currently only support Q8_0 output on little endian systems.
output_choices.append("q8_0")
parser = argparse.ArgumentParser(description="Convert a LLaMA model to a GGML compatible file")
parser.add_argument("--awq-path", type=Path, help="Path to scale awq cache file", default=None)
parser.add_argument("--dump", action="store_true", help="don't convert, just show what's in the model")
parser.add_argument("--dump-single", action="store_true", help="don't convert, just show what's in a single model file")
parser.add_argument("--vocab-only", action="store_true", help="extract only the vocab")
@ -1393,18 +1392,6 @@ def main(args_in: list[str] | None = None) -> None:
parser.add_argument("--skip-unknown", action="store_true", help="skip unknown tensor names instead of failing")
args = parser.parse_args(args_in)
if args.awq_path:
sys.path.insert(1, str(Path(__file__).parent / 'awq-py'))
from awq.apply_awq import add_scale_weights # type: ignore[import-not-found]
tmp_model_path = args.model / "weighted_model"
if tmp_model_path.is_dir():
print(f"{tmp_model_path} exists as a weighted model.")
else:
tmp_model_path.mkdir(parents=True, exist_ok=True)
print("Saving new weighted model ...")
add_scale_weights(str(args.model), str(args.awq_path), str(tmp_model_path))
print(f"Saved weighted model at {tmp_model_path}.")
args.model = tmp_model_path
if args.dump_single:
model_plus = lazy_load_file(args.model)

1
examples/CMakeLists.txt
View File

@ -20,6 +20,7 @@ else()
add_subdirectory(convert-llama2c-to-ggml)
add_subdirectory(embedding)
add_subdirectory(finetune)
add_subdirectory(gritlm)
add_subdirectory(infill)
add_subdirectory(llama-bench)
add_subdirectory(llava)

13
examples/batched-bench/batched-bench.cpp
View File

@ -105,6 +105,9 @@ int main(int argc, char ** argv) {
ctx_params.n_threads = params.n_threads;
ctx_params.n_threads_batch = params.n_threads_batch == -1 ? params.n_threads : params.n_threads_batch;
// ensure enough sequences are available
ctx_params.n_parallel = *std::max_element(n_pl.begin(), n_pl.end());
llama_context * ctx = llama_new_context_with_model(model, ctx_params);
if (ctx == NULL) {
@ -174,10 +177,10 @@ int main(int argc, char ** argv) {
llama_batch_clear(batch);
const int n_tokens = is_pp_shared ? pp : pl*pp;
for (int i = 0; i < n_tokens; ++i) {
llama_batch_add(batch, 0, i, { 0 }, false);
for (int i = 0; i < pp; ++i) {
for (int j = 0; j < (is_pp_shared ? 1 : pl); ++j) {
llama_batch_add(batch, 0, i, { j }, false);
}
}
batch.logits[batch.n_tokens - 1] = true;
@ -192,7 +195,7 @@ int main(int argc, char ** argv) {
if (is_pp_shared) {
for (int32_t i = 1; i < pl; ++i) {
llama_kv_cache_seq_cp(ctx, 0, i, 0, pp);
llama_kv_cache_seq_cp(ctx, 0, i, -1, -1);
}
}

3
examples/batched/batched.cpp
View File

@ -80,6 +80,7 @@ int main(int argc, char ** argv) {
ctx_params.seed = 1234;
ctx_params.n_ctx = n_kv_req;
ctx_params.n_batch = std::max(n_len, n_parallel);
ctx_params.n_parallel = n_parallel;
ctx_params.n_threads = params.n_threads;
ctx_params.n_threads_batch = params.n_threads_batch == -1 ? params.n_threads : params.n_threads_batch;
@ -132,7 +133,7 @@ int main(int argc, char ** argv) {
// assign the system KV cache to all parallel sequences
// this way, the parallel sequences will "reuse" the prompt tokens without having to copy them
for (int32_t i = 1; i < n_parallel; ++i) {
llama_kv_cache_seq_cp(ctx, 0, i, 0, batch.n_tokens);
llama_kv_cache_seq_cp(ctx, 0, i, -1, -1);
}
if (n_parallel > 1) {

6
examples/benchmark/benchmark-matmult.cpp
View File

@ -189,12 +189,10 @@ int main(int argc, char ** argv) {
int32_t nelements = sizex*sizey;
std::vector<int64_t> hist_cur(1 << 4, 0);
// Set up a the benchmark matrices
// printf("Creating new tensor q11 & Running quantize\n");
struct ggml_tensor * q11 = ggml_new_tensor_2d(ctx, qtype, sizex, sizey);
ggml_quantize_chunk(qtype, (const float *) m11->data, q11->data, 0, nelements/m11->ne[0], m11->ne[0], hist_cur.data(), nullptr);
ggml_quantize_chunk(qtype, (const float *) m11->data, q11->data, 0, nelements/m11->ne[0], m11->ne[0], nullptr);
// Set up a the compute graph
// printf("Creating new tensor q31\n");
@ -207,7 +205,7 @@ int main(int argc, char ** argv) {
// Set up a second graph computation to make sure we override the CPU cache lines
// printf("Creating new tensor q12 & Running quantize\n");
struct ggml_tensor * q12 = ggml_new_tensor_2d(ctx, qtype, sizex, sizey);
ggml_quantize_chunk(qtype, (const float *) m12->data, q12->data, 0, nelements/m12->ne[0], m12->ne[0], hist_cur.data(), nullptr);
ggml_quantize_chunk(qtype, (const float *) m12->data, q12->data, 0, nelements/m12->ne[0], m12->ne[0], nullptr);
// printf("Creating new tensor q32\n");
struct ggml_tensor * q32 = ggml_mul_mat(ctx, q12, m2);

14
examples/embedding/embedding.cpp
View File

@ -23,17 +23,6 @@ static void batch_add_seq(llama_batch & batch, const std::vector<int32_t> & toke
}
}
static void normalize(const float * vec, float * out, int n) {
float norm = 0;
for (int i = 0; i < n; i++) {
norm += vec[i] * vec[i];
}
norm = sqrt(norm);
for (int i = 0; i < n; i++) {
out[i] = vec[i] / norm;
}
}
static void batch_decode(llama_context * ctx, llama_batch & batch, float * output, int n_seq, int n_embd) {
// clear previous kv_cache values (irrelevant for embeddings)
llama_kv_cache_clear(ctx);
@ -44,7 +33,6 @@ static void batch_decode(llama_context * ctx, llama_batch & batch, float * outpu
fprintf(stderr, "%s : failed to decode\n", __func__);
}
// normalize on copy
for (int i = 0; i < batch.n_tokens; i++) {
if (!batch.logits[i]) {
continue;
@ -61,7 +49,7 @@ static void batch_decode(llama_context * ctx, llama_batch & batch, float * outpu
}
float * out = output + batch.seq_id[i][0] * n_embd;
normalize(embd, out, n_embd);
llama_embd_normalize(embd, out, n_embd);
}
}

5
examples/gritlm/CMakeLists.txt Normal file
View File

@ -0,0 +1,5 @@
set(TARGET gritlm)
add_executable(${TARGET} gritlm.cpp)
install(TARGETS ${TARGET} RUNTIME)
target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
target_compile_features(${TARGET} PRIVATE cxx_std_11)

229
examples/gritlm/gritlm.cpp Normal file
View File

@ -0,0 +1,229 @@
#include "common.h"
#include "llama.h"
#include <string>
#include <vector>
// #define GRIT_DEBUG
static float dot_product(const std::vector<float> & v1, const std::vector<float> & v2) {
float dot = 0.0f;
for (uint64_t i = 0; i < v1.size(); ++i) {
dot += v1[i] * v2[i];
}
return dot;
}
static float norm(const std::vector<float> & v) {
return std::sqrt(dot_product(v, v));
}
static float cosine_similarity(const std::vector<float> & v1, const std::vector<float> & v2) {
return dot_product(v1, v2) / (norm(v1) * norm(v2));
}
static std::vector<std::vector<float>> encode(llama_context * ctx, const std::vector<std::string> & sentences, const std::string & instruction) {
std::vector<std::vector<float>> result;
const llama_model * mdl = llama_get_model(ctx);
llama_batch batch = llama_batch_init(llama_n_batch(ctx), 0, 1);
for (uint64_t i = 0; i < sentences.size(); i++) {
llama_batch_clear(batch);
const std::string input_string = instruction + sentences[i];
std::vector<llama_token> inputs = llama_tokenize(mdl, input_string, true, false);
const int32_t n_toks = inputs.size();
// GritLM seems to have EOS = ""
// https://github.com/ContextualAI/gritlm/blob/92025b16534712b31b3c4aaaf069350e222bd5f8/gritlm/gritlm.py#L18
// inputs.push_back(llama_token_eos(mdl));
// we want to ignore instruction tokens for mean pooling
const int32_t n_inst = llama_tokenize(mdl, instruction, true, false).size();
#ifdef GRIT_DEBUG
// debug tokens - should be matching as referenced in the GritLM sample
std::for_each(inputs.begin(), inputs.end(), [&ctx](llama_token t) {
std::printf("[%u:%s]", t, llama_token_to_piece(ctx, t).c_str());
});
std::printf("\n");
#endif
// add input to batch (this increments n_tokens)
for (int32_t j = 0; j < n_toks; j++) {
llama_batch_add(batch, inputs[j], j, { 0 }, j >= n_inst);
}
// clear previous kv_cache values (irrelevant for embeddings)
llama_kv_cache_clear(ctx);
llama_set_causal_attn(ctx, false);
// run model
llama_decode(ctx, batch);
// get embedding dimensions
uint64_t n_embd = llama_n_embd(mdl);
// allocate embedding output
std::vector<float> emb_unorm(n_embd, 0.0f);
// sum up all token embeddings
for (int32_t k = n_inst; k < n_toks; k++) {
float * emb = llama_get_embeddings_ith(ctx, k);
for (uint64_t j = 0; j < n_embd; j++) {
emb_unorm[j] += emb[j];
}
}
// divide by number of tokens (mean pooling)
{
const uint64_t n_sent = n_toks - n_inst;
for (uint64_t j = 0; j < n_embd; j++) {
emb_unorm[j] /= n_sent;
}
}
std::vector<float> emb_norm(emb_unorm.size());
llama_embd_normalize(emb_unorm.data(), emb_norm.data(), n_embd);
result.push_back(emb_norm);
#ifdef GRIT_DEBUG
// print out emb_norm
std::printf("embedding %ld: ", i);
for (uint64_t j = 0; j < n_embd; j++) {
std::printf("%.5f ", emb_norm[j]);
}
std::printf("\n\n");
#endif
}
llama_batch_free(batch);
return result;
}
static std::string generate(llama_context * ctx, const std::string & prompt, bool stream) {
std::string result;
const llama_model * mdl = llama_get_model(ctx);
llama_token eos_token = llama_token_eos(mdl);
llama_kv_cache_clear(ctx);
llama_set_causal_attn(ctx, true);
llama_batch bat = llama_batch_init(llama_n_batch(ctx), 0, 1);
std::vector<llama_token> inputs = llama_tokenize(mdl, prompt, false, true);
int32_t i_current_token = 0;
while (true) {
llama_batch_clear(bat);
auto n_inputs = (int32_t)inputs.size();
for (int32_t i = 0; i < n_inputs; i++) {
llama_batch_add(bat, inputs[i], i_current_token++, { 0 }, i == n_inputs - 1);
}
inputs.clear();
llama_decode(ctx, bat);
auto logits = llama_get_logits_ith(ctx, bat.n_tokens - 1);
auto candidates = std::vector<llama_token_data>(llama_n_vocab(mdl));
auto n_candidates = (int32_t)candidates.size();
for (int32_t token = 0; token < n_candidates; token++) {
candidates[token] = llama_token_data{ token, logits[token], 0.0f };
}
auto candidates_p = llama_token_data_array{ candidates.data(), candidates.size(), false };
llama_token token = llama_sample_token_greedy(ctx, &candidates_p);
if (token == eos_token) {
break;
}
std::string piece = llama_token_to_piece(ctx, token);
if (stream) {
std::printf("%s", piece.c_str());
std::fflush(stdout);
}
inputs.push_back(token);
result += piece;
}
if (stream) {
std::printf("\n");
}
llama_batch_free(bat);
return result;
}
static std::string gritlm_instruction(const std::string & instruction) {
return !instruction.empty() ? "<|user|>\n" + instruction + "\n<|embed|>\n" : "<|embed|>\n";
}
int main(int argc, char * argv[]) {
gpt_params params;
if (!gpt_params_parse(argc, argv, params)) {
return 1;
}
llama_model_params mparams = llama_model_params_from_gpt_params(params);
llama_context_params cparams = llama_context_params_from_gpt_params(params);
llama_backend_init();
llama_model * mdl = llama_load_model_from_file(params.model.c_str(), mparams);
// create new context - set to embedding mode
cparams.embeddings = true;
llama_context * ctx = llama_new_context_with_model(mdl, cparams);
// ### Embedding/Representation ###
// samples taken from: https://github.com/ContextualAI/gritlm#basic
{
const std::string instruction = "Given a scientific paper title, retrieve the paper's abstract";
const std::vector<std::string> queries = {
"Bitcoin: A Peer-to-Peer Electronic Cash System",
"Generative Representational Instruction Tuning",
};
const std::vector<std::string> documents = {
"A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution. Digital signatures provide part of the solution, but the main benefits are lost if a trusted third party is still required to prevent double-spending. We propose a solution to the double-spending problem using a peer-to-peer network. The network timestamps transactions by hashing them into an ongoing chain of hash-based proof-of-work, forming a record that cannot be changed without redoing the proof-of-work. The longest chain not only serves as proof of the sequence of events witnessed, but proof that it came from the largest pool of CPU power. As long as a majority of CPU power is controlled by nodes that are not cooperating to attack the network, they'll generate the longest chain and outpace attackers. The network itself requires minimal structure. Messages are broadcast on a best effort basis, and nodes can leave and rejoin the network at will, accepting the longest proof-of-work chain as proof of what happened while they were gone.",
"All text-based language problems can be reduced to either generation or embedding. Current models only perform well at one or the other. We introduce generative representational instruction tuning (GRIT) whereby a large language model is trained to handle both generative and embedding tasks by distinguishing between them through instructions. Compared to other open models, our resulting GritLM 7B sets a new state of the art on the Massive Text Embedding Benchmark (MTEB) and outperforms all models up to its size on a range of generative tasks. By scaling up further, GritLM 8X7B outperforms all open generative language models that we tried while still being among the best embedding models. Notably, we find that GRIT matches training on only generative or embedding data, thus we can unify both at no performance loss. Among other benefits, the unification via GRIT speeds up Retrieval-Augmented Generation (RAG) by > 60% for long documents, by no longer requiring separate retrieval and generation models. Models, code, etc. are freely available at https://github.com/ContextualAI/gritlm.",
};
// No need to add instruction for retrieval documents
const std::vector<std::vector<float>> d_rep = encode(ctx, documents, gritlm_instruction(""));
const std::vector<std::vector<float>> q_rep = encode(ctx, queries, gritlm_instruction(instruction));
const float cosine_sim_q0_d0 = cosine_similarity(q_rep[0], d_rep[0]);
const float cosine_sim_q0_d1 = cosine_similarity(q_rep[0], d_rep[1]);
const float cosine_sim_q1_d0 = cosine_similarity(q_rep[1], d_rep[0]);
const float cosine_sim_q1_d1 = cosine_similarity(q_rep[1], d_rep[1]);
std::printf("Cosine similarity between \"%.50s\" and \"%.50s\" is: %.3f\n", queries[0].c_str(), documents[0].c_str(), cosine_sim_q0_d0);
std::printf("Cosine similarity between \"%.50s\" and \"%.50s\" is: %.3f\n", queries[0].c_str(), documents[1].c_str(), cosine_sim_q0_d1);
std::printf("Cosine similarity between \"%.50s\" and \"%.50s\" is: %.3f\n", queries[1].c_str(), documents[0].c_str(), cosine_sim_q1_d0);
std::printf("Cosine similarity between \"%.50s\" and \"%.50s\" is: %.3f\n", queries[1].c_str(), documents[1].c_str(), cosine_sim_q1_d1);
}
// ### Generation ###
// GritLM models are not finetuned with system prompts, as you can just include system-like instructions together with your user instruction
{
const std::string prompt = "<|user|>\nPlease write me a poem about my recent hike of Mt. Fuji at midnight in the style of Shakespeare.\n<|assistant|>\n";
std::string response = generate(ctx, prompt, true);
}
llama_free(ctx);
llama_free_model(mdl);
llama_backend_free();
return 0;
}

30
examples/llama-bench/llama-bench.cpp
View File

@ -173,6 +173,7 @@ struct cmd_params {
std::vector<bool> no_kv_offload;
std::vector<std::vector<float>> tensor_split;
std::vector<bool> use_mmap;
std::vector<bool> embeddings;
int reps;
bool verbose;
output_formats output_format;
@ -192,6 +193,7 @@ static const cmd_params cmd_params_defaults = {
/* no_kv_offload */ {false},
/* tensor_split */ {std::vector<float>(llama_max_devices(), 0.0f)},
/* use_mmap */ {true},
/* embeddings */ {false},
/* reps */ 5,
/* verbose */ false,
/* output_format */ MARKDOWN
@ -214,6 +216,7 @@ static void print_usage(int /* argc */, char ** argv) {
printf(" -mg, --main-gpu <i> (default: %s)\n", join(cmd_params_defaults.main_gpu, ",").c_str());
printf(" -nkvo, --no-kv-offload <0|1> (default: %s)\n", join(cmd_params_defaults.no_kv_offload, ",").c_str());
printf(" -mmp, --mmap <0|1> (default: %s)\n", join(cmd_params_defaults.use_mmap, ",").c_str());
printf(" -embd, --embeddings <0|1> (default: %s)\n", join(cmd_params_defaults.embeddings, ",").c_str());
printf(" -ts, --tensor_split <ts0/ts1/..> (default: 0)\n");
printf(" -r, --repetitions <n> (default: %d)\n", cmd_params_defaults.reps);
printf(" -o, --output <csv|json|md|sql> (default: %s)\n", output_format_str(cmd_params_defaults.output_format));
@ -382,6 +385,13 @@ static cmd_params parse_cmd_params(int argc, char ** argv) {
}
auto p = split<bool>(argv[i], split_delim);
params.use_mmap.insert(params.use_mmap.end(), p.begin(), p.end());
} else if (arg == "-embd" || arg == "--embeddings") {
if (++i >= argc) {
invalid_param = true;
break;
}
auto p = split<bool>(argv[i], split_delim);
params.embeddings.insert(params.embeddings.end(), p.begin(), p.end());
} else if (arg == "-ts" || arg == "--tensor-split") {
if (++i >= argc) {
invalid_param = true;
@ -453,6 +463,7 @@ static cmd_params parse_cmd_params(int argc, char ** argv) {
if (params.no_kv_offload.empty()){ params.no_kv_offload = cmd_params_defaults.no_kv_offload; }
if (params.tensor_split.empty()) { params.tensor_split = cmd_params_defaults.tensor_split; }
if (params.use_mmap.empty()) { params.use_mmap = cmd_params_defaults.use_mmap; }
if (params.embeddings.empty()) { params.embeddings = cmd_params_defaults.embeddings; }
if (params.n_threads.empty()) { params.n_threads = cmd_params_defaults.n_threads; }
return params;
@ -472,6 +483,7 @@ struct cmd_params_instance {
bool no_kv_offload;
std::vector<float> tensor_split;
bool use_mmap;
bool embeddings;
llama_model_params to_llama_mparams() const {
llama_model_params mparams = llama_model_default_params();
@ -502,6 +514,7 @@ struct cmd_params_instance {
cparams.type_k = type_k;
cparams.type_v = type_v;
cparams.offload_kqv = !no_kv_offload;
cparams.embeddings = embeddings;
return cparams;
}
@ -517,6 +530,7 @@ static std::vector<cmd_params_instance> get_cmd_params_instances(const cmd_param
for (const auto & mg : params.main_gpu)
for (const auto & ts : params.tensor_split)
for (const auto & mmp : params.use_mmap)
for (const auto & embd : params.embeddings)
for (const auto & nb : params.n_batch)
for (const auto & tk : params.type_k)
for (const auto & tv : params.type_v)
@ -540,6 +554,7 @@ static std::vector<cmd_params_instance> get_cmd_params_instances(const cmd_param
/* .no_kv_offload= */ nkvo,
/* .tensor_split = */ ts,
/* .use_mmap = */ mmp,
/* .embeddings = */ embd,
};
instances.push_back(instance);
}
@ -562,6 +577,7 @@ static std::vector<cmd_params_instance> get_cmd_params_instances(const cmd_param
/* .no_kv_offload= */ nkvo,
/* .tensor_split = */ ts,
/* .use_mmap = */ mmp,
/* .embeddings = */ embd,
};
instances.push_back(instance);
}
@ -597,6 +613,7 @@ struct test {
bool no_kv_offload;
std::vector<float> tensor_split;
bool use_mmap;
bool embeddings;
int n_prompt;
int n_gen;
std::string test_time;
@ -619,6 +636,7 @@ struct test {
no_kv_offload = inst.no_kv_offload;
tensor_split = inst.tensor_split;
use_mmap = inst.use_mmap;
embeddings = inst.embeddings;
n_prompt = inst.n_prompt;
n_gen = inst.n_gen;
// RFC 3339 date-time format
@ -690,7 +708,7 @@ struct test {
"n_batch", "n_threads", "type_k", "type_v",
"n_gpu_layers", "split_mode",
"main_gpu", "no_kv_offload",
"tensor_split", "use_mmap",
"tensor_split", "use_mmap", "embeddings",
"n_prompt", "n_gen", "test_time",
"avg_ns", "stddev_ns",
"avg_ts", "stddev_ts"
@ -710,7 +728,7 @@ struct test {
}
if (field == "cuda" || field == "opencl" || field == "vulkan" || field == "kompute" || field == "metal" ||
field == "gpu_blas" || field == "blas" || field == "sycl" ||field == "f16_kv" || field == "no_kv_offload" ||
field == "use_mmap") {
field == "use_mmap" || field == "embeddings") {
return BOOL;
}
if (field == "avg_ts" || field == "stddev_ts") {
@ -744,7 +762,7 @@ struct test {
std::to_string(n_batch), std::to_string(n_threads), ggml_type_name(type_k), ggml_type_name(type_v),
std::to_string(n_gpu_layers), split_mode_str(split_mode),
std::to_string(main_gpu), std::to_string(no_kv_offload),
tensor_split_str, std::to_string(use_mmap),
tensor_split_str, std::to_string(use_mmap), std::to_string(embeddings),
std::to_string(n_prompt), std::to_string(n_gen), test_time,
std::to_string(avg_ns()), std::to_string(stdev_ns()),
std::to_string(avg_ts()), std::to_string(stdev_ts())
@ -914,6 +932,9 @@ struct markdown_printer : public printer {
if (field == "use_mmap") {
return "mmap";
}
if (field == "embeddings") {
return "embd";
}
if (field == "tensor_split") {
return "ts";
}
@ -957,6 +978,9 @@ struct markdown_printer : public printer {
if (params.use_mmap.size() > 1 || params.use_mmap != cmd_params_defaults.use_mmap) {
fields.emplace_back("use_mmap");
}
if (params.embeddings.size() > 1 || params.embeddings != cmd_params_defaults.embeddings) {
fields.emplace_back("embeddings");
}
fields.emplace_back("test");
fields.emplace_back("t/s");

53
examples/llama.android/app/src/main/cpp/llama-android.cpp
View File

@ -33,6 +33,45 @@ jclass la_int_var;
jmethodID la_int_var_value;
jmethodID la_int_var_inc;
std::string cached_token_chars;
bool is_valid_utf8(const char * string) {
if (!string) {
return true;
}
const unsigned char * bytes = (const unsigned char *)string;
int num;
while (*bytes != 0x00) {
if ((*bytes & 0x80) == 0x00) {
// U+0000 to U+007F
num = 1;
} else if ((*bytes & 0xE0) == 0xC0) {
// U+0080 to U+07FF
num = 2;
} else if ((*bytes & 0xF0) == 0xE0) {
// U+0800 to U+FFFF
num = 3;
} else if ((*bytes & 0xF8) == 0xF0) {
// U+10000 to U+10FFFF
num = 4;
} else {
return false;
}
bytes += 1;
for (int i = 1; i < num; ++i) {
if ((*bytes & 0xC0) != 0x80) {
return false;
}
bytes += 1;
}
}
return true;
}
static void log_callback(ggml_log_level level, const char * fmt, void * data) {
if (level == GGML_LOG_LEVEL_ERROR) __android_log_print(ANDROID_LOG_ERROR, TAG, fmt, data);
else if (level == GGML_LOG_LEVEL_INFO) __android_log_print(ANDROID_LOG_INFO, TAG, fmt, data);
@ -295,6 +334,8 @@ Java_com_example_llama_Llm_completion_1init(
jint n_len
) {
cached_token_chars.clear();
const auto text = env->GetStringUTFChars(jtext, 0);
const auto context = reinterpret_cast<llama_context *>(context_pointer);
const auto batch = reinterpret_cast<llama_batch *>(batch_pointer);
@ -372,8 +413,16 @@ Java_com_example_llama_Llm_completion_1loop(
}
auto new_token_chars = llama_token_to_piece(context, new_token_id);
LOGi("new_token_chars: `%s`", new_token_chars.c_str());
auto new_token = env->NewStringUTF(new_token_chars.c_str());
cached_token_chars += new_token_chars;
jstring new_token = nullptr;
if (is_valid_utf8(cached_token_chars.c_str())) {
new_token = env->NewStringUTF(cached_token_chars.c_str());
LOGi("cached: %s, new_token_chars: `%s`, id: %d", cached_token_chars.c_str(), new_token_chars.c_str(), new_token_id);
cached_token_chars.clear();
} else {
new_token = env->NewStringUTF("");
}
llama_batch_clear(*batch);
llama_batch_add(*batch, new_token_id, n_cur, { 0 }, true);

4
examples/llama.android/app/src/main/java/com/example/llama/Llm.kt
View File

@ -71,7 +71,7 @@ class Llm {
batch: Long,
nLen: Int,
ncur: IntVar
): String
): String?
private external fun kv_cache_clear(context: Long)
@ -115,7 +115,7 @@ class Llm {
val ncur = IntVar(completion_init(state.context, state.batch, message, nlen))
while (ncur.value <= nlen) {
val str = completion_loop(state.context, state.batch, nlen, ncur)
if (str.isEmpty()) {
if (str == null) {
break
}
emit(str)

55
examples/llava/clip.cpp
View File

@ -1862,7 +1862,6 @@ bool clip_model_quantize(const char * fname_inp, const char * fname_out, const i
std::vector<uint8_t> work(512);
std::vector<float> conv_buf(512);
std::vector<int64_t> hist_all(1 << 4, 0);
size_t total_size_org = 0;
size_t total_size_new = 0;
@ -1917,48 +1916,7 @@ bool clip_model_quantize(const char * fname_inp, const char * fname_out, const i
}
new_data = work.data();
std::vector<int64_t> hist_cur(1 << 4, 0);
switch (new_type) {
case GGML_TYPE_Q4_0: {
new_size = ggml_quantize_q4_0(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q4_1: {
new_size = ggml_quantize_q4_1(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q5_0: {
new_size = ggml_quantize_q5_0(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q5_1: {
new_size = ggml_quantize_q5_1(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q8_0: {
new_size = ggml_quantize_q8_0(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q2_K: {
new_size = ggml_quantize_q2_K(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q3_K: {
new_size = ggml_quantize_q3_K(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q4_K: {
new_size = ggml_quantize_q4_K(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q5_K: {
new_size = ggml_quantize_q5_K(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
case GGML_TYPE_Q6_K: {
new_size = ggml_quantize_q6_K(f32_data, new_data, n_elms, cur->ne[0], hist_cur.data());
} break;
default: {
fprintf(stderr, "%s: unsupported quantization type %d\n", __func__, new_type);
return false;
}
}
for (size_t j = 0; j < hist_cur.size(); ++j) {
hist_all[j] += hist_cur[j];
}
new_size = ggml_quantize_chunk(new_type, f32_data, new_data, 0, n_elms/cur->ne[0], cur->ne[0], nullptr);
} else {
new_type = cur->type;
new_data = cur->data;
@ -1993,17 +1951,6 @@ bool clip_model_quantize(const char * fname_inp, const char * fname_out, const i
{
printf("%s: original size = %8.2f MB\n", __func__, total_size_org / 1024.0 / 1024.0);
printf("%s: quantized size = %8.2f MB\n", __func__, total_size_new / 1024.0 / 1024.0);
int64_t sum_all = 0;
for (size_t i = 0; i < hist_all.size(); ++i) {
sum_all += hist_all[i];
}
printf("%s: hist: ", __func__);
for (size_t i = 0; i < hist_all.size(); ++i) {
printf("%5.3f ", hist_all[i] / (float)sum_all);
}
printf("\n");
}
return true;

20
examples/parallel/parallel.cpp
View File

@ -107,6 +107,9 @@ int main(int argc, char ** argv) {
// number of simultaneous "clients" to simulate
const int32_t n_clients = params.n_parallel;
// dedicate one sequence to the system prompt
params.n_parallel += 1;
// requests to simulate
const int32_t n_seq = params.n_sequences;
@ -196,8 +199,8 @@ int main(int argc, char ** argv) {
}
// assign the system KV cache to all parallel sequences
for (int32_t i = 1; i < n_clients; ++i) {
llama_kv_cache_seq_cp(ctx, 0, i, 0, n_tokens_system);
for (int32_t i = 1; i <= n_clients; ++i) {
llama_kv_cache_seq_cp(ctx, 0, i, -1, -1);
}
LOG_TEE("\n");
@ -221,15 +224,17 @@ int main(int argc, char ** argv) {
client.i_batch = batch.n_tokens;
llama_batch_add(batch, client.sampled, n_tokens_system + client.n_prompt + client.n_decoded, { client.id }, true);
llama_batch_add(batch, client.sampled, n_tokens_system + client.n_prompt + client.n_decoded, { client.id + 1 }, true);
client.n_decoded += 1;
}
if (batch.n_tokens == 0) {
// all sequences have ended - clear the entire KV cache
for (int i = 0; i < n_clients; ++i) {
llama_kv_cache_seq_rm(ctx, i, n_tokens_system, -1);
for (int i = 1; i <= n_clients; ++i) {
llama_kv_cache_seq_rm(ctx, i, -1, -1);
// but keep the system prompt
llama_kv_cache_seq_cp(ctx, 0, i, -1, -1);
}
LOG_TEE("%s: clearing the KV cache\n", __func__);
@ -255,7 +260,7 @@ int main(int argc, char ** argv) {
tokens_prompt = ::llama_tokenize(ctx, client.prompt, false);
for (size_t i = 0; i < tokens_prompt.size(); ++i) {
llama_batch_add(batch, tokens_prompt[i], i + n_tokens_system, { client.id }, false);
llama_batch_add(batch, tokens_prompt[i], i + n_tokens_system, { client.id + 1 }, false);
}
// extract the logits only for the last token
@ -366,7 +371,8 @@ int main(int argc, char ** argv) {
}
// delete only the generated part of the sequence, i.e. keep the system prompt in the cache
llama_kv_cache_seq_rm(ctx, client.id, n_tokens_system, -1);
llama_kv_cache_seq_rm(ctx, client.id + 1, -1, -1);
llama_kv_cache_seq_cp(ctx, 0, client.id + 1, -1, -1);
const auto t_main_end = ggml_time_us();

168
examples/perplexity/perplexity.cpp
View File

@ -442,7 +442,7 @@ static results_perplexity perplexity_v2(llama_context * ctx, const gpt_params &
return {tokens, std::exp(nll / count), logit_history, prob_history};
}
static results_perplexity perplexity(llama_context * ctx, const gpt_params & params) {
static results_perplexity perplexity(llama_context * ctx, const gpt_params & params, const int32_t n_ctx) {
if (params.ppl_stride > 0) {
return perplexity_v2(ctx, params);
}
@ -453,7 +453,6 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
// BOS tokens will be added for each chunk before eval
const bool add_bos = llama_should_add_bos_token(llama_get_model(ctx));
const int n_ctx = llama_n_ctx(ctx);
std::ofstream logits_stream;
if (!params.logits_file.empty()) {
@ -499,13 +498,19 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
double nll2 = 0.0;
const int num_batches = (n_ctx + n_batch - 1) / n_batch;
const int n_seq = std::max(1, n_batch / n_ctx);
GGML_ASSERT(n_batch < n_ctx || n_batch % n_ctx == 0);
GGML_ASSERT(params.n_ctx == n_seq * n_ctx);
llama_batch batch = llama_batch_init(std::min(n_batch, n_ctx*n_seq), 0, 1);
std::vector<float> logits;
if (num_batches > 1) {
logits.reserve((size_t)n_ctx * n_vocab);
}
fprintf(stderr, "%s: calculating perplexity over %d chunks, batch_size=%d\n", __func__, n_chunk, n_batch);
fprintf(stderr, "%s: calculating perplexity over %d chunks, n_ctx=%d, batch_size=%d, n_seq=%d\n", __func__, n_chunk, n_ctx, n_batch, n_seq);
std::vector<std::thread> workers(std::thread::hardware_concurrency() - 1);
@ -518,54 +523,6 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
log_probs.resize(n_ctx * nv);
}
for (int i = 0; i < n_chunk; ++i) {
const int start = i * n_ctx;
const int end = start + n_ctx;
const auto t_start = std::chrono::high_resolution_clock::now();
// clear the KV cache
llama_kv_cache_clear(ctx);
for (int j = 0; j < num_batches; ++j) {
const int batch_start = start + j * n_batch;
const int batch_size = std::min(end - batch_start, n_batch);
// save original token and restore it after eval
const auto token_org = tokens[batch_start];
// add BOS token for the first batch of each chunk
if (add_bos && j == 0) {
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))) {
fprintf(stderr, "%s : failed to eval\n", __func__);
return {tokens, -1, logit_history, prob_history};
}
// restore the original token in case it was set to BOS
tokens[batch_start] = token_org;
if (num_batches > 1) {
const auto * batch_logits = llama_get_logits(ctx);
logits.insert(logits.end(), batch_logits, batch_logits + batch_size * n_vocab);
}
}
const auto t_end = std::chrono::high_resolution_clock::now();
if (i == 0) {
const float t_total = std::chrono::duration<float>(t_end - t_start).count();
fprintf(stderr, "%s: %.2f seconds per pass - ETA ", __func__, t_total);
int total_seconds = (int)(t_total * n_chunk);
if (total_seconds >= 60*60) {
fprintf(stderr, "%d hours ", total_seconds / (60*60));
total_seconds = total_seconds % (60*60);
}
fprintf(stderr, "%.2f minutes\n", total_seconds / 60.0);
}
// We get the logits for all the tokens in the context window (params.n_ctx)
// from llama_eval above. Now, based on https://huggingface.co/docs/transformers/perplexity,
// calculate the perplexity over the last half of the window (so the model always has
@ -579,25 +536,98 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
// last 256 tokens. Then, we split the input up into context window size chunks to
// process the entire prompt.
const int first = n_ctx/2;
const float * all_logits = num_batches > 1 ? logits.data() : llama_get_logits(ctx);
for (int i = 0; i < n_chunk; i += n_seq) {
const int start = i * n_ctx;
const int end = start + n_ctx;
const int n_seq_batch = std::min(n_seq, n_chunk - i);
const auto t_start = std::chrono::high_resolution_clock::now();
// clear the KV cache
llama_kv_cache_clear(ctx);
for (int j = 0; j < num_batches; ++j) {
const int batch_start = start + j * n_batch;
const int batch_size = std::min(end - batch_start, n_batch);
batch.n_tokens = 0;
for (int seq = 0; seq < n_seq_batch; seq++) {
int seq_start = batch_start + seq*n_ctx;
// save original token and restore it after eval
const auto token_org = tokens[seq_start];
// add BOS token for the first batch of each chunk
if (add_bos && j == 0) {
tokens[seq_start] = llama_token_bos(llama_get_model(ctx));
}
for (int k = 0; k < batch_size; ++k) {
const int idx = seq*n_ctx + k;
batch.token[idx] = tokens[seq_start + k];
batch.pos[idx] = j*n_batch + k;
batch.n_seq_id[idx] = 1;
batch.seq_id[idx][0] = seq;
batch.logits[idx] = batch.pos[idx] >= first ? 1 : 0;
}
batch.n_tokens += batch_size;
// restore the original token in case it was set to BOS
tokens[seq_start] = token_org;
}
if (llama_decode(ctx, batch)) {
fprintf(stderr, "%s : failed to eval\n", __func__);
return {tokens, -1, logit_history, prob_history};
}
if (num_batches > 1) {
const auto * batch_logits = llama_get_logits(ctx);
logits.insert(logits.end(), batch_logits, batch_logits + batch_size * n_vocab);
}
}
const auto t_end = std::chrono::high_resolution_clock::now();
if (i == 0) {
const float t_total = std::chrono::duration<float>(t_end - t_start).count();
fprintf(stderr, "%s: %.2f seconds per pass - ETA ", __func__, t_total);
int total_seconds = (int)(t_total*n_chunk/n_seq);
if (total_seconds >= 60*60) {
fprintf(stderr, "%d hours ", total_seconds / (60*60));
total_seconds = total_seconds % (60*60);
}
fprintf(stderr, "%.2f minutes\n", total_seconds / 60.0);
}
for (int seq = 0; seq < n_seq_batch; seq++) {
const float * all_logits = num_batches > 1 ? logits.data() : llama_get_logits_ith(ctx, seq*n_ctx);
llama_token * tokens_data = tokens.data() + start + seq*n_ctx + first;
if (!params.logits_file.empty()) {
process_logits(logits_stream, n_vocab, all_logits + first*n_vocab, tokens.data() + start + first, n_ctx - 1 - first,
process_logits(logits_stream, n_vocab, all_logits + first*n_vocab,
tokens_data, n_ctx - 1 - first,
workers, log_probs, nll, nll2);
} else {
process_logits(n_vocab, all_logits + first*n_vocab, tokens.data() + start + first, n_ctx - 1 - first,
workers, nll, nll2, logit_history.data() + start + first, prob_history.data() + start + first);
process_logits(n_vocab, all_logits + first*n_vocab,
tokens_data, n_ctx - 1 - first,
workers, nll, nll2,
logit_history.data() + start + seq*n_ctx + first,
prob_history.data() + start + seq*n_ctx + first);
}
count += n_ctx - first - 1;
// perplexity is e^(average negative log-likelihood)
if (params.ppl_output_type == 0) {
printf("[%d]%.4lf,", i + 1, std::exp(nll / count));
printf("[%d]%.4lf,", i + seq + 1, std::exp(nll / count));
} else {
double av = nll/count;
double av2 = nll2/count - av*av;
if (av2 > 0) av2 = sqrt(av2/(count-1));
printf("%8d %.4lf %4lf %4lf\n", i*n_ctx, std::exp(nll / count), av, av2);
}
}
fflush(stdout);
logits.clear();
@ -615,6 +645,8 @@ static results_perplexity perplexity(llama_context * ctx, const gpt_params & par
printf("Unexpected negative standard deviation of log(prob)\n");
}
llama_batch_free(batch);
return {tokens, ppl, logit_history, prob_history};
}
@ -809,7 +841,7 @@ static void hellaswag_score(llama_context * ctx, const gpt_params & params) {
const int n_batch = params.n_batch;
const int max_tasks_per_batch = 32;
const int max_seq = 4*max_tasks_per_batch;
const int max_seq = std::min(4*max_tasks_per_batch, (int) llama_n_max_seq(ctx));
llama_batch batch = llama_batch_init(n_ctx, 0, max_seq);
@ -1086,7 +1118,7 @@ static void winogrande_score(llama_context * ctx, const gpt_params & params) {
const int n_batch = params.n_batch;
const int max_tasks_per_batch = 128;
const int max_seq = 2*max_tasks_per_batch;
const int max_seq = std::min(2*max_tasks_per_batch, (int) llama_n_max_seq(ctx));
llama_batch batch = llama_batch_init(n_ctx, 0, max_seq);
@ -1438,7 +1470,7 @@ static void multiple_choice_score(llama_context * ctx, const gpt_params & params
const int n_batch = params.n_batch;
const int max_tasks_per_batch = 32;
const int max_seq = 4*max_tasks_per_batch;
const int max_seq = std::min(4*max_tasks_per_batch, (int) llama_n_max_seq(ctx));
llama_batch batch = llama_batch_init(n_ctx, 0, max_seq);
@ -1782,13 +1814,24 @@ static void kl_divergence(llama_context * ctx, const gpt_params & params) {
int main(int argc, char ** argv) {
gpt_params params;
params.n_batch = 512;
if (!gpt_params_parse(argc, argv, params)) {
return 1;
}
params.logits_all = true;
const int32_t n_ctx = params.n_ctx;
const bool ppl = !params.hellaswag && !params.winogrande && !params.multiple_choice && !params.kl_divergence;
if (ppl) {
int n_seq = std::max(1, params.n_batch / n_ctx);
int32_t n_kv = n_seq * n_ctx;
params.n_parallel = n_seq;
params.n_ctx = n_kv;
params.n_batch = std::min(params.n_batch, n_kv);
} else {
params.n_batch = std::min(params.n_batch, params.n_ctx);
}
if (params.ppl_stride > 0) {
fprintf(stderr, "Will perform strided perplexity calculation -> adjusting context size from %d to %d\n",
@ -1815,6 +1858,9 @@ int main(int argc, char ** argv) {
llama_model * model;
llama_context * ctx;
// ensure there's at least enough seq_ids for HellaSwag
params.n_parallel = std::max(4, params.n_parallel);
// load the model and apply lora adapter, if any
std::tie(model, ctx) = llama_init_from_gpt_params(params);
if (model == NULL) {
@ -1844,7 +1890,7 @@ int main(int argc, char ** argv) {
} else if (params.kl_divergence) {
kl_divergence(ctx, params);
} else {
results = perplexity(ctx, params);
results = perplexity(ctx, params, n_ctx);
}
llama_print_timings(ctx);

2
examples/server-embd.py
View File

@ -13,7 +13,7 @@ async def main():
model_url = "http://127.0.0.1:6900"
responses: list[requests.Response] = await asyncio.gather(*[requests_post_async(
url= f"{model_url}/embedding",
json= {"content": str(i)*1024}
json= {"content": str(0)*1024}
) for i in range(n)])
for response in responses:

10
examples/server/CMakeLists.txt
View File

@ -1,12 +1,18 @@
set(TARGET server)
option(LLAMA_SERVER_VERBOSE "Build verbose logging option for Server" ON)
option(LLAMA_SERVER_SSL "Build SSL support for the server" OFF)
include_directories(${CMAKE_CURRENT_SOURCE_DIR})
add_executable(${TARGET} server.cpp oai.hpp utils.hpp json.hpp httplib.h)
add_executable(${TARGET} server.cpp utils.hpp json.hpp httplib.h)
install(TARGETS ${TARGET} RUNTIME)
target_compile_definitions(${TARGET} PRIVATE
SERVER_VERBOSE=$<BOOL:${LLAMA_SERVER_VERBOSE}>
)
target_link_libraries(${TARGET} PRIVATE common llava ${CMAKE_THREAD_LIBS_INIT})
target_link_libraries(${TARGET} PRIVATE common ${CMAKE_THREAD_LIBS_INIT})
if (LLAMA_SERVER_SSL)
find_package(OpenSSL REQUIRED)
target_link_libraries(${TARGET} PRIVATE OpenSSL::SSL OpenSSL::Crypto)
target_compile_definitions(${TARGET} PRIVATE CPPHTTPLIB_OPENSSL_SUPPORT)
endif()
if (WIN32)
TARGET_LINK_LIBRARIES(${TARGET} PRIVATE ws2_32)
endif()

40
examples/server/README.md
View File

@ -42,7 +42,7 @@ see https://github.com/ggerganov/llama.cpp/issues/1437
- `-to N`, `--timeout N`: Server read/write timeout in seconds. Default `600`.
- `--host`: Set the hostname or ip address to listen. Default `127.0.0.1`.
- `--port`: Set the port to listen. Default: `8080`.
- `--path`: path from which to serve static files (default examples/server/public)
- `--path`: path from which to serve static files (default: disabled)
- `--api-key`: Set an api key for request authorization. By default the server responds to every request. With an api key set, the requests must have the Authorization header set with the api key as Bearer token. May be used multiple times to enable multiple valid keys.
- `--api-key-file`: path to file containing api keys delimited by new lines. If set, requests must include one of the keys for access. May be used in conjunction with `--api-key`'s.
- `--embedding`: Enable embedding extraction, Default: disabled.
@ -59,6 +59,10 @@ see https://github.com/ggerganov/llama.cpp/issues/1437
- `--log-disable`: Output logs to stdout only, default: enabled.
- `--log-format FORMAT`: Define the log output to FORMAT: json or text (default: json)
**If compiled with `LLAMA_SERVER_SSL=ON`**
- `--ssl-key-file FNAME`: path to file a PEM-encoded SSL private key
- `--ssl-cert-file FNAME`: path to file a PEM-encoded SSL certificate
## Build
server is build alongside everything else from the root of the project
@ -75,6 +79,28 @@ server is build alongside everything else from the root of the project
cmake --build . --config Release
```
## Build with SSL
server can also be built with SSL support using OpenSSL 3
- Using `make`:
```bash
# NOTE: For non-system openssl, use the following:
# CXXFLAGS="-I /path/to/openssl/include"
# LDFLAGS="-L /path/to/openssl/lib"
make LLAMA_SERVER_SSL=true server
```
- Using `CMake`:
```bash
mkdir build
cd build
cmake .. -DLLAMA_SERVER_SSL=ON
make server
```
## Quick Start
To get started right away, run the following command, making sure to use the correct path for the model you have:
@ -169,7 +195,11 @@ node index.js
*Options:*
`prompt`: Provide the prompt for this completion as a string or as an array of strings or numbers representing tokens. Internally, the prompt is compared to the previous completion and only the "unseen" suffix is evaluated. If the prompt is a string or an array with the first element given as a string, a `bos` token is inserted in the front like `main` does.
`prompt`: Provide the prompt for this completion as a string or as an array of strings or numbers representing tokens. Internally, if `cache_prompt` is `true`, the prompt is compared to the previous completion and only the "unseen" suffix is evaluated. A `BOS` token is inserted at the start, if all of the following conditions are true:
- The prompt is a string or an array with the first element given as a string
- The model's `tokenizer.ggml.add_bos_token` metadata is `true`
- The system prompt is empty
`temperature`: Adjust the randomness of the generated text (default: 0.8).
@ -282,7 +312,7 @@ Notice that each `probs` is an array of length `n_probs`.
`content`: Set the text to tokenize.
Note that the special `BOS` token is not added in front of the text and also a space character is not inserted automatically as it is for `/completion`.
Note that a special `BOS` token is never inserted.
- **POST** `/detokenize`: Convert tokens to text.
@ -436,7 +466,7 @@ Notice that each `probs` is an array of length `n_probs`.
"next_token": {
"has_next_token": true,
"n_remain": -1,
"num_tokens_predicted": 0,
"n_decoded": 0,
"stopped_eos": false,
"stopped_limit": false,
"stopped_word": false,
@ -532,7 +562,7 @@ The HTTP server supports OAI-like API
### Extending or building alternative Web Front End
The default location for the static files is `examples/server/public`. You can extend the front end by running the server binary with `--path` set to `./your-directory` and importing `/completion.js` to get access to the llamaComplete() method.
You can extend the front end by running the server binary with `--path` set to `./your-directory` and importing `/completion.js` to get access to the llamaComplete() method.
Read the documentation in `/completion.js` to see convenient ways to access llama.

88
examples/server/bench/README.md Normal file
View File

@ -0,0 +1,88 @@
### Server benchmark tools
Benchmark is using [k6](https://k6.io/).
##### Install k6
Follow instruction from: https://k6.io/docs/get-started/installation/
Example for ubuntu:
```shell
snap install k6
```
#### Download a dataset
This dataset was originally proposed in [vLLM benchmarks](https://github.com/vllm-project/vllm/blob/main/benchmarks/README.md).
```shell
wget https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json
```
#### Download a model
Example for PHI-2
```shell
../../../scripts/hf.sh --repo ggml-org/models --file phi-2/ggml-model-q4_0.gguf
```
#### Start the server
The server must answer OAI Chat completion requests on `http://localhost:8080/v1` or according to the environment variable `SERVER_BENCH_URL`.
Example:
```shell
server --host localhost --port 8080 \
--model ggml-model-q4_0.gguf \
--cont-batching \
--metrics \
--parallel 8 \
--batch-size 512 \
--ctx-size 4096 \
--log-format text \
-ngl 33
```
#### Run the benchmark
For 500 chat completions request with 8 concurrent users during maximum 10 minutes, run:
```shell
k6 run script.js --duration 10m --iterations 500 --vus 8
```
The benchmark values can be overridden with:
- `SERVER_BENCH_URL` server url prefix for chat completions, default `http://localhost:8080/v1`
- `SERVER_BENCH_N_PROMPTS` total prompts to randomly select in the benchmark, default `480`
- `SERVER_BENCH_MODEL_ALIAS` model alias to pass in the completion request, default `my-model`
- `SERVER_BENCH_MAX_TOKENS` max tokens to predict, default: `512`
- `SERVER_BENCH_DATASET` path to the benchmark dataset file
- `SERVER_BENCH_MAX_PROMPT_TOKENS` maximum prompt tokens to filter out in the dataset: default `1024`
- `SERVER_BENCH_MAX_CONTEXT` maximum context size of the completions request to filter out in the dataset: prompt + predicted tokens, default `2048`
Note: the local tokenizer is just a string space split, real number of tokens will differ.
Or with [k6 options](https://k6.io/docs/using-k6/k6-options/reference/):
```shell
SERVER_BENCH_N_PROMPTS=500 k6 run script.js --duration 10m --iterations 500 --vus 8
```
To [debug http request](https://k6.io/docs/using-k6/http-debugging/) use `--http-debug="full"`.
#### Metrics
Following metrics are available computed from the OAI chat completions response `usage`:
- `llamacpp_tokens_second` Trend of `usage.total_tokens / request duration`
- `llamacpp_prompt_tokens` Trend of `usage.prompt_tokens`
- `llamacpp_prompt_tokens_total_counter` Counter of `usage.prompt_tokens`
- `llamacpp_completion_tokens` Trend of `usage.completion_tokens`
- `llamacpp_completion_tokens_total_counter` Counter of `usage.completion_tokens`
- `llamacpp_completions_truncated_rate` Rate of completions truncated, i.e. if `finish_reason === 'length'`
- `llamacpp_completions_stop_rate` Rate of completions stopped by the model, i.e. if `finish_reason === 'stop'`
The script will fail if too many completions are truncated, see `llamacpp_completions_truncated_rate`.
K6 metrics might be compared against [server metrics](../README.md), with:
```shell
curl http://localhost:8080/metrics
```

120
examples/server/bench/script.js Normal file
View File

@ -0,0 +1,120 @@
import http from 'k6/http'
import {check, sleep} from 'k6'
import {SharedArray} from 'k6/data'
import {Counter, Rate, Trend} from 'k6/metrics'
import exec from 'k6/execution';
// Server chat completions prefix
const server_url = __ENV.SERVER_BENCH_URL ? __ENV.SERVER_BENCH_URL : 'http://localhost:8080/v1'
// Number of total prompts in the dataset - default 10m / 10 seconds/request * number of users
const n_prompt = __ENV.SERVER_BENCH_N_PROMPTS ? parseInt(__ENV.SERVER_BENCH_N_PROMPTS) : 600 / 10 * 8
// Model name to request
const model = __ENV.SERVER_BENCH_MODEL_ALIAS ? __ENV.SERVER_BENCH_MODEL_ALIAS : 'my-model'
// Dataset path
const dataset_path = __ENV.SERVER_BENCH_DATASET ? __ENV.SERVER_BENCH_DATASET : './ShareGPT_V3_unfiltered_cleaned_split.json'
// Max tokens to predict
const max_tokens = __ENV.SERVER_BENCH_MAX_TOKENS ? parseInt(__ENV.SERVER_BENCH_MAX_TOKENS) : 512
// Max prompt tokens
const n_prompt_tokens = __ENV.SERVER_BENCH_MAX_PROMPT_TOKENS ? parseInt(__ENV.SERVER_BENCH_MAX_PROMPT_TOKENS) : 1024
// Max slot context
const n_ctx_slot = __ENV.SERVER_BENCH_MAX_CONTEXT ? parseInt(__ENV.SERVER_BENCH_MAX_CONTEXT) : 2048
export function setup() {
console.info(`Benchmark config: server_url=${server_url} n_prompt=${n_prompt} model=${model} dataset_path=${dataset_path} max_tokens=${max_tokens}`)
}
const data = new SharedArray('conversations', function () {
const tokenizer = (message) => message.split(/[\s,'".?]/)
return JSON.parse(open(dataset_path))
// Filter out the conversations with less than 2 turns.
.filter(data => data["conversations"].length >= 2)
.filter(data => data["conversations"][0]["from"] === "human")
.map(data => {
return {
prompt: data["conversations"][0]["value"],
n_prompt_tokens: tokenizer(data["conversations"][0]["value"]).length,
n_completion_tokens: tokenizer(data["conversations"][1]["value"]).length,
}
})
// Filter out too short sequences
.filter(conv => conv.n_prompt_tokens >= 4 && conv.n_completion_tokens >= 4)
// Filter out too long sequences.
.filter(conv => conv.n_prompt_tokens <= n_prompt_tokens && conv.n_prompt_tokens + conv.n_completion_tokens <= n_ctx_slot)
// Keep only first n prompts
.slice(0, n_prompt)
})
const llamacpp_prompt_tokens = new Trend('llamacpp_prompt_tokens')
const llamacpp_completion_tokens = new Trend('llamacpp_completion_tokens')
const llamacpp_tokens_second = new Trend('llamacpp_tokens_second')
const llamacpp_prompt_tokens_total_counter = new Counter('llamacpp_prompt_tokens_total_counter')
const llamacpp_completion_tokens_total_counter = new Counter('llamacpp_completion_tokens_total_counter')
const llamacpp_completions_truncated_rate = new Rate('llamacpp_completions_truncated_rate')
const llamacpp_completions_stop_rate = new Rate('llamacpp_completions_stop_rate')
export const options = {
thresholds: {
llamacpp_completions_truncated_rate: [
// more than 80% of truncated input will abort the test
{threshold: 'rate < 0.8', abortOnFail: true, delayAbortEval: '1m'},
],
},
duration: '10m',
vus: 8,
}
export default function () {
const conversation = data[exec.scenario.iterationInInstance % data.length]
const payload = {
"messages": [
{
"role": "system",
"content": "You are ChatGPT, an AI assistant.",
},
{
"role": "user",
"content": conversation.prompt,
}
],
"model": model,
"stream": false,
"max_tokens": max_tokens
}
const body = JSON.stringify(payload)
let res = http.post(`${server_url}/chat/completions`, body, {
headers: {'Content-Type': 'application/json'},
timeout: '300s'
})
check(res, {'success completion': (r) => r.status === 200})
if (res.status === 200) {
const completions = res.json()
llamacpp_prompt_tokens.add(completions.usage.prompt_tokens)
llamacpp_prompt_tokens_total_counter.add(completions.usage.prompt_tokens)
llamacpp_completion_tokens.add(completions.usage.completion_tokens)
llamacpp_completion_tokens_total_counter.add(completions.usage.completion_tokens)
llamacpp_completions_truncated_rate.add(completions.choices[0].finish_reason === 'length')
llamacpp_completions_stop_rate.add(completions.choices[0].finish_reason === 'stop')
llamacpp_tokens_second.add(completions.usage.total_tokens / res.timings.duration * 1.e3)
} else {
console.error(`response: ${res.body} request=${payload}`)
}
sleep(0.3)
}

225
examples/server/oai.hpp
View File

@ -1,225 +0,0 @@
#pragma once
#include <string>
#include <vector>
#include <set>
#include <mutex>
#include <condition_variable>
#include <unordered_map>
#include "json.hpp"
#include "utils.hpp"
#define DEFAULT_OAICOMPAT_MODEL "gpt-3.5-turbo-0613"
using json = nlohmann::json;
inline static json oaicompat_completion_params_parse(
const struct llama_model * model,
const json &body, /* openai api json semantics */
const std::string &chat_template)
{
json llama_params;
llama_params["__oaicompat"] = true;
// Map OpenAI parameters to llama.cpp parameters
//
// For parameters that are defined by the OpenAI documentation (e.g.
// temperature), we explicitly specify OpenAI's intended default; we
// need to do that because sometimes OpenAI disagrees with llama.cpp
//
// https://platform.openai.com/docs/api-reference/chat/create
llama_sampling_params default_sparams;
llama_params["model"] = json_value(body, "model", std::string("unknown"));
llama_params["prompt"] = format_chat(model, chat_template, body["messages"]);
llama_params["cache_prompt"] = json_value(body, "cache_prompt", false);
llama_params["temperature"] = json_value(body, "temperature", 0.0);
llama_params["top_k"] = json_value(body, "top_k", default_sparams.top_k);
llama_params["top_p"] = json_value(body, "top_p", 1.0);
llama_params["n_predict"] = json_value(body, "max_tokens", -1);
llama_params["logit_bias"] = json_value(body, "logit_bias",json::object());
llama_params["frequency_penalty"] = json_value(body, "frequency_penalty", 0.0);
llama_params["presence_penalty"] = json_value(body, "presence_penalty", 0.0);
llama_params["seed"] = json_value(body, "seed", LLAMA_DEFAULT_SEED);
llama_params["stream"] = json_value(body, "stream", false);
llama_params["mirostat"] = json_value(body, "mirostat", default_sparams.mirostat);
llama_params["mirostat_tau"] = json_value(body, "mirostat_tau", default_sparams.mirostat_tau);
llama_params["mirostat_eta"] = json_value(body, "mirostat_eta", default_sparams.mirostat_eta);
llama_params["penalize_nl"] = json_value(body, "penalize_nl", default_sparams.penalize_nl);
llama_params["typical_p"] = json_value(body, "typical_p", default_sparams.typical_p);
llama_params["repeat_last_n"] = json_value(body, "repeat_last_n", default_sparams.penalty_last_n);
llama_params["ignore_eos"] = json_value(body, "ignore_eos", false);
llama_params["tfs_z"] = json_value(body, "tfs_z", default_sparams.tfs_z);
if (body.count("grammar") != 0) {
llama_params["grammar"] = json_value(body, "grammar", json::object());
}
// Handle 'stop' field
if (body.contains("stop") && body["stop"].is_string()) {
llama_params["stop"] = json::array({body["stop"].get<std::string>()});
} else {
llama_params["stop"] = json_value(body, "stop", json::array());
}
// Ensure there is ChatML-specific end sequence among stop words
llama_params["stop"].push_back("<|im_end|>");
return llama_params;
}
inline static json format_final_response_oaicompat(const json &request, const task_result &response, bool streaming = false)
{
json result = response.result_json;
bool stopped_word = result.count("stopped_word") != 0;
bool stopped_eos = json_value(result, "stopped_eos", false);
int num_tokens_predicted = json_value(result, "tokens_predicted", 0);
int num_prompt_tokens = json_value(result, "tokens_evaluated", 0);
std::string content = json_value(result, "content", std::string(""));
std::string finish_reason = "length";
if (stopped_word || stopped_eos) {
finish_reason = "stop";
}
json choices =
streaming ? json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"delta", json::object()}}})
: json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"message", json{{"content", content},
{"role", "assistant"}}}}});
std::time_t t = std::time(0);
json res =
json{{"choices", choices},
{"created", t},
{"model",
json_value(request, "model", std::string(DEFAULT_OAICOMPAT_MODEL))},
{"object", streaming ? "chat.completion.chunk" : "chat.completion"},
{"usage",
json{{"completion_tokens", num_tokens_predicted},
{"prompt_tokens", num_prompt_tokens},
{"total_tokens", num_tokens_predicted + num_prompt_tokens}}},
{"id", gen_chatcmplid()}};
if (server_verbose) {
res["__verbose"] = result;
}
if (result.contains("completion_probabilities")) {
res["completion_probabilities"] = json_value(result, "completion_probabilities", json::array());
}
return res;
}
// return value is vector as there is one case where we might need to generate two responses
inline static std::vector<json> format_partial_response_oaicompat(const task_result &response) {
json result = response.result_json;
if (!result.contains("model") || !result.contains("oaicompat_token_ctr")) {
return std::vector<json>({response.result_json});
}
bool first = json_value(result, "oaicompat_token_ctr", 0) == 0;
std::string modelname = json_value(result, "model", std::string(DEFAULT_OAICOMPAT_MODEL));
bool stopped_word = json_value(result, "stopped_word", false);
bool stopped_eos = json_value(result, "stopped_eos", false);
bool stopped_limit = json_value(result, "stopped_limit", false);
std::string content = json_value(result, "content", std::string(""));
std::string finish_reason;
if (stopped_word || stopped_eos) {
finish_reason = "stop";
}
if (stopped_limit) {
finish_reason = "length";
}
std::time_t t = std::time(0);
json choices;
if (!finish_reason.empty()) {
choices = json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"delta", json::object()}}});
} else {
if (first) {
if (content.empty()) {
choices = json::array({json{{"finish_reason", nullptr},
{"index", 0},
{"delta", json{{"role", "assistant"}}}}});
} else {
// We have to send this as two updates to conform to openai behavior
json initial_ret = json{{"choices", json::array({json{
{"finish_reason", nullptr},
{"index", 0},
{"delta", json{
{"role", "assistant"}
}}}})},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}};
json second_ret = json{
{"choices", json::array({json{{"finish_reason", nullptr},
{"index", 0},
{"delta", json{
{"content", content}}}
}})},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}};
return std::vector<json>({initial_ret, second_ret});
}
} else {
// Some idiosyncrasy in task processing logic makes several trailing calls
// with empty content, we ignore these at the calee site.
if (content.empty()) {
return std::vector<json>({json::object()});
}
choices = json::array({json{
{"finish_reason", nullptr},
{"index", 0},
{"delta",
json{
{"content", content},
}},
}});
}
}
json ret = json{{"choices", choices},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}};
return std::vector<json>({ret});
}
inline static json format_embeddings_response_oaicompat(const json &request, const json &embeddings)
{
json res =
json{
{"model", json_value(request, "model", std::string(DEFAULT_OAICOMPAT_MODEL))},
{"object", "list"},
{"usage",
json{{"prompt_tokens", 0},
{"total_tokens", 0}}},
{"data", embeddings}
};
return res;
}

3433
examples/server/server.cpp
View File

File diff suppressed because it is too large Load Diff

94
examples/server/tests/features/embeddings.feature Normal file
View File

@ -0,0 +1,94 @@
@llama.cpp
@embeddings
Feature: llama.cpp server
Background: Server startup
Given a server listening on localhost:8080
And a model file bert-bge-small/ggml-model-f16.gguf from HF repo ggml-org/models
And a model alias bert-bge-small
And 42 as server seed
And 2 slots
And 1024 as batch size
And 2048 KV cache size
And embeddings extraction
Then the server is starting
Then the server is healthy
Scenario: Embedding
When embeddings are computed for:
"""
What is the capital of Bulgaria ?
"""
Then embeddings are generated
Scenario: OAI Embeddings compatibility
Given a model bert-bge-small
When an OAI compatible embeddings computation request for:
"""
What is the capital of Spain ?
"""
Then embeddings are generated
Scenario: OAI Embeddings compatibility with multiple inputs
Given a model bert-bge-small
Given a prompt:
"""
In which country Paris is located ?
"""
And a prompt:
"""
Is Madrid the capital of Spain ?
"""
When an OAI compatible embeddings computation request for multiple inputs
Then embeddings are generated
Scenario: Multi users embeddings
Given a prompt:
"""
Write a very long story about AI.
"""
And a prompt:
"""
Write another very long music lyrics.
"""
And a prompt:
"""
Write a very long poem.
"""
And a prompt:
"""
Write a very long joke.
"""
Given concurrent embedding requests
Then the server is busy
Then the server is idle
Then all embeddings are generated
Scenario: Multi users OAI compatibility embeddings
Given a prompt:
"""
In which country Paris is located ?
"""
And a prompt:
"""
Is Madrid the capital of Spain ?
"""
And a prompt:
"""
What is the biggest US city ?
"""
And a prompt:
"""
What is the capital of Bulgaria ?
"""
And a model bert-bge-small
Given concurrent OAI embedding requests
Then the server is busy
Then the server is idle
Then all embeddings are generated
Scenario: All embeddings should be the same
Given 10 fixed prompts
And a model bert-bge-small
Given concurrent OAI embedding requests
Then all embeddings are the same

54
examples/server/tests/features/environment.py
View File

@ -1,9 +1,10 @@
import errno
import os
import socket
import subprocess
import time
from contextlib import closing
from signal import SIGKILL
import signal
def before_scenario(context, scenario):
@ -29,41 +30,68 @@ def after_scenario(context, scenario):
for line in f:
print(line)
if not is_server_listening(context.server_fqdn, context.server_port):
print("\x1b[33;101mERROR: Server stopped listening\x1b[0m")
print("\x1b[33;101mERROR: Server stopped listening\x1b[0m\n")
if not pid_exists(context.server_process.pid):
assert False, f"Server not running pid={context.server_process.pid} ..."
print(f"stopping server pid={context.server_process.pid} ...")
context.server_process.kill()
server_graceful_shutdown(context)
# Wait few for socket to free up
time.sleep(0.05)
attempts = 0
while is_server_listening(context.server_fqdn, context.server_port):
print(f"stopping server pid={context.server_process.pid} ...")
os.kill(context.server_process.pid, SIGKILL)
while pid_exists(context.server_process.pid) or is_server_listening(context.server_fqdn, context.server_port):
server_kill(context)
time.sleep(0.1)
attempts += 1
if attempts > 5:
print(f"Server dangling exits, killing all {context.server_path} ...")
process = subprocess.run(['killall', '-9', context.server_path],
stderr=subprocess.PIPE,
universal_newlines=True)
server_kill_hard(context)
def server_graceful_shutdown(context):
print(f"shutting down server pid={context.server_process.pid} ...\n")
if os.name == 'nt':
os.kill(context.server_process.pid, signal.CTRL_C_EVENT)
else:
os.kill(context.server_process.pid, signal.SIGINT)
def server_kill(context):
print(f"killing server pid={context.server_process.pid} ...\n")
context.server_process.kill()
def server_kill_hard(context):
pid = context.server_process.pid
path = context.server_path
print(f"Server dangling exits, hard killing force {pid}={path}...\n")
if os.name == 'nt':
process = subprocess.check_output(['taskkill', '/F', '/pid', str(pid)]).decode()
print(process)
else:
os.kill(-pid, signal.SIGKILL)
def is_server_listening(server_fqdn, server_port):
with closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as sock:
result = sock.connect_ex((server_fqdn, server_port))
return result == 0
_is_server_listening = result == 0
if _is_server_listening:
print(f"server is listening on {server_fqdn}:{server_port}...\n")
return _is_server_listening
def pid_exists(pid):
"""Check whether pid exists in the current process table."""
import errno
if pid < 0:
return False
if os.name == 'nt':
output = subprocess.check_output(['TASKLIST', '/FI', f'pid eq {pid}']).decode()
print(output)
return "No tasks are running" not in output
else:
try:
os.kill(pid, 0)
except OSError as e:

51
examples/server/tests/features/parallel.feature
View File

@ -6,10 +6,9 @@ Feature: Parallel
Given a server listening on localhost:8080
And a model file tinyllamas/stories260K.gguf from HF repo ggml-org/models
And 42 as server seed
And 512 as batch size
And 64 KV cache size
And 128 as batch size
And 256 KV cache size
And 2 slots
And embeddings extraction
And continuous batching
Then the server is starting
Then the server is healthy
@ -77,6 +76,7 @@ Feature: Parallel
| disabled | 128 |
| enabled | 64 |
Scenario: Multi users with total number of tokens to predict exceeds the KV Cache size #3969
Given a prompt:
"""
@ -99,48 +99,3 @@ Feature: Parallel
Then the server is busy
Then the server is idle
Then all prompts are predicted
Scenario: Multi users embeddings
Given a prompt:
"""
Write a very long story about AI.
"""
And a prompt:
"""
Write another very long music lyrics.
"""
And a prompt:
"""
Write a very long poem.
"""
And a prompt:
"""
Write a very long joke.
"""
Given concurrent embedding requests
Then the server is busy
Then the server is idle
Then all embeddings are generated
Scenario: Multi users OAI compatibility embeddings
Given a prompt:
"""
In which country Paris is located ?
"""
And a prompt:
"""
Is Madrid the capital of Spain ?
"""
And a prompt:
"""
What is the biggest US city ?
"""
And a prompt:
"""
What is the capital of Bulgaria ?
"""
And a model tinyllama-2
Given concurrent OAI embedding requests
Then the server is busy
Then the server is idle
Then all embeddings are generated

1
examples/server/tests/features/security.feature
View File

@ -39,6 +39,7 @@ Feature: Security
Scenario Outline: CORS Options
Given a user api key llama.cpp
When an OPTIONS request is sent from <origin>
Then CORS header <cors_header> is set to <cors_header_value>

68
examples/server/tests/features/server.feature
View File

@ -10,11 +10,10 @@ Feature: llama.cpp server
# KV Cache corresponds to the total amount of tokens
# that can be stored across all independent sequences: #4130
# see --ctx-size and #5568
And 32 KV cache size
And 512 as batch size
And 1 slots
And embeddings extraction
And 32 server max tokens to predict
And 256 KV cache size
And 32 as batch size
And 2 slots
And 64 server max tokens to predict
And prometheus compatible metrics exposed
Then the server is starting
Then the server is healthy
@ -23,17 +22,35 @@ Feature: llama.cpp server
Then the server is ready
And all slots are idle
Scenario Outline: Completion
Given a prompt <prompt>
And <n_predict> max tokens to predict
And a completion request with no api error
Then <n_predicted> tokens are predicted matching <re_content>
And the completion is <truncated> truncated
And <n_prompt> prompt tokens are processed
And prometheus metrics are exposed
And metric llamacpp:tokens_predicted is <n_predicted>
Examples: Prompts
| prompt | n_predict | re_content | n_predicted |
| I believe the meaning of life is | 8 | (read\|going)+ | 8 |
| Write a joke about AI | 64 | (park\|friends\|scared\|always)+ | 32 |
| prompt | n_predict | re_content | n_prompt | n_predicted | truncated |
| I believe the meaning of life is | 8 | (read\|going)+ | 18 | 8 | not |
| Write a joke about AI from a very long prompt which will not be truncated | 256 | (princesses\|everyone\|kids)+ | 46 | 64 | not |
Scenario: Completion prompt truncated
Given a prompt:
"""
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.
Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.
Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
"""
And a completion request with no api error
Then 64 tokens are predicted matching fun|Annaks|popcorns|pictry
And the completion is truncated
And 109 prompt tokens are processed
Scenario Outline: OAI Compatibility
Given a model <model>
@ -43,39 +60,14 @@ Feature: llama.cpp server
And streaming is <enable_streaming>
Given an OAI compatible chat completions request with no api error
Then <n_predicted> tokens are predicted matching <re_content>
And <n_prompt> prompt tokens are processed
And the completion is <truncated> truncated
Examples: Prompts
| model | system_prompt | user_prompt | max_tokens | re_content | n_predicted | enable_streaming |
| llama-2 | Book | What is the best book | 8 | (Mom\|what)+ | 8 | disabled |
| codellama70b | You are a coding assistant. | Write the fibonacci function in c++. | 64 | (thanks\|happy\|bird)+ | 32 | enabled |
| model | system_prompt | user_prompt | max_tokens | re_content | n_prompt | n_predicted | enable_streaming | truncated |
| llama-2 | Book | What is the best book | 8 | (Here\|what)+ | 77 | 8 | disabled | not |
| codellama70b | You are a coding assistant. | Write the fibonacci function in c++. | 128 | (thanks\|happy\|bird)+ | -1 | 64 | enabled | |
Scenario: Embedding
When embeddings are computed for:
"""
What is the capital of Bulgaria ?
"""
Then embeddings are generated
Scenario: OAI Embeddings compatibility
Given a model tinyllama-2
When an OAI compatible embeddings computation request for:
"""
What is the capital of Spain ?
"""
Then embeddings are generated
Scenario: OAI Embeddings compatibility with multiple inputs
Given a model tinyllama-2
Given a prompt:
"""
In which country Paris is located ?
"""
And a prompt:
"""
Is Madrid the capital of Spain ?
"""
When an OAI compatible embeddings computation request for multiple inputs
Then embeddings are generated
Scenario: Tokenize / Detokenize
When tokenizing:

291
examples/server/tests/features/steps/steps.py
View File

@ -10,6 +10,7 @@ from contextlib import closing
from re import RegexFlag
import aiohttp
import numpy as np
import openai
from behave import step
from behave.api.async_step import async_run_until_complete
@ -17,13 +18,16 @@ from huggingface_hub import hf_hub_download
from prometheus_client import parser
@step(u"a server listening on {server_fqdn}:{server_port}")
@step("a server listening on {server_fqdn}:{server_port}")
def step_server_config(context, server_fqdn, server_port):
context.server_fqdn = server_fqdn
context.server_port = int(server_port)
if 'PORT' in os.environ:
context.server_port = int(os.environ['PORT'])
print(f"$PORT set, overriding server port with to {context.server_port}")
if 'FQDN' in os.environ:
context.server_fqdn = os.environ['FQDN']
print(f"$FQDN set, overriding server fqdn with to {context.server_fqdn}")
context.base_url = f'http://{context.server_fqdn}:{context.server_port}'
@ -34,6 +38,7 @@ def step_server_config(context, server_fqdn, server_port):
context.n_ga_w = None
context.n_gpu_layer = None
context.n_predict = None
context.n_prompts = 0
context.n_server_predict = None
context.n_slots = None
context.prompt_prefix = None
@ -52,24 +57,24 @@ def step_server_config(context, server_fqdn, server_port):
context.prompts = []
@step(u'a model file {hf_file} from HF repo {hf_repo}')
@step('a model file {hf_file} from HF repo {hf_repo}')
def step_download_hf_model(context, hf_file, hf_repo):
context.model_file = hf_hub_download(repo_id=hf_repo, filename=hf_file)
if context.debug:
print(f"model file: {context.model_file}\n")
@step(u'a model alias {model_alias}')
@step('a model alias {model_alias}')
def step_model_alias(context, model_alias):
context.model_alias = model_alias
@step(u'{seed:d} as server seed')
@step('{seed:d} as server seed')
def step_seed(context, seed):
context.server_seed = seed
@step(u'{ngl:d} GPU offloaded layers')
@step('{ngl:d} GPU offloaded layers')
def step_n_gpu_layer(context, ngl):
if 'N_GPU_LAYERS' in os.environ:
new_ngl = int(os.environ['N_GPU_LAYERS'])
@ -79,37 +84,37 @@ def step_n_gpu_layer(context, ngl):
context.n_gpu_layer = ngl
@step(u'{n_ctx:d} KV cache size')
@step('{n_ctx:d} KV cache size')
def step_n_ctx(context, n_ctx):
context.n_ctx = n_ctx
@step(u'{n_slots:d} slots')
@step('{n_slots:d} slots')
def step_n_slots(context, n_slots):
context.n_slots = n_slots
@step(u'{n_predict:d} server max tokens to predict')
@step('{n_predict:d} server max tokens to predict')
def step_server_n_predict(context, n_predict):
context.n_server_predict = n_predict
@step(u'continuous batching')
@step('continuous batching')
def step_server_continuous_batching(context):
context.server_continuous_batching = True
@step(u'embeddings extraction')
@step('embeddings extraction')
def step_server_embeddings(context):
context.server_embeddings = True
@step(u'prometheus compatible metrics exposed')
@step('prometheus compatible metrics exposed')
def step_server_metrics(context):
context.server_metrics = True
@step(u"the server is starting")
@step("the server is starting")
def step_start_server(context):
start_server_background(context)
attempts = 0
@ -126,7 +131,7 @@ def step_start_server(context):
time.sleep(0.1)
@step(u"the server is {expecting_status}")
@step("the server is {expecting_status}")
@async_run_until_complete
async def step_wait_for_the_server_to_be_started(context, expecting_status):
match expecting_status:
@ -155,7 +160,7 @@ async def step_wait_for_the_server_to_be_started(context, expecting_status):
assert False, "unknown status"
@step(u'all slots are {expected_slot_status_string}')
@step('all slots are {expected_slot_status_string}')
@async_run_until_complete
async def step_all_slots_status(context, expected_slot_status_string):
match expected_slot_status_string:
@ -171,7 +176,7 @@ async def step_all_slots_status(context, expected_slot_status_string):
await request_slots_status(context, expected_slots)
@step(u'a completion request with {api_error} api error')
@step('a completion request with {api_error} api error')
@async_run_until_complete
async def step_request_completion(context, api_error):
expect_api_error = api_error == 'raised'
@ -189,108 +194,133 @@ async def step_request_completion(context, api_error):
assert completion == 401, f"completion must be an 401 status code: {completion}"
@step(u'{predicted_n:d} tokens are predicted matching {re_content}')
@step('{predicted_n:d} tokens are predicted matching {re_content}')
def step_n_tokens_predicted_with_content(context, predicted_n, re_content):
assert_n_tokens_predicted(context.tasks_result.pop(), predicted_n, re_content)
context.completion = context.tasks_result.pop()
assert_n_tokens_predicted(context.completion, predicted_n, re_content)
@step(u'{predicted_n:d} tokens are predicted')
@step('{predicted_n:d} tokens are predicted')
def step_n_tokens_predicted(context, predicted_n):
assert_n_tokens_predicted(context.tasks_result.pop(), predicted_n)
context.completion = context.tasks_result.pop()
assert_n_tokens_predicted(context.completion, predicted_n)
@step(u'a user prompt {user_prompt}')
@step('the completion is truncated')
def step_assert_completion_truncated(context):
step_assert_completion_truncated(context, '')
@step('the completion is {truncated} truncated')
def step_assert_completion_truncated(context, truncated):
truncated = truncated != "not"
assert context.completion['truncated'] == truncated, f'{context.completion}'
@step('{n_prompt:d} prompt tokens are processed')
def step_impl(context, n_prompt):
assert n_prompt < 0 or n_prompt == context.completion['timings']['prompt_n'], f"n_prompt={context.completion['timings']['prompt_n']}"
@step('a user prompt {user_prompt}')
def step_user_prompt(context, user_prompt):
context.prompts.append(user_prompt)
context.n_prompts = len(context.prompts)
@step(u'a system prompt {system_prompt}')
@step('a system prompt {system_prompt}')
def step_system_prompt(context, system_prompt):
context.system_prompt = system_prompt
@step(u'a model {model}')
@step('a model {model}')
def step_model(context, model):
context.model = model
@step(u'{max_tokens:d} max tokens to predict')
@step('{max_tokens:d} max tokens to predict')
def step_max_tokens(context, max_tokens):
context.n_predict = max_tokens
@step(u'streaming is {enable_streaming}')
@step('streaming is {enable_streaming}')
def step_streaming(context, enable_streaming):
context.enable_streaming = enable_streaming == 'enabled'
@step(u'a user api key {user_api_key}')
@step('a user api key {user_api_key}')
def step_user_api_key(context, user_api_key):
context.user_api_key = user_api_key
@step(u'no user api key')
@step('no user api key')
def step_no_user_api_key(context):
context.user_api_key = None
@step(u'a user api key ')
@step('a user api key ')
def step_no_user_api_key_space(context):
context.user_api_key = None
@step(u'a server api key {server_api_key}')
@step('a server api key {server_api_key}')
def step_server_api_key(context, server_api_key):
context.server_api_key = server_api_key
@step(u'{n_junk:d} as number of junk')
@step('{n_junk:d} as number of junk')
def step_n_junk(context, n_junk):
context.n_junk = n_junk
@step(u'{n_batch:d} as batch size')
@step('{n_batch:d} as batch size')
def step_n_batch(context, n_batch):
context.n_batch = n_batch
@step(u'{seed:d} as seed')
@step('{seed:d} as seed')
def step_seed(context, seed):
context.seed = seed
@step(u'a prefix prompt')
@step('a prefix prompt')
def step_prompt_prefix(context):
context.prompt_prefix = context.text
context.prompt_prefix = context_text(context)
@step(u'a junk suffix prompt')
@step('a junk suffix prompt')
def step_prompt_junk_suffix(context):
context.prompt_junk_suffix = context.text
context.prompt_junk_suffix = context_text(context)
@step(u'a suffix prompt')
@step('a suffix prompt')
def step_prompt_suffix(context):
context.prompt_suffix = context.text
context.prompt_suffix = context_text(context)
@step(u'{n_ga:d} group attention factor'
u' to extend context size through self-extend')
@step('{n_ga:d} group attention factor'
' to extend context size through self-extend')
def step_impl(context, n_ga):
context.n_ga = n_ga
@step(u'{n_ga_w:d} group attention width to extend context size through self-extend')
@step('{n_ga_w:d} group attention width to extend context size through self-extend')
def step_impl(context, n_ga_w):
context.n_ga_w = n_ga_w
@step(u'a passkey prompt template')
@step('a passkey prompt template')
def step_prompt_passkey(context):
context.prompt_passkey = context.text
context.prompt_passkey = context_text(context)
@step(u'a "{passkey}" passkey challenge prompt with the passkey inserted every {i_pos:d} junk')
@step('{n_prompts:d} fixed prompts')
def step_fixed_prompts(context, n_prompts):
context.prompts.extend([str(0)*(context.n_batch if context.n_batch is not None else 512) for i in range(n_prompts)])
context.n_prompts = n_prompts
@step('a "{passkey}" passkey challenge prompt with the passkey inserted every {i_pos:d} junk')
def step_prompt_passkey(context, passkey, i_pos):
prompt = ""
for i in range(context.n_junk):
@ -301,9 +331,10 @@ def step_prompt_passkey(context, passkey, i_pos):
passkey_highlight = "\x1b[33m" + passkey + "\x1b[0m"
print(f"Passkey challenge:\n```{prompt.replace(passkey, passkey_highlight)}```\n")
context.prompts.append(context.prompt_prefix + prompt + context.prompt_suffix)
context.n_prompts = len(context.prompts)
@step(u'an OAI compatible chat completions request with {api_error} api error')
@step('an OAI compatible chat completions request with {api_error} api error')
@async_run_until_complete
async def step_oai_chat_completions(context, api_error):
if context.debug:
@ -338,17 +369,19 @@ async def step_oai_chat_completions(context, api_error):
print(f"Completion response: {completion}")
@step(u'a prompt')
@step('a prompt')
def step_a_prompt(context):
context.prompts.append(context.text)
context.prompts.append(context_text(context))
context.n_prompts = len(context.prompts)
@step(u'a prompt {prompt}')
@step('a prompt {prompt}')
def step_a_prompt_prompt(context, prompt):
context.prompts.append(prompt)
context.n_prompts = len(context.prompts)
@step(u'concurrent completion requests')
@step('concurrent completion requests')
@async_run_until_complete()
async def step_concurrent_completion_requests(context):
await concurrent_requests(context,
@ -364,7 +397,7 @@ async def step_concurrent_completion_requests(context):
'user_api_key') else None)
@step(u'concurrent OAI completions requests')
@step('concurrent OAI completions requests')
@async_run_until_complete
async def step_oai_chat_completions(context):
await concurrent_requests(context, oai_chat_completions,
@ -384,7 +417,7 @@ async def step_oai_chat_completions(context):
if hasattr(context, 'user_api_key') else None)
@step(u'concurrent OAI completions requests no v1')
@step('concurrent OAI completions requests no v1')
@async_run_until_complete
async def step_oai_chat_completions(context):
await concurrent_requests(context, oai_chat_completions,
@ -407,13 +440,13 @@ async def step_oai_chat_completions(context):
if hasattr(context, 'user_api_key') else None)
@step(u'all prompts are predicted')
@step('all prompts are predicted')
@async_run_until_complete
async def step_all_prompts_are_predicted(context):
await all_prompts_are_predicted(context)
@step(u'all prompts are predicted with {n_expected_predicted:d} tokens')
@step('all prompts are predicted with {n_expected_predicted:d} tokens')
@async_run_until_complete
async def step_all_prompts_are_predicted_with_n_tokens(context, n_expected_predicted):
await all_prompts_are_predicted(context, n_expected_predicted)
@ -427,44 +460,68 @@ async def all_prompts_are_predicted(context, expected_predicted_n=None):
assert len(context.concurrent_tasks) == 0, f"{len(context.concurrent_tasks)} pending requests"
@step(u'embeddings are computed for')
@step('embeddings are computed for')
@async_run_until_complete
async def step_compute_embedding(context):
context.embeddings = await request_embedding(context.text, base_url=context.base_url)
context.n_prompts = 1
context.embeddings = await request_embedding(context_text(context), base_url=context.base_url)
@step(u'embeddings are generated')
@step('all embeddings are the same')
@async_run_until_complete
async def step_all_embeddings_are_the_same(context):
n_embedding_requests = await gather_tasks_results(context)
assert n_embedding_requests > 0
embeddings = []
for i in range(n_embedding_requests):
embedding = context.tasks_result.pop().pop()
embeddings.append(embedding)
assert_embeddings(embedding)
n = len(embeddings)
for i in range(n-1):
for j in range(i+1, n):
embedding1 = np.array(embeddings[i])
embedding2 = np.array(embeddings[j])
if context.debug:
print(f"embedding1: {embedding1[-8:]}\n")
print(f"embedding2: {embedding2[-8:]}\n")
similarity = np.dot(embedding1, embedding2) / (np.linalg.norm(embedding1) * np.linalg.norm(embedding2))
msg = f"Similarity between {i} and {j}: {similarity:.10f}"
if context.debug:
print(f"{msg}\n")
assert np.isclose(similarity, 1.0, rtol=1e-05, atol=1e-08, equal_nan=False), msg
@step('embeddings are generated')
def step_assert_embeddings(context):
if len(context.prompts) == 0:
assert_embeddings(context.embeddings)
else:
assert len(context.embeddings) == len(context.prompts), (f"unexpected response:\n"
f"context.prompts={context.prompts}\n"
assert context.n_prompts == len(context.embeddings), (f"unexpected response:\n"
f"context.n_prompts={context.n_prompts}\n"
f"context.embeddings={context.embeddings}")
for embedding in context.embeddings:
context.prompts.pop()
assert_embeddings(embedding)
@step(u'an OAI compatible embeddings computation request for')
@step('an OAI compatible embeddings computation request for')
@async_run_until_complete
async def step_oai_compute_embeddings(context):
context.embeddings = await request_oai_embeddings(context.text,
context.n_prompts = 1
context.embeddings = await request_oai_embeddings(context_text(context),
base_url=context.base_url,
user_api_key=context.user_api_key,
model=context.model)
@step(u'an OAI compatible embeddings computation request for multiple inputs')
@step('an OAI compatible embeddings computation request for multiple inputs')
@async_run_until_complete
async def step_oai_compute_embeddings_multiple_inputs(context):
context.embeddings = await request_oai_embeddings(context.prompts,
base_url=context.base_url,
user_api_key=context.user_api_key,
model=context.model)
context.prompts.clear()
@step(u'concurrent embedding requests')
@step('concurrent embedding requests')
@async_run_until_complete()
async def step_concurrent_embedding_requests(context):
await concurrent_requests(context,
@ -473,7 +530,7 @@ async def step_concurrent_embedding_requests(context):
base_url=context.base_url)
@step(u'concurrent OAI embedding requests')
@step('concurrent OAI embedding requests')
@async_run_until_complete()
async def step_concurrent_oai_embedding_requests(context):
await concurrent_requests(context,
@ -484,19 +541,19 @@ async def step_concurrent_oai_embedding_requests(context):
model=context.model)
@step(u'all embeddings are generated')
@step('all embeddings are generated')
@async_run_until_complete()
async def all_embeddings_are_generated(context):
n_embedding_requests = await gather_tasks_results(context)
assert n_embedding_requests > 0
assert n_embedding_requests == context.n_prompts
for i in range(n_embedding_requests):
assert_embeddings(context.tasks_result.pop())
assert_embeddings(context.tasks_result.pop().pop())
@step(u'tokenizing')
@step('tokenizing')
@async_run_until_complete
async def step_tokenize(context):
context.tokenized_text = context.text
context.tokenized_text = context_text(context)
async with aiohttp.ClientSession() as session:
async with session.post(f'{context.base_url}/tokenize',
json={
@ -507,7 +564,7 @@ async def step_tokenize(context):
context.tokens = tokenize_json['tokens']
@step(u'tokens can be detokenize')
@step('tokens can be detokenize')
@async_run_until_complete
async def step_detokenize(context):
assert len(context.tokens) > 0
@ -522,22 +579,23 @@ async def step_detokenize(context):
assert context.tokenized_text == detokenize_json['content'].strip()
@step(u'an OPTIONS request is sent from {origin}')
@step('an OPTIONS request is sent from {origin}')
@async_run_until_complete
async def step_options_request(context, origin):
async with aiohttp.ClientSession() as session:
headers = {'Authorization': f'Bearer {context.user_api_key}', 'Origin': origin}
async with session.options(f'{context.base_url}/v1/chat/completions',
headers={"Origin": origin}) as response:
headers=headers) as response:
assert response.status == 200
context.options_response = response
@step(u'CORS header {cors_header} is set to {cors_header_value}')
@step('CORS header {cors_header} is set to {cors_header_value}')
def step_check_options_header_value(context, cors_header, cors_header_value):
assert context.options_response.headers[cors_header] == cors_header_value
@step(u'prometheus metrics are exposed')
@step('prometheus metrics are exposed')
@async_run_until_complete
async def step_prometheus_metrics_exported(context):
async with aiohttp.ClientSession() as session:
@ -548,15 +606,25 @@ async def step_prometheus_metrics_exported(context):
metric_exported = False
if context.debug:
print(f"/metrics answer:\n{metrics_raw}\n")
context.metrics = {}
for metric in parser.text_string_to_metric_families(metrics_raw):
match metric.name:
case "llamacpp:kv_cache_usage_ratio":
assert len(metric.samples) > 0
metric_exported = True
context.metrics[metric.name] = metric
assert int(metrics_response.headers["Process-Start-Time-Unix"]) > 0, "no header process start time"
assert metric_exported, "No metrics exported"
@step(u'available models')
@step('metric {metric_name} is {metric_value:d}')
def step_assert_metric_value(context, metric_name, metric_value):
if metric_name not in context.metrics:
assert False, f"no metric {metric_name} in {context.metrics.keys()}"
assert context.metrics[metric_name].samples[0].value == metric_value, f"metric: {context.metrics[metric_name]}"
@step('available models')
def step_available_models(context):
# openai client always expects an api_key
openai.api_key = context.user_api_key if context.user_api_key is not None else 'nope'
@ -564,14 +632,14 @@ def step_available_models(context):
context.models = openai.Model.list().data
@step(u'{n_model:d} models are supported')
@step('{n_model:d} models are supported')
def step_supported_models(context, n_model):
if context.debug:
print("server models available:", context.models)
assert len(context.models) == n_model
@step(u'model {i_model:d} is {param} {preposition} {param_value}')
@step('model {i_model:d} is {param} {preposition} {param_value}')
def step_supported_models(context, i_model, param, preposition, param_value):
assert i_model < len(context.models)
model = context.models[i_model]
@ -588,11 +656,11 @@ def step_supported_models(context, i_model, param, preposition, param_value):
async def concurrent_requests(context, f_completion, *args, **kwargs):
n_prompts = len(context.prompts)
context.n_prompts = len(context.prompts)
if context.debug:
print(f"starting {n_prompts} concurrent completion requests...")
assert n_prompts > 0
for prompt_no in range(n_prompts):
print(f"starting {context.n_prompts} concurrent completion requests...")
assert context.n_prompts > 0
for prompt_no in range(context.n_prompts):
shifted_args = [context.prompts.pop(), *args]
context.concurrent_tasks.append(asyncio.create_task(f_completion(*shifted_args, **kwargs)))
await asyncio.sleep(0.1)
@ -674,7 +742,8 @@ async def oai_chat_completions(user_prompt,
completion_response = {
'content': '',
'timings': {
'predicted_n': 0
'predicted_n': 0,
'prompt_n': 0
}
}
if async_client:
@ -715,7 +784,8 @@ async def oai_chat_completions(user_prompt,
completion_response = {
'content': chat_completion_raw['choices'][0]['message'],
'timings': {
'predicted_n': chat_completion_raw['usage']['completion_tokens']
'predicted_n': chat_completion_raw['usage']['completion_tokens'],
'prompt_n': chat_completion_raw['usage']['prompt_tokens']
}
}
else:
@ -744,13 +814,16 @@ async def oai_chat_completions(user_prompt,
if 'content' in delta:
completion_response['content'] += delta['content']
completion_response['timings']['predicted_n'] += 1
completion_response['truncated'] = chunk.choices[0].finish_reason != 'stop'
else:
assert len(chat_completion.choices) == 1
completion_response = {
'content': chat_completion.choices[0].message.content,
'timings': {
'predicted_n': chat_completion.usage.completion_tokens
}
'predicted_n': chat_completion.usage.completion_tokens,
'prompt_n': chat_completion.usage.prompt_tokens
},
'truncated': chat_completion.choices[0].finish_reason != 'stop'
}
if debug:
print("OAI response formatted to llama.cpp:", completion_response)
@ -765,7 +838,7 @@ async def request_embedding(content, base_url=None):
}) as response:
assert response.status == 200
response_json = await response.json()
return response_json['embedding']
return [response_json['embedding']]
async def request_oai_embeddings(input,
@ -775,6 +848,7 @@ async def request_oai_embeddings(input,
user_api_key = user_api_key if user_api_key is not None else 'nope'
if async_client:
origin = 'llama.cpp'
headers=[]
if user_api_key is not None:
headers = {'Authorization': f'Bearer {user_api_key}', 'Origin': origin}
async with aiohttp.ClientSession() as session:
@ -783,14 +857,21 @@ async def request_oai_embeddings(input,
"input": input,
"model": model,
},
headers=headers) as response:
headers=headers,
timeout=3600) as response:
assert response.status == 200, f"received status code not expected: {response.status}"
assert response.headers['Access-Control-Allow-Origin'] == origin
assert response.headers['Content-Type'] == "application/json; charset=utf-8"
response_json = await response.json()
assert response_json['model'] == model, f"invalid model received: {response_json['model']}"
assert response_json['object'] == 'list'
return response_json['data']
if isinstance(input, collections.abc.Sequence):
embeddings = []
for an_oai_embeddings in response_json['data']:
embeddings.append(an_oai_embeddings['embedding'])
else:
embeddings = [response_json['data']['embedding']]
return embeddings
else:
openai.api_key = user_api_key
openai.api_base = f'{base_url}/v1'
@ -804,7 +885,7 @@ async def request_oai_embeddings(input,
for an_oai_embeddings in oai_embeddings.data:
embeddings.append(an_oai_embeddings.embedding)
else:
embeddings = oai_embeddings.data.embedding
embeddings = [oai_embeddings.data.embedding]
return embeddings
@ -833,7 +914,6 @@ def assert_n_tokens_predicted(completion_response, expected_predicted_n=None, re
f' {n_predicted} <> {expected_predicted_n}')
async def gather_tasks_results(context):
n_tasks = len(context.concurrent_tasks)
if context.debug:
@ -899,6 +979,8 @@ def assert_embeddings(embeddings):
assert len(embeddings) > 0
embeddings_computed = False
for emb in embeddings:
if not isinstance(emb, float):
assert False, f"Bad embeddings: {embeddings}"
if emb != 0:
embeddings_computed = True
assert embeddings_computed, f"Embeddings: {embeddings}"
@ -926,12 +1008,22 @@ async def completions_seed(context):
else context.server_seed if hasattr(context, 'server_seed') else None
def context_text(context):
return context.text.replace('\r', '')
def start_server_background(context):
if os.name == 'nt':
context.server_path = '../../../build/bin/Release/server.exe'
else:
context.server_path = '../../../build/bin/server'
if 'LLAMA_SERVER_BIN_PATH' in os.environ:
context.server_path = os.environ['LLAMA_SERVER_BIN_PATH']
server_listen_addr = context.server_fqdn
if os.name == 'nt':
server_listen_addr = '0.0.0.0'
server_args = [
'--host', context.server_fqdn,
'--host', server_listen_addr,
'--port', context.server_port,
'--model', context.model_file
]
@ -964,7 +1056,16 @@ def start_server_background(context):
if 'SERVER_LOG_FORMAT_JSON' not in os.environ:
server_args.extend(['--log-format', "text"])
print(f"starting server with: {context.server_path} {server_args}\n")
flags = 0
if 'nt' == os.name:
flags |= subprocess.DETACHED_PROCESS
flags |= subprocess.CREATE_NEW_PROCESS_GROUP
flags |= subprocess.CREATE_NO_WINDOW
pkwargs = {
'creationflags': flags,
}
context.server_process = subprocess.Popen(
[str(arg) for arg in [context.server_path, *server_args]],
close_fds=True)
print(f"server pid={context.server_process.pid}")
**pkwargs)
print(f"server pid={context.server_process.pid}, behave pid={os.getpid()}")

1
examples/server/tests/requirements.txt
View File

@ -1,5 +1,6 @@
aiohttp~=3.9.3
behave~=1.2.6
huggingface_hub~=0.20.3
numpy~=1.24.4
openai~=0.25.0
prometheus-client~=0.20.0

691
examples/server/utils.hpp
View File

@ -1,15 +1,16 @@
#pragma once
#include <string>
#include <vector>
#include <set>
#include <mutex>
#include <condition_variable>
#include <unordered_map>
#include "llama.h"
#include "common.h"
#include "json.hpp"
#include "../llava/clip.h"
#include <string>
#include <vector>
#include <sstream>
#include <random>
#define DEFAULT_OAICOMPAT_MODEL "gpt-3.5-turbo-0613"
using json = nlohmann::json;
@ -37,61 +38,13 @@ extern bool server_log_json;
#define LOG_WARNING(MSG, ...) server_log("WARN", __func__, __LINE__, MSG, __VA_ARGS__)
#define LOG_INFO( MSG, ...) server_log("INFO", __func__, __LINE__, MSG, __VA_ARGS__)
enum server_state {
SERVER_STATE_LOADING_MODEL, // Server is starting up, model not fully loaded yet
SERVER_STATE_READY, // Server is ready and model is loaded
SERVER_STATE_ERROR // An error occurred, load_model failed
};
enum task_type {
TASK_TYPE_COMPLETION,
TASK_TYPE_CANCEL,
TASK_TYPE_NEXT_RESPONSE,
TASK_TYPE_METRICS
};
struct task_server {
int id = -1; // to be filled by llama_server_queue
int target_id;
task_type type;
json data;
bool infill_mode = false;
bool embedding_mode = false;
int multitask_id = -1;
};
struct task_result {
int id;
int multitask_id = -1;
bool stop;
bool error;
json result_json;
};
struct task_multi {
int id;
std::set<int> subtasks_remaining{};
std::vector<task_result> results{};
};
// completion token output with probabilities
struct completion_token_output {
struct token_prob
{
llama_token tok;
float prob;
};
std::vector<token_prob> probs;
llama_token tok;
std::string text_to_send;
};
struct token_translator {
llama_context * ctx;
std::string operator()(llama_token tok) const { return llama_token_to_piece(ctx, tok); }
std::string operator()(const completion_token_output &cto) const { return (*this)(cto.tok); }
};
template <typename T>
static T json_value(const json &body, const std::string &key, const T &default_value) {
// Fallback null to default value
return body.contains(key) && !body.at(key).is_null()
? body.value(key, default_value)
: default_value;
}
static inline void server_log(const char *level, const char *function, int line, const char *message, const nlohmann::ordered_json &extra) {
std::stringstream ss_tid;
@ -102,18 +55,18 @@ static inline void server_log(const char *level, const char *function, int line,
};
if (server_log_json) {
log.merge_patch(
{
log.merge_patch( {
{"level", level},
{"function", function},
{"line", line},
{"msg", message},
});
if (!extra.empty()) {
log.merge_patch(extra);
}
std::cout << log.dump(-1, ' ', false, json::error_handler_t::replace) << "\n" << std::flush;
printf("%s\n", log.dump(-1, ' ', false, json::error_handler_t::replace).c_str());
} else {
char buf[1024];
snprintf(buf, 1024, "%4s [%24s] %s", level, function, message);
@ -136,22 +89,13 @@ static inline void server_log(const char *level, const char *function, int line,
}
//
// server utils
// chat template utils
//
template <typename T>
static T json_value(const json &body, const std::string &key, const T &default_value) {
// Fallback null to default value
return body.contains(key) && !body.at(key).is_null()
? body.value(key, default_value)
: default_value;
}
// Check if the template supplied via "--chat-template" is supported or not. Returns true if it's valid
inline bool verify_custom_template(const std::string & tmpl) {
llama_chat_message chat[] = {{"user", "test"}};
std::vector<char> buf(1);
int res = llama_chat_apply_template(nullptr, tmpl.c_str(), chat, 1, true, buf.data(), buf.size());
int res = llama_chat_apply_template(nullptr, tmpl.c_str(), chat, 1, true, nullptr, 0);
return res >= 0;
}
@ -163,7 +107,7 @@ inline std::string format_chat(const struct llama_model * model, const std::stri
std::vector<llama_chat_message> chat(messages.size());
for (size_t i = 0; i < messages.size(); ++i) {
auto &curr_msg = messages[i];
const auto & curr_msg = messages[i];
str[i*2 + 0] = json_value(curr_msg, "role", std::string(""));
str[i*2 + 1] = json_value(curr_msg, "content", std::string(""));
alloc_size += str[i*2 + 1].length();
@ -183,261 +127,13 @@ inline std::string format_chat(const struct llama_model * model, const std::stri
res = llama_chat_apply_template(model, ptr_tmpl, chat.data(), chat.size(), true, buf.data(), buf.size());
}
std::string formatted_chat(buf.data(), res);
const std::string formatted_chat(buf.data(), res);
LOG_VERBOSE("formatted_chat", {{"text", formatted_chat.c_str()}});
return formatted_chat;
}
//
// work queue utils
//
struct llama_server_queue {
int id = 0;
std::mutex mutex_tasks;
bool running;
// queues
std::vector<task_server> queue_tasks;
std::vector<task_server> queue_tasks_deferred;
std::vector<task_multi> queue_multitasks;
std::condition_variable condition_tasks;
// callback functions
std::function<void(task_server&)> callback_new_task;
std::function<void(task_multi&)> callback_finish_multitask;
std::function<void(void)> callback_run_slots;
// Add a new task to the end of the queue
int post(task_server task) {
std::unique_lock<std::mutex> lock(mutex_tasks);
if (task.id == -1) {
task.id = id++;
LOG_VERBOSE("new task id", {{"new_id", task.id}});
}
queue_tasks.push_back(std::move(task));
condition_tasks.notify_one();
return task.id;
}
// Add a new task, but defer until one slot is available
void defer(task_server task) {
std::unique_lock<std::mutex> lock(mutex_tasks);
queue_tasks_deferred.push_back(std::move(task));
}
// Get the next id for creating anew task
int get_new_id() {
std::unique_lock<std::mutex> lock(mutex_tasks);
int new_id = id++;
LOG_VERBOSE("new task id", {{"new_id", new_id}});
return new_id;
}
// Register function to process a new task
void on_new_task(std::function<void(task_server&)> callback) {
callback_new_task = callback;
}
// Register function to process a multitask when it is finished
void on_finish_multitask(std::function<void(task_multi&)> callback) {
callback_finish_multitask = callback;
}
// Register the function to be called when all slots data is ready to be processed
void on_run_slots(std::function<void(void)> callback) {
callback_run_slots = callback;
}
// Call when the state of one slot is changed
void notify_slot_changed() {
// move deferred tasks back to main loop
std::unique_lock<std::mutex> lock(mutex_tasks);
for (auto & task : queue_tasks_deferred) {
queue_tasks.push_back(std::move(task));
}
queue_tasks_deferred.clear();
}
// end the start_loop routine
void terminate() {
{
std::unique_lock<std::mutex> lock(mutex_tasks);
running = false;
}
condition_tasks.notify_all();
}
/**
* Main loop consists of these steps:
* - Wait until a new task arrives
* - Process the task (i.e. maybe copy data into slot)
* - Check if multitask is finished
* - Run all slots
*/
void start_loop() {
running = true;
while (true) {
LOG_VERBOSE("new task may arrive", {});
{
while (true)
{
std::unique_lock<std::mutex> lock(mutex_tasks);
if (queue_tasks.empty()) {
lock.unlock();
break;
}
task_server task = queue_tasks.front();
queue_tasks.erase(queue_tasks.begin());
lock.unlock();
LOG_VERBOSE("callback_new_task", {{"task_id", task.id}});
callback_new_task(task);
}
LOG_VERBOSE("update_multitasks", {});
// check if we have any finished multitasks
auto queue_iterator = queue_multitasks.begin();
while (queue_iterator != queue_multitasks.end())
{
if (queue_iterator->subtasks_remaining.empty())
{
// all subtasks done == multitask is done
task_multi current_multitask = *queue_iterator;
callback_finish_multitask(current_multitask);
// remove this multitask
queue_iterator = queue_multitasks.erase(queue_iterator);
}
else
{
++queue_iterator;
}
}
// all tasks in the current loop is processed, slots data is now ready
LOG_VERBOSE("callback_run_slots", {});
callback_run_slots();
}
LOG_VERBOSE("wait for new task", {});
// wait for new task
{
std::unique_lock<std::mutex> lock(mutex_tasks);
if (queue_tasks.empty()) {
if (!running) {
LOG_VERBOSE("ending start_loop", {});
return;
}
condition_tasks.wait(lock, [&]{
return (!queue_tasks.empty() || !running);
});
}
}
}
}
//
// functions to manage multitasks
//
// add a multitask by specifying the id of all subtask (subtask is a task_server)
void add_multitask(int multitask_id, std::vector<int>& sub_ids)
{
std::lock_guard<std::mutex> lock(mutex_tasks);
task_multi multi;
multi.id = multitask_id;
std::copy(sub_ids.begin(), sub_ids.end(), std::inserter(multi.subtasks_remaining, multi.subtasks_remaining.end()));
queue_multitasks.push_back(multi);
}
// updatethe remaining subtasks, while appending results to multitask
void update_multitask(int multitask_id, int subtask_id, task_result& result)
{
std::lock_guard<std::mutex> lock(mutex_tasks);
for (auto& multitask : queue_multitasks)
{
if (multitask.id == multitask_id)
{
multitask.subtasks_remaining.erase(subtask_id);
multitask.results.push_back(result);
}
}
}
};
struct llama_server_response {
typedef std::function<void(int, int, task_result&)> callback_multitask_t;
callback_multitask_t callback_update_multitask;
// for keeping track of all tasks waiting for the result
std::set<int> waiting_task_ids;
// the main result queue
std::vector<task_result> queue_results;
std::mutex mutex_results;
std::condition_variable condition_results;
// add the task_id to the list of tasks waiting for response
void add_waiting_task_id(int task_id) {
LOG_VERBOSE("waiting for task id", {{"task_id", task_id}});
std::unique_lock<std::mutex> lock(mutex_results);
waiting_task_ids.insert(task_id);
}
// when the request is finished, we can remove task associated with it
void remove_waiting_task_id(int task_id) {
LOG_VERBOSE("remove waiting for task id", {{"task_id", task_id}});
std::unique_lock<std::mutex> lock(mutex_results);
waiting_task_ids.erase(task_id);
}
// This function blocks the thread until there is a response for this task_id
task_result recv(int task_id) {
while (true)
{
std::unique_lock<std::mutex> lock(mutex_results);
condition_results.wait(lock, [&]{
return !queue_results.empty();
});
for (int i = 0; i < (int) queue_results.size(); i++)
{
if (queue_results[i].id == task_id)
{
assert(queue_results[i].multitask_id == -1);
task_result res = queue_results[i];
queue_results.erase(queue_results.begin() + i);
return res;
}
}
}
// should never reach here
}
// Register the function to update multitask
void on_multitask_update(callback_multitask_t callback) {
callback_update_multitask = callback;
}
// Send a new result to a waiting task_id
void send(task_result result) {
std::unique_lock<std::mutex> lock(mutex_results);
LOG_VERBOSE("send new result", {{"task_id", result.id}});
for (auto& task_id : waiting_task_ids) {
// LOG_TEE("waiting task id %i \n", task_id);
// for now, tasks that have associated parent multitasks just get erased once multitask picks up the result
if (result.multitask_id == task_id)
{
LOG_VERBOSE("callback_update_multitask", {{"task_id", task_id}});
callback_update_multitask(task_id, result.id, result);
continue;
}
if (result.id == task_id)
{
LOG_VERBOSE("queue_results.push_back", {{"task_id", task_id}});
queue_results.push_back(result);
condition_results.notify_all();
return;
}
}
}
};
//
// base64 utils (TODO: move to common in the future)
//
@ -447,13 +143,11 @@ static const std::string base64_chars =
"abcdefghijklmnopqrstuvwxyz"
"0123456789+/";
static inline bool is_base64(uint8_t c)
{
static inline bool is_base64(uint8_t c) {
return (isalnum(c) || (c == '+') || (c == '/'));
}
static inline std::vector<uint8_t> base64_decode(const std::string & encoded_string)
{
static inline std::vector<uint8_t> base64_decode(const std::string & encoded_string) {
int i = 0;
int j = 0;
int in_ = 0;
@ -465,13 +159,10 @@ static inline std::vector<uint8_t> base64_decode(const std::string & encoded_str
std::vector<uint8_t> ret;
while (in_len-- && (encoded_string[in_] != '=') && is_base64(encoded_string[in_]))
{
while (in_len-- && (encoded_string[in_] != '=') && is_base64(encoded_string[in_])) {
char_array_4[i++] = encoded_string[in_]; in_++;
if (i == 4)
{
for (i = 0; i <4; i++)
{
if (i == 4) {
for (i = 0; i < 4; i++) {
char_array_4[i] = base64_chars.find(char_array_4[i]);
}
@ -479,23 +170,20 @@ static inline std::vector<uint8_t> base64_decode(const std::string & encoded_str
char_array_3[1] = ((char_array_4[1] & 0xf) << 4) + ((char_array_4[2] & 0x3c) >> 2);
char_array_3[2] = ((char_array_4[2] & 0x3) << 6) + char_array_4[3];
for (i = 0; (i < 3); i++)
{
for (i = 0; (i < 3); i++) {
ret.push_back(char_array_3[i]);
}
i = 0;
}
}
if (i)
{
for (j = i; j <4; j++)
{
if (i) {
for (j = i; j < 4; j++) {
char_array_4[j] = 0;
}
for (j = 0; j <4; j++)
{
for (j = 0; j < 4; j++) {
char_array_4[j] = base64_chars.find(char_array_4[j]);
}
@ -503,8 +191,7 @@ static inline std::vector<uint8_t> base64_decode(const std::string & encoded_str
char_array_3[1] = ((char_array_4[1] & 0xf) << 4) + ((char_array_4[2] & 0x3c) >> 2);
char_array_3[2] = ((char_array_4[2] & 0x3) << 6) + char_array_4[3];
for (j = 0; (j < i - 1); j++)
{
for (j = 0; j < i - 1; j++) {
ret.push_back(char_array_3[j]);
}
}
@ -516,8 +203,7 @@ static inline std::vector<uint8_t> base64_decode(const std::string & encoded_str
// random string / id
//
static std::string random_string()
{
static std::string random_string() {
static const std::string str("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz");
std::random_device rd;
@ -532,10 +218,10 @@ static std::string random_string()
return result;
}
static std::string gen_chatcmplid()
{
static std::string gen_chatcmplid() {
std::stringstream chatcmplid;
chatcmplid << "chatcmpl-" << random_string();
return chatcmplid.str();
}
@ -543,91 +229,316 @@ static std::string gen_chatcmplid()
// other common utils
//
static size_t common_part(const std::vector<llama_token> &a, const std::vector<llama_token> &b)
{
static size_t common_part(const std::vector<llama_token> & a, const std::vector<llama_token> & b) {
size_t i;
for (i = 0; i < a.size() && i < b.size() && a[i] == b[i]; i++)
{
}
for (i = 0; i < a.size() && i < b.size() && a[i] == b[i]; i++) {}
return i;
}
static bool ends_with(const std::string &str, const std::string &suffix)
{
return str.size() >= suffix.size() &&
0 == str.compare(str.size() - suffix.size(), suffix.size(), suffix);
static bool ends_with(const std::string & str, const std::string & suffix) {
return str.size() >= suffix.size() && 0 == str.compare(str.size() - suffix.size(), suffix.size(), suffix);
}
static size_t find_partial_stop_string(const std::string &stop,
const std::string &text)
{
if (!text.empty() && !stop.empty())
{
static size_t find_partial_stop_string(const std::string &stop, const std::string &text) {
if (!text.empty() && !stop.empty()) {
const char text_last_char = text.back();
for (int64_t char_index = stop.size() - 1; char_index >= 0; char_index--)
{
if (stop[char_index] == text_last_char)
{
for (int64_t char_index = stop.size() - 1; char_index >= 0; char_index--) {
if (stop[char_index] == text_last_char) {
const std::string current_partial = stop.substr(0, char_index + 1);
if (ends_with(text, current_partial))
{
if (ends_with(text, current_partial)) {
return text.size() - char_index - 1;
}
}
}
}
return std::string::npos;
}
// TODO: reuse llama_detokenize
template <class Iter>
static std::string tokens_to_str(llama_context *ctx, Iter begin, Iter end)
{
static std::string tokens_to_str(llama_context * ctx, Iter begin, Iter end) {
std::string ret;
for (; begin != end; ++begin)
{
for (; begin != end; ++begin) {
ret += llama_token_to_piece(ctx, *begin);
}
return ret;
}
// format incomplete utf-8 multibyte character for output
static std::string tokens_to_output_formatted_string(const llama_context *ctx, const llama_token token)
{
static std::string tokens_to_output_formatted_string(const llama_context * ctx, const llama_token token) {
std::string out = token == -1 ? "" : llama_token_to_piece(ctx, token);
// if the size is 1 and first bit is 1, meaning it's a partial character
// (size > 1 meaning it's already a known token)
if (out.size() == 1 && (out[0] & 0x80) == 0x80)
{
if (out.size() == 1 && (out[0] & 0x80) == 0x80) {
std::stringstream ss;
ss << std::hex << (out[0] & 0xff);
std::string res(ss.str());
out = "byte: \\x" + res;
}
return out;
}
struct completion_token_output {
llama_token tok;
std::string text_to_send;
struct token_prob {
llama_token tok;
float prob;
};
std::vector<token_prob> probs;
};
// convert a vector of completion_token_output to json
static json probs_vector_to_json(const llama_context *ctx, const std::vector<completion_token_output> &probs)
{
static json probs_vector_to_json(const llama_context * ctx, const std::vector<completion_token_output> & probs) {
json out = json::array();
for (const auto &prob : probs)
{
for (const auto & prob : probs) {
json probs_for_token = json::array();
for (const auto &p : prob.probs)
{
std::string tok_str = tokens_to_output_formatted_string(ctx, p.tok);
probs_for_token.push_back(json
{
for (const auto & p : prob.probs) {
const std::string tok_str = tokens_to_output_formatted_string(ctx, p.tok);
probs_for_token.push_back(json {
{"tok_str", tok_str},
{"prob", p.prob},
});
}
std::string tok_str = tokens_to_output_formatted_string(ctx, prob.tok);
out.push_back(json{
const std::string tok_str = tokens_to_output_formatted_string(ctx, prob.tok);
out.push_back(json {
{"content", tok_str},
{"probs", probs_for_token},
});
}
return out;
}
//
// OAI utils
//
static json oaicompat_completion_params_parse(
const struct llama_model * model,
const json & body, /* openai api json semantics */
const std::string & chat_template) {
json llama_params;
llama_params["__oaicompat"] = true;
// Map OpenAI parameters to llama.cpp parameters
//
// For parameters that are defined by the OpenAI documentation (e.g.
// temperature), we explicitly specify OpenAI's intended default; we
// need to do that because sometimes OpenAI disagrees with llama.cpp
//
// https://platform.openai.com/docs/api-reference/chat/create
llama_sampling_params default_sparams;
llama_params["model"] = json_value(body, "model", std::string("unknown"));
llama_params["prompt"] = format_chat(model, chat_template, body["messages"]);
llama_params["cache_prompt"] = json_value(body, "cache_prompt", false);
llama_params["temperature"] = json_value(body, "temperature", 0.0);
llama_params["top_k"] = json_value(body, "top_k", default_sparams.top_k);
llama_params["top_p"] = json_value(body, "top_p", 1.0);
llama_params["n_predict"] = json_value(body, "max_tokens", -1);
llama_params["logit_bias"] = json_value(body, "logit_bias", json::object());
llama_params["frequency_penalty"] = json_value(body, "frequency_penalty", 0.0);
llama_params["presence_penalty"] = json_value(body, "presence_penalty", 0.0);
llama_params["seed"] = json_value(body, "seed", LLAMA_DEFAULT_SEED);
llama_params["stream"] = json_value(body, "stream", false);
llama_params["mirostat"] = json_value(body, "mirostat", default_sparams.mirostat);
llama_params["mirostat_tau"] = json_value(body, "mirostat_tau", default_sparams.mirostat_tau);
llama_params["mirostat_eta"] = json_value(body, "mirostat_eta", default_sparams.mirostat_eta);
llama_params["penalize_nl"] = json_value(body, "penalize_nl", default_sparams.penalize_nl);
llama_params["typical_p"] = json_value(body, "typical_p", default_sparams.typical_p);
llama_params["repeat_last_n"] = json_value(body, "repeat_last_n", default_sparams.penalty_last_n);
llama_params["ignore_eos"] = json_value(body, "ignore_eos", false);
llama_params["tfs_z"] = json_value(body, "tfs_z", default_sparams.tfs_z);
if (body.count("grammar") != 0) {
llama_params["grammar"] = json_value(body, "grammar", json::object());
}
// Handle 'stop' field
if (body.contains("stop") && body["stop"].is_string()) {
llama_params["stop"] = json::array({body["stop"].get<std::string>()});
} else {
llama_params["stop"] = json_value(body, "stop", json::array());
}
// Ensure there is ChatML-specific end sequence among stop words
llama_params["stop"].push_back("<|im_end|>");
return llama_params;
}
static json format_final_response_oaicompat(const json & request, json result, bool streaming = false) {
bool stopped_word = result.count("stopped_word") != 0;
bool stopped_eos = json_value(result, "stopped_eos", false);
int num_tokens_predicted = json_value(result, "tokens_predicted", 0);
int num_prompt_tokens = json_value(result, "tokens_evaluated", 0);
std::string content = json_value(result, "content", std::string(""));
std::string finish_reason = "length";
if (stopped_word || stopped_eos) {
finish_reason = "stop";
}
json choices =
streaming ? json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"delta", json::object()}}})
: json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"message", json{{"content", content},
{"role", "assistant"}}}}});
std::time_t t = std::time(0);
json res = json {
{"choices", choices},
{"created", t},
{"model",
json_value(request, "model", std::string(DEFAULT_OAICOMPAT_MODEL))},
{"object", streaming ? "chat.completion.chunk" : "chat.completion"},
{"usage", json {
{"completion_tokens", num_tokens_predicted},
{"prompt_tokens", num_prompt_tokens},
{"total_tokens", num_tokens_predicted + num_prompt_tokens}
}},
{"id", gen_chatcmplid()}
};
if (server_verbose) {
res["__verbose"] = result;
}
if (result.contains("completion_probabilities")) {
res["completion_probabilities"] = json_value(result, "completion_probabilities", json::array());
}
return res;
}
// return value is vector as there is one case where we might need to generate two responses
static std::vector<json> format_partial_response_oaicompat(json result) {
if (!result.contains("model") || !result.contains("oaicompat_token_ctr")) {
return std::vector<json>({result});
}
bool first = json_value(result, "oaicompat_token_ctr", 0) == 0;
std::string modelname = json_value(result, "model", std::string(DEFAULT_OAICOMPAT_MODEL));
bool stopped_word = json_value(result, "stopped_word", false);
bool stopped_eos = json_value(result, "stopped_eos", false);
bool stopped_limit = json_value(result, "stopped_limit", false);
std::string content = json_value(result, "content", std::string(""));
std::string finish_reason;
if (stopped_word || stopped_eos) {
finish_reason = "stop";
}
if (stopped_limit) {
finish_reason = "length";
}
std::time_t t = std::time(0);
json choices;
if (!finish_reason.empty()) {
choices = json::array({json{{"finish_reason", finish_reason},
{"index", 0},
{"delta", json::object()}}});
} else {
if (first) {
if (content.empty()) {
choices = json::array({json{{"finish_reason", nullptr},
{"index", 0},
{"delta", json{{"role", "assistant"}}}}});
} else {
// We have to send this as two updates to conform to openai behavior
json initial_ret = json{{"choices", json::array({json{
{"finish_reason", nullptr},
{"index", 0},
{"delta", json{
{"role", "assistant"}
}}}})},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}};
json second_ret = json{
{"choices", json::array({json{{"finish_reason", nullptr},
{"index", 0},
{"delta", json{
{"content", content}}}
}})},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}};
return std::vector<json>({initial_ret, second_ret});
}
} else {
// Some idiosyncrasy in task processing logic makes several trailing calls
// with empty content, we ignore these at the calee site.
if (content.empty()) {
return std::vector<json>({json::object()});
}
choices = json::array({json{
{"finish_reason", nullptr},
{"index", 0},
{"delta",
json{
{"content", content},
}},
}});
}
}
json ret = json {
{"choices", choices},
{"created", t},
{"id", gen_chatcmplid()},
{"model", modelname},
{"object", "chat.completion.chunk"}
};
return std::vector<json>({ret});
}
static json format_embeddings_response_oaicompat(const json & request, const json & embeddings) {
json res = json {
{"model", json_value(request, "model", std::string(DEFAULT_OAICOMPAT_MODEL))},
{"object", "list"},
{"usage", json {
{"prompt_tokens", 0},
{"total_tokens", 0}
}},
{"data", embeddings}
};
return res;
}
static json format_tokenizer_response(const std::vector<llama_token> & tokens) {
return json {
{"tokens", tokens}
};
}
static json format_detokenized_response(const std::string & content) {
return json {
{"content", content}
};
}

6
flake.lock
View File

@ -20,11 +20,11 @@
},
"nixpkgs": {
"locked": {
"lastModified": 1709237383,
"narHash": "sha256-cy6ArO4k5qTx+l5o+0mL9f5fa86tYUX3ozE1S+Txlds=",
"lastModified": 1709703039,
"narHash": "sha256-6hqgQ8OK6gsMu1VtcGKBxKQInRLHtzulDo9Z5jxHEFY=",
"owner": "NixOS",
"repo": "nixpkgs",
"rev": "1536926ef5621b09bba54035ae2bb6d806d72ac8",
"rev": "9df3e30ce24fd28c7b3e2de0d986769db5d6225d",
"type": "github"
},
"original": {

779
ggml-common.h Normal file
View File

@ -0,0 +1,779 @@
#pragma once
#if defined(GGML_COMMON_IMPL_C)
#include <stdint.h>
#define GGML_TABLE_BEGIN(type, name, size) static const type name[size] = {
#define GGML_TABLE_END() };
#define GGML_COMMON_IMPL
#elif defined(GGML_COMMON_IMPL_METAL)
#include <metal_stdlib>
#define GGML_TABLE_BEGIN(type, name, size) static const constant type name[size] = {
#define GGML_TABLE_END() };
#define GGML_COMMON_IMPL
#elif defined(GGML_COMMON_IMPL_CUDA)
#include <cstdint>
#define GGML_TABLE_BEGIN(type, name, size) static const __device__ type name[size] = {
#define GGML_TABLE_END() };
#define GGML_COMMON_IMPL
#elif defined(GGML_COMMON_IMPL_SYCL)
#include <cstdint>
#define GGML_TABLE_BEGIN(type, name, size) static dpct::global_memory<const type, 1> name(sycl::range<1>(size), {
#define GGML_TABLE_END() });
#define GGML_COMMON_IMPL
#endif
#if defined(GGML_COMMON_IMPL)
GGML_TABLE_BEGIN(uint8_t, kmask_iq2xs, 8)
1, 2, 4, 8, 16, 32, 64, 128
GGML_TABLE_END()
GGML_TABLE_BEGIN(uint8_t, ksigns_iq2xs, 128)
0, 129, 130, 3, 132, 5, 6, 135, 136, 9, 10, 139, 12, 141, 142, 15,
144, 17, 18, 147, 20, 149, 150, 23, 24, 153, 154, 27, 156, 29, 30, 159,
160, 33, 34, 163, 36, 165, 166, 39, 40, 169, 170, 43, 172, 45, 46, 175,
48, 177, 178, 51, 180, 53, 54, 183, 184, 57, 58, 187, 60, 189, 190, 63,
192, 65, 66, 195, 68, 197, 198, 71, 72, 201, 202, 75, 204, 77, 78, 207,
80, 209, 210, 83, 212, 85, 86, 215, 216, 89, 90, 219, 92, 221, 222, 95,
96, 225, 226, 99, 228, 101, 102, 231, 232, 105, 106, 235, 108, 237, 238, 111,
240, 113, 114, 243, 116, 245, 246, 119, 120, 249, 250, 123, 252, 125, 126, 255,
GGML_TABLE_END()
//#if __CUDA_ARCH__ >= MIN_CC_DP4A // lowest compute capability for integer intrinsics
GGML_TABLE_BEGIN(uint64_t, ksigns64, 128)
0x0000000000000000, 0xff000000000000ff, 0xff0000000000ff00, 0x000000000000ffff,
0xff00000000ff0000, 0x0000000000ff00ff, 0x0000000000ffff00, 0xff00000000ffffff,
0xff000000ff000000, 0x00000000ff0000ff, 0x00000000ff00ff00, 0xff000000ff00ffff,
0x00000000ffff0000, 0xff000000ffff00ff, 0xff000000ffffff00, 0x00000000ffffffff,
0xff0000ff00000000, 0x000000ff000000ff, 0x000000ff0000ff00, 0xff0000ff0000ffff,
0x000000ff00ff0000, 0xff0000ff00ff00ff, 0xff0000ff00ffff00, 0x000000ff00ffffff,
0x000000ffff000000, 0xff0000ffff0000ff, 0xff0000ffff00ff00, 0x000000ffff00ffff,
0xff0000ffffff0000, 0x000000ffffff00ff, 0x000000ffffffff00, 0xff0000ffffffffff,
0xff00ff0000000000, 0x0000ff00000000ff, 0x0000ff000000ff00, 0xff00ff000000ffff,
0x0000ff0000ff0000, 0xff00ff0000ff00ff, 0xff00ff0000ffff00, 0x0000ff0000ffffff,
0x0000ff00ff000000, 0xff00ff00ff0000ff, 0xff00ff00ff00ff00, 0x0000ff00ff00ffff,
0xff00ff00ffff0000, 0x0000ff00ffff00ff, 0x0000ff00ffffff00, 0xff00ff00ffffffff,
0x0000ffff00000000, 0xff00ffff000000ff, 0xff00ffff0000ff00, 0x0000ffff0000ffff,
0xff00ffff00ff0000, 0x0000ffff00ff00ff, 0x0000ffff00ffff00, 0xff00ffff00ffffff,
0xff00ffffff000000, 0x0000ffffff0000ff, 0x0000ffffff00ff00, 0xff00ffffff00ffff,
0x0000ffffffff0000, 0xff00ffffffff00ff, 0xff00ffffffffff00, 0x0000ffffffffffff,
0xffff000000000000, 0x00ff0000000000ff, 0x00ff00000000ff00, 0xffff00000000ffff,
0x00ff000000ff0000, 0xffff000000ff00ff, 0xffff000000ffff00, 0x00ff000000ffffff,
0x00ff0000ff000000, 0xffff0000ff0000ff, 0xffff0000ff00ff00, 0x00ff0000ff00ffff,
0xffff0000ffff0000, 0x00ff0000ffff00ff, 0x00ff0000ffffff00, 0xffff0000ffffffff,
0x00ff00ff00000000, 0xffff00ff000000ff, 0xffff00ff0000ff00, 0x00ff00ff0000ffff,
0xffff00ff00ff0000, 0x00ff00ff00ff00ff, 0x00ff00ff00ffff00, 0xffff00ff00ffffff,
0xffff00ffff000000, 0x00ff00ffff0000ff, 0x00ff00ffff00ff00, 0xffff00ffff00ffff,
0x00ff00ffffff0000, 0xffff00ffffff00ff, 0xffff00ffffffff00, 0x00ff00ffffffffff,
0x00ffff0000000000, 0xffffff00000000ff, 0xffffff000000ff00, 0x00ffff000000ffff,
0xffffff0000ff0000, 0x00ffff0000ff00ff, 0x00ffff0000ffff00, 0xffffff0000ffffff,
0xffffff00ff000000, 0x00ffff00ff0000ff, 0x00ffff00ff00ff00, 0xffffff00ff00ffff,
0x00ffff00ffff0000, 0xffffff00ffff00ff, 0xffffff00ffffff00, 0x00ffff00ffffffff,
0xffffffff00000000, 0x00ffffff000000ff, 0x00ffffff0000ff00, 0xffffffff0000ffff,
0x00ffffff00ff0000, 0xffffffff00ff00ff, 0xffffffff00ffff00, 0x00ffffff00ffffff,
0x00ffffffff000000, 0xffffffffff0000ff, 0xffffffffff00ff00, 0x00ffffffff00ffff,
0xffffffffffff0000, 0x00ffffffffff00ff, 0x00ffffffffffff00, 0xffffffffffffffff,
GGML_TABLE_END()
//#endif
GGML_TABLE_BEGIN(uint64_t, iq2xxs_grid, 256)
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x08080808082b0808,
0x08080808082b082b, 0x08080808082b2b08, 0x08080808082b2b2b, 0x0808080819080819,
0x0808080819081908, 0x0808080819190808, 0x0808080819192b08, 0x08080808192b0819,
0x08080808192b1908, 0x080808082b080808, 0x080808082b08082b, 0x080808082b082b2b,
0x080808082b2b082b, 0x0808081908080819, 0x0808081908081908, 0x0808081908190808,
0x0808081908191919, 0x0808081919080808, 0x080808192b081908, 0x080808192b192b08,
0x0808082b08080808, 0x0808082b0808082b, 0x0808082b082b082b, 0x0808082b2b08082b,
0x0808190808080819, 0x0808190808081908, 0x0808190808190808, 0x08081908082b0819,
0x08081908082b1908, 0x0808190819080808, 0x080819081908082b, 0x0808190819082b08,
0x08081908192b0808, 0x080819082b080819, 0x080819082b081908, 0x080819082b190808,
0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b, 0x0808191908082b08,
0x08081919082b0808, 0x080819191908192b, 0x08081919192b2b19, 0x080819192b080808,
0x080819192b190819, 0x0808192b08082b19, 0x0808192b08190808, 0x0808192b19080808,
0x0808192b2b081908, 0x0808192b2b2b1908, 0x08082b0808080808, 0x08082b0808081919,
0x08082b0808082b08, 0x08082b0808191908, 0x08082b08082b2b08, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b081919082b, 0x08082b082b082b08,
0x08082b1908081908, 0x08082b1919080808, 0x08082b2b0808082b, 0x08082b2b08191908,
0x0819080808080819, 0x0819080808081908, 0x0819080808190808, 0x08190808082b0819,
0x0819080819080808, 0x08190808192b0808, 0x081908082b081908, 0x081908082b190808,
0x081908082b191919, 0x0819081908080808, 0x0819081908082b08, 0x08190819082b0808,
0x0819081919190808, 0x0819081919192b2b, 0x081908192b080808, 0x0819082b082b1908,
0x0819082b19081919, 0x0819190808080808, 0x0819190808082b08, 0x08191908082b0808,
0x08191908082b1919, 0x0819190819082b19, 0x081919082b080808, 0x0819191908192b08,
0x08191919192b082b, 0x0819192b08080808, 0x0819192b0819192b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b0808190808, 0x08192b0819080808, 0x08192b082b080819,
0x08192b1908080808, 0x08192b1908081919, 0x08192b192b2b0808, 0x08192b2b19190819,
0x082b080808080808, 0x082b08080808082b, 0x082b080808082b2b, 0x082b080819081908,
0x082b0808192b0819, 0x082b08082b080808, 0x082b08082b08082b, 0x082b0819082b2b19,
0x082b081919082b08, 0x082b082b08080808, 0x082b082b0808082b, 0x082b190808080819,
0x082b190808081908, 0x082b190808190808, 0x082b190819080808, 0x082b19081919192b,
0x082b191908080808, 0x082b191919080819, 0x082b1919192b1908, 0x082b192b2b190808,
0x082b2b0808082b08, 0x082b2b08082b0808, 0x082b2b082b191908, 0x082b2b2b19081908,
0x1908080808080819, 0x1908080808081908, 0x1908080808190808, 0x1908080808192b08,
0x19080808082b0819, 0x19080808082b1908, 0x1908080819080808, 0x1908080819082b08,
0x190808081919192b, 0x19080808192b0808, 0x190808082b080819, 0x190808082b081908,
0x190808082b190808, 0x1908081908080808, 0x19080819082b0808, 0x19080819192b0819,
0x190808192b080808, 0x190808192b081919, 0x1908082b08080819, 0x1908082b08190808,
0x1908082b19082b08, 0x1908082b1919192b, 0x1908082b192b2b08, 0x1908190808080808,
0x1908190808082b08, 0x19081908082b0808, 0x190819082b080808, 0x190819082b192b19,
0x190819190819082b, 0x19081919082b1908, 0x1908192b08080808, 0x19082b0808080819,
0x19082b0808081908, 0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919,
0x19082b1908080808, 0x19082b1919192b08, 0x19082b19192b0819, 0x19082b192b08082b,
0x19082b2b19081919, 0x19082b2b2b190808, 0x1919080808080808, 0x1919080808082b08,
0x1919080808190819, 0x1919080808192b19, 0x19190808082b0808, 0x191908082b080808,
0x191908082b082b08, 0x1919081908081908, 0x191908191908082b, 0x191908192b2b1908,
0x1919082b2b190819, 0x191919082b190808, 0x191919082b19082b, 0x1919191908082b2b,
0x1919192b08080819, 0x1919192b19191908, 0x19192b0808080808, 0x19192b0808190819,
0x19192b0808192b19, 0x19192b08192b1908, 0x19192b1919080808, 0x19192b2b08082b08,
0x192b080808081908, 0x192b080808190808, 0x192b080819080808, 0x192b0808192b2b08,
0x192b081908080808, 0x192b081919191919, 0x192b082b08192b08, 0x192b082b192b0808,
0x192b190808080808, 0x192b190808081919, 0x192b191908190808, 0x192b19190819082b,
0x192b19192b081908, 0x192b2b081908082b, 0x2b08080808080808, 0x2b0808080808082b,
0x2b08080808082b2b, 0x2b08080819080819, 0x2b0808082b08082b, 0x2b08081908081908,
0x2b08081908192b08, 0x2b08081919080808, 0x2b08082b08190819, 0x2b08190808080819,
0x2b08190808081908, 0x2b08190808190808, 0x2b08190808191919, 0x2b08190819080808,
0x2b081908192b0808, 0x2b08191908080808, 0x2b0819191908192b, 0x2b0819192b191908,
0x2b08192b08082b19, 0x2b08192b19080808, 0x2b08192b192b0808, 0x2b082b080808082b,
0x2b082b1908081908, 0x2b082b2b08190819, 0x2b19080808081908, 0x2b19080808190808,
0x2b190808082b1908, 0x2b19080819080808, 0x2b1908082b2b0819, 0x2b1908190819192b,
0x2b1908192b080808, 0x2b19082b19081919, 0x2b19190808080808, 0x2b191908082b082b,
0x2b19190819081908, 0x2b19191919190819, 0x2b192b082b080819, 0x2b192b19082b0808,
0x2b2b08080808082b, 0x2b2b080819190808, 0x2b2b08082b081919, 0x2b2b081908082b19,
0x2b2b082b08080808, 0x2b2b190808192b08, 0x2b2b2b0819190808, 0x2b2b2b1908081908,
GGML_TABLE_END()
GGML_TABLE_BEGIN(uint64_t, iq2xs_grid, 512)
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x080808082b080808,
0x080808082b08082b, 0x080808082b081919, 0x080808082b082b08, 0x080808082b190819,
0x080808082b191908, 0x080808082b192b19, 0x080808082b2b0808, 0x0808081908080819,
0x0808081908081908, 0x080808190808192b, 0x0808081908082b19, 0x0808081908190808,
0x080808190819082b, 0x0808081908191919, 0x0808081908192b08, 0x0808081908192b2b,
0x08080819082b0819, 0x08080819082b1908, 0x0808081919080808, 0x080808191908082b,
0x0808081919081919, 0x0808081919082b08, 0x0808081919190819, 0x0808081919191908,
0x08080819192b0808, 0x08080819192b2b08, 0x080808192b080819, 0x080808192b081908,
0x080808192b190808, 0x0808082b08080808, 0x0808082b0808082b, 0x0808082b08081919,
0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908, 0x0808082b082b0808,
0x0808082b19080819, 0x0808082b19081908, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b082b2b, 0x0808190808080819, 0x0808190808081908,
0x080819080808192b, 0x0808190808082b19, 0x0808190808190808, 0x080819080819082b,
0x0808190808191919, 0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908,
0x0808190819080808, 0x080819081908082b, 0x0808190819081919, 0x0808190819082b08,
0x0808190819190819, 0x0808190819191908, 0x080819081919192b, 0x08081908192b0808,
0x080819082b080819, 0x080819082b081908, 0x080819082b190808, 0x0808191908080808,
0x080819190808082b, 0x0808191908081919, 0x0808191908082b08, 0x0808191908190819,
0x0808191908191908, 0x08081919082b0808, 0x0808191919080819, 0x0808191919081908,
0x0808191919190808, 0x08081919192b0819, 0x080819192b080808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b08190808, 0x0808192b082b192b, 0x0808192b19080808,
0x0808192b1908082b, 0x0808192b2b081908, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808082b2b, 0x08082b0808190819,
0x08082b0808191908, 0x08082b08082b0808, 0x08082b08082b1919, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b0819192b08, 0x08082b082b080808,
0x08082b082b2b0808, 0x08082b082b2b2b2b, 0x08082b1908080819, 0x08082b1908081908,
0x08082b1908190808, 0x08082b1919080808, 0x08082b192b080819, 0x08082b192b082b19,
0x08082b2b08080808, 0x08082b2b082b0808, 0x08082b2b082b2b08, 0x08082b2b2b19192b,
0x08082b2b2b2b0808, 0x0819080808080819, 0x0819080808081908, 0x081908080808192b,
0x0819080808082b19, 0x0819080808190808, 0x081908080819082b, 0x0819080808191919,
0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908, 0x0819080819080808,
0x081908081908082b, 0x0819080819081919, 0x0819080819082b08, 0x0819080819190819,
0x0819080819191908, 0x08190808192b0808, 0x08190808192b2b2b, 0x081908082b080819,
0x081908082b081908, 0x081908082b190808, 0x0819081908080808, 0x081908190808082b,
0x0819081908081919, 0x0819081908082b08, 0x0819081908190819, 0x0819081908191908,
0x08190819082b0808, 0x0819081919080819, 0x0819081919081908, 0x0819081919190808,
0x081908192b080808, 0x081908192b191908, 0x081908192b19192b, 0x0819082b08080819,
0x0819082b08081908, 0x0819082b0808192b, 0x0819082b08190808, 0x0819082b19080808,
0x0819082b192b0808, 0x0819190808080808, 0x081919080808082b, 0x0819190808081919,
0x0819190808082b08, 0x0819190808190819, 0x0819190808191908, 0x08191908082b0808,
0x0819190819080819, 0x0819190819081908, 0x0819190819082b19, 0x0819190819190808,
0x08191908192b1908, 0x081919082b080808, 0x0819191908080819, 0x0819191908081908,
0x0819191908190808, 0x0819191919080808, 0x0819192b08080808, 0x0819192b08191908,
0x0819192b19082b19, 0x08192b0808080819, 0x08192b0808081908, 0x08192b0808190808,
0x08192b080819082b, 0x08192b0819080808, 0x08192b0819191908, 0x08192b082b08192b,
0x08192b1908080808, 0x08192b1908081919, 0x08192b19192b192b, 0x08192b2b19190819,
0x08192b2b2b2b2b19, 0x082b080808080808, 0x082b08080808082b, 0x082b080808081919,
0x082b080808082b08, 0x082b080808082b2b, 0x082b080808190819, 0x082b080808191908,
0x082b0808082b0808, 0x082b080819080819, 0x082b080819081908, 0x082b080819190808,
0x082b08082b080808, 0x082b08082b2b0808, 0x082b081908080819, 0x082b081908081908,
0x082b081908190808, 0x082b081919080808, 0x082b081919082b08, 0x082b0819192b1919,
0x082b082b08080808, 0x082b082b082b082b, 0x082b082b2b080808, 0x082b082b2b2b2b08,
0x082b190808080819, 0x082b190808081908, 0x082b190808190808, 0x082b1908082b2b19,
0x082b190819080808, 0x082b191908080808, 0x082b191919080819, 0x082b19191919082b,
0x082b19192b192b19, 0x082b192b08080819, 0x082b192b08192b2b, 0x082b192b2b2b192b,
0x082b2b0808080808, 0x082b2b0808082b08, 0x082b2b0808082b2b, 0x082b2b08082b0808,
0x082b2b0819191919, 0x082b2b082b082b08, 0x082b2b082b2b082b, 0x082b2b19192b2b08,
0x082b2b192b190808, 0x082b2b2b08082b08, 0x082b2b2b082b0808, 0x082b2b2b2b08082b,
0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819, 0x1908080808081908,
0x190808080808192b, 0x1908080808082b19, 0x1908080808190808, 0x190808080819082b,
0x1908080808191919, 0x1908080808192b08, 0x19080808082b0819, 0x19080808082b1908,
0x1908080819080808, 0x190808081908082b, 0x1908080819081919, 0x1908080819082b08,
0x1908080819082b2b, 0x1908080819190819, 0x1908080819191908, 0x19080808192b0808,
0x19080808192b1919, 0x190808082b080819, 0x190808082b081908, 0x190808082b190808,
0x1908081908080808, 0x190808190808082b, 0x1908081908081919, 0x1908081908082b08,
0x1908081908190819, 0x1908081908191908, 0x19080819082b0808, 0x1908081919080819,
0x1908081919081908, 0x1908081919190808, 0x190808192b080808, 0x190808192b081919,
0x190808192b2b082b, 0x1908082b08080819, 0x1908082b08081908, 0x1908082b08190808,
0x1908082b0819082b, 0x1908082b082b2b19, 0x1908082b19080808, 0x1908190808080808,
0x190819080808082b, 0x1908190808081919, 0x1908190808082b08, 0x1908190808190819,
0x1908190808191908, 0x1908190808192b19, 0x19081908082b0808, 0x1908190819080819,
0x1908190819081908, 0x1908190819190808, 0x190819082b080808, 0x190819082b191908,
0x1908191908080819, 0x1908191908081908, 0x1908191908190808, 0x19081919082b1908,
0x1908191919080808, 0x190819192b192b2b, 0x1908192b08080808, 0x1908192b08082b2b,
0x1908192b19081908, 0x1908192b19190808, 0x19082b0808080819, 0x19082b0808081908,
0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919, 0x19082b0819191908,
0x19082b08192b082b, 0x19082b1908080808, 0x19082b1908190819, 0x19082b1919081908,
0x19082b1919190808, 0x19082b19192b2b19, 0x19082b2b08081908, 0x1919080808080808,
0x191908080808082b, 0x1919080808081919, 0x1919080808082b08, 0x1919080808190819,
0x1919080808191908, 0x19190808082b0808, 0x19190808082b2b08, 0x1919080819080819,
0x1919080819081908, 0x1919080819190808, 0x191908082b080808, 0x1919081908080819,
0x1919081908081908, 0x1919081908190808, 0x1919081908191919, 0x1919081919080808,
0x191908191908082b, 0x1919082b08080808, 0x1919082b19081908, 0x1919082b2b2b2b2b,
0x1919190808080819, 0x1919190808081908, 0x1919190808190808, 0x19191908082b0819,
0x1919190819080808, 0x19191908192b0808, 0x191919082b080819, 0x191919082b2b0819,
0x1919191908080808, 0x1919191908082b08, 0x191919192b080808, 0x191919192b082b08,
0x1919192b082b0819, 0x1919192b192b2b08, 0x1919192b2b2b0819, 0x19192b0808080808,
0x19192b0808191908, 0x19192b0819080819, 0x19192b0819190808, 0x19192b082b192b19,
0x19192b1908192b2b, 0x19192b1919080808, 0x19192b191908082b, 0x19192b2b2b081919,
0x192b080808080819, 0x192b080808081908, 0x192b080808190808, 0x192b080819080808,
0x192b080819191908, 0x192b0808192b082b, 0x192b08082b08192b, 0x192b08082b2b2b19,
0x192b081908080808, 0x192b082b082b1908, 0x192b082b19082b2b, 0x192b082b2b19082b,
0x192b190808080808, 0x192b19080819192b, 0x192b191908190808, 0x192b191919080808,
0x192b191919081919, 0x192b19192b2b1908, 0x192b2b0808080819, 0x192b2b08192b2b2b,
0x192b2b19082b1919, 0x192b2b2b0808192b, 0x192b2b2b19191908, 0x192b2b2b192b082b,
0x2b08080808080808, 0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08,
0x2b08080808190819, 0x2b08080808191908, 0x2b080808082b0808, 0x2b080808082b2b2b,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808082b080808,
0x2b0808082b08082b, 0x2b0808082b2b2b08, 0x2b0808082b2b2b2b, 0x2b08081908080819,
0x2b08081908081908, 0x2b0808190808192b, 0x2b08081908190808, 0x2b08081919080808,
0x2b08081919190819, 0x2b08081919192b19, 0x2b08082b08080808, 0x2b08082b082b0808,
0x2b08082b2b080808, 0x2b08082b2b08082b, 0x2b08082b2b2b0808, 0x2b08082b2b2b2b08,
0x2b08190808080819, 0x2b08190808081908, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190819080808, 0x2b081908192b0808, 0x2b0819082b082b19,
0x2b08191908080808, 0x2b08191919081908, 0x2b0819192b2b1919, 0x2b08192b08192b08,
0x2b08192b192b2b2b, 0x2b082b0808080808, 0x2b082b0808082b08, 0x2b082b08082b1919,
0x2b082b0819192b2b, 0x2b082b082b080808, 0x2b082b082b08082b, 0x2b082b082b2b2b08,
0x2b082b190808192b, 0x2b082b2b082b082b, 0x2b082b2b2b080808, 0x2b082b2b2b082b08,
0x2b082b2b2b19192b, 0x2b082b2b2b2b2b08, 0x2b19080808080819, 0x2b19080808081908,
0x2b19080808190808, 0x2b19080819080808, 0x2b1908081919192b, 0x2b1908082b081908,
0x2b19081908080808, 0x2b190819082b082b, 0x2b190819192b1908, 0x2b19082b1919192b,
0x2b19082b2b082b19, 0x2b19190808080808, 0x2b19190808081919, 0x2b19190819081908,
0x2b19190819190808, 0x2b19190819192b08, 0x2b191919082b2b19, 0x2b1919192b190808,
0x2b1919192b19082b, 0x2b19192b19080819, 0x2b192b0819190819, 0x2b192b082b2b192b,
0x2b192b1919082b19, 0x2b192b2b08191919, 0x2b192b2b192b0808, 0x2b2b080808080808,
0x2b2b08080808082b, 0x2b2b080808082b08, 0x2b2b080808082b2b, 0x2b2b0808082b0808,
0x2b2b0808082b2b2b, 0x2b2b08082b2b0808, 0x2b2b081919190819, 0x2b2b081919192b19,
0x2b2b08192b2b192b, 0x2b2b082b08080808, 0x2b2b082b0808082b, 0x2b2b082b08082b08,
0x2b2b082b082b2b2b, 0x2b2b082b2b080808, 0x2b2b082b2b2b0808, 0x2b2b190819080808,
0x2b2b19082b191919, 0x2b2b192b192b1919, 0x2b2b192b2b192b08, 0x2b2b2b0808082b2b,
0x2b2b2b08082b0808, 0x2b2b2b08082b082b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b0808,
0x2b2b2b082b2b2b08, 0x2b2b2b1908081908, 0x2b2b2b192b081908, 0x2b2b2b192b08192b,
0x2b2b2b2b082b2b08, 0x2b2b2b2b082b2b2b, 0x2b2b2b2b2b190819, 0x2b2b2b2b2b2b2b2b,
GGML_TABLE_END()
GGML_TABLE_BEGIN(uint64_t, iq2s_grid, 1024)
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x08080808192b192b,
0x08080808192b2b19, 0x080808082b080808, 0x080808082b08082b, 0x080808082b081919,
0x080808082b082b08, 0x080808082b190819, 0x080808082b191908, 0x080808082b2b0808,
0x080808082b2b1919, 0x080808082b2b2b2b, 0x0808081908080819, 0x0808081908081908,
0x080808190808192b, 0x0808081908082b19, 0x0808081908190808, 0x080808190819082b,
0x0808081908191919, 0x0808081908192b08, 0x08080819082b0819, 0x08080819082b1908,
0x0808081919080808, 0x080808191908082b, 0x0808081919081919, 0x0808081919082b08,
0x0808081919190819, 0x0808081919191908, 0x080808191919192b, 0x0808081919192b19,
0x08080819192b0808, 0x08080819192b1919, 0x08080819192b2b08, 0x080808192b080819,
0x080808192b081908, 0x080808192b190808, 0x080808192b19082b, 0x080808192b191919,
0x080808192b2b0819, 0x080808192b2b1908, 0x0808082b08080808, 0x0808082b0808082b,
0x0808082b08081919, 0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908,
0x0808082b082b0808, 0x0808082b082b2b2b, 0x0808082b19080819, 0x0808082b19081908,
0x0808082b1908192b, 0x0808082b19082b19, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b081919, 0x0808082b2b082b2b, 0x0808082b2b191908,
0x0808082b2b2b082b, 0x0808190808080819, 0x0808190808081908, 0x080819080808192b,
0x0808190808082b19, 0x0808190808190808, 0x080819080819082b, 0x0808190808191919,
0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908, 0x08081908082b192b,
0x08081908082b2b19, 0x0808190819080808, 0x080819081908082b, 0x0808190819081919,
0x0808190819082b08, 0x0808190819082b2b, 0x0808190819190819, 0x0808190819191908,
0x080819081919192b, 0x0808190819192b19, 0x08081908192b0808, 0x08081908192b082b,
0x08081908192b1919, 0x080819082b080819, 0x080819082b081908, 0x080819082b08192b,
0x080819082b082b19, 0x080819082b190808, 0x080819082b191919, 0x080819082b192b08,
0x080819082b2b0819, 0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b,
0x0808191908081919, 0x0808191908082b08, 0x0808191908082b2b, 0x0808191908190819,
0x0808191908191908, 0x080819190819192b, 0x0808191908192b19, 0x08081919082b0808,
0x08081919082b1919, 0x08081919082b2b08, 0x0808191919080819, 0x0808191919081908,
0x080819191908192b, 0x0808191919082b19, 0x0808191919190808, 0x080819191919082b,
0x0808191919191919, 0x0808191919192b08, 0x08081919192b0819, 0x08081919192b1908,
0x080819192b080808, 0x080819192b08082b, 0x080819192b081919, 0x080819192b082b08,
0x080819192b190819, 0x080819192b191908, 0x080819192b2b0808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b0808192b, 0x0808192b08082b19, 0x0808192b08190808,
0x0808192b08191919, 0x0808192b19080808, 0x0808192b19081919, 0x0808192b19082b08,
0x0808192b19190819, 0x0808192b19191908, 0x0808192b192b0808, 0x0808192b2b080819,
0x0808192b2b081908, 0x0808192b2b190808, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808190819, 0x08082b0808191908,
0x08082b080819192b, 0x08082b0808192b19, 0x08082b08082b0808, 0x08082b08082b1919,
0x08082b08082b2b2b, 0x08082b0819080819, 0x08082b0819081908, 0x08082b081908192b,
0x08082b0819082b19, 0x08082b0819190808, 0x08082b081919082b, 0x08082b0819191919,
0x08082b0819192b08, 0x08082b08192b0819, 0x08082b08192b1908, 0x08082b082b080808,
0x08082b082b081919, 0x08082b082b191908, 0x08082b082b2b2b2b, 0x08082b1908080819,
0x08082b1908081908, 0x08082b1908190808, 0x08082b190819082b, 0x08082b1908191919,
0x08082b1908192b08, 0x08082b19082b0819, 0x08082b1919080808, 0x08082b1919081919,
0x08082b1919082b08, 0x08082b1919190819, 0x08082b1919191908, 0x08082b19192b0808,
0x08082b192b080819, 0x08082b192b190808, 0x08082b2b08080808, 0x08082b2b08190819,
0x08082b2b08191908, 0x08082b2b082b082b, 0x08082b2b082b2b08, 0x08082b2b082b2b2b,
0x08082b2b19190808, 0x08082b2b2b192b19, 0x0819080808080819, 0x0819080808081908,
0x081908080808192b, 0x0819080808082b19, 0x0819080808190808, 0x081908080819082b,
0x0819080808191919, 0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908,
0x08190808082b192b, 0x0819080819080808, 0x081908081908082b, 0x0819080819081919,
0x0819080819082b08, 0x0819080819190819, 0x0819080819191908, 0x081908081919192b,
0x0819080819192b19, 0x08190808192b0808, 0x08190808192b082b, 0x08190808192b1919,
0x08190808192b2b08, 0x081908082b080819, 0x081908082b081908, 0x081908082b08192b,
0x081908082b190808, 0x081908082b191919, 0x081908082b192b08, 0x081908082b2b0819,
0x081908082b2b1908, 0x0819081908080808, 0x081908190808082b, 0x0819081908081919,
0x0819081908082b08, 0x0819081908082b2b, 0x0819081908190819, 0x0819081908191908,
0x081908190819192b, 0x0819081908192b19, 0x08190819082b0808, 0x08190819082b082b,
0x08190819082b1919, 0x08190819082b2b08, 0x0819081919080819, 0x0819081919081908,
0x081908191908192b, 0x0819081919082b19, 0x0819081919190808, 0x081908191919082b,
0x0819081919191919, 0x0819081919192b08, 0x08190819192b0819, 0x08190819192b1908,
0x081908192b080808, 0x081908192b08082b, 0x081908192b081919, 0x081908192b082b08,
0x081908192b190819, 0x081908192b191908, 0x0819082b08080819, 0x0819082b08081908,
0x0819082b08082b19, 0x0819082b08190808, 0x0819082b08191919, 0x0819082b082b0819,
0x0819082b082b1908, 0x0819082b19080808, 0x0819082b19081919, 0x0819082b19190819,
0x0819082b19191908, 0x0819082b2b080819, 0x0819082b2b081908, 0x0819082b2b190808,
0x0819190808080808, 0x081919080808082b, 0x0819190808081919, 0x0819190808082b08,
0x0819190808190819, 0x0819190808191908, 0x081919080819192b, 0x0819190808192b19,
0x08191908082b0808, 0x08191908082b1919, 0x08191908082b2b08, 0x0819190819080819,
0x0819190819081908, 0x081919081908192b, 0x0819190819082b19, 0x0819190819190808,
0x081919081919082b, 0x0819190819191919, 0x0819190819192b08, 0x08191908192b0819,
0x08191908192b1908, 0x081919082b080808, 0x081919082b08082b, 0x081919082b081919,
0x081919082b082b08, 0x081919082b190819, 0x081919082b191908, 0x081919082b2b0808,
0x0819191908080819, 0x0819191908081908, 0x081919190808192b, 0x0819191908082b19,
0x0819191908190808, 0x081919190819082b, 0x0819191908191919, 0x0819191908192b08,
0x08191919082b0819, 0x08191919082b1908, 0x0819191919080808, 0x081919191908082b,
0x0819191919081919, 0x0819191919082b08, 0x0819191919190819, 0x0819191919191908,
0x08191919192b0808, 0x081919192b080819, 0x081919192b081908, 0x081919192b190808,
0x0819192b08080808, 0x0819192b08081919, 0x0819192b08082b08, 0x0819192b08190819,
0x0819192b08191908, 0x0819192b082b0808, 0x0819192b19080819, 0x0819192b19081908,
0x0819192b19190808, 0x0819192b2b080808, 0x0819192b2b2b2b2b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b080808192b, 0x08192b0808082b19, 0x08192b0808190808,
0x08192b0808191919, 0x08192b0808192b08, 0x08192b08082b0819, 0x08192b0819080808,
0x08192b081908082b, 0x08192b0819081919, 0x08192b0819082b08, 0x08192b0819190819,
0x08192b0819191908, 0x08192b08192b0808, 0x08192b082b080819, 0x08192b082b081908,
0x08192b1908080808, 0x08192b190808082b, 0x08192b1908081919, 0x08192b1908082b08,
0x08192b1908190819, 0x08192b1908191908, 0x08192b19082b0808, 0x08192b1919080819,
0x08192b1919081908, 0x08192b1919190808, 0x08192b19192b2b19, 0x08192b192b2b082b,
0x08192b2b08081908, 0x08192b2b08190808, 0x08192b2b19080808, 0x08192b2b1919192b,
0x082b080808080808, 0x082b08080808082b, 0x082b080808081919, 0x082b080808082b08,
0x082b080808190819, 0x082b080808191908, 0x082b08080819192b, 0x082b080808192b19,
0x082b0808082b0808, 0x082b0808082b1919, 0x082b0808082b2b2b, 0x082b080819080819,
0x082b080819081908, 0x082b080819190808, 0x082b08081919082b, 0x082b080819191919,
0x082b0808192b1908, 0x082b08082b080808, 0x082b08082b082b2b, 0x082b08082b191908,
0x082b08082b2b2b2b, 0x082b081908080819, 0x082b081908081908, 0x082b081908190808,
0x082b08190819082b, 0x082b081908191919, 0x082b0819082b0819, 0x082b081919080808,
0x082b08191908082b, 0x082b081919081919, 0x082b081919190819, 0x082b081919191908,
0x082b0819192b0808, 0x082b08192b080819, 0x082b08192b081908, 0x082b08192b190808,
0x082b082b08080808, 0x082b082b08082b2b, 0x082b082b082b082b, 0x082b082b082b2b08,
0x082b082b082b2b2b, 0x082b082b19081908, 0x082b082b19190808, 0x082b082b2b082b08,
0x082b082b2b082b2b, 0x082b082b2b2b2b08, 0x082b190808080819, 0x082b190808081908,
0x082b19080808192b, 0x082b190808082b19, 0x082b190808190808, 0x082b190808191919,
0x082b190808192b08, 0x082b1908082b0819, 0x082b1908082b1908, 0x082b190819080808,
0x082b19081908082b, 0x082b190819081919, 0x082b190819082b08, 0x082b190819190819,
0x082b190819191908, 0x082b1908192b0808, 0x082b19082b080819, 0x082b19082b081908,
0x082b19082b190808, 0x082b191908080808, 0x082b191908081919, 0x082b191908082b08,
0x082b191908190819, 0x082b191908191908, 0x082b1919082b0808, 0x082b191919080819,
0x082b191919081908, 0x082b191919190808, 0x082b1919192b192b, 0x082b19192b080808,
0x082b192b08080819, 0x082b192b08081908, 0x082b192b08190808, 0x082b192b19080808,
0x082b192b19192b19, 0x082b2b0808080808, 0x082b2b0808081919, 0x082b2b0808190819,
0x082b2b0808191908, 0x082b2b0819080819, 0x082b2b0819081908, 0x082b2b0819190808,
0x082b2b082b082b2b, 0x082b2b082b2b2b2b, 0x082b2b1908080819, 0x082b2b1908081908,
0x082b2b1908190808, 0x082b2b192b191919, 0x082b2b2b08082b2b, 0x082b2b2b082b082b,
0x082b2b2b192b1908, 0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819,
0x1908080808081908, 0x190808080808192b, 0x1908080808082b19, 0x1908080808190808,
0x190808080819082b, 0x1908080808191919, 0x1908080808192b08, 0x1908080808192b2b,
0x19080808082b0819, 0x19080808082b1908, 0x19080808082b192b, 0x1908080819080808,
0x190808081908082b, 0x1908080819081919, 0x1908080819082b08, 0x1908080819082b2b,
0x1908080819190819, 0x1908080819191908, 0x190808081919192b, 0x1908080819192b19,
0x19080808192b0808, 0x19080808192b082b, 0x19080808192b1919, 0x190808082b080819,
0x190808082b081908, 0x190808082b190808, 0x190808082b191919, 0x190808082b192b08,
0x190808082b2b0819, 0x190808082b2b1908, 0x1908081908080808, 0x190808190808082b,
0x1908081908081919, 0x1908081908082b08, 0x1908081908190819, 0x1908081908191908,
0x190808190819192b, 0x1908081908192b19, 0x19080819082b0808, 0x19080819082b082b,
0x19080819082b1919, 0x1908081919080819, 0x1908081919081908, 0x190808191908192b,
0x1908081919082b19, 0x1908081919190808, 0x190808191919082b, 0x1908081919191919,
0x1908081919192b08, 0x19080819192b0819, 0x19080819192b1908, 0x190808192b080808,
0x190808192b08082b, 0x190808192b081919, 0x190808192b082b08, 0x190808192b190819,
0x190808192b191908, 0x190808192b2b0808, 0x1908082b08080819, 0x1908082b08081908,
0x1908082b08190808, 0x1908082b0819082b, 0x1908082b08191919, 0x1908082b08192b08,
0x1908082b082b1908, 0x1908082b19080808, 0x1908082b19081919, 0x1908082b19082b08,
0x1908082b19190819, 0x1908082b19191908, 0x1908082b192b0808, 0x1908082b2b080819,
0x1908082b2b081908, 0x1908190808080808, 0x190819080808082b, 0x1908190808081919,
0x1908190808082b08, 0x1908190808082b2b, 0x1908190808190819, 0x1908190808191908,
0x190819080819192b, 0x1908190808192b19, 0x19081908082b0808, 0x19081908082b082b,
0x19081908082b1919, 0x19081908082b2b08, 0x1908190819080819, 0x1908190819081908,
0x190819081908192b, 0x1908190819082b19, 0x1908190819190808, 0x190819081919082b,
0x1908190819191919, 0x1908190819192b08, 0x19081908192b0819, 0x19081908192b1908,
0x190819082b080808, 0x190819082b08082b, 0x190819082b081919, 0x190819082b082b08,
0x190819082b190819, 0x190819082b191908, 0x190819082b2b0808, 0x1908191908080819,
0x1908191908081908, 0x190819190808192b, 0x1908191908082b19, 0x1908191908190808,
0x190819190819082b, 0x1908191908191919, 0x1908191908192b08, 0x19081919082b0819,
0x19081919082b1908, 0x1908191919080808, 0x190819191908082b, 0x1908191919081919,
0x1908191919082b08, 0x1908191919190819, 0x1908191919191908, 0x19081919192b0808,
0x19081919192b2b2b, 0x190819192b080819, 0x190819192b081908, 0x190819192b190808,
0x1908192b08080808, 0x1908192b0808082b, 0x1908192b08081919, 0x1908192b08082b08,
0x1908192b08190819, 0x1908192b08191908, 0x1908192b082b0808, 0x1908192b19080819,
0x1908192b19081908, 0x1908192b19190808, 0x1908192b2b080808, 0x1908192b2b2b1919,
0x19082b0808080819, 0x19082b0808081908, 0x19082b0808082b19, 0x19082b0808190808,
0x19082b080819082b, 0x19082b0808191919, 0x19082b0808192b08, 0x19082b08082b0819,
0x19082b08082b1908, 0x19082b0819080808, 0x19082b081908082b, 0x19082b0819081919,
0x19082b0819082b08, 0x19082b0819190819, 0x19082b0819191908, 0x19082b08192b0808,
0x19082b082b081908, 0x19082b082b190808, 0x19082b1908080808, 0x19082b190808082b,
0x19082b1908081919, 0x19082b1908082b08, 0x19082b1908190819, 0x19082b1908191908,
0x19082b19082b0808, 0x19082b1919080819, 0x19082b1919081908, 0x19082b1919190808,
0x19082b192b080808, 0x19082b192b19192b, 0x19082b2b08080819, 0x19082b2b08081908,
0x19082b2b08190808, 0x19082b2b19080808, 0x1919080808080808, 0x191908080808082b,
0x1919080808081919, 0x1919080808082b08, 0x1919080808190819, 0x1919080808191908,
0x191908080819192b, 0x1919080808192b19, 0x19190808082b0808, 0x19190808082b082b,
0x19190808082b1919, 0x19190808082b2b08, 0x1919080819080819, 0x1919080819081908,
0x191908081908192b, 0x1919080819082b19, 0x1919080819190808, 0x191908081919082b,
0x1919080819191919, 0x1919080819192b08, 0x19190808192b0819, 0x19190808192b1908,
0x191908082b080808, 0x191908082b08082b, 0x191908082b081919, 0x191908082b082b08,
0x191908082b190819, 0x191908082b191908, 0x1919081908080819, 0x1919081908081908,
0x191908190808192b, 0x1919081908082b19, 0x1919081908190808, 0x191908190819082b,
0x1919081908191919, 0x1919081908192b08, 0x19190819082b0819, 0x19190819082b1908,
0x1919081919080808, 0x191908191908082b, 0x1919081919081919, 0x1919081919082b08,
0x1919081919190819, 0x1919081919191908, 0x19190819192b0808, 0x191908192b080819,
0x191908192b081908, 0x191908192b190808, 0x1919082b08080808, 0x1919082b08081919,
0x1919082b08082b08, 0x1919082b08190819, 0x1919082b08191908, 0x1919082b082b0808,
0x1919082b19080819, 0x1919082b19081908, 0x1919082b19190808, 0x1919082b192b2b19,
0x1919082b2b080808, 0x1919190808080819, 0x1919190808081908, 0x191919080808192b,
0x1919190808082b19, 0x1919190808190808, 0x191919080819082b, 0x1919190808191919,
0x1919190808192b08, 0x19191908082b0819, 0x19191908082b1908, 0x1919190819080808,
0x191919081908082b, 0x1919190819081919, 0x1919190819082b08, 0x1919190819190819,
0x1919190819191908, 0x19191908192b0808, 0x191919082b080819, 0x191919082b081908,
0x191919082b190808, 0x1919191908080808, 0x191919190808082b, 0x1919191908081919,
0x1919191908082b08, 0x1919191908190819, 0x1919191908191908, 0x19191919082b0808,
0x1919191919080819, 0x1919191919081908, 0x1919191919190808, 0x191919192b080808,
0x1919192b08080819, 0x1919192b08081908, 0x1919192b08190808, 0x1919192b082b192b,
0x1919192b19080808, 0x19192b0808080808, 0x19192b080808082b, 0x19192b0808081919,
0x19192b0808082b08, 0x19192b0808190819, 0x19192b0808191908, 0x19192b08082b0808,
0x19192b0819080819, 0x19192b0819081908, 0x19192b0819190808, 0x19192b0819192b2b,
0x19192b082b080808, 0x19192b1908080819, 0x19192b1908081908, 0x19192b1908190808,
0x19192b1919080808, 0x19192b2b08080808, 0x19192b2b08192b19, 0x19192b2b2b081919,
0x19192b2b2b2b2b08, 0x192b080808080819, 0x192b080808081908, 0x192b08080808192b,
0x192b080808190808, 0x192b08080819082b, 0x192b080808191919, 0x192b080808192b08,
0x192b0808082b0819, 0x192b0808082b1908, 0x192b080819080808, 0x192b080819081919,
0x192b080819082b08, 0x192b080819190819, 0x192b080819191908, 0x192b0808192b0808,
0x192b08082b081908, 0x192b08082b190808, 0x192b081908080808, 0x192b08190808082b,
0x192b081908081919, 0x192b081908082b08, 0x192b081908190819, 0x192b081908191908,
0x192b0819082b0808, 0x192b081919080819, 0x192b081919081908, 0x192b081919190808,
0x192b08192b080808, 0x192b08192b192b19, 0x192b082b08081908, 0x192b082b08190808,
0x192b082b19080808, 0x192b082b1919192b, 0x192b082b2b2b0819, 0x192b190808080808,
0x192b190808081919, 0x192b190808082b08, 0x192b190808190819, 0x192b190808191908,
0x192b1908082b0808, 0x192b190819080819, 0x192b190819081908, 0x192b190819190808,
0x192b19082b080808, 0x192b191908080819, 0x192b191908081908, 0x192b191908190808,
0x192b191919080808, 0x192b191919082b2b, 0x192b1919192b2b08, 0x192b19192b19082b,
0x192b192b08080808, 0x192b192b2b191908, 0x192b2b0808080819, 0x192b2b0808081908,
0x192b2b0808190808, 0x192b2b08192b1919, 0x192b2b082b192b08, 0x192b2b1908080808,
0x192b2b19082b2b2b, 0x192b2b2b1908082b, 0x192b2b2b2b2b0819, 0x2b08080808080808,
0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08, 0x2b08080808190819,
0x2b08080808191908, 0x2b08080808192b19, 0x2b080808082b0808, 0x2b080808082b1919,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808081919082b,
0x2b08080819191919, 0x2b08080819192b08, 0x2b080808192b0819, 0x2b0808082b080808,
0x2b0808082b081919, 0x2b0808082b190819, 0x2b0808082b191908, 0x2b08081908080819,
0x2b08081908081908, 0x2b08081908082b19, 0x2b08081908190808, 0x2b0808190819082b,
0x2b08081908191919, 0x2b08081908192b08, 0x2b080819082b0819, 0x2b080819082b1908,
0x2b08081919080808, 0x2b0808191908082b, 0x2b08081919081919, 0x2b08081919082b08,
0x2b08081919190819, 0x2b08081919191908, 0x2b0808192b080819, 0x2b0808192b081908,
0x2b0808192b190808, 0x2b0808192b2b2b19, 0x2b08082b08080808, 0x2b08082b08081919,
0x2b08082b08082b2b, 0x2b08082b08190819, 0x2b08082b08191908, 0x2b08082b19080819,
0x2b08082b19081908, 0x2b08082b19190808, 0x2b08190808080819, 0x2b08190808081908,
0x2b0819080808192b, 0x2b08190808082b19, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190808192b08, 0x2b081908082b0819, 0x2b08190819080808,
0x2b0819081908082b, 0x2b08190819081919, 0x2b08190819082b08, 0x2b08190819190819,
0x2b08190819191908, 0x2b081908192b0808, 0x2b0819082b080819, 0x2b0819082b081908,
0x2b0819082b190808, 0x2b08191908080808, 0x2b0819190808082b, 0x2b08191908081919,
0x2b08191908082b08, 0x2b08191908190819, 0x2b08191908191908, 0x2b081919082b0808,
0x2b08191919080819, 0x2b08191919081908, 0x2b08191919190808, 0x2b0819192b080808,
0x2b0819192b082b2b, 0x2b08192b08080819, 0x2b08192b08081908, 0x2b08192b08190808,
0x2b08192b082b2b19, 0x2b08192b19080808, 0x2b082b0808080808, 0x2b082b0808081919,
0x2b082b0808190819, 0x2b082b0808191908, 0x2b082b0819080819, 0x2b082b0819081908,
0x2b082b0819190808, 0x2b082b082b2b082b, 0x2b082b1908080819, 0x2b082b1908081908,
0x2b082b1919080808, 0x2b082b19192b1919, 0x2b082b2b082b082b, 0x2b082b2b19192b08,
0x2b082b2b19192b2b, 0x2b082b2b2b08082b, 0x2b082b2b2b2b082b, 0x2b19080808080819,
0x2b19080808081908, 0x2b19080808082b19, 0x2b19080808190808, 0x2b1908080819082b,
0x2b19080808191919, 0x2b19080808192b08, 0x2b190808082b1908, 0x2b19080819080808,
0x2b1908081908082b, 0x2b19080819081919, 0x2b19080819082b08, 0x2b19080819190819,
0x2b19080819191908, 0x2b190808192b0808, 0x2b1908082b080819, 0x2b1908082b081908,
0x2b1908082b190808, 0x2b19081908080808, 0x2b19081908081919, 0x2b19081908190819,
0x2b19081908191908, 0x2b19081919080819, 0x2b19081919081908, 0x2b19081919190808,
0x2b19081919192b2b, 0x2b19082b08080819, 0x2b19082b08081908, 0x2b19082b08190808,
0x2b19082b19080808, 0x2b19082b2b2b192b, 0x2b19190808080808, 0x2b1919080808082b,
0x2b19190808081919, 0x2b19190808082b08, 0x2b19190808190819, 0x2b19190808191908,
0x2b191908082b0808, 0x2b19190819080819, 0x2b19190819081908, 0x2b19190819190808,
0x2b1919082b080808, 0x2b1919082b19192b, 0x2b19191908080819, 0x2b19191908081908,
0x2b19191908190808, 0x2b19191919080808, 0x2b1919192b192b08, 0x2b1919192b2b0819,
0x2b19192b08080808, 0x2b19192b1908192b, 0x2b19192b192b1908, 0x2b192b0808080819,
0x2b192b0808081908, 0x2b192b0808190808, 0x2b192b08082b192b, 0x2b192b0819080808,
0x2b192b082b2b2b19, 0x2b192b1908080808, 0x2b192b1919082b19, 0x2b192b191919082b,
0x2b192b2b2b190808, 0x2b2b080808080808, 0x2b2b080808081919, 0x2b2b080808082b2b,
0x2b2b080808191908, 0x2b2b0808082b082b, 0x2b2b0808082b2b2b, 0x2b2b080819080819,
0x2b2b080819081908, 0x2b2b080819190808, 0x2b2b08082b2b082b, 0x2b2b08082b2b2b2b,
0x2b2b081919080808, 0x2b2b0819192b1919, 0x2b2b082b0808082b, 0x2b2b082b08082b2b,
0x2b2b082b082b082b, 0x2b2b082b082b2b08, 0x2b2b082b082b2b2b, 0x2b2b082b2b08082b,
0x2b2b082b2b082b08, 0x2b2b082b2b082b2b, 0x2b2b082b2b2b2b08, 0x2b2b190808080819,
0x2b2b190808081908, 0x2b2b190808190808, 0x2b2b190819080808, 0x2b2b19082b082b19,
0x2b2b19082b2b1908, 0x2b2b191908080808, 0x2b2b191908192b19, 0x2b2b192b19190819,
0x2b2b2b0808082b2b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b082b, 0x2b2b2b1919191908,
0x2b2b2b192b08192b, 0x2b2b2b2b08082b08, 0x2b2b2b2b08082b2b, 0x2b2b2b2b082b0808,
0x2b2b2b2b082b082b, 0x2b2b2b2b082b2b08, 0x2b2b2b2b2b082b08, 0x2b2b2b2b2b2b2b2b,
GGML_TABLE_END()
GGML_TABLE_BEGIN(uint32_t, iq3xxs_grid, 256)
0x04040404, 0x04040414, 0x04040424, 0x04040c0c, 0x04040c1c, 0x04040c3e, 0x04041404, 0x04041414,
0x04041c0c, 0x04042414, 0x04043e1c, 0x04043e2c, 0x040c040c, 0x040c041c, 0x040c0c04, 0x040c0c14,
0x040c140c, 0x040c142c, 0x040c1c04, 0x040c1c14, 0x040c240c, 0x040c2c24, 0x040c3e04, 0x04140404,
0x04140414, 0x04140424, 0x04140c0c, 0x04141404, 0x04141414, 0x04141c0c, 0x04141c1c, 0x04141c3e,
0x04142c0c, 0x04142c3e, 0x04143e2c, 0x041c040c, 0x041c043e, 0x041c0c04, 0x041c0c14, 0x041c142c,
0x041c3e04, 0x04240c1c, 0x04241c3e, 0x04242424, 0x04242c3e, 0x04243e1c, 0x04243e2c, 0x042c040c,
0x042c043e, 0x042c1c14, 0x042c2c14, 0x04341c2c, 0x04343424, 0x043e0c04, 0x043e0c24, 0x043e0c34,
0x043e241c, 0x043e340c, 0x0c04040c, 0x0c04041c, 0x0c040c04, 0x0c040c14, 0x0c04140c, 0x0c04141c,
0x0c041c04, 0x0c041c14, 0x0c041c24, 0x0c04243e, 0x0c042c04, 0x0c0c0404, 0x0c0c0414, 0x0c0c0c0c,
0x0c0c1404, 0x0c0c1414, 0x0c14040c, 0x0c14041c, 0x0c140c04, 0x0c140c14, 0x0c14140c, 0x0c141c04,
0x0c143e14, 0x0c1c0404, 0x0c1c0414, 0x0c1c1404, 0x0c1c1c0c, 0x0c1c2434, 0x0c1c3434, 0x0c24040c,
0x0c24042c, 0x0c242c04, 0x0c2c1404, 0x0c2c1424, 0x0c2c2434, 0x0c2c3e0c, 0x0c34042c, 0x0c3e1414,
0x0c3e2404, 0x14040404, 0x14040414, 0x14040c0c, 0x14040c1c, 0x14041404, 0x14041414, 0x14041434,
0x14041c0c, 0x14042414, 0x140c040c, 0x140c041c, 0x140c042c, 0x140c0c04, 0x140c0c14, 0x140c140c,
0x140c1c04, 0x140c341c, 0x140c343e, 0x140c3e04, 0x14140404, 0x14140414, 0x14140c0c, 0x14140c3e,
0x14141404, 0x14141414, 0x14141c3e, 0x14142404, 0x14142c2c, 0x141c040c, 0x141c0c04, 0x141c0c24,
0x141c3e04, 0x141c3e24, 0x14241c2c, 0x14242c1c, 0x142c041c, 0x142c143e, 0x142c240c, 0x142c3e24,
0x143e040c, 0x143e041c, 0x143e0c34, 0x143e242c, 0x1c04040c, 0x1c040c04, 0x1c040c14, 0x1c04140c,
0x1c04141c, 0x1c042c04, 0x1c04342c, 0x1c043e14, 0x1c0c0404, 0x1c0c0414, 0x1c0c1404, 0x1c0c1c0c,
0x1c0c2424, 0x1c0c2434, 0x1c14040c, 0x1c14041c, 0x1c140c04, 0x1c14142c, 0x1c142c14, 0x1c143e14,
0x1c1c0c0c, 0x1c1c1c1c, 0x1c241c04, 0x1c24243e, 0x1c243e14, 0x1c2c0404, 0x1c2c0434, 0x1c2c1414,
0x1c2c2c2c, 0x1c340c24, 0x1c341c34, 0x1c34341c, 0x1c3e1c1c, 0x1c3e3404, 0x24040424, 0x24040c3e,
0x24041c2c, 0x24041c3e, 0x24042c1c, 0x24042c3e, 0x240c3e24, 0x24141404, 0x24141c3e, 0x24142404,
0x24143404, 0x24143434, 0x241c043e, 0x241c242c, 0x24240424, 0x24242c0c, 0x24243424, 0x242c142c,
0x242c241c, 0x242c3e04, 0x243e042c, 0x243e0c04, 0x243e0c14, 0x243e1c04, 0x2c040c14, 0x2c04240c,
0x2c043e04, 0x2c0c0404, 0x2c0c0434, 0x2c0c1434, 0x2c0c2c2c, 0x2c140c24, 0x2c141c14, 0x2c143e14,
0x2c1c0414, 0x2c1c2c1c, 0x2c240c04, 0x2c24141c, 0x2c24143e, 0x2c243e14, 0x2c2c0414, 0x2c2c1c0c,
0x2c342c04, 0x2c3e1424, 0x2c3e2414, 0x34041424, 0x34042424, 0x34042434, 0x34043424, 0x340c140c,
0x340c340c, 0x34140c3e, 0x34143424, 0x341c1c04, 0x341c1c34, 0x34242424, 0x342c042c, 0x342c2c14,
0x34341c1c, 0x343e041c, 0x343e140c, 0x3e04041c, 0x3e04042c, 0x3e04043e, 0x3e040c04, 0x3e041c14,
0x3e042c14, 0x3e0c1434, 0x3e0c2404, 0x3e140c14, 0x3e14242c, 0x3e142c14, 0x3e1c0404, 0x3e1c0c2c,
0x3e1c1c1c, 0x3e1c3404, 0x3e24140c, 0x3e24240c, 0x3e2c0404, 0x3e2c0414, 0x3e2c1424, 0x3e341c04,
GGML_TABLE_END()
GGML_TABLE_BEGIN(uint32_t, iq3s_grid, 512)
0x01010101, 0x01010103, 0x01010105, 0x0101010b, 0x0101010f, 0x01010301, 0x01010303, 0x01010305,
0x01010309, 0x0101030d, 0x01010501, 0x01010503, 0x0101050b, 0x01010707, 0x01010901, 0x01010905,
0x0101090b, 0x0101090f, 0x01010b03, 0x01010b07, 0x01010d01, 0x01010d05, 0x01010f03, 0x01010f09,
0x01010f0f, 0x01030101, 0x01030103, 0x01030105, 0x01030109, 0x01030301, 0x01030303, 0x0103030b,
0x01030501, 0x01030507, 0x0103050f, 0x01030703, 0x0103070b, 0x01030909, 0x01030d03, 0x01030d0b,
0x01030f05, 0x01050101, 0x01050103, 0x0105010b, 0x0105010f, 0x01050301, 0x01050307, 0x0105030d,
0x01050503, 0x0105050b, 0x01050701, 0x01050709, 0x01050905, 0x0105090b, 0x0105090f, 0x01050b03,
0x01050b07, 0x01050f01, 0x01050f07, 0x01070107, 0x01070303, 0x0107030b, 0x01070501, 0x01070505,
0x01070703, 0x01070707, 0x0107070d, 0x01070909, 0x01070b01, 0x01070b05, 0x01070d0f, 0x01070f03,
0x01070f0b, 0x01090101, 0x01090307, 0x0109030f, 0x01090503, 0x01090509, 0x01090705, 0x01090901,
0x01090907, 0x01090b03, 0x01090f01, 0x010b0105, 0x010b0109, 0x010b0501, 0x010b0505, 0x010b050d,
0x010b0707, 0x010b0903, 0x010b090b, 0x010b090f, 0x010b0d0d, 0x010b0f07, 0x010d010d, 0x010d0303,
0x010d0307, 0x010d0703, 0x010d0b05, 0x010d0f03, 0x010f0101, 0x010f0105, 0x010f0109, 0x010f0501,
0x010f0505, 0x010f050d, 0x010f0707, 0x010f0b01, 0x010f0b09, 0x03010101, 0x03010103, 0x03010105,
0x03010109, 0x03010301, 0x03010303, 0x03010307, 0x0301030b, 0x0301030f, 0x03010501, 0x03010505,
0x03010703, 0x03010709, 0x0301070d, 0x03010b09, 0x03010b0d, 0x03010d03, 0x03010f05, 0x03030101,
0x03030103, 0x03030107, 0x0303010d, 0x03030301, 0x03030309, 0x03030503, 0x03030701, 0x03030707,
0x03030903, 0x03030b01, 0x03030b05, 0x03030f01, 0x03030f0d, 0x03050101, 0x03050305, 0x0305030b,
0x0305030f, 0x03050501, 0x03050509, 0x03050705, 0x03050901, 0x03050907, 0x03050b0b, 0x03050d01,
0x03050f05, 0x03070103, 0x03070109, 0x0307010f, 0x03070301, 0x03070307, 0x03070503, 0x0307050f,
0x03070701, 0x03070709, 0x03070903, 0x03070d05, 0x03070f01, 0x03090107, 0x0309010b, 0x03090305,
0x03090309, 0x03090703, 0x03090707, 0x03090905, 0x0309090d, 0x03090b01, 0x03090b09, 0x030b0103,
0x030b0301, 0x030b0307, 0x030b0503, 0x030b0701, 0x030b0705, 0x030b0b03, 0x030d0501, 0x030d0509,
0x030d050f, 0x030d0909, 0x030d090d, 0x030f0103, 0x030f0107, 0x030f0301, 0x030f0305, 0x030f0503,
0x030f070b, 0x030f0903, 0x030f0d05, 0x030f0f01, 0x05010101, 0x05010103, 0x05010107, 0x0501010b,
0x0501010f, 0x05010301, 0x05010305, 0x05010309, 0x0501030d, 0x05010503, 0x05010507, 0x0501050f,
0x05010701, 0x05010705, 0x05010903, 0x05010907, 0x0501090b, 0x05010b01, 0x05010b05, 0x05010d0f,
0x05010f01, 0x05010f07, 0x05010f0b, 0x05030101, 0x05030105, 0x05030301, 0x05030307, 0x0503030f,
0x05030505, 0x0503050b, 0x05030703, 0x05030709, 0x05030905, 0x05030b03, 0x05050103, 0x05050109,
0x0505010f, 0x05050503, 0x05050507, 0x05050701, 0x0505070f, 0x05050903, 0x05050b07, 0x05050b0f,
0x05050f03, 0x05050f09, 0x05070101, 0x05070105, 0x0507010b, 0x05070303, 0x05070505, 0x05070509,
0x05070703, 0x05070707, 0x05070905, 0x05070b01, 0x05070d0d, 0x05090103, 0x0509010f, 0x05090501,
0x05090507, 0x05090705, 0x0509070b, 0x05090903, 0x05090f05, 0x05090f0b, 0x050b0109, 0x050b0303,
0x050b0505, 0x050b070f, 0x050b0901, 0x050b0b07, 0x050b0f01, 0x050d0101, 0x050d0105, 0x050d010f,
0x050d0503, 0x050d0b0b, 0x050d0d03, 0x050f010b, 0x050f0303, 0x050f050d, 0x050f0701, 0x050f0907,
0x050f0b01, 0x07010105, 0x07010303, 0x07010307, 0x0701030b, 0x0701030f, 0x07010505, 0x07010703,
0x07010707, 0x0701070b, 0x07010905, 0x07010909, 0x0701090f, 0x07010b03, 0x07010d07, 0x07010f03,
0x07030103, 0x07030107, 0x0703010b, 0x07030309, 0x07030503, 0x07030507, 0x07030901, 0x07030d01,
0x07030f05, 0x07030f0d, 0x07050101, 0x07050305, 0x07050501, 0x07050705, 0x07050709, 0x07050b01,
0x07070103, 0x07070301, 0x07070309, 0x07070503, 0x07070507, 0x0707050f, 0x07070701, 0x07070903,
0x07070907, 0x0707090f, 0x07070b0b, 0x07070f07, 0x07090107, 0x07090303, 0x0709030d, 0x07090505,
0x07090703, 0x07090b05, 0x07090d01, 0x07090d09, 0x070b0103, 0x070b0301, 0x070b0305, 0x070b050b,
0x070b0705, 0x070b0909, 0x070b0b0d, 0x070b0f07, 0x070d030d, 0x070d0903, 0x070f0103, 0x070f0107,
0x070f0501, 0x070f0505, 0x070f070b, 0x09010101, 0x09010109, 0x09010305, 0x09010501, 0x09010509,
0x0901050f, 0x09010705, 0x09010903, 0x09010b01, 0x09010f01, 0x09030105, 0x0903010f, 0x09030303,
0x09030307, 0x09030505, 0x09030701, 0x0903070b, 0x09030907, 0x09030b03, 0x09030b0b, 0x09050103,
0x09050107, 0x09050301, 0x0905030b, 0x09050503, 0x09050707, 0x09050901, 0x09050b0f, 0x09050d05,
0x09050f01, 0x09070109, 0x09070303, 0x09070307, 0x09070501, 0x09070505, 0x09070703, 0x0907070b,
0x09090101, 0x09090105, 0x09090509, 0x0909070f, 0x09090901, 0x09090f03, 0x090b010b, 0x090b010f,
0x090b0503, 0x090b0d05, 0x090d0307, 0x090d0709, 0x090d0d01, 0x090f0301, 0x090f030b, 0x090f0701,
0x090f0907, 0x090f0b03, 0x0b010105, 0x0b010301, 0x0b010309, 0x0b010505, 0x0b010901, 0x0b010909,
0x0b01090f, 0x0b010b05, 0x0b010d0d, 0x0b010f09, 0x0b030103, 0x0b030107, 0x0b03010b, 0x0b030305,
0x0b030503, 0x0b030705, 0x0b030f05, 0x0b050101, 0x0b050303, 0x0b050507, 0x0b050701, 0x0b05070d,
0x0b050b07, 0x0b070105, 0x0b07010f, 0x0b070301, 0x0b07050f, 0x0b070909, 0x0b070b03, 0x0b070d0b,
0x0b070f07, 0x0b090103, 0x0b090109, 0x0b090501, 0x0b090705, 0x0b09090d, 0x0b0b0305, 0x0b0b050d,
0x0b0b0b03, 0x0b0b0b07, 0x0b0d0905, 0x0b0f0105, 0x0b0f0109, 0x0b0f0505, 0x0d010303, 0x0d010307,
0x0d01030b, 0x0d010703, 0x0d010707, 0x0d010d01, 0x0d030101, 0x0d030501, 0x0d03050f, 0x0d030d09,
0x0d050305, 0x0d050709, 0x0d050905, 0x0d050b0b, 0x0d050d05, 0x0d050f01, 0x0d070101, 0x0d070309,
0x0d070503, 0x0d070901, 0x0d09050b, 0x0d090907, 0x0d090d05, 0x0d0b0101, 0x0d0b0107, 0x0d0b0709,
0x0d0b0d01, 0x0d0d010b, 0x0d0d0901, 0x0d0f0303, 0x0d0f0307, 0x0f010101, 0x0f010109, 0x0f01010f,
0x0f010501, 0x0f010505, 0x0f01070d, 0x0f010901, 0x0f010b09, 0x0f010d05, 0x0f030105, 0x0f030303,
0x0f030509, 0x0f030907, 0x0f03090b, 0x0f050103, 0x0f050109, 0x0f050301, 0x0f05030d, 0x0f050503,
0x0f050701, 0x0f050b03, 0x0f070105, 0x0f070705, 0x0f07070b, 0x0f070b07, 0x0f090103, 0x0f09010b,
0x0f090307, 0x0f090501, 0x0f090b01, 0x0f0b0505, 0x0f0b0905, 0x0f0d0105, 0x0f0d0703, 0x0f0f0101,
GGML_TABLE_END()
#define NGRID_IQ2XXS 512
GGML_TABLE_BEGIN(uint64_t, iq1s_grid, NGRID_IQ2XXS)
0xffffffffffff0101, 0xffffffffff01ff00, 0xffffffffff010100, 0xffffffff00000000,
0xffffffff01ff00ff, 0xffffffff01ff0001, 0xffffffff0101ffff, 0xffffffff0101ff01,
0xffffff00ff000000, 0xffffff000000ff00, 0xffffff00000000ff, 0xffffff0000000100,
0xffffff0000010000, 0xffffff0001000000, 0xffffff01ffff00ff, 0xffffff01ff01ff00,
0xffffff01ff010100, 0xffffff0100000001, 0xffffff0101ffff00, 0xffffff0101ff0101,
0xffffff0101010100, 0xffff00ffff00ff01, 0xffff00ffff0000ff, 0xffff00ff00ff0100,
0xffff00ff0100ff00, 0xffff00ff010001ff, 0xffff0000ff0101ff, 0xffff000000ffff00,
0xffff000000000000, 0xffff00000001ff01, 0xffff000001000101, 0xffff0000010100ff,
0xffff0001ffff0100, 0xffff00010000ff00, 0xffff000100010101, 0xffff000101000000,
0xffff01ffffff0000, 0xffff01ffff01ffff, 0xffff01ffff010100, 0xffff01ff00000000,
0xffff01ff01ffffff, 0xffff01ff01ff0001, 0xffff01ff0101ffff, 0xffff01ff01010001,
0xffff0100ffffff01, 0xffff01000000ffff, 0xffff010000000100, 0xffff010001ff01ff,
0xffff010001000000, 0xffff0101ff000000, 0xffff0101000101ff, 0xffff010101ffff01,
0xffff01010101ff00, 0xff00ffffff000000, 0xff00ffff00ffff00, 0xff00ffff00000001,
0xff00ffff000001ff, 0xff00ffff01010000, 0xff00ff00ffff0000, 0xff00ff00ff00ff00,
0xff00ff00ff0000ff, 0xff00ff00ff000100, 0xff00ff00ff010001, 0xff00ff0000ff0001,
0xff00ff000000ffff, 0xff00ff0000000000, 0xff00ff000001ff00, 0xff00ff0000010100,
0xff00ff0001ff0000, 0xff00ff000100ff00, 0xff00ff0001000100, 0xff00ff01ff000000,
0xff00ff0100ff0000, 0xff00ff01000001ff, 0xff00ff0101010001, 0xff0000ff00000000,
0xff0000ff0001ff00, 0xff0000ff00010100, 0xff000000ffff0101, 0xff000000ff000000,
0xff000000ff01ff00, 0xff00000000ff0000, 0xff0000000000ff00, 0xff000000000000ff,
0xff00000000000000, 0xff00000000000001, 0xff00000000000100, 0xff0000000001ffff,
0xff00000000010000, 0xff00000001000000, 0xff00000001010100, 0xff000001ff00ff01,
0xff000001ff0100ff, 0xff00000100000000, 0xff0000010001ff00, 0xff00000101ff0100,
0xff0000010100ff00, 0xff0001ff00ff00ff, 0xff0001ff00000101, 0xff0001ff000100ff,
0xff0001ff01000000, 0xff000100ff0001ff, 0xff0001000000ff01, 0xff00010000000000,
0xff00010000010001, 0xff00010000010100, 0xff00010001ffff00, 0xff00010001ff0101,
0xff00010001010000, 0xff000101ffffffff, 0xff000101ff000101, 0xff00010101ff00ff,
0xff00010101000001, 0xff000101010100ff, 0xff01ffffff000101, 0xff01ffffff01ffff,
0xff01ffffff01ff01, 0xff01ffffff0101ff, 0xff01ffff00000000, 0xff01ffff01ff0001,
0xff01ffff0101ff01, 0xff01ff00ff000000, 0xff01ff0000ff0100, 0xff01ff000000ff01,
0xff01ff0000010000, 0xff01ff00010000ff, 0xff01ff01ff01ff00, 0xff01ff0100000101,
0xff0100ffffff0000, 0xff0100ffff010000, 0xff0100ff01ff00ff, 0xff0100ff01000100,
0xff0100ff010100ff, 0xff010000ffffff01, 0xff01000000000000, 0xff0100000101ff00,
0xff010001ffff00ff, 0xff010001ff000100, 0xff01000100ffff00, 0xff01000100010001,
0xff01000101ff0001, 0xff010001010001ff, 0xff0101ffffffffff, 0xff0101ffff01ffff,
0xff0101ffff010101, 0xff0101ff0000ff00, 0xff0101ff01010001, 0xff010100ff000000,
0xff010100ff01ff01, 0xff01010000ff0001, 0xff01010000000100, 0xff01010001000000,
0xff0101010100ffff, 0x00ffffff0000ff01, 0x00ffffff000000ff, 0x00ffffff00000100,
0x00ffffff00010000, 0x00ffff00ffff0001, 0x00ffff00ff0000ff, 0x00ffff00ff000100,
0x00ffff0000000000, 0x00ffff0001000100, 0x00ffff0001010001, 0x00ffff01ff00ff01,
0x00ffff0100ff0100, 0x00ffff010000ff00, 0x00ffff01000100ff, 0x00ffff0101ff00ff,
0x00ffff010101ff00, 0x00ff00ffffffffff, 0x00ff00ffffff01ff, 0x00ff00ffff000101,
0x00ff00ff00000000, 0x00ff00ff000101ff, 0x00ff00ff01010101, 0x00ff0000ff000000,
0x00ff0000ff01ffff, 0x00ff000000ff0000, 0x00ff00000000ff00, 0x00ff0000000000ff,
0x00ff000000000000, 0x00ff000000000001, 0x00ff000000000100, 0x00ff000000010000,
0x00ff000001ffff01, 0x00ff000001000000, 0x00ff0001ff000101, 0x00ff000100ffffff,
0x00ff000100000000, 0x00ff0001010001ff, 0x00ff01ffff000000, 0x00ff01ff0001ff00,
0x00ff01ff01ff0100, 0x00ff0100ff01ff01, 0x00ff010000ff00ff, 0x00ff010000ff0101,
0x00ff010000000000, 0x00ff010000010101, 0x00ff01000100ff00, 0x00ff010001010000,
0x00ff0101ffffff00, 0x00ff01010000ff01, 0x00ff010100000100, 0x00ff010101ff0000,
0x0000ffffffff0100, 0x0000ffffff00ff00, 0x0000ffffff0000ff, 0x0000ffffff010000,
0x0000ffff00000000, 0x0000ffff00010101, 0x0000ffff01ffff01, 0x0000ffff01000100,
0x0000ff00ff000000, 0x0000ff00ff01ff00, 0x0000ff00ff0101ff, 0x0000ff0000ff0000,
0x0000ff000000ff00, 0x0000ff00000000ff, 0x0000ff0000000000, 0x0000ff0000000001,
0x0000ff0000000100, 0x0000ff0000010000, 0x0000ff0001ffffff, 0x0000ff0001ff01ff,
0x0000ff0001000000, 0x0000ff000101ffff, 0x0000ff01ffff0101, 0x0000ff01ff010000,
0x0000ff0100000000, 0x0000ff0101000101, 0x000000ffffff0001, 0x000000ffff000000,
0x000000ff00ff0000, 0x000000ff0000ff00, 0x000000ff000000ff, 0x000000ff00000000,
0x000000ff00000001, 0x000000ff00000100, 0x000000ff00010000, 0x000000ff01000000,
0x000000ff0101ff00, 0x00000000ffff0000, 0x00000000ff00ff00, 0x00000000ff0000ff,
0x00000000ff000000, 0x00000000ff000001, 0x00000000ff000100, 0x00000000ff010000,
0x0000000000ffff00, 0x0000000000ff00ff, 0x0000000000ff0000, 0x0000000000ff0001,
0x0000000000ff0100, 0x000000000000ffff, 0x000000000000ff00, 0x000000000000ff01,
0x00000000000000ff, 0x0000000000000001, 0x00000000000001ff, 0x0000000000000100,
0x0000000000000101, 0x000000000001ff00, 0x00000000000100ff, 0x0000000000010000,
0x0000000000010001, 0x0000000000010100, 0x0000000001ff0000, 0x000000000100ff00,
0x00000000010000ff, 0x0000000001000000, 0x0000000001000001, 0x0000000001000100,
0x0000000001010000, 0x00000001ffff01ff, 0x00000001ff000000, 0x0000000100ff0000,
0x000000010000ff00, 0x00000001000000ff, 0x0000000100000000, 0x0000000100000001,
0x0000000100000100, 0x0000000100010000, 0x0000000101000000, 0x000001ffff00ff00,
0x000001ffff010001, 0x000001ffff0101ff, 0x000001ff00ffff01, 0x000001ff0000ffff,
0x000001ff00000000, 0x000001ff010000ff, 0x000001ff01010100, 0x00000100ffff0100,
0x00000100ff000000, 0x0000010000ff0000, 0x000001000000ff00, 0x00000100000000ff,
0x0000010000000000, 0x0000010000000001, 0x0000010000000100, 0x0000010000010000,
0x0000010001000000, 0x000001000101ff01, 0x00000101ffff0001, 0x00000101ff01ffff,
0x0000010100000000, 0x0000010101010100, 0x0001ffffff000000, 0x0001ffff00ffffff,
0x0001ffff00000100, 0x0001ffff0001ff00, 0x0001ffff01000000, 0x0001ff00ffffff00,
0x0001ff00ffff01ff, 0x0001ff00ff010000, 0x0001ff0000000000, 0x0001ff0000010001,
0x0001ff0001ff0000, 0x0001ff0001010100, 0x0001ff01ff0000ff, 0x0001ff01ff000001,
0x0001ff0100ffffff, 0x0001ff010001ffff, 0x0001ff01000101ff, 0x0001ff010100ff01,
0x000100ffff00ffff, 0x000100ffff00ff01, 0x000100ffff000100, 0x000100ff00000000,
0x000100ff000101ff, 0x000100ff01ff0101, 0x000100ff0100ffff, 0x000100ff01010101,
0x00010000ff000000, 0x00010000ff010100, 0x0001000000ff0000, 0x000100000000ff00,
0x00010000000000ff, 0x0001000000000000, 0x0001000000000001, 0x0001000000000100,
0x0001000000010000, 0x0001000001ffff01, 0x0001000001000000, 0x0001000100ff0101,
0x0001000100000000, 0x00010001010100ff, 0x000101ffffff01ff, 0x000101ffffff0101,
0x000101ff00010000, 0x000101ff01ff0000, 0x000101ff0100ff01, 0x00010100ffff0000,
0x0001010000000000, 0x000101000001ffff, 0x0001010000010101, 0x00010100010001ff,
0x00010101ff00ff00, 0x00010101ff010001, 0x0001010100ffffff, 0x0001010100ff01ff,
0x00010101000101ff, 0x0001010101ff0000, 0x000101010100ff01, 0x0001010101000101,
0x01ffffffffff0101, 0x01ffffffff01ffff, 0x01ffffffff01ff01, 0x01ffffffff0101ff,
0x01ffffffff010101, 0x01ffffff00000000, 0x01ffffff01ff01ff, 0x01ffffff01000101,
0x01ffffff0101ff01, 0x01ffffff010100ff, 0x01ffff000000ff00, 0x01ffff0000000001,
0x01ffff00000001ff, 0x01ffff0000010000, 0x01ffff0001ff0000, 0x01ffff01ffffffff,
0x01ffff01ffff01ff, 0x01ffff01ff000000, 0x01ffff01ff01ffff, 0x01ffff01ff0101ff,
0x01ffff010100ffff, 0x01ff00ffffff0000, 0x01ff00ffff010000, 0x01ff00ff00ffff01,
0x01ff0000ff0000ff, 0x01ff000000000000, 0x01ff00000001ff01, 0x01ff000001ffffff,
0x01ff000001010100, 0x01ff0001ffffff01, 0x01ff0001ff010001, 0x01ff000101ff0100,
0x01ff000101000001, 0x01ff0001010100ff, 0x01ff01ffff00ffff, 0x01ff01ff00010001,
0x01ff01ff01000000, 0x01ff01ff010101ff, 0x01ff0100ff000001, 0x01ff010000ffff00,
0x01ff010000000100, 0x01ff010001ff01ff, 0x01ff01000101ffff, 0x01ff0101ffff00ff,
0x01ff0101ffff0101, 0x01ff0101ff0101ff, 0x01ff010100010000, 0x0100ffff00ff00ff,
0x0100ffff00ff0001, 0x0100ffff00000100, 0x0100ffff0100ff00, 0x0100ff00ffff0000,
0x0100ff00ff00ffff, 0x0100ff00ff00ff01, 0x0100ff00ff000100, 0x0100ff00ff010000,
0x0100ff0000000000, 0x0100ff00000100ff, 0x0100ff0001ff0101, 0x0100ff0001010101,
0x0100ff0100ff00ff, 0x0100ff0100ff0001, 0x0100ff0100000100, 0x0100ff0100010001,
0x0100ff0101000000, 0x010000ffff00ff00, 0x010000ff0000ffff, 0x010000ff00000000,
0x010000ff010001ff, 0x010000ff01010001, 0x01000000ffffff00, 0x01000000ffff0101,
0x01000000ff000000, 0x01000000ff0100ff, 0x01000000ff010101, 0x0100000000ff0000,
0x010000000000ff00, 0x01000000000000ff, 0x0100000000000000, 0x0100000000000001,
0x0100000000000100, 0x0100000000010000, 0x0100000001000000, 0x0100000100000000,
0x01000001000101ff, 0x0100000101ffff01, 0x010001ffff000101, 0x010001ff00ff0100,
0x010001ff0000ff00, 0x010001ff000100ff, 0x010001ff01ffffff, 0x01000100ffff0000,
0x01000100ff0001ff, 0x0100010000000000, 0x010001000001ff00, 0x0100010001ff0000,
0x01000100010000ff, 0x0100010001000101, 0x01000101ff00ff01, 0x0100010100ff0100,
0x010001010000ffff, 0x0100010101010001, 0x0101ffffffff0101, 0x0101ffffff0001ff,
0x0101ffffff01ffff, 0x0101ffffff010101, 0x0101ffff00000000, 0x0101ffff0101ffff,
0x0101ffff010101ff, 0x0101ff00ff000000, 0x0101ff0000ff0100, 0x0101ff000000ff00,
0x0101ff0000010000, 0x0101ff00010000ff, 0x0101ff0001000001, 0x0101ff01ff010101,
0x0101ff0100000000, 0x0101ff010101ff00, 0x010100ffffff0000, 0x010100ffff010000,
0x010100ff00ff01ff, 0x010100ff000000ff, 0x010100ff00000101, 0x010100ff01ffff00,
0x01010000ffffff01, 0x01010000ff000100, 0x01010000ff01ff01, 0x0101000000000000,
0x01010000000100ff, 0x010100000101ff01, 0x01010001ffff0000, 0x01010001ff00ffff,
0x01010001ff010000, 0x0101000101ffffff, 0x0101000101ff01ff, 0x0101000101010101,
0x010101ffff01ffff, 0x010101ff00000000, 0x010101ff0001ff01, 0x010101ff0101ffff,
0x010101ff010101ff, 0x01010100ffffffff, 0x01010100ff000001, 0x010101000000ff00,
0x0101010001010000, 0x0101010100ff0001, 0x010101010001ff01, 0x010101010101ffff,
GGML_TABLE_END()
#endif // GGML_COMMON_IMPL

743
ggml-cuda.cu
View File

@ -2,6 +2,9 @@
#include "ggml.h"
#include "ggml-backend-impl.h"
#define GGML_COMMON_IMPL_CUDA
#include "ggml-common.h"
#include <algorithm>
#include <assert.h>
#include <atomic>
@ -1569,746 +1572,6 @@ static __global__ void dequantize_block_q6_K(const void * __restrict__ vx, dst_t
#endif
}
static const __device__ uint64_t iq2xxs_grid[256] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x08080808082b0808,
0x08080808082b082b, 0x08080808082b2b08, 0x08080808082b2b2b, 0x0808080819080819,
0x0808080819081908, 0x0808080819190808, 0x0808080819192b08, 0x08080808192b0819,
0x08080808192b1908, 0x080808082b080808, 0x080808082b08082b, 0x080808082b082b2b,
0x080808082b2b082b, 0x0808081908080819, 0x0808081908081908, 0x0808081908190808,
0x0808081908191919, 0x0808081919080808, 0x080808192b081908, 0x080808192b192b08,
0x0808082b08080808, 0x0808082b0808082b, 0x0808082b082b082b, 0x0808082b2b08082b,
0x0808190808080819, 0x0808190808081908, 0x0808190808190808, 0x08081908082b0819,
0x08081908082b1908, 0x0808190819080808, 0x080819081908082b, 0x0808190819082b08,
0x08081908192b0808, 0x080819082b080819, 0x080819082b081908, 0x080819082b190808,
0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b, 0x0808191908082b08,
0x08081919082b0808, 0x080819191908192b, 0x08081919192b2b19, 0x080819192b080808,
0x080819192b190819, 0x0808192b08082b19, 0x0808192b08190808, 0x0808192b19080808,
0x0808192b2b081908, 0x0808192b2b2b1908, 0x08082b0808080808, 0x08082b0808081919,
0x08082b0808082b08, 0x08082b0808191908, 0x08082b08082b2b08, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b081919082b, 0x08082b082b082b08,
0x08082b1908081908, 0x08082b1919080808, 0x08082b2b0808082b, 0x08082b2b08191908,
0x0819080808080819, 0x0819080808081908, 0x0819080808190808, 0x08190808082b0819,
0x0819080819080808, 0x08190808192b0808, 0x081908082b081908, 0x081908082b190808,
0x081908082b191919, 0x0819081908080808, 0x0819081908082b08, 0x08190819082b0808,
0x0819081919190808, 0x0819081919192b2b, 0x081908192b080808, 0x0819082b082b1908,
0x0819082b19081919, 0x0819190808080808, 0x0819190808082b08, 0x08191908082b0808,
0x08191908082b1919, 0x0819190819082b19, 0x081919082b080808, 0x0819191908192b08,
0x08191919192b082b, 0x0819192b08080808, 0x0819192b0819192b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b0808190808, 0x08192b0819080808, 0x08192b082b080819,
0x08192b1908080808, 0x08192b1908081919, 0x08192b192b2b0808, 0x08192b2b19190819,
0x082b080808080808, 0x082b08080808082b, 0x082b080808082b2b, 0x082b080819081908,
0x082b0808192b0819, 0x082b08082b080808, 0x082b08082b08082b, 0x082b0819082b2b19,
0x082b081919082b08, 0x082b082b08080808, 0x082b082b0808082b, 0x082b190808080819,
0x082b190808081908, 0x082b190808190808, 0x082b190819080808, 0x082b19081919192b,
0x082b191908080808, 0x082b191919080819, 0x082b1919192b1908, 0x082b192b2b190808,
0x082b2b0808082b08, 0x082b2b08082b0808, 0x082b2b082b191908, 0x082b2b2b19081908,
0x1908080808080819, 0x1908080808081908, 0x1908080808190808, 0x1908080808192b08,
0x19080808082b0819, 0x19080808082b1908, 0x1908080819080808, 0x1908080819082b08,
0x190808081919192b, 0x19080808192b0808, 0x190808082b080819, 0x190808082b081908,
0x190808082b190808, 0x1908081908080808, 0x19080819082b0808, 0x19080819192b0819,
0x190808192b080808, 0x190808192b081919, 0x1908082b08080819, 0x1908082b08190808,
0x1908082b19082b08, 0x1908082b1919192b, 0x1908082b192b2b08, 0x1908190808080808,
0x1908190808082b08, 0x19081908082b0808, 0x190819082b080808, 0x190819082b192b19,
0x190819190819082b, 0x19081919082b1908, 0x1908192b08080808, 0x19082b0808080819,
0x19082b0808081908, 0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919,
0x19082b1908080808, 0x19082b1919192b08, 0x19082b19192b0819, 0x19082b192b08082b,
0x19082b2b19081919, 0x19082b2b2b190808, 0x1919080808080808, 0x1919080808082b08,
0x1919080808190819, 0x1919080808192b19, 0x19190808082b0808, 0x191908082b080808,
0x191908082b082b08, 0x1919081908081908, 0x191908191908082b, 0x191908192b2b1908,
0x1919082b2b190819, 0x191919082b190808, 0x191919082b19082b, 0x1919191908082b2b,
0x1919192b08080819, 0x1919192b19191908, 0x19192b0808080808, 0x19192b0808190819,
0x19192b0808192b19, 0x19192b08192b1908, 0x19192b1919080808, 0x19192b2b08082b08,
0x192b080808081908, 0x192b080808190808, 0x192b080819080808, 0x192b0808192b2b08,
0x192b081908080808, 0x192b081919191919, 0x192b082b08192b08, 0x192b082b192b0808,
0x192b190808080808, 0x192b190808081919, 0x192b191908190808, 0x192b19190819082b,
0x192b19192b081908, 0x192b2b081908082b, 0x2b08080808080808, 0x2b0808080808082b,
0x2b08080808082b2b, 0x2b08080819080819, 0x2b0808082b08082b, 0x2b08081908081908,
0x2b08081908192b08, 0x2b08081919080808, 0x2b08082b08190819, 0x2b08190808080819,
0x2b08190808081908, 0x2b08190808190808, 0x2b08190808191919, 0x2b08190819080808,
0x2b081908192b0808, 0x2b08191908080808, 0x2b0819191908192b, 0x2b0819192b191908,
0x2b08192b08082b19, 0x2b08192b19080808, 0x2b08192b192b0808, 0x2b082b080808082b,
0x2b082b1908081908, 0x2b082b2b08190819, 0x2b19080808081908, 0x2b19080808190808,
0x2b190808082b1908, 0x2b19080819080808, 0x2b1908082b2b0819, 0x2b1908190819192b,
0x2b1908192b080808, 0x2b19082b19081919, 0x2b19190808080808, 0x2b191908082b082b,
0x2b19190819081908, 0x2b19191919190819, 0x2b192b082b080819, 0x2b192b19082b0808,
0x2b2b08080808082b, 0x2b2b080819190808, 0x2b2b08082b081919, 0x2b2b081908082b19,
0x2b2b082b08080808, 0x2b2b190808192b08, 0x2b2b2b0819190808, 0x2b2b2b1908081908,
};
static const __device__ uint64_t iq2xs_grid[512] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x080808082b080808,
0x080808082b08082b, 0x080808082b081919, 0x080808082b082b08, 0x080808082b190819,
0x080808082b191908, 0x080808082b192b19, 0x080808082b2b0808, 0x0808081908080819,
0x0808081908081908, 0x080808190808192b, 0x0808081908082b19, 0x0808081908190808,
0x080808190819082b, 0x0808081908191919, 0x0808081908192b08, 0x0808081908192b2b,
0x08080819082b0819, 0x08080819082b1908, 0x0808081919080808, 0x080808191908082b,
0x0808081919081919, 0x0808081919082b08, 0x0808081919190819, 0x0808081919191908,
0x08080819192b0808, 0x08080819192b2b08, 0x080808192b080819, 0x080808192b081908,
0x080808192b190808, 0x0808082b08080808, 0x0808082b0808082b, 0x0808082b08081919,
0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908, 0x0808082b082b0808,
0x0808082b19080819, 0x0808082b19081908, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b082b2b, 0x0808190808080819, 0x0808190808081908,
0x080819080808192b, 0x0808190808082b19, 0x0808190808190808, 0x080819080819082b,
0x0808190808191919, 0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908,
0x0808190819080808, 0x080819081908082b, 0x0808190819081919, 0x0808190819082b08,
0x0808190819190819, 0x0808190819191908, 0x080819081919192b, 0x08081908192b0808,
0x080819082b080819, 0x080819082b081908, 0x080819082b190808, 0x0808191908080808,
0x080819190808082b, 0x0808191908081919, 0x0808191908082b08, 0x0808191908190819,
0x0808191908191908, 0x08081919082b0808, 0x0808191919080819, 0x0808191919081908,
0x0808191919190808, 0x08081919192b0819, 0x080819192b080808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b08190808, 0x0808192b082b192b, 0x0808192b19080808,
0x0808192b1908082b, 0x0808192b2b081908, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808082b2b, 0x08082b0808190819,
0x08082b0808191908, 0x08082b08082b0808, 0x08082b08082b1919, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b0819192b08, 0x08082b082b080808,
0x08082b082b2b0808, 0x08082b082b2b2b2b, 0x08082b1908080819, 0x08082b1908081908,
0x08082b1908190808, 0x08082b1919080808, 0x08082b192b080819, 0x08082b192b082b19,
0x08082b2b08080808, 0x08082b2b082b0808, 0x08082b2b082b2b08, 0x08082b2b2b19192b,
0x08082b2b2b2b0808, 0x0819080808080819, 0x0819080808081908, 0x081908080808192b,
0x0819080808082b19, 0x0819080808190808, 0x081908080819082b, 0x0819080808191919,
0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908, 0x0819080819080808,
0x081908081908082b, 0x0819080819081919, 0x0819080819082b08, 0x0819080819190819,
0x0819080819191908, 0x08190808192b0808, 0x08190808192b2b2b, 0x081908082b080819,
0x081908082b081908, 0x081908082b190808, 0x0819081908080808, 0x081908190808082b,
0x0819081908081919, 0x0819081908082b08, 0x0819081908190819, 0x0819081908191908,
0x08190819082b0808, 0x0819081919080819, 0x0819081919081908, 0x0819081919190808,
0x081908192b080808, 0x081908192b191908, 0x081908192b19192b, 0x0819082b08080819,
0x0819082b08081908, 0x0819082b0808192b, 0x0819082b08190808, 0x0819082b19080808,
0x0819082b192b0808, 0x0819190808080808, 0x081919080808082b, 0x0819190808081919,
0x0819190808082b08, 0x0819190808190819, 0x0819190808191908, 0x08191908082b0808,
0x0819190819080819, 0x0819190819081908, 0x0819190819082b19, 0x0819190819190808,
0x08191908192b1908, 0x081919082b080808, 0x0819191908080819, 0x0819191908081908,
0x0819191908190808, 0x0819191919080808, 0x0819192b08080808, 0x0819192b08191908,
0x0819192b19082b19, 0x08192b0808080819, 0x08192b0808081908, 0x08192b0808190808,
0x08192b080819082b, 0x08192b0819080808, 0x08192b0819191908, 0x08192b082b08192b,
0x08192b1908080808, 0x08192b1908081919, 0x08192b19192b192b, 0x08192b2b19190819,
0x08192b2b2b2b2b19, 0x082b080808080808, 0x082b08080808082b, 0x082b080808081919,
0x082b080808082b08, 0x082b080808082b2b, 0x082b080808190819, 0x082b080808191908,
0x082b0808082b0808, 0x082b080819080819, 0x082b080819081908, 0x082b080819190808,
0x082b08082b080808, 0x082b08082b2b0808, 0x082b081908080819, 0x082b081908081908,
0x082b081908190808, 0x082b081919080808, 0x082b081919082b08, 0x082b0819192b1919,
0x082b082b08080808, 0x082b082b082b082b, 0x082b082b2b080808, 0x082b082b2b2b2b08,
0x082b190808080819, 0x082b190808081908, 0x082b190808190808, 0x082b1908082b2b19,
0x082b190819080808, 0x082b191908080808, 0x082b191919080819, 0x082b19191919082b,
0x082b19192b192b19, 0x082b192b08080819, 0x082b192b08192b2b, 0x082b192b2b2b192b,
0x082b2b0808080808, 0x082b2b0808082b08, 0x082b2b0808082b2b, 0x082b2b08082b0808,
0x082b2b0819191919, 0x082b2b082b082b08, 0x082b2b082b2b082b, 0x082b2b19192b2b08,
0x082b2b192b190808, 0x082b2b2b08082b08, 0x082b2b2b082b0808, 0x082b2b2b2b08082b,
0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819, 0x1908080808081908,
0x190808080808192b, 0x1908080808082b19, 0x1908080808190808, 0x190808080819082b,
0x1908080808191919, 0x1908080808192b08, 0x19080808082b0819, 0x19080808082b1908,
0x1908080819080808, 0x190808081908082b, 0x1908080819081919, 0x1908080819082b08,
0x1908080819082b2b, 0x1908080819190819, 0x1908080819191908, 0x19080808192b0808,
0x19080808192b1919, 0x190808082b080819, 0x190808082b081908, 0x190808082b190808,
0x1908081908080808, 0x190808190808082b, 0x1908081908081919, 0x1908081908082b08,
0x1908081908190819, 0x1908081908191908, 0x19080819082b0808, 0x1908081919080819,
0x1908081919081908, 0x1908081919190808, 0x190808192b080808, 0x190808192b081919,
0x190808192b2b082b, 0x1908082b08080819, 0x1908082b08081908, 0x1908082b08190808,
0x1908082b0819082b, 0x1908082b082b2b19, 0x1908082b19080808, 0x1908190808080808,
0x190819080808082b, 0x1908190808081919, 0x1908190808082b08, 0x1908190808190819,
0x1908190808191908, 0x1908190808192b19, 0x19081908082b0808, 0x1908190819080819,
0x1908190819081908, 0x1908190819190808, 0x190819082b080808, 0x190819082b191908,
0x1908191908080819, 0x1908191908081908, 0x1908191908190808, 0x19081919082b1908,
0x1908191919080808, 0x190819192b192b2b, 0x1908192b08080808, 0x1908192b08082b2b,
0x1908192b19081908, 0x1908192b19190808, 0x19082b0808080819, 0x19082b0808081908,
0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919, 0x19082b0819191908,
0x19082b08192b082b, 0x19082b1908080808, 0x19082b1908190819, 0x19082b1919081908,
0x19082b1919190808, 0x19082b19192b2b19, 0x19082b2b08081908, 0x1919080808080808,
0x191908080808082b, 0x1919080808081919, 0x1919080808082b08, 0x1919080808190819,
0x1919080808191908, 0x19190808082b0808, 0x19190808082b2b08, 0x1919080819080819,
0x1919080819081908, 0x1919080819190808, 0x191908082b080808, 0x1919081908080819,
0x1919081908081908, 0x1919081908190808, 0x1919081908191919, 0x1919081919080808,
0x191908191908082b, 0x1919082b08080808, 0x1919082b19081908, 0x1919082b2b2b2b2b,
0x1919190808080819, 0x1919190808081908, 0x1919190808190808, 0x19191908082b0819,
0x1919190819080808, 0x19191908192b0808, 0x191919082b080819, 0x191919082b2b0819,
0x1919191908080808, 0x1919191908082b08, 0x191919192b080808, 0x191919192b082b08,
0x1919192b082b0819, 0x1919192b192b2b08, 0x1919192b2b2b0819, 0x19192b0808080808,
0x19192b0808191908, 0x19192b0819080819, 0x19192b0819190808, 0x19192b082b192b19,
0x19192b1908192b2b, 0x19192b1919080808, 0x19192b191908082b, 0x19192b2b2b081919,
0x192b080808080819, 0x192b080808081908, 0x192b080808190808, 0x192b080819080808,
0x192b080819191908, 0x192b0808192b082b, 0x192b08082b08192b, 0x192b08082b2b2b19,
0x192b081908080808, 0x192b082b082b1908, 0x192b082b19082b2b, 0x192b082b2b19082b,
0x192b190808080808, 0x192b19080819192b, 0x192b191908190808, 0x192b191919080808,
0x192b191919081919, 0x192b19192b2b1908, 0x192b2b0808080819, 0x192b2b08192b2b2b,
0x192b2b19082b1919, 0x192b2b2b0808192b, 0x192b2b2b19191908, 0x192b2b2b192b082b,
0x2b08080808080808, 0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08,
0x2b08080808190819, 0x2b08080808191908, 0x2b080808082b0808, 0x2b080808082b2b2b,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808082b080808,
0x2b0808082b08082b, 0x2b0808082b2b2b08, 0x2b0808082b2b2b2b, 0x2b08081908080819,
0x2b08081908081908, 0x2b0808190808192b, 0x2b08081908190808, 0x2b08081919080808,
0x2b08081919190819, 0x2b08081919192b19, 0x2b08082b08080808, 0x2b08082b082b0808,
0x2b08082b2b080808, 0x2b08082b2b08082b, 0x2b08082b2b2b0808, 0x2b08082b2b2b2b08,
0x2b08190808080819, 0x2b08190808081908, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190819080808, 0x2b081908192b0808, 0x2b0819082b082b19,
0x2b08191908080808, 0x2b08191919081908, 0x2b0819192b2b1919, 0x2b08192b08192b08,
0x2b08192b192b2b2b, 0x2b082b0808080808, 0x2b082b0808082b08, 0x2b082b08082b1919,
0x2b082b0819192b2b, 0x2b082b082b080808, 0x2b082b082b08082b, 0x2b082b082b2b2b08,
0x2b082b190808192b, 0x2b082b2b082b082b, 0x2b082b2b2b080808, 0x2b082b2b2b082b08,
0x2b082b2b2b19192b, 0x2b082b2b2b2b2b08, 0x2b19080808080819, 0x2b19080808081908,
0x2b19080808190808, 0x2b19080819080808, 0x2b1908081919192b, 0x2b1908082b081908,
0x2b19081908080808, 0x2b190819082b082b, 0x2b190819192b1908, 0x2b19082b1919192b,
0x2b19082b2b082b19, 0x2b19190808080808, 0x2b19190808081919, 0x2b19190819081908,
0x2b19190819190808, 0x2b19190819192b08, 0x2b191919082b2b19, 0x2b1919192b190808,
0x2b1919192b19082b, 0x2b19192b19080819, 0x2b192b0819190819, 0x2b192b082b2b192b,
0x2b192b1919082b19, 0x2b192b2b08191919, 0x2b192b2b192b0808, 0x2b2b080808080808,
0x2b2b08080808082b, 0x2b2b080808082b08, 0x2b2b080808082b2b, 0x2b2b0808082b0808,
0x2b2b0808082b2b2b, 0x2b2b08082b2b0808, 0x2b2b081919190819, 0x2b2b081919192b19,
0x2b2b08192b2b192b, 0x2b2b082b08080808, 0x2b2b082b0808082b, 0x2b2b082b08082b08,
0x2b2b082b082b2b2b, 0x2b2b082b2b080808, 0x2b2b082b2b2b0808, 0x2b2b190819080808,
0x2b2b19082b191919, 0x2b2b192b192b1919, 0x2b2b192b2b192b08, 0x2b2b2b0808082b2b,
0x2b2b2b08082b0808, 0x2b2b2b08082b082b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b0808,
0x2b2b2b082b2b2b08, 0x2b2b2b1908081908, 0x2b2b2b192b081908, 0x2b2b2b192b08192b,
0x2b2b2b2b082b2b08, 0x2b2b2b2b082b2b2b, 0x2b2b2b2b2b190819, 0x2b2b2b2b2b2b2b2b,
};
static const __device__ uint64_t iq2s_grid[1024] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x08080808192b192b,
0x08080808192b2b19, 0x080808082b080808, 0x080808082b08082b, 0x080808082b081919,
0x080808082b082b08, 0x080808082b190819, 0x080808082b191908, 0x080808082b2b0808,
0x080808082b2b1919, 0x080808082b2b2b2b, 0x0808081908080819, 0x0808081908081908,
0x080808190808192b, 0x0808081908082b19, 0x0808081908190808, 0x080808190819082b,
0x0808081908191919, 0x0808081908192b08, 0x08080819082b0819, 0x08080819082b1908,
0x0808081919080808, 0x080808191908082b, 0x0808081919081919, 0x0808081919082b08,
0x0808081919190819, 0x0808081919191908, 0x080808191919192b, 0x0808081919192b19,
0x08080819192b0808, 0x08080819192b1919, 0x08080819192b2b08, 0x080808192b080819,
0x080808192b081908, 0x080808192b190808, 0x080808192b19082b, 0x080808192b191919,
0x080808192b2b0819, 0x080808192b2b1908, 0x0808082b08080808, 0x0808082b0808082b,
0x0808082b08081919, 0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908,
0x0808082b082b0808, 0x0808082b082b2b2b, 0x0808082b19080819, 0x0808082b19081908,
0x0808082b1908192b, 0x0808082b19082b19, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b081919, 0x0808082b2b082b2b, 0x0808082b2b191908,
0x0808082b2b2b082b, 0x0808190808080819, 0x0808190808081908, 0x080819080808192b,
0x0808190808082b19, 0x0808190808190808, 0x080819080819082b, 0x0808190808191919,
0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908, 0x08081908082b192b,
0x08081908082b2b19, 0x0808190819080808, 0x080819081908082b, 0x0808190819081919,
0x0808190819082b08, 0x0808190819082b2b, 0x0808190819190819, 0x0808190819191908,
0x080819081919192b, 0x0808190819192b19, 0x08081908192b0808, 0x08081908192b082b,
0x08081908192b1919, 0x080819082b080819, 0x080819082b081908, 0x080819082b08192b,
0x080819082b082b19, 0x080819082b190808, 0x080819082b191919, 0x080819082b192b08,
0x080819082b2b0819, 0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b,
0x0808191908081919, 0x0808191908082b08, 0x0808191908082b2b, 0x0808191908190819,
0x0808191908191908, 0x080819190819192b, 0x0808191908192b19, 0x08081919082b0808,
0x08081919082b1919, 0x08081919082b2b08, 0x0808191919080819, 0x0808191919081908,
0x080819191908192b, 0x0808191919082b19, 0x0808191919190808, 0x080819191919082b,
0x0808191919191919, 0x0808191919192b08, 0x08081919192b0819, 0x08081919192b1908,
0x080819192b080808, 0x080819192b08082b, 0x080819192b081919, 0x080819192b082b08,
0x080819192b190819, 0x080819192b191908, 0x080819192b2b0808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b0808192b, 0x0808192b08082b19, 0x0808192b08190808,
0x0808192b08191919, 0x0808192b19080808, 0x0808192b19081919, 0x0808192b19082b08,
0x0808192b19190819, 0x0808192b19191908, 0x0808192b192b0808, 0x0808192b2b080819,
0x0808192b2b081908, 0x0808192b2b190808, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808190819, 0x08082b0808191908,
0x08082b080819192b, 0x08082b0808192b19, 0x08082b08082b0808, 0x08082b08082b1919,
0x08082b08082b2b2b, 0x08082b0819080819, 0x08082b0819081908, 0x08082b081908192b,
0x08082b0819082b19, 0x08082b0819190808, 0x08082b081919082b, 0x08082b0819191919,
0x08082b0819192b08, 0x08082b08192b0819, 0x08082b08192b1908, 0x08082b082b080808,
0x08082b082b081919, 0x08082b082b191908, 0x08082b082b2b2b2b, 0x08082b1908080819,
0x08082b1908081908, 0x08082b1908190808, 0x08082b190819082b, 0x08082b1908191919,
0x08082b1908192b08, 0x08082b19082b0819, 0x08082b1919080808, 0x08082b1919081919,
0x08082b1919082b08, 0x08082b1919190819, 0x08082b1919191908, 0x08082b19192b0808,
0x08082b192b080819, 0x08082b192b190808, 0x08082b2b08080808, 0x08082b2b08190819,
0x08082b2b08191908, 0x08082b2b082b082b, 0x08082b2b082b2b08, 0x08082b2b082b2b2b,
0x08082b2b19190808, 0x08082b2b2b192b19, 0x0819080808080819, 0x0819080808081908,
0x081908080808192b, 0x0819080808082b19, 0x0819080808190808, 0x081908080819082b,
0x0819080808191919, 0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908,
0x08190808082b192b, 0x0819080819080808, 0x081908081908082b, 0x0819080819081919,
0x0819080819082b08, 0x0819080819190819, 0x0819080819191908, 0x081908081919192b,
0x0819080819192b19, 0x08190808192b0808, 0x08190808192b082b, 0x08190808192b1919,
0x08190808192b2b08, 0x081908082b080819, 0x081908082b081908, 0x081908082b08192b,
0x081908082b190808, 0x081908082b191919, 0x081908082b192b08, 0x081908082b2b0819,
0x081908082b2b1908, 0x0819081908080808, 0x081908190808082b, 0x0819081908081919,
0x0819081908082b08, 0x0819081908082b2b, 0x0819081908190819, 0x0819081908191908,
0x081908190819192b, 0x0819081908192b19, 0x08190819082b0808, 0x08190819082b082b,
0x08190819082b1919, 0x08190819082b2b08, 0x0819081919080819, 0x0819081919081908,
0x081908191908192b, 0x0819081919082b19, 0x0819081919190808, 0x081908191919082b,
0x0819081919191919, 0x0819081919192b08, 0x08190819192b0819, 0x08190819192b1908,
0x081908192b080808, 0x081908192b08082b, 0x081908192b081919, 0x081908192b082b08,
0x081908192b190819, 0x081908192b191908, 0x0819082b08080819, 0x0819082b08081908,
0x0819082b08082b19, 0x0819082b08190808, 0x0819082b08191919, 0x0819082b082b0819,
0x0819082b082b1908, 0x0819082b19080808, 0x0819082b19081919, 0x0819082b19190819,
0x0819082b19191908, 0x0819082b2b080819, 0x0819082b2b081908, 0x0819082b2b190808,
0x0819190808080808, 0x081919080808082b, 0x0819190808081919, 0x0819190808082b08,
0x0819190808190819, 0x0819190808191908, 0x081919080819192b, 0x0819190808192b19,
0x08191908082b0808, 0x08191908082b1919, 0x08191908082b2b08, 0x0819190819080819,
0x0819190819081908, 0x081919081908192b, 0x0819190819082b19, 0x0819190819190808,
0x081919081919082b, 0x0819190819191919, 0x0819190819192b08, 0x08191908192b0819,
0x08191908192b1908, 0x081919082b080808, 0x081919082b08082b, 0x081919082b081919,
0x081919082b082b08, 0x081919082b190819, 0x081919082b191908, 0x081919082b2b0808,
0x0819191908080819, 0x0819191908081908, 0x081919190808192b, 0x0819191908082b19,
0x0819191908190808, 0x081919190819082b, 0x0819191908191919, 0x0819191908192b08,
0x08191919082b0819, 0x08191919082b1908, 0x0819191919080808, 0x081919191908082b,
0x0819191919081919, 0x0819191919082b08, 0x0819191919190819, 0x0819191919191908,
0x08191919192b0808, 0x081919192b080819, 0x081919192b081908, 0x081919192b190808,
0x0819192b08080808, 0x0819192b08081919, 0x0819192b08082b08, 0x0819192b08190819,
0x0819192b08191908, 0x0819192b082b0808, 0x0819192b19080819, 0x0819192b19081908,
0x0819192b19190808, 0x0819192b2b080808, 0x0819192b2b2b2b2b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b080808192b, 0x08192b0808082b19, 0x08192b0808190808,
0x08192b0808191919, 0x08192b0808192b08, 0x08192b08082b0819, 0x08192b0819080808,
0x08192b081908082b, 0x08192b0819081919, 0x08192b0819082b08, 0x08192b0819190819,
0x08192b0819191908, 0x08192b08192b0808, 0x08192b082b080819, 0x08192b082b081908,
0x08192b1908080808, 0x08192b190808082b, 0x08192b1908081919, 0x08192b1908082b08,
0x08192b1908190819, 0x08192b1908191908, 0x08192b19082b0808, 0x08192b1919080819,
0x08192b1919081908, 0x08192b1919190808, 0x08192b19192b2b19, 0x08192b192b2b082b,
0x08192b2b08081908, 0x08192b2b08190808, 0x08192b2b19080808, 0x08192b2b1919192b,
0x082b080808080808, 0x082b08080808082b, 0x082b080808081919, 0x082b080808082b08,
0x082b080808190819, 0x082b080808191908, 0x082b08080819192b, 0x082b080808192b19,
0x082b0808082b0808, 0x082b0808082b1919, 0x082b0808082b2b2b, 0x082b080819080819,
0x082b080819081908, 0x082b080819190808, 0x082b08081919082b, 0x082b080819191919,
0x082b0808192b1908, 0x082b08082b080808, 0x082b08082b082b2b, 0x082b08082b191908,
0x082b08082b2b2b2b, 0x082b081908080819, 0x082b081908081908, 0x082b081908190808,
0x082b08190819082b, 0x082b081908191919, 0x082b0819082b0819, 0x082b081919080808,
0x082b08191908082b, 0x082b081919081919, 0x082b081919190819, 0x082b081919191908,
0x082b0819192b0808, 0x082b08192b080819, 0x082b08192b081908, 0x082b08192b190808,
0x082b082b08080808, 0x082b082b08082b2b, 0x082b082b082b082b, 0x082b082b082b2b08,
0x082b082b082b2b2b, 0x082b082b19081908, 0x082b082b19190808, 0x082b082b2b082b08,
0x082b082b2b082b2b, 0x082b082b2b2b2b08, 0x082b190808080819, 0x082b190808081908,
0x082b19080808192b, 0x082b190808082b19, 0x082b190808190808, 0x082b190808191919,
0x082b190808192b08, 0x082b1908082b0819, 0x082b1908082b1908, 0x082b190819080808,
0x082b19081908082b, 0x082b190819081919, 0x082b190819082b08, 0x082b190819190819,
0x082b190819191908, 0x082b1908192b0808, 0x082b19082b080819, 0x082b19082b081908,
0x082b19082b190808, 0x082b191908080808, 0x082b191908081919, 0x082b191908082b08,
0x082b191908190819, 0x082b191908191908, 0x082b1919082b0808, 0x082b191919080819,
0x082b191919081908, 0x082b191919190808, 0x082b1919192b192b, 0x082b19192b080808,
0x082b192b08080819, 0x082b192b08081908, 0x082b192b08190808, 0x082b192b19080808,
0x082b192b19192b19, 0x082b2b0808080808, 0x082b2b0808081919, 0x082b2b0808190819,
0x082b2b0808191908, 0x082b2b0819080819, 0x082b2b0819081908, 0x082b2b0819190808,
0x082b2b082b082b2b, 0x082b2b082b2b2b2b, 0x082b2b1908080819, 0x082b2b1908081908,
0x082b2b1908190808, 0x082b2b192b191919, 0x082b2b2b08082b2b, 0x082b2b2b082b082b,
0x082b2b2b192b1908, 0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819,
0x1908080808081908, 0x190808080808192b, 0x1908080808082b19, 0x1908080808190808,
0x190808080819082b, 0x1908080808191919, 0x1908080808192b08, 0x1908080808192b2b,
0x19080808082b0819, 0x19080808082b1908, 0x19080808082b192b, 0x1908080819080808,
0x190808081908082b, 0x1908080819081919, 0x1908080819082b08, 0x1908080819082b2b,
0x1908080819190819, 0x1908080819191908, 0x190808081919192b, 0x1908080819192b19,
0x19080808192b0808, 0x19080808192b082b, 0x19080808192b1919, 0x190808082b080819,
0x190808082b081908, 0x190808082b190808, 0x190808082b191919, 0x190808082b192b08,
0x190808082b2b0819, 0x190808082b2b1908, 0x1908081908080808, 0x190808190808082b,
0x1908081908081919, 0x1908081908082b08, 0x1908081908190819, 0x1908081908191908,
0x190808190819192b, 0x1908081908192b19, 0x19080819082b0808, 0x19080819082b082b,
0x19080819082b1919, 0x1908081919080819, 0x1908081919081908, 0x190808191908192b,
0x1908081919082b19, 0x1908081919190808, 0x190808191919082b, 0x1908081919191919,
0x1908081919192b08, 0x19080819192b0819, 0x19080819192b1908, 0x190808192b080808,
0x190808192b08082b, 0x190808192b081919, 0x190808192b082b08, 0x190808192b190819,
0x190808192b191908, 0x190808192b2b0808, 0x1908082b08080819, 0x1908082b08081908,
0x1908082b08190808, 0x1908082b0819082b, 0x1908082b08191919, 0x1908082b08192b08,
0x1908082b082b1908, 0x1908082b19080808, 0x1908082b19081919, 0x1908082b19082b08,
0x1908082b19190819, 0x1908082b19191908, 0x1908082b192b0808, 0x1908082b2b080819,
0x1908082b2b081908, 0x1908190808080808, 0x190819080808082b, 0x1908190808081919,
0x1908190808082b08, 0x1908190808082b2b, 0x1908190808190819, 0x1908190808191908,
0x190819080819192b, 0x1908190808192b19, 0x19081908082b0808, 0x19081908082b082b,
0x19081908082b1919, 0x19081908082b2b08, 0x1908190819080819, 0x1908190819081908,
0x190819081908192b, 0x1908190819082b19, 0x1908190819190808, 0x190819081919082b,
0x1908190819191919, 0x1908190819192b08, 0x19081908192b0819, 0x19081908192b1908,
0x190819082b080808, 0x190819082b08082b, 0x190819082b081919, 0x190819082b082b08,
0x190819082b190819, 0x190819082b191908, 0x190819082b2b0808, 0x1908191908080819,
0x1908191908081908, 0x190819190808192b, 0x1908191908082b19, 0x1908191908190808,
0x190819190819082b, 0x1908191908191919, 0x1908191908192b08, 0x19081919082b0819,
0x19081919082b1908, 0x1908191919080808, 0x190819191908082b, 0x1908191919081919,
0x1908191919082b08, 0x1908191919190819, 0x1908191919191908, 0x19081919192b0808,
0x19081919192b2b2b, 0x190819192b080819, 0x190819192b081908, 0x190819192b190808,
0x1908192b08080808, 0x1908192b0808082b, 0x1908192b08081919, 0x1908192b08082b08,
0x1908192b08190819, 0x1908192b08191908, 0x1908192b082b0808, 0x1908192b19080819,
0x1908192b19081908, 0x1908192b19190808, 0x1908192b2b080808, 0x1908192b2b2b1919,
0x19082b0808080819, 0x19082b0808081908, 0x19082b0808082b19, 0x19082b0808190808,
0x19082b080819082b, 0x19082b0808191919, 0x19082b0808192b08, 0x19082b08082b0819,
0x19082b08082b1908, 0x19082b0819080808, 0x19082b081908082b, 0x19082b0819081919,
0x19082b0819082b08, 0x19082b0819190819, 0x19082b0819191908, 0x19082b08192b0808,
0x19082b082b081908, 0x19082b082b190808, 0x19082b1908080808, 0x19082b190808082b,
0x19082b1908081919, 0x19082b1908082b08, 0x19082b1908190819, 0x19082b1908191908,
0x19082b19082b0808, 0x19082b1919080819, 0x19082b1919081908, 0x19082b1919190808,
0x19082b192b080808, 0x19082b192b19192b, 0x19082b2b08080819, 0x19082b2b08081908,
0x19082b2b08190808, 0x19082b2b19080808, 0x1919080808080808, 0x191908080808082b,
0x1919080808081919, 0x1919080808082b08, 0x1919080808190819, 0x1919080808191908,
0x191908080819192b, 0x1919080808192b19, 0x19190808082b0808, 0x19190808082b082b,
0x19190808082b1919, 0x19190808082b2b08, 0x1919080819080819, 0x1919080819081908,
0x191908081908192b, 0x1919080819082b19, 0x1919080819190808, 0x191908081919082b,
0x1919080819191919, 0x1919080819192b08, 0x19190808192b0819, 0x19190808192b1908,
0x191908082b080808, 0x191908082b08082b, 0x191908082b081919, 0x191908082b082b08,
0x191908082b190819, 0x191908082b191908, 0x1919081908080819, 0x1919081908081908,
0x191908190808192b, 0x1919081908082b19, 0x1919081908190808, 0x191908190819082b,
0x1919081908191919, 0x1919081908192b08, 0x19190819082b0819, 0x19190819082b1908,
0x1919081919080808, 0x191908191908082b, 0x1919081919081919, 0x1919081919082b08,
0x1919081919190819, 0x1919081919191908, 0x19190819192b0808, 0x191908192b080819,
0x191908192b081908, 0x191908192b190808, 0x1919082b08080808, 0x1919082b08081919,
0x1919082b08082b08, 0x1919082b08190819, 0x1919082b08191908, 0x1919082b082b0808,
0x1919082b19080819, 0x1919082b19081908, 0x1919082b19190808, 0x1919082b192b2b19,
0x1919082b2b080808, 0x1919190808080819, 0x1919190808081908, 0x191919080808192b,
0x1919190808082b19, 0x1919190808190808, 0x191919080819082b, 0x1919190808191919,
0x1919190808192b08, 0x19191908082b0819, 0x19191908082b1908, 0x1919190819080808,
0x191919081908082b, 0x1919190819081919, 0x1919190819082b08, 0x1919190819190819,
0x1919190819191908, 0x19191908192b0808, 0x191919082b080819, 0x191919082b081908,
0x191919082b190808, 0x1919191908080808, 0x191919190808082b, 0x1919191908081919,
0x1919191908082b08, 0x1919191908190819, 0x1919191908191908, 0x19191919082b0808,
0x1919191919080819, 0x1919191919081908, 0x1919191919190808, 0x191919192b080808,
0x1919192b08080819, 0x1919192b08081908, 0x1919192b08190808, 0x1919192b082b192b,
0x1919192b19080808, 0x19192b0808080808, 0x19192b080808082b, 0x19192b0808081919,
0x19192b0808082b08, 0x19192b0808190819, 0x19192b0808191908, 0x19192b08082b0808,
0x19192b0819080819, 0x19192b0819081908, 0x19192b0819190808, 0x19192b0819192b2b,
0x19192b082b080808, 0x19192b1908080819, 0x19192b1908081908, 0x19192b1908190808,
0x19192b1919080808, 0x19192b2b08080808, 0x19192b2b08192b19, 0x19192b2b2b081919,
0x19192b2b2b2b2b08, 0x192b080808080819, 0x192b080808081908, 0x192b08080808192b,
0x192b080808190808, 0x192b08080819082b, 0x192b080808191919, 0x192b080808192b08,
0x192b0808082b0819, 0x192b0808082b1908, 0x192b080819080808, 0x192b080819081919,
0x192b080819082b08, 0x192b080819190819, 0x192b080819191908, 0x192b0808192b0808,
0x192b08082b081908, 0x192b08082b190808, 0x192b081908080808, 0x192b08190808082b,
0x192b081908081919, 0x192b081908082b08, 0x192b081908190819, 0x192b081908191908,
0x192b0819082b0808, 0x192b081919080819, 0x192b081919081908, 0x192b081919190808,
0x192b08192b080808, 0x192b08192b192b19, 0x192b082b08081908, 0x192b082b08190808,
0x192b082b19080808, 0x192b082b1919192b, 0x192b082b2b2b0819, 0x192b190808080808,
0x192b190808081919, 0x192b190808082b08, 0x192b190808190819, 0x192b190808191908,
0x192b1908082b0808, 0x192b190819080819, 0x192b190819081908, 0x192b190819190808,
0x192b19082b080808, 0x192b191908080819, 0x192b191908081908, 0x192b191908190808,
0x192b191919080808, 0x192b191919082b2b, 0x192b1919192b2b08, 0x192b19192b19082b,
0x192b192b08080808, 0x192b192b2b191908, 0x192b2b0808080819, 0x192b2b0808081908,
0x192b2b0808190808, 0x192b2b08192b1919, 0x192b2b082b192b08, 0x192b2b1908080808,
0x192b2b19082b2b2b, 0x192b2b2b1908082b, 0x192b2b2b2b2b0819, 0x2b08080808080808,
0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08, 0x2b08080808190819,
0x2b08080808191908, 0x2b08080808192b19, 0x2b080808082b0808, 0x2b080808082b1919,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808081919082b,
0x2b08080819191919, 0x2b08080819192b08, 0x2b080808192b0819, 0x2b0808082b080808,
0x2b0808082b081919, 0x2b0808082b190819, 0x2b0808082b191908, 0x2b08081908080819,
0x2b08081908081908, 0x2b08081908082b19, 0x2b08081908190808, 0x2b0808190819082b,
0x2b08081908191919, 0x2b08081908192b08, 0x2b080819082b0819, 0x2b080819082b1908,
0x2b08081919080808, 0x2b0808191908082b, 0x2b08081919081919, 0x2b08081919082b08,
0x2b08081919190819, 0x2b08081919191908, 0x2b0808192b080819, 0x2b0808192b081908,
0x2b0808192b190808, 0x2b0808192b2b2b19, 0x2b08082b08080808, 0x2b08082b08081919,
0x2b08082b08082b2b, 0x2b08082b08190819, 0x2b08082b08191908, 0x2b08082b19080819,
0x2b08082b19081908, 0x2b08082b19190808, 0x2b08190808080819, 0x2b08190808081908,
0x2b0819080808192b, 0x2b08190808082b19, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190808192b08, 0x2b081908082b0819, 0x2b08190819080808,
0x2b0819081908082b, 0x2b08190819081919, 0x2b08190819082b08, 0x2b08190819190819,
0x2b08190819191908, 0x2b081908192b0808, 0x2b0819082b080819, 0x2b0819082b081908,
0x2b0819082b190808, 0x2b08191908080808, 0x2b0819190808082b, 0x2b08191908081919,
0x2b08191908082b08, 0x2b08191908190819, 0x2b08191908191908, 0x2b081919082b0808,
0x2b08191919080819, 0x2b08191919081908, 0x2b08191919190808, 0x2b0819192b080808,
0x2b0819192b082b2b, 0x2b08192b08080819, 0x2b08192b08081908, 0x2b08192b08190808,
0x2b08192b082b2b19, 0x2b08192b19080808, 0x2b082b0808080808, 0x2b082b0808081919,
0x2b082b0808190819, 0x2b082b0808191908, 0x2b082b0819080819, 0x2b082b0819081908,
0x2b082b0819190808, 0x2b082b082b2b082b, 0x2b082b1908080819, 0x2b082b1908081908,
0x2b082b1919080808, 0x2b082b19192b1919, 0x2b082b2b082b082b, 0x2b082b2b19192b08,
0x2b082b2b19192b2b, 0x2b082b2b2b08082b, 0x2b082b2b2b2b082b, 0x2b19080808080819,
0x2b19080808081908, 0x2b19080808082b19, 0x2b19080808190808, 0x2b1908080819082b,
0x2b19080808191919, 0x2b19080808192b08, 0x2b190808082b1908, 0x2b19080819080808,
0x2b1908081908082b, 0x2b19080819081919, 0x2b19080819082b08, 0x2b19080819190819,
0x2b19080819191908, 0x2b190808192b0808, 0x2b1908082b080819, 0x2b1908082b081908,
0x2b1908082b190808, 0x2b19081908080808, 0x2b19081908081919, 0x2b19081908190819,
0x2b19081908191908, 0x2b19081919080819, 0x2b19081919081908, 0x2b19081919190808,
0x2b19081919192b2b, 0x2b19082b08080819, 0x2b19082b08081908, 0x2b19082b08190808,
0x2b19082b19080808, 0x2b19082b2b2b192b, 0x2b19190808080808, 0x2b1919080808082b,
0x2b19190808081919, 0x2b19190808082b08, 0x2b19190808190819, 0x2b19190808191908,
0x2b191908082b0808, 0x2b19190819080819, 0x2b19190819081908, 0x2b19190819190808,
0x2b1919082b080808, 0x2b1919082b19192b, 0x2b19191908080819, 0x2b19191908081908,
0x2b19191908190808, 0x2b19191919080808, 0x2b1919192b192b08, 0x2b1919192b2b0819,
0x2b19192b08080808, 0x2b19192b1908192b, 0x2b19192b192b1908, 0x2b192b0808080819,
0x2b192b0808081908, 0x2b192b0808190808, 0x2b192b08082b192b, 0x2b192b0819080808,
0x2b192b082b2b2b19, 0x2b192b1908080808, 0x2b192b1919082b19, 0x2b192b191919082b,
0x2b192b2b2b190808, 0x2b2b080808080808, 0x2b2b080808081919, 0x2b2b080808082b2b,
0x2b2b080808191908, 0x2b2b0808082b082b, 0x2b2b0808082b2b2b, 0x2b2b080819080819,
0x2b2b080819081908, 0x2b2b080819190808, 0x2b2b08082b2b082b, 0x2b2b08082b2b2b2b,
0x2b2b081919080808, 0x2b2b0819192b1919, 0x2b2b082b0808082b, 0x2b2b082b08082b2b,
0x2b2b082b082b082b, 0x2b2b082b082b2b08, 0x2b2b082b082b2b2b, 0x2b2b082b2b08082b,
0x2b2b082b2b082b08, 0x2b2b082b2b082b2b, 0x2b2b082b2b2b2b08, 0x2b2b190808080819,
0x2b2b190808081908, 0x2b2b190808190808, 0x2b2b190819080808, 0x2b2b19082b082b19,
0x2b2b19082b2b1908, 0x2b2b191908080808, 0x2b2b191908192b19, 0x2b2b192b19190819,
0x2b2b2b0808082b2b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b082b, 0x2b2b2b1919191908,
0x2b2b2b192b08192b, 0x2b2b2b2b08082b08, 0x2b2b2b2b08082b2b, 0x2b2b2b2b082b0808,
0x2b2b2b2b082b082b, 0x2b2b2b2b082b2b08, 0x2b2b2b2b2b082b08, 0x2b2b2b2b2b2b2b2b,
};
static const __device__ uint32_t iq3xxs_grid[256] = {
0x04040404, 0x04040414, 0x04040424, 0x04040c0c, 0x04040c1c, 0x04040c3e, 0x04041404, 0x04041414,
0x04041c0c, 0x04042414, 0x04043e1c, 0x04043e2c, 0x040c040c, 0x040c041c, 0x040c0c04, 0x040c0c14,
0x040c140c, 0x040c142c, 0x040c1c04, 0x040c1c14, 0x040c240c, 0x040c2c24, 0x040c3e04, 0x04140404,
0x04140414, 0x04140424, 0x04140c0c, 0x04141404, 0x04141414, 0x04141c0c, 0x04141c1c, 0x04141c3e,
0x04142c0c, 0x04142c3e, 0x04143e2c, 0x041c040c, 0x041c043e, 0x041c0c04, 0x041c0c14, 0x041c142c,
0x041c3e04, 0x04240c1c, 0x04241c3e, 0x04242424, 0x04242c3e, 0x04243e1c, 0x04243e2c, 0x042c040c,
0x042c043e, 0x042c1c14, 0x042c2c14, 0x04341c2c, 0x04343424, 0x043e0c04, 0x043e0c24, 0x043e0c34,
0x043e241c, 0x043e340c, 0x0c04040c, 0x0c04041c, 0x0c040c04, 0x0c040c14, 0x0c04140c, 0x0c04141c,
0x0c041c04, 0x0c041c14, 0x0c041c24, 0x0c04243e, 0x0c042c04, 0x0c0c0404, 0x0c0c0414, 0x0c0c0c0c,
0x0c0c1404, 0x0c0c1414, 0x0c14040c, 0x0c14041c, 0x0c140c04, 0x0c140c14, 0x0c14140c, 0x0c141c04,
0x0c143e14, 0x0c1c0404, 0x0c1c0414, 0x0c1c1404, 0x0c1c1c0c, 0x0c1c2434, 0x0c1c3434, 0x0c24040c,
0x0c24042c, 0x0c242c04, 0x0c2c1404, 0x0c2c1424, 0x0c2c2434, 0x0c2c3e0c, 0x0c34042c, 0x0c3e1414,
0x0c3e2404, 0x14040404, 0x14040414, 0x14040c0c, 0x14040c1c, 0x14041404, 0x14041414, 0x14041434,
0x14041c0c, 0x14042414, 0x140c040c, 0x140c041c, 0x140c042c, 0x140c0c04, 0x140c0c14, 0x140c140c,
0x140c1c04, 0x140c341c, 0x140c343e, 0x140c3e04, 0x14140404, 0x14140414, 0x14140c0c, 0x14140c3e,
0x14141404, 0x14141414, 0x14141c3e, 0x14142404, 0x14142c2c, 0x141c040c, 0x141c0c04, 0x141c0c24,
0x141c3e04, 0x141c3e24, 0x14241c2c, 0x14242c1c, 0x142c041c, 0x142c143e, 0x142c240c, 0x142c3e24,
0x143e040c, 0x143e041c, 0x143e0c34, 0x143e242c, 0x1c04040c, 0x1c040c04, 0x1c040c14, 0x1c04140c,
0x1c04141c, 0x1c042c04, 0x1c04342c, 0x1c043e14, 0x1c0c0404, 0x1c0c0414, 0x1c0c1404, 0x1c0c1c0c,
0x1c0c2424, 0x1c0c2434, 0x1c14040c, 0x1c14041c, 0x1c140c04, 0x1c14142c, 0x1c142c14, 0x1c143e14,
0x1c1c0c0c, 0x1c1c1c1c, 0x1c241c04, 0x1c24243e, 0x1c243e14, 0x1c2c0404, 0x1c2c0434, 0x1c2c1414,
0x1c2c2c2c, 0x1c340c24, 0x1c341c34, 0x1c34341c, 0x1c3e1c1c, 0x1c3e3404, 0x24040424, 0x24040c3e,
0x24041c2c, 0x24041c3e, 0x24042c1c, 0x24042c3e, 0x240c3e24, 0x24141404, 0x24141c3e, 0x24142404,
0x24143404, 0x24143434, 0x241c043e, 0x241c242c, 0x24240424, 0x24242c0c, 0x24243424, 0x242c142c,
0x242c241c, 0x242c3e04, 0x243e042c, 0x243e0c04, 0x243e0c14, 0x243e1c04, 0x2c040c14, 0x2c04240c,
0x2c043e04, 0x2c0c0404, 0x2c0c0434, 0x2c0c1434, 0x2c0c2c2c, 0x2c140c24, 0x2c141c14, 0x2c143e14,
0x2c1c0414, 0x2c1c2c1c, 0x2c240c04, 0x2c24141c, 0x2c24143e, 0x2c243e14, 0x2c2c0414, 0x2c2c1c0c,
0x2c342c04, 0x2c3e1424, 0x2c3e2414, 0x34041424, 0x34042424, 0x34042434, 0x34043424, 0x340c140c,
0x340c340c, 0x34140c3e, 0x34143424, 0x341c1c04, 0x341c1c34, 0x34242424, 0x342c042c, 0x342c2c14,
0x34341c1c, 0x343e041c, 0x343e140c, 0x3e04041c, 0x3e04042c, 0x3e04043e, 0x3e040c04, 0x3e041c14,
0x3e042c14, 0x3e0c1434, 0x3e0c2404, 0x3e140c14, 0x3e14242c, 0x3e142c14, 0x3e1c0404, 0x3e1c0c2c,
0x3e1c1c1c, 0x3e1c3404, 0x3e24140c, 0x3e24240c, 0x3e2c0404, 0x3e2c0414, 0x3e2c1424, 0x3e341c04,
};
static const __device__ uint32_t iq3s_grid[512] = {
0x01010101, 0x01010103, 0x01010105, 0x0101010b, 0x0101010f, 0x01010301, 0x01010303, 0x01010305,
0x01010309, 0x0101030d, 0x01010501, 0x01010503, 0x0101050b, 0x01010707, 0x01010901, 0x01010905,
0x0101090b, 0x0101090f, 0x01010b03, 0x01010b07, 0x01010d01, 0x01010d05, 0x01010f03, 0x01010f09,
0x01010f0f, 0x01030101, 0x01030103, 0x01030105, 0x01030109, 0x01030301, 0x01030303, 0x0103030b,
0x01030501, 0x01030507, 0x0103050f, 0x01030703, 0x0103070b, 0x01030909, 0x01030d03, 0x01030d0b,
0x01030f05, 0x01050101, 0x01050103, 0x0105010b, 0x0105010f, 0x01050301, 0x01050307, 0x0105030d,
0x01050503, 0x0105050b, 0x01050701, 0x01050709, 0x01050905, 0x0105090b, 0x0105090f, 0x01050b03,
0x01050b07, 0x01050f01, 0x01050f07, 0x01070107, 0x01070303, 0x0107030b, 0x01070501, 0x01070505,
0x01070703, 0x01070707, 0x0107070d, 0x01070909, 0x01070b01, 0x01070b05, 0x01070d0f, 0x01070f03,
0x01070f0b, 0x01090101, 0x01090307, 0x0109030f, 0x01090503, 0x01090509, 0x01090705, 0x01090901,
0x01090907, 0x01090b03, 0x01090f01, 0x010b0105, 0x010b0109, 0x010b0501, 0x010b0505, 0x010b050d,
0x010b0707, 0x010b0903, 0x010b090b, 0x010b090f, 0x010b0d0d, 0x010b0f07, 0x010d010d, 0x010d0303,
0x010d0307, 0x010d0703, 0x010d0b05, 0x010d0f03, 0x010f0101, 0x010f0105, 0x010f0109, 0x010f0501,
0x010f0505, 0x010f050d, 0x010f0707, 0x010f0b01, 0x010f0b09, 0x03010101, 0x03010103, 0x03010105,
0x03010109, 0x03010301, 0x03010303, 0x03010307, 0x0301030b, 0x0301030f, 0x03010501, 0x03010505,
0x03010703, 0x03010709, 0x0301070d, 0x03010b09, 0x03010b0d, 0x03010d03, 0x03010f05, 0x03030101,
0x03030103, 0x03030107, 0x0303010d, 0x03030301, 0x03030309, 0x03030503, 0x03030701, 0x03030707,
0x03030903, 0x03030b01, 0x03030b05, 0x03030f01, 0x03030f0d, 0x03050101, 0x03050305, 0x0305030b,
0x0305030f, 0x03050501, 0x03050509, 0x03050705, 0x03050901, 0x03050907, 0x03050b0b, 0x03050d01,
0x03050f05, 0x03070103, 0x03070109, 0x0307010f, 0x03070301, 0x03070307, 0x03070503, 0x0307050f,
0x03070701, 0x03070709, 0x03070903, 0x03070d05, 0x03070f01, 0x03090107, 0x0309010b, 0x03090305,
0x03090309, 0x03090703, 0x03090707, 0x03090905, 0x0309090d, 0x03090b01, 0x03090b09, 0x030b0103,
0x030b0301, 0x030b0307, 0x030b0503, 0x030b0701, 0x030b0705, 0x030b0b03, 0x030d0501, 0x030d0509,
0x030d050f, 0x030d0909, 0x030d090d, 0x030f0103, 0x030f0107, 0x030f0301, 0x030f0305, 0x030f0503,
0x030f070b, 0x030f0903, 0x030f0d05, 0x030f0f01, 0x05010101, 0x05010103, 0x05010107, 0x0501010b,
0x0501010f, 0x05010301, 0x05010305, 0x05010309, 0x0501030d, 0x05010503, 0x05010507, 0x0501050f,
0x05010701, 0x05010705, 0x05010903, 0x05010907, 0x0501090b, 0x05010b01, 0x05010b05, 0x05010d0f,
0x05010f01, 0x05010f07, 0x05010f0b, 0x05030101, 0x05030105, 0x05030301, 0x05030307, 0x0503030f,
0x05030505, 0x0503050b, 0x05030703, 0x05030709, 0x05030905, 0x05030b03, 0x05050103, 0x05050109,
0x0505010f, 0x05050503, 0x05050507, 0x05050701, 0x0505070f, 0x05050903, 0x05050b07, 0x05050b0f,
0x05050f03, 0x05050f09, 0x05070101, 0x05070105, 0x0507010b, 0x05070303, 0x05070505, 0x05070509,
0x05070703, 0x05070707, 0x05070905, 0x05070b01, 0x05070d0d, 0x05090103, 0x0509010f, 0x05090501,
0x05090507, 0x05090705, 0x0509070b, 0x05090903, 0x05090f05, 0x05090f0b, 0x050b0109, 0x050b0303,
0x050b0505, 0x050b070f, 0x050b0901, 0x050b0b07, 0x050b0f01, 0x050d0101, 0x050d0105, 0x050d010f,
0x050d0503, 0x050d0b0b, 0x050d0d03, 0x050f010b, 0x050f0303, 0x050f050d, 0x050f0701, 0x050f0907,
0x050f0b01, 0x07010105, 0x07010303, 0x07010307, 0x0701030b, 0x0701030f, 0x07010505, 0x07010703,
0x07010707, 0x0701070b, 0x07010905, 0x07010909, 0x0701090f, 0x07010b03, 0x07010d07, 0x07010f03,
0x07030103, 0x07030107, 0x0703010b, 0x07030309, 0x07030503, 0x07030507, 0x07030901, 0x07030d01,
0x07030f05, 0x07030f0d, 0x07050101, 0x07050305, 0x07050501, 0x07050705, 0x07050709, 0x07050b01,
0x07070103, 0x07070301, 0x07070309, 0x07070503, 0x07070507, 0x0707050f, 0x07070701, 0x07070903,
0x07070907, 0x0707090f, 0x07070b0b, 0x07070f07, 0x07090107, 0x07090303, 0x0709030d, 0x07090505,
0x07090703, 0x07090b05, 0x07090d01, 0x07090d09, 0x070b0103, 0x070b0301, 0x070b0305, 0x070b050b,
0x070b0705, 0x070b0909, 0x070b0b0d, 0x070b0f07, 0x070d030d, 0x070d0903, 0x070f0103, 0x070f0107,
0x070f0501, 0x070f0505, 0x070f070b, 0x09010101, 0x09010109, 0x09010305, 0x09010501, 0x09010509,
0x0901050f, 0x09010705, 0x09010903, 0x09010b01, 0x09010f01, 0x09030105, 0x0903010f, 0x09030303,
0x09030307, 0x09030505, 0x09030701, 0x0903070b, 0x09030907, 0x09030b03, 0x09030b0b, 0x09050103,
0x09050107, 0x09050301, 0x0905030b, 0x09050503, 0x09050707, 0x09050901, 0x09050b0f, 0x09050d05,
0x09050f01, 0x09070109, 0x09070303, 0x09070307, 0x09070501, 0x09070505, 0x09070703, 0x0907070b,
0x09090101, 0x09090105, 0x09090509, 0x0909070f, 0x09090901, 0x09090f03, 0x090b010b, 0x090b010f,
0x090b0503, 0x090b0d05, 0x090d0307, 0x090d0709, 0x090d0d01, 0x090f0301, 0x090f030b, 0x090f0701,
0x090f0907, 0x090f0b03, 0x0b010105, 0x0b010301, 0x0b010309, 0x0b010505, 0x0b010901, 0x0b010909,
0x0b01090f, 0x0b010b05, 0x0b010d0d, 0x0b010f09, 0x0b030103, 0x0b030107, 0x0b03010b, 0x0b030305,
0x0b030503, 0x0b030705, 0x0b030f05, 0x0b050101, 0x0b050303, 0x0b050507, 0x0b050701, 0x0b05070d,
0x0b050b07, 0x0b070105, 0x0b07010f, 0x0b070301, 0x0b07050f, 0x0b070909, 0x0b070b03, 0x0b070d0b,
0x0b070f07, 0x0b090103, 0x0b090109, 0x0b090501, 0x0b090705, 0x0b09090d, 0x0b0b0305, 0x0b0b050d,
0x0b0b0b03, 0x0b0b0b07, 0x0b0d0905, 0x0b0f0105, 0x0b0f0109, 0x0b0f0505, 0x0d010303, 0x0d010307,
0x0d01030b, 0x0d010703, 0x0d010707, 0x0d010d01, 0x0d030101, 0x0d030501, 0x0d03050f, 0x0d030d09,
0x0d050305, 0x0d050709, 0x0d050905, 0x0d050b0b, 0x0d050d05, 0x0d050f01, 0x0d070101, 0x0d070309,
0x0d070503, 0x0d070901, 0x0d09050b, 0x0d090907, 0x0d090d05, 0x0d0b0101, 0x0d0b0107, 0x0d0b0709,
0x0d0b0d01, 0x0d0d010b, 0x0d0d0901, 0x0d0f0303, 0x0d0f0307, 0x0f010101, 0x0f010109, 0x0f01010f,
0x0f010501, 0x0f010505, 0x0f01070d, 0x0f010901, 0x0f010b09, 0x0f010d05, 0x0f030105, 0x0f030303,
0x0f030509, 0x0f030907, 0x0f03090b, 0x0f050103, 0x0f050109, 0x0f050301, 0x0f05030d, 0x0f050503,
0x0f050701, 0x0f050b03, 0x0f070105, 0x0f070705, 0x0f07070b, 0x0f070b07, 0x0f090103, 0x0f09010b,
0x0f090307, 0x0f090501, 0x0f090b01, 0x0f0b0505, 0x0f0b0905, 0x0f0d0105, 0x0f0d0703, 0x0f0f0101,
};
static const __device__ uint64_t iq1s_grid[512] = {
0xffffffffffff0101, 0xffffffffff01ff00, 0xffffffffff010100, 0xffffffff00000000,
0xffffffff01ff00ff, 0xffffffff01ff0001, 0xffffffff0101ffff, 0xffffffff0101ff01,
0xffffff00ff000000, 0xffffff000000ff00, 0xffffff00000000ff, 0xffffff0000000100,
0xffffff0000010000, 0xffffff0001000000, 0xffffff01ffff00ff, 0xffffff01ff01ff00,
0xffffff01ff010100, 0xffffff0100000001, 0xffffff0101ffff00, 0xffffff0101ff0101,
0xffffff0101010100, 0xffff00ffff00ff01, 0xffff00ffff0000ff, 0xffff00ff00ff0100,
0xffff00ff0100ff00, 0xffff00ff010001ff, 0xffff0000ff0101ff, 0xffff000000ffff00,
0xffff000000000000, 0xffff00000001ff01, 0xffff000001000101, 0xffff0000010100ff,
0xffff0001ffff0100, 0xffff00010000ff00, 0xffff000100010101, 0xffff000101000000,
0xffff01ffffff0000, 0xffff01ffff01ffff, 0xffff01ffff010100, 0xffff01ff00000000,
0xffff01ff01ffffff, 0xffff01ff01ff0001, 0xffff01ff0101ffff, 0xffff01ff01010001,
0xffff0100ffffff01, 0xffff01000000ffff, 0xffff010000000100, 0xffff010001ff01ff,
0xffff010001000000, 0xffff0101ff000000, 0xffff0101000101ff, 0xffff010101ffff01,
0xffff01010101ff00, 0xff00ffffff000000, 0xff00ffff00ffff00, 0xff00ffff00000001,
0xff00ffff000001ff, 0xff00ffff01010000, 0xff00ff00ffff0000, 0xff00ff00ff00ff00,
0xff00ff00ff0000ff, 0xff00ff00ff000100, 0xff00ff00ff010001, 0xff00ff0000ff0001,
0xff00ff000000ffff, 0xff00ff0000000000, 0xff00ff000001ff00, 0xff00ff0000010100,
0xff00ff0001ff0000, 0xff00ff000100ff00, 0xff00ff0001000100, 0xff00ff01ff000000,
0xff00ff0100ff0000, 0xff00ff01000001ff, 0xff00ff0101010001, 0xff0000ff00000000,
0xff0000ff0001ff00, 0xff0000ff00010100, 0xff000000ffff0101, 0xff000000ff000000,
0xff000000ff01ff00, 0xff00000000ff0000, 0xff0000000000ff00, 0xff000000000000ff,
0xff00000000000000, 0xff00000000000001, 0xff00000000000100, 0xff0000000001ffff,
0xff00000000010000, 0xff00000001000000, 0xff00000001010100, 0xff000001ff00ff01,
0xff000001ff0100ff, 0xff00000100000000, 0xff0000010001ff00, 0xff00000101ff0100,
0xff0000010100ff00, 0xff0001ff00ff00ff, 0xff0001ff00000101, 0xff0001ff000100ff,
0xff0001ff01000000, 0xff000100ff0001ff, 0xff0001000000ff01, 0xff00010000000000,
0xff00010000010001, 0xff00010000010100, 0xff00010001ffff00, 0xff00010001ff0101,
0xff00010001010000, 0xff000101ffffffff, 0xff000101ff000101, 0xff00010101ff00ff,
0xff00010101000001, 0xff000101010100ff, 0xff01ffffff000101, 0xff01ffffff01ffff,
0xff01ffffff01ff01, 0xff01ffffff0101ff, 0xff01ffff00000000, 0xff01ffff01ff0001,
0xff01ffff0101ff01, 0xff01ff00ff000000, 0xff01ff0000ff0100, 0xff01ff000000ff01,
0xff01ff0000010000, 0xff01ff00010000ff, 0xff01ff01ff01ff00, 0xff01ff0100000101,
0xff0100ffffff0000, 0xff0100ffff010000, 0xff0100ff01ff00ff, 0xff0100ff01000100,
0xff0100ff010100ff, 0xff010000ffffff01, 0xff01000000000000, 0xff0100000101ff00,
0xff010001ffff00ff, 0xff010001ff000100, 0xff01000100ffff00, 0xff01000100010001,
0xff01000101ff0001, 0xff010001010001ff, 0xff0101ffffffffff, 0xff0101ffff01ffff,
0xff0101ffff010101, 0xff0101ff0000ff00, 0xff0101ff01010001, 0xff010100ff000000,
0xff010100ff01ff01, 0xff01010000ff0001, 0xff01010000000100, 0xff01010001000000,
0xff0101010100ffff, 0x00ffffff0000ff01, 0x00ffffff000000ff, 0x00ffffff00000100,
0x00ffffff00010000, 0x00ffff00ffff0001, 0x00ffff00ff0000ff, 0x00ffff00ff000100,
0x00ffff0000000000, 0x00ffff0001000100, 0x00ffff0001010001, 0x00ffff01ff00ff01,
0x00ffff0100ff0100, 0x00ffff010000ff00, 0x00ffff01000100ff, 0x00ffff0101ff00ff,
0x00ffff010101ff00, 0x00ff00ffffffffff, 0x00ff00ffffff01ff, 0x00ff00ffff000101,
0x00ff00ff00000000, 0x00ff00ff000101ff, 0x00ff00ff01010101, 0x00ff0000ff000000,
0x00ff0000ff01ffff, 0x00ff000000ff0000, 0x00ff00000000ff00, 0x00ff0000000000ff,
0x00ff000000000000, 0x00ff000000000001, 0x00ff000000000100, 0x00ff000000010000,
0x00ff000001ffff01, 0x00ff000001000000, 0x00ff0001ff000101, 0x00ff000100ffffff,
0x00ff000100000000, 0x00ff0001010001ff, 0x00ff01ffff000000, 0x00ff01ff0001ff00,
0x00ff01ff01ff0100, 0x00ff0100ff01ff01, 0x00ff010000ff00ff, 0x00ff010000ff0101,
0x00ff010000000000, 0x00ff010000010101, 0x00ff01000100ff00, 0x00ff010001010000,
0x00ff0101ffffff00, 0x00ff01010000ff01, 0x00ff010100000100, 0x00ff010101ff0000,
0x0000ffffffff0100, 0x0000ffffff00ff00, 0x0000ffffff0000ff, 0x0000ffffff010000,
0x0000ffff00000000, 0x0000ffff00010101, 0x0000ffff01ffff01, 0x0000ffff01000100,
0x0000ff00ff000000, 0x0000ff00ff01ff00, 0x0000ff00ff0101ff, 0x0000ff0000ff0000,
0x0000ff000000ff00, 0x0000ff00000000ff, 0x0000ff0000000000, 0x0000ff0000000001,
0x0000ff0000000100, 0x0000ff0000010000, 0x0000ff0001ffffff, 0x0000ff0001ff01ff,
0x0000ff0001000000, 0x0000ff000101ffff, 0x0000ff01ffff0101, 0x0000ff01ff010000,
0x0000ff0100000000, 0x0000ff0101000101, 0x000000ffffff0001, 0x000000ffff000000,
0x000000ff00ff0000, 0x000000ff0000ff00, 0x000000ff000000ff, 0x000000ff00000000,
0x000000ff00000001, 0x000000ff00000100, 0x000000ff00010000, 0x000000ff01000000,
0x000000ff0101ff00, 0x00000000ffff0000, 0x00000000ff00ff00, 0x00000000ff0000ff,
0x00000000ff000000, 0x00000000ff000001, 0x00000000ff000100, 0x00000000ff010000,
0x0000000000ffff00, 0x0000000000ff00ff, 0x0000000000ff0000, 0x0000000000ff0001,
0x0000000000ff0100, 0x000000000000ffff, 0x000000000000ff00, 0x000000000000ff01,
0x00000000000000ff, 0x0000000000000001, 0x00000000000001ff, 0x0000000000000100,
0x0000000000000101, 0x000000000001ff00, 0x00000000000100ff, 0x0000000000010000,
0x0000000000010001, 0x0000000000010100, 0x0000000001ff0000, 0x000000000100ff00,
0x00000000010000ff, 0x0000000001000000, 0x0000000001000001, 0x0000000001000100,
0x0000000001010000, 0x00000001ffff01ff, 0x00000001ff000000, 0x0000000100ff0000,
0x000000010000ff00, 0x00000001000000ff, 0x0000000100000000, 0x0000000100000001,
0x0000000100000100, 0x0000000100010000, 0x0000000101000000, 0x000001ffff00ff00,
0x000001ffff010001, 0x000001ffff0101ff, 0x000001ff00ffff01, 0x000001ff0000ffff,
0x000001ff00000000, 0x000001ff010000ff, 0x000001ff01010100, 0x00000100ffff0100,
0x00000100ff000000, 0x0000010000ff0000, 0x000001000000ff00, 0x00000100000000ff,
0x0000010000000000, 0x0000010000000001, 0x0000010000000100, 0x0000010000010000,
0x0000010001000000, 0x000001000101ff01, 0x00000101ffff0001, 0x00000101ff01ffff,
0x0000010100000000, 0x0000010101010100, 0x0001ffffff000000, 0x0001ffff00ffffff,
0x0001ffff00000100, 0x0001ffff0001ff00, 0x0001ffff01000000, 0x0001ff00ffffff00,
0x0001ff00ffff01ff, 0x0001ff00ff010000, 0x0001ff0000000000, 0x0001ff0000010001,
0x0001ff0001ff0000, 0x0001ff0001010100, 0x0001ff01ff0000ff, 0x0001ff01ff000001,
0x0001ff0100ffffff, 0x0001ff010001ffff, 0x0001ff01000101ff, 0x0001ff010100ff01,
0x000100ffff00ffff, 0x000100ffff00ff01, 0x000100ffff000100, 0x000100ff00000000,
0x000100ff000101ff, 0x000100ff01ff0101, 0x000100ff0100ffff, 0x000100ff01010101,
0x00010000ff000000, 0x00010000ff010100, 0x0001000000ff0000, 0x000100000000ff00,
0x00010000000000ff, 0x0001000000000000, 0x0001000000000001, 0x0001000000000100,
0x0001000000010000, 0x0001000001ffff01, 0x0001000001000000, 0x0001000100ff0101,
0x0001000100000000, 0x00010001010100ff, 0x000101ffffff01ff, 0x000101ffffff0101,
0x000101ff00010000, 0x000101ff01ff0000, 0x000101ff0100ff01, 0x00010100ffff0000,
0x0001010000000000, 0x000101000001ffff, 0x0001010000010101, 0x00010100010001ff,
0x00010101ff00ff00, 0x00010101ff010001, 0x0001010100ffffff, 0x0001010100ff01ff,
0x00010101000101ff, 0x0001010101ff0000, 0x000101010100ff01, 0x0001010101000101,
0x01ffffffffff0101, 0x01ffffffff01ffff, 0x01ffffffff01ff01, 0x01ffffffff0101ff,
0x01ffffffff010101, 0x01ffffff00000000, 0x01ffffff01ff01ff, 0x01ffffff01000101,
0x01ffffff0101ff01, 0x01ffffff010100ff, 0x01ffff000000ff00, 0x01ffff0000000001,
0x01ffff00000001ff, 0x01ffff0000010000, 0x01ffff0001ff0000, 0x01ffff01ffffffff,
0x01ffff01ffff01ff, 0x01ffff01ff000000, 0x01ffff01ff01ffff, 0x01ffff01ff0101ff,
0x01ffff010100ffff, 0x01ff00ffffff0000, 0x01ff00ffff010000, 0x01ff00ff00ffff01,
0x01ff0000ff0000ff, 0x01ff000000000000, 0x01ff00000001ff01, 0x01ff000001ffffff,
0x01ff000001010100, 0x01ff0001ffffff01, 0x01ff0001ff010001, 0x01ff000101ff0100,
0x01ff000101000001, 0x01ff0001010100ff, 0x01ff01ffff00ffff, 0x01ff01ff00010001,
0x01ff01ff01000000, 0x01ff01ff010101ff, 0x01ff0100ff000001, 0x01ff010000ffff00,
0x01ff010000000100, 0x01ff010001ff01ff, 0x01ff01000101ffff, 0x01ff0101ffff00ff,
0x01ff0101ffff0101, 0x01ff0101ff0101ff, 0x01ff010100010000, 0x0100ffff00ff00ff,
0x0100ffff00ff0001, 0x0100ffff00000100, 0x0100ffff0100ff00, 0x0100ff00ffff0000,
0x0100ff00ff00ffff, 0x0100ff00ff00ff01, 0x0100ff00ff000100, 0x0100ff00ff010000,
0x0100ff0000000000, 0x0100ff00000100ff, 0x0100ff0001ff0101, 0x0100ff0001010101,
0x0100ff0100ff00ff, 0x0100ff0100ff0001, 0x0100ff0100000100, 0x0100ff0100010001,
0x0100ff0101000000, 0x010000ffff00ff00, 0x010000ff0000ffff, 0x010000ff00000000,
0x010000ff010001ff, 0x010000ff01010001, 0x01000000ffffff00, 0x01000000ffff0101,
0x01000000ff000000, 0x01000000ff0100ff, 0x01000000ff010101, 0x0100000000ff0000,
0x010000000000ff00, 0x01000000000000ff, 0x0100000000000000, 0x0100000000000001,
0x0100000000000100, 0x0100000000010000, 0x0100000001000000, 0x0100000100000000,
0x01000001000101ff, 0x0100000101ffff01, 0x010001ffff000101, 0x010001ff00ff0100,
0x010001ff0000ff00, 0x010001ff000100ff, 0x010001ff01ffffff, 0x01000100ffff0000,
0x01000100ff0001ff, 0x0100010000000000, 0x010001000001ff00, 0x0100010001ff0000,
0x01000100010000ff, 0x0100010001000101, 0x01000101ff00ff01, 0x0100010100ff0100,
0x010001010000ffff, 0x0100010101010001, 0x0101ffffffff0101, 0x0101ffffff0001ff,
0x0101ffffff01ffff, 0x0101ffffff010101, 0x0101ffff00000000, 0x0101ffff0101ffff,
0x0101ffff010101ff, 0x0101ff00ff000000, 0x0101ff0000ff0100, 0x0101ff000000ff00,
0x0101ff0000010000, 0x0101ff00010000ff, 0x0101ff0001000001, 0x0101ff01ff010101,
0x0101ff0100000000, 0x0101ff010101ff00, 0x010100ffffff0000, 0x010100ffff010000,
0x010100ff00ff01ff, 0x010100ff000000ff, 0x010100ff00000101, 0x010100ff01ffff00,
0x01010000ffffff01, 0x01010000ff000100, 0x01010000ff01ff01, 0x0101000000000000,
0x01010000000100ff, 0x010100000101ff01, 0x01010001ffff0000, 0x01010001ff00ffff,
0x01010001ff010000, 0x0101000101ffffff, 0x0101000101ff01ff, 0x0101000101010101,
0x010101ffff01ffff, 0x010101ff00000000, 0x010101ff0001ff01, 0x010101ff0101ffff,
0x010101ff010101ff, 0x01010100ffffffff, 0x01010100ff000001, 0x010101000000ff00,
0x0101010001010000, 0x0101010100ff0001, 0x010101010001ff01, 0x010101010101ffff,
};
static const __device__ uint8_t ksigns_iq2xs[128] = {
0, 129, 130, 3, 132, 5, 6, 135, 136, 9, 10, 139, 12, 141, 142, 15,
144, 17, 18, 147, 20, 149, 150, 23, 24, 153, 154, 27, 156, 29, 30, 159,
160, 33, 34, 163, 36, 165, 166, 39, 40, 169, 170, 43, 172, 45, 46, 175,
48, 177, 178, 51, 180, 53, 54, 183, 184, 57, 58, 187, 60, 189, 190, 63,
192, 65, 66, 195, 68, 197, 198, 71, 72, 201, 202, 75, 204, 77, 78, 207,
80, 209, 210, 83, 212, 85, 86, 215, 216, 89, 90, 219, 92, 221, 222, 95,
96, 225, 226, 99, 228, 101, 102, 231, 232, 105, 106, 235, 108, 237, 238, 111,
240, 113, 114, 243, 116, 245, 246, 119, 120, 249, 250, 123, 252, 125, 126, 255,
};
//#if __CUDA_ARCH__ >= MIN_CC_DP4A // lowest compute capability for integer intrinsics
static const __device__ uint64_t ksigns64[128] = {
0x0000000000000000, 0xff000000000000ff, 0xff0000000000ff00, 0x000000000000ffff,
0xff00000000ff0000, 0x0000000000ff00ff, 0x0000000000ffff00, 0xff00000000ffffff,
0xff000000ff000000, 0x00000000ff0000ff, 0x00000000ff00ff00, 0xff000000ff00ffff,
0x00000000ffff0000, 0xff000000ffff00ff, 0xff000000ffffff00, 0x00000000ffffffff,
0xff0000ff00000000, 0x000000ff000000ff, 0x000000ff0000ff00, 0xff0000ff0000ffff,
0x000000ff00ff0000, 0xff0000ff00ff00ff, 0xff0000ff00ffff00, 0x000000ff00ffffff,
0x000000ffff000000, 0xff0000ffff0000ff, 0xff0000ffff00ff00, 0x000000ffff00ffff,
0xff0000ffffff0000, 0x000000ffffff00ff, 0x000000ffffffff00, 0xff0000ffffffffff,
0xff00ff0000000000, 0x0000ff00000000ff, 0x0000ff000000ff00, 0xff00ff000000ffff,
0x0000ff0000ff0000, 0xff00ff0000ff00ff, 0xff00ff0000ffff00, 0x0000ff0000ffffff,
0x0000ff00ff000000, 0xff00ff00ff0000ff, 0xff00ff00ff00ff00, 0x0000ff00ff00ffff,
0xff00ff00ffff0000, 0x0000ff00ffff00ff, 0x0000ff00ffffff00, 0xff00ff00ffffffff,
0x0000ffff00000000, 0xff00ffff000000ff, 0xff00ffff0000ff00, 0x0000ffff0000ffff,
0xff00ffff00ff0000, 0x0000ffff00ff00ff, 0x0000ffff00ffff00, 0xff00ffff00ffffff,
0xff00ffffff000000, 0x0000ffffff0000ff, 0x0000ffffff00ff00, 0xff00ffffff00ffff,
0x0000ffffffff0000, 0xff00ffffffff00ff, 0xff00ffffffffff00, 0x0000ffffffffffff,
0xffff000000000000, 0x00ff0000000000ff, 0x00ff00000000ff00, 0xffff00000000ffff,
0x00ff000000ff0000, 0xffff000000ff00ff, 0xffff000000ffff00, 0x00ff000000ffffff,
0x00ff0000ff000000, 0xffff0000ff0000ff, 0xffff0000ff00ff00, 0x00ff0000ff00ffff,
0xffff0000ffff0000, 0x00ff0000ffff00ff, 0x00ff0000ffffff00, 0xffff0000ffffffff,
0x00ff00ff00000000, 0xffff00ff000000ff, 0xffff00ff0000ff00, 0x00ff00ff0000ffff,
0xffff00ff00ff0000, 0x00ff00ff00ff00ff, 0x00ff00ff00ffff00, 0xffff00ff00ffffff,
0xffff00ffff000000, 0x00ff00ffff0000ff, 0x00ff00ffff00ff00, 0xffff00ffff00ffff,
0x00ff00ffffff0000, 0xffff00ffffff00ff, 0xffff00ffffffff00, 0x00ff00ffffffffff,
0x00ffff0000000000, 0xffffff00000000ff, 0xffffff000000ff00, 0x00ffff000000ffff,
0xffffff0000ff0000, 0x00ffff0000ff00ff, 0x00ffff0000ffff00, 0xffffff0000ffffff,
0xffffff00ff000000, 0x00ffff00ff0000ff, 0x00ffff00ff00ff00, 0xffffff00ff00ffff,
0x00ffff00ffff0000, 0xffffff00ffff00ff, 0xffffff00ffffff00, 0x00ffff00ffffffff,
0xffffffff00000000, 0x00ffffff000000ff, 0x00ffffff0000ff00, 0xffffffff0000ffff,
0x00ffffff00ff0000, 0xffffffff00ff00ff, 0xffffffff00ffff00, 0x00ffffff00ffffff,
0x00ffffffff000000, 0xffffffffff0000ff, 0xffffffffff00ff00, 0x00ffffffff00ffff,
0xffffffffffff0000, 0x00ffffffffff00ff, 0x00ffffffffffff00, 0xffffffffffffffff,
};
//#endif
static const __device__ uint8_t kmask_iq2xs[8] = {1, 2, 4, 8, 16, 32, 64, 128};
inline bool ggml_cuda_supports_mmq(enum ggml_type type) {
switch (type) {
case GGML_TYPE_Q4_0:

6
ggml-metal.m
View File

@ -1642,8 +1642,8 @@ static enum ggml_status ggml_metal_graph_compute(
// TODO: make this more general
GGML_ASSERT(n_as <= 8);
// max size of the src1ids array in the kernel stack
GGML_ASSERT(ne11 <= 512);
// max size of the src1ids array in the kernel shared buffer
GGML_ASSERT(ne11 <= 4096);
const int64_t ne20 = src2 ? src2->ne[0] : 0;
const int64_t ne21 = src2 ? src2->ne[1] : 0;
@ -1741,7 +1741,7 @@ static enum ggml_status ggml_metal_graph_compute(
[encoder setBuffer:id_src_cur offset:offs_src_cur atIndex:19 + j];
}
[encoder setThreadgroupMemoryLength:8192 atIndex:0];
[encoder setThreadgroupMemoryLength:GGML_PAD(8192 + 2*ne11, 16) atIndex:0];
[encoder dispatchThreadgroups:MTLSizeMake((ne11 + 31)/32, (ne21 + 63)/64, n_as*ne12*ne13) threadsPerThreadgroup:MTLSizeMake(128, 1, 1)];
} else {

713
ggml-metal.metal
View File

@ -1,5 +1,8 @@
#include <metal_stdlib>
#define GGML_COMMON_IMPL_METAL
#include "ggml-common.h"
using namespace metal;
#define MAX(x, y) ((x) > (y) ? (x) : (y))
@ -3638,710 +3641,6 @@ kernel void kernel_mul_mv_q6_K_f32(
// ======================= "True" 2-bit
constexpr constant static uint64_t iq2xxs_grid[256] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x08080808082b0808,
0x08080808082b082b, 0x08080808082b2b08, 0x08080808082b2b2b, 0x0808080819080819,
0x0808080819081908, 0x0808080819190808, 0x0808080819192b08, 0x08080808192b0819,
0x08080808192b1908, 0x080808082b080808, 0x080808082b08082b, 0x080808082b082b2b,
0x080808082b2b082b, 0x0808081908080819, 0x0808081908081908, 0x0808081908190808,
0x0808081908191919, 0x0808081919080808, 0x080808192b081908, 0x080808192b192b08,
0x0808082b08080808, 0x0808082b0808082b, 0x0808082b082b082b, 0x0808082b2b08082b,
0x0808190808080819, 0x0808190808081908, 0x0808190808190808, 0x08081908082b0819,
0x08081908082b1908, 0x0808190819080808, 0x080819081908082b, 0x0808190819082b08,
0x08081908192b0808, 0x080819082b080819, 0x080819082b081908, 0x080819082b190808,
0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b, 0x0808191908082b08,
0x08081919082b0808, 0x080819191908192b, 0x08081919192b2b19, 0x080819192b080808,
0x080819192b190819, 0x0808192b08082b19, 0x0808192b08190808, 0x0808192b19080808,
0x0808192b2b081908, 0x0808192b2b2b1908, 0x08082b0808080808, 0x08082b0808081919,
0x08082b0808082b08, 0x08082b0808191908, 0x08082b08082b2b08, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b081919082b, 0x08082b082b082b08,
0x08082b1908081908, 0x08082b1919080808, 0x08082b2b0808082b, 0x08082b2b08191908,
0x0819080808080819, 0x0819080808081908, 0x0819080808190808, 0x08190808082b0819,
0x0819080819080808, 0x08190808192b0808, 0x081908082b081908, 0x081908082b190808,
0x081908082b191919, 0x0819081908080808, 0x0819081908082b08, 0x08190819082b0808,
0x0819081919190808, 0x0819081919192b2b, 0x081908192b080808, 0x0819082b082b1908,
0x0819082b19081919, 0x0819190808080808, 0x0819190808082b08, 0x08191908082b0808,
0x08191908082b1919, 0x0819190819082b19, 0x081919082b080808, 0x0819191908192b08,
0x08191919192b082b, 0x0819192b08080808, 0x0819192b0819192b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b0808190808, 0x08192b0819080808, 0x08192b082b080819,
0x08192b1908080808, 0x08192b1908081919, 0x08192b192b2b0808, 0x08192b2b19190819,
0x082b080808080808, 0x082b08080808082b, 0x082b080808082b2b, 0x082b080819081908,
0x082b0808192b0819, 0x082b08082b080808, 0x082b08082b08082b, 0x082b0819082b2b19,
0x082b081919082b08, 0x082b082b08080808, 0x082b082b0808082b, 0x082b190808080819,
0x082b190808081908, 0x082b190808190808, 0x082b190819080808, 0x082b19081919192b,
0x082b191908080808, 0x082b191919080819, 0x082b1919192b1908, 0x082b192b2b190808,
0x082b2b0808082b08, 0x082b2b08082b0808, 0x082b2b082b191908, 0x082b2b2b19081908,
0x1908080808080819, 0x1908080808081908, 0x1908080808190808, 0x1908080808192b08,
0x19080808082b0819, 0x19080808082b1908, 0x1908080819080808, 0x1908080819082b08,
0x190808081919192b, 0x19080808192b0808, 0x190808082b080819, 0x190808082b081908,
0x190808082b190808, 0x1908081908080808, 0x19080819082b0808, 0x19080819192b0819,
0x190808192b080808, 0x190808192b081919, 0x1908082b08080819, 0x1908082b08190808,
0x1908082b19082b08, 0x1908082b1919192b, 0x1908082b192b2b08, 0x1908190808080808,
0x1908190808082b08, 0x19081908082b0808, 0x190819082b080808, 0x190819082b192b19,
0x190819190819082b, 0x19081919082b1908, 0x1908192b08080808, 0x19082b0808080819,
0x19082b0808081908, 0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919,
0x19082b1908080808, 0x19082b1919192b08, 0x19082b19192b0819, 0x19082b192b08082b,
0x19082b2b19081919, 0x19082b2b2b190808, 0x1919080808080808, 0x1919080808082b08,
0x1919080808190819, 0x1919080808192b19, 0x19190808082b0808, 0x191908082b080808,
0x191908082b082b08, 0x1919081908081908, 0x191908191908082b, 0x191908192b2b1908,
0x1919082b2b190819, 0x191919082b190808, 0x191919082b19082b, 0x1919191908082b2b,
0x1919192b08080819, 0x1919192b19191908, 0x19192b0808080808, 0x19192b0808190819,
0x19192b0808192b19, 0x19192b08192b1908, 0x19192b1919080808, 0x19192b2b08082b08,
0x192b080808081908, 0x192b080808190808, 0x192b080819080808, 0x192b0808192b2b08,
0x192b081908080808, 0x192b081919191919, 0x192b082b08192b08, 0x192b082b192b0808,
0x192b190808080808, 0x192b190808081919, 0x192b191908190808, 0x192b19190819082b,
0x192b19192b081908, 0x192b2b081908082b, 0x2b08080808080808, 0x2b0808080808082b,
0x2b08080808082b2b, 0x2b08080819080819, 0x2b0808082b08082b, 0x2b08081908081908,
0x2b08081908192b08, 0x2b08081919080808, 0x2b08082b08190819, 0x2b08190808080819,
0x2b08190808081908, 0x2b08190808190808, 0x2b08190808191919, 0x2b08190819080808,
0x2b081908192b0808, 0x2b08191908080808, 0x2b0819191908192b, 0x2b0819192b191908,
0x2b08192b08082b19, 0x2b08192b19080808, 0x2b08192b192b0808, 0x2b082b080808082b,
0x2b082b1908081908, 0x2b082b2b08190819, 0x2b19080808081908, 0x2b19080808190808,
0x2b190808082b1908, 0x2b19080819080808, 0x2b1908082b2b0819, 0x2b1908190819192b,
0x2b1908192b080808, 0x2b19082b19081919, 0x2b19190808080808, 0x2b191908082b082b,
0x2b19190819081908, 0x2b19191919190819, 0x2b192b082b080819, 0x2b192b19082b0808,
0x2b2b08080808082b, 0x2b2b080819190808, 0x2b2b08082b081919, 0x2b2b081908082b19,
0x2b2b082b08080808, 0x2b2b190808192b08, 0x2b2b2b0819190808, 0x2b2b2b1908081908,
};
constexpr constant static uint64_t iq2xs_grid[512] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x080808082b080808,
0x080808082b08082b, 0x080808082b081919, 0x080808082b082b08, 0x080808082b190819,
0x080808082b191908, 0x080808082b192b19, 0x080808082b2b0808, 0x0808081908080819,
0x0808081908081908, 0x080808190808192b, 0x0808081908082b19, 0x0808081908190808,
0x080808190819082b, 0x0808081908191919, 0x0808081908192b08, 0x0808081908192b2b,
0x08080819082b0819, 0x08080819082b1908, 0x0808081919080808, 0x080808191908082b,
0x0808081919081919, 0x0808081919082b08, 0x0808081919190819, 0x0808081919191908,
0x08080819192b0808, 0x08080819192b2b08, 0x080808192b080819, 0x080808192b081908,
0x080808192b190808, 0x0808082b08080808, 0x0808082b0808082b, 0x0808082b08081919,
0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908, 0x0808082b082b0808,
0x0808082b19080819, 0x0808082b19081908, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b082b2b, 0x0808190808080819, 0x0808190808081908,
0x080819080808192b, 0x0808190808082b19, 0x0808190808190808, 0x080819080819082b,
0x0808190808191919, 0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908,
0x0808190819080808, 0x080819081908082b, 0x0808190819081919, 0x0808190819082b08,
0x0808190819190819, 0x0808190819191908, 0x080819081919192b, 0x08081908192b0808,
0x080819082b080819, 0x080819082b081908, 0x080819082b190808, 0x0808191908080808,
0x080819190808082b, 0x0808191908081919, 0x0808191908082b08, 0x0808191908190819,
0x0808191908191908, 0x08081919082b0808, 0x0808191919080819, 0x0808191919081908,
0x0808191919190808, 0x08081919192b0819, 0x080819192b080808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b08190808, 0x0808192b082b192b, 0x0808192b19080808,
0x0808192b1908082b, 0x0808192b2b081908, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808082b2b, 0x08082b0808190819,
0x08082b0808191908, 0x08082b08082b0808, 0x08082b08082b1919, 0x08082b0819080819,
0x08082b0819081908, 0x08082b0819190808, 0x08082b0819192b08, 0x08082b082b080808,
0x08082b082b2b0808, 0x08082b082b2b2b2b, 0x08082b1908080819, 0x08082b1908081908,
0x08082b1908190808, 0x08082b1919080808, 0x08082b192b080819, 0x08082b192b082b19,
0x08082b2b08080808, 0x08082b2b082b0808, 0x08082b2b082b2b08, 0x08082b2b2b19192b,
0x08082b2b2b2b0808, 0x0819080808080819, 0x0819080808081908, 0x081908080808192b,
0x0819080808082b19, 0x0819080808190808, 0x081908080819082b, 0x0819080808191919,
0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908, 0x0819080819080808,
0x081908081908082b, 0x0819080819081919, 0x0819080819082b08, 0x0819080819190819,
0x0819080819191908, 0x08190808192b0808, 0x08190808192b2b2b, 0x081908082b080819,
0x081908082b081908, 0x081908082b190808, 0x0819081908080808, 0x081908190808082b,
0x0819081908081919, 0x0819081908082b08, 0x0819081908190819, 0x0819081908191908,
0x08190819082b0808, 0x0819081919080819, 0x0819081919081908, 0x0819081919190808,
0x081908192b080808, 0x081908192b191908, 0x081908192b19192b, 0x0819082b08080819,
0x0819082b08081908, 0x0819082b0808192b, 0x0819082b08190808, 0x0819082b19080808,
0x0819082b192b0808, 0x0819190808080808, 0x081919080808082b, 0x0819190808081919,
0x0819190808082b08, 0x0819190808190819, 0x0819190808191908, 0x08191908082b0808,
0x0819190819080819, 0x0819190819081908, 0x0819190819082b19, 0x0819190819190808,
0x08191908192b1908, 0x081919082b080808, 0x0819191908080819, 0x0819191908081908,
0x0819191908190808, 0x0819191919080808, 0x0819192b08080808, 0x0819192b08191908,
0x0819192b19082b19, 0x08192b0808080819, 0x08192b0808081908, 0x08192b0808190808,
0x08192b080819082b, 0x08192b0819080808, 0x08192b0819191908, 0x08192b082b08192b,
0x08192b1908080808, 0x08192b1908081919, 0x08192b19192b192b, 0x08192b2b19190819,
0x08192b2b2b2b2b19, 0x082b080808080808, 0x082b08080808082b, 0x082b080808081919,
0x082b080808082b08, 0x082b080808082b2b, 0x082b080808190819, 0x082b080808191908,
0x082b0808082b0808, 0x082b080819080819, 0x082b080819081908, 0x082b080819190808,
0x082b08082b080808, 0x082b08082b2b0808, 0x082b081908080819, 0x082b081908081908,
0x082b081908190808, 0x082b081919080808, 0x082b081919082b08, 0x082b0819192b1919,
0x082b082b08080808, 0x082b082b082b082b, 0x082b082b2b080808, 0x082b082b2b2b2b08,
0x082b190808080819, 0x082b190808081908, 0x082b190808190808, 0x082b1908082b2b19,
0x082b190819080808, 0x082b191908080808, 0x082b191919080819, 0x082b19191919082b,
0x082b19192b192b19, 0x082b192b08080819, 0x082b192b08192b2b, 0x082b192b2b2b192b,
0x082b2b0808080808, 0x082b2b0808082b08, 0x082b2b0808082b2b, 0x082b2b08082b0808,
0x082b2b0819191919, 0x082b2b082b082b08, 0x082b2b082b2b082b, 0x082b2b19192b2b08,
0x082b2b192b190808, 0x082b2b2b08082b08, 0x082b2b2b082b0808, 0x082b2b2b2b08082b,
0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819, 0x1908080808081908,
0x190808080808192b, 0x1908080808082b19, 0x1908080808190808, 0x190808080819082b,
0x1908080808191919, 0x1908080808192b08, 0x19080808082b0819, 0x19080808082b1908,
0x1908080819080808, 0x190808081908082b, 0x1908080819081919, 0x1908080819082b08,
0x1908080819082b2b, 0x1908080819190819, 0x1908080819191908, 0x19080808192b0808,
0x19080808192b1919, 0x190808082b080819, 0x190808082b081908, 0x190808082b190808,
0x1908081908080808, 0x190808190808082b, 0x1908081908081919, 0x1908081908082b08,
0x1908081908190819, 0x1908081908191908, 0x19080819082b0808, 0x1908081919080819,
0x1908081919081908, 0x1908081919190808, 0x190808192b080808, 0x190808192b081919,
0x190808192b2b082b, 0x1908082b08080819, 0x1908082b08081908, 0x1908082b08190808,
0x1908082b0819082b, 0x1908082b082b2b19, 0x1908082b19080808, 0x1908190808080808,
0x190819080808082b, 0x1908190808081919, 0x1908190808082b08, 0x1908190808190819,
0x1908190808191908, 0x1908190808192b19, 0x19081908082b0808, 0x1908190819080819,
0x1908190819081908, 0x1908190819190808, 0x190819082b080808, 0x190819082b191908,
0x1908191908080819, 0x1908191908081908, 0x1908191908190808, 0x19081919082b1908,
0x1908191919080808, 0x190819192b192b2b, 0x1908192b08080808, 0x1908192b08082b2b,
0x1908192b19081908, 0x1908192b19190808, 0x19082b0808080819, 0x19082b0808081908,
0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919, 0x19082b0819191908,
0x19082b08192b082b, 0x19082b1908080808, 0x19082b1908190819, 0x19082b1919081908,
0x19082b1919190808, 0x19082b19192b2b19, 0x19082b2b08081908, 0x1919080808080808,
0x191908080808082b, 0x1919080808081919, 0x1919080808082b08, 0x1919080808190819,
0x1919080808191908, 0x19190808082b0808, 0x19190808082b2b08, 0x1919080819080819,
0x1919080819081908, 0x1919080819190808, 0x191908082b080808, 0x1919081908080819,
0x1919081908081908, 0x1919081908190808, 0x1919081908191919, 0x1919081919080808,
0x191908191908082b, 0x1919082b08080808, 0x1919082b19081908, 0x1919082b2b2b2b2b,
0x1919190808080819, 0x1919190808081908, 0x1919190808190808, 0x19191908082b0819,
0x1919190819080808, 0x19191908192b0808, 0x191919082b080819, 0x191919082b2b0819,
0x1919191908080808, 0x1919191908082b08, 0x191919192b080808, 0x191919192b082b08,
0x1919192b082b0819, 0x1919192b192b2b08, 0x1919192b2b2b0819, 0x19192b0808080808,
0x19192b0808191908, 0x19192b0819080819, 0x19192b0819190808, 0x19192b082b192b19,
0x19192b1908192b2b, 0x19192b1919080808, 0x19192b191908082b, 0x19192b2b2b081919,
0x192b080808080819, 0x192b080808081908, 0x192b080808190808, 0x192b080819080808,
0x192b080819191908, 0x192b0808192b082b, 0x192b08082b08192b, 0x192b08082b2b2b19,
0x192b081908080808, 0x192b082b082b1908, 0x192b082b19082b2b, 0x192b082b2b19082b,
0x192b190808080808, 0x192b19080819192b, 0x192b191908190808, 0x192b191919080808,
0x192b191919081919, 0x192b19192b2b1908, 0x192b2b0808080819, 0x192b2b08192b2b2b,
0x192b2b19082b1919, 0x192b2b2b0808192b, 0x192b2b2b19191908, 0x192b2b2b192b082b,
0x2b08080808080808, 0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08,
0x2b08080808190819, 0x2b08080808191908, 0x2b080808082b0808, 0x2b080808082b2b2b,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808082b080808,
0x2b0808082b08082b, 0x2b0808082b2b2b08, 0x2b0808082b2b2b2b, 0x2b08081908080819,
0x2b08081908081908, 0x2b0808190808192b, 0x2b08081908190808, 0x2b08081919080808,
0x2b08081919190819, 0x2b08081919192b19, 0x2b08082b08080808, 0x2b08082b082b0808,
0x2b08082b2b080808, 0x2b08082b2b08082b, 0x2b08082b2b2b0808, 0x2b08082b2b2b2b08,
0x2b08190808080819, 0x2b08190808081908, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190819080808, 0x2b081908192b0808, 0x2b0819082b082b19,
0x2b08191908080808, 0x2b08191919081908, 0x2b0819192b2b1919, 0x2b08192b08192b08,
0x2b08192b192b2b2b, 0x2b082b0808080808, 0x2b082b0808082b08, 0x2b082b08082b1919,
0x2b082b0819192b2b, 0x2b082b082b080808, 0x2b082b082b08082b, 0x2b082b082b2b2b08,
0x2b082b190808192b, 0x2b082b2b082b082b, 0x2b082b2b2b080808, 0x2b082b2b2b082b08,
0x2b082b2b2b19192b, 0x2b082b2b2b2b2b08, 0x2b19080808080819, 0x2b19080808081908,
0x2b19080808190808, 0x2b19080819080808, 0x2b1908081919192b, 0x2b1908082b081908,
0x2b19081908080808, 0x2b190819082b082b, 0x2b190819192b1908, 0x2b19082b1919192b,
0x2b19082b2b082b19, 0x2b19190808080808, 0x2b19190808081919, 0x2b19190819081908,
0x2b19190819190808, 0x2b19190819192b08, 0x2b191919082b2b19, 0x2b1919192b190808,
0x2b1919192b19082b, 0x2b19192b19080819, 0x2b192b0819190819, 0x2b192b082b2b192b,
0x2b192b1919082b19, 0x2b192b2b08191919, 0x2b192b2b192b0808, 0x2b2b080808080808,
0x2b2b08080808082b, 0x2b2b080808082b08, 0x2b2b080808082b2b, 0x2b2b0808082b0808,
0x2b2b0808082b2b2b, 0x2b2b08082b2b0808, 0x2b2b081919190819, 0x2b2b081919192b19,
0x2b2b08192b2b192b, 0x2b2b082b08080808, 0x2b2b082b0808082b, 0x2b2b082b08082b08,
0x2b2b082b082b2b2b, 0x2b2b082b2b080808, 0x2b2b082b2b2b0808, 0x2b2b190819080808,
0x2b2b19082b191919, 0x2b2b192b192b1919, 0x2b2b192b2b192b08, 0x2b2b2b0808082b2b,
0x2b2b2b08082b0808, 0x2b2b2b08082b082b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b0808,
0x2b2b2b082b2b2b08, 0x2b2b2b1908081908, 0x2b2b2b192b081908, 0x2b2b2b192b08192b,
0x2b2b2b2b082b2b08, 0x2b2b2b2b082b2b2b, 0x2b2b2b2b2b190819, 0x2b2b2b2b2b2b2b2b,
};
constexpr constant static uint64_t iq2s_grid[1024] = {
0x0808080808080808, 0x080808080808082b, 0x0808080808081919, 0x0808080808082b08,
0x0808080808082b2b, 0x0808080808190819, 0x0808080808191908, 0x080808080819192b,
0x0808080808192b19, 0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908, 0x080808081908192b,
0x0808080819082b19, 0x0808080819190808, 0x080808081919082b, 0x0808080819191919,
0x0808080819192b08, 0x08080808192b0819, 0x08080808192b1908, 0x08080808192b192b,
0x08080808192b2b19, 0x080808082b080808, 0x080808082b08082b, 0x080808082b081919,
0x080808082b082b08, 0x080808082b190819, 0x080808082b191908, 0x080808082b2b0808,
0x080808082b2b1919, 0x080808082b2b2b2b, 0x0808081908080819, 0x0808081908081908,
0x080808190808192b, 0x0808081908082b19, 0x0808081908190808, 0x080808190819082b,
0x0808081908191919, 0x0808081908192b08, 0x08080819082b0819, 0x08080819082b1908,
0x0808081919080808, 0x080808191908082b, 0x0808081919081919, 0x0808081919082b08,
0x0808081919190819, 0x0808081919191908, 0x080808191919192b, 0x0808081919192b19,
0x08080819192b0808, 0x08080819192b1919, 0x08080819192b2b08, 0x080808192b080819,
0x080808192b081908, 0x080808192b190808, 0x080808192b19082b, 0x080808192b191919,
0x080808192b2b0819, 0x080808192b2b1908, 0x0808082b08080808, 0x0808082b0808082b,
0x0808082b08081919, 0x0808082b08082b08, 0x0808082b08190819, 0x0808082b08191908,
0x0808082b082b0808, 0x0808082b082b2b2b, 0x0808082b19080819, 0x0808082b19081908,
0x0808082b1908192b, 0x0808082b19082b19, 0x0808082b19190808, 0x0808082b19191919,
0x0808082b2b080808, 0x0808082b2b081919, 0x0808082b2b082b2b, 0x0808082b2b191908,
0x0808082b2b2b082b, 0x0808190808080819, 0x0808190808081908, 0x080819080808192b,
0x0808190808082b19, 0x0808190808190808, 0x080819080819082b, 0x0808190808191919,
0x0808190808192b08, 0x08081908082b0819, 0x08081908082b1908, 0x08081908082b192b,
0x08081908082b2b19, 0x0808190819080808, 0x080819081908082b, 0x0808190819081919,
0x0808190819082b08, 0x0808190819082b2b, 0x0808190819190819, 0x0808190819191908,
0x080819081919192b, 0x0808190819192b19, 0x08081908192b0808, 0x08081908192b082b,
0x08081908192b1919, 0x080819082b080819, 0x080819082b081908, 0x080819082b08192b,
0x080819082b082b19, 0x080819082b190808, 0x080819082b191919, 0x080819082b192b08,
0x080819082b2b0819, 0x080819082b2b1908, 0x0808191908080808, 0x080819190808082b,
0x0808191908081919, 0x0808191908082b08, 0x0808191908082b2b, 0x0808191908190819,
0x0808191908191908, 0x080819190819192b, 0x0808191908192b19, 0x08081919082b0808,
0x08081919082b1919, 0x08081919082b2b08, 0x0808191919080819, 0x0808191919081908,
0x080819191908192b, 0x0808191919082b19, 0x0808191919190808, 0x080819191919082b,
0x0808191919191919, 0x0808191919192b08, 0x08081919192b0819, 0x08081919192b1908,
0x080819192b080808, 0x080819192b08082b, 0x080819192b081919, 0x080819192b082b08,
0x080819192b190819, 0x080819192b191908, 0x080819192b2b0808, 0x0808192b08080819,
0x0808192b08081908, 0x0808192b0808192b, 0x0808192b08082b19, 0x0808192b08190808,
0x0808192b08191919, 0x0808192b19080808, 0x0808192b19081919, 0x0808192b19082b08,
0x0808192b19190819, 0x0808192b19191908, 0x0808192b192b0808, 0x0808192b2b080819,
0x0808192b2b081908, 0x0808192b2b190808, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808190819, 0x08082b0808191908,
0x08082b080819192b, 0x08082b0808192b19, 0x08082b08082b0808, 0x08082b08082b1919,
0x08082b08082b2b2b, 0x08082b0819080819, 0x08082b0819081908, 0x08082b081908192b,
0x08082b0819082b19, 0x08082b0819190808, 0x08082b081919082b, 0x08082b0819191919,
0x08082b0819192b08, 0x08082b08192b0819, 0x08082b08192b1908, 0x08082b082b080808,
0x08082b082b081919, 0x08082b082b191908, 0x08082b082b2b2b2b, 0x08082b1908080819,
0x08082b1908081908, 0x08082b1908190808, 0x08082b190819082b, 0x08082b1908191919,
0x08082b1908192b08, 0x08082b19082b0819, 0x08082b1919080808, 0x08082b1919081919,
0x08082b1919082b08, 0x08082b1919190819, 0x08082b1919191908, 0x08082b19192b0808,
0x08082b192b080819, 0x08082b192b190808, 0x08082b2b08080808, 0x08082b2b08190819,
0x08082b2b08191908, 0x08082b2b082b082b, 0x08082b2b082b2b08, 0x08082b2b082b2b2b,
0x08082b2b19190808, 0x08082b2b2b192b19, 0x0819080808080819, 0x0819080808081908,
0x081908080808192b, 0x0819080808082b19, 0x0819080808190808, 0x081908080819082b,
0x0819080808191919, 0x0819080808192b08, 0x08190808082b0819, 0x08190808082b1908,
0x08190808082b192b, 0x0819080819080808, 0x081908081908082b, 0x0819080819081919,
0x0819080819082b08, 0x0819080819190819, 0x0819080819191908, 0x081908081919192b,
0x0819080819192b19, 0x08190808192b0808, 0x08190808192b082b, 0x08190808192b1919,
0x08190808192b2b08, 0x081908082b080819, 0x081908082b081908, 0x081908082b08192b,
0x081908082b190808, 0x081908082b191919, 0x081908082b192b08, 0x081908082b2b0819,
0x081908082b2b1908, 0x0819081908080808, 0x081908190808082b, 0x0819081908081919,
0x0819081908082b08, 0x0819081908082b2b, 0x0819081908190819, 0x0819081908191908,
0x081908190819192b, 0x0819081908192b19, 0x08190819082b0808, 0x08190819082b082b,
0x08190819082b1919, 0x08190819082b2b08, 0x0819081919080819, 0x0819081919081908,
0x081908191908192b, 0x0819081919082b19, 0x0819081919190808, 0x081908191919082b,
0x0819081919191919, 0x0819081919192b08, 0x08190819192b0819, 0x08190819192b1908,
0x081908192b080808, 0x081908192b08082b, 0x081908192b081919, 0x081908192b082b08,
0x081908192b190819, 0x081908192b191908, 0x0819082b08080819, 0x0819082b08081908,
0x0819082b08082b19, 0x0819082b08190808, 0x0819082b08191919, 0x0819082b082b0819,
0x0819082b082b1908, 0x0819082b19080808, 0x0819082b19081919, 0x0819082b19190819,
0x0819082b19191908, 0x0819082b2b080819, 0x0819082b2b081908, 0x0819082b2b190808,
0x0819190808080808, 0x081919080808082b, 0x0819190808081919, 0x0819190808082b08,
0x0819190808190819, 0x0819190808191908, 0x081919080819192b, 0x0819190808192b19,
0x08191908082b0808, 0x08191908082b1919, 0x08191908082b2b08, 0x0819190819080819,
0x0819190819081908, 0x081919081908192b, 0x0819190819082b19, 0x0819190819190808,
0x081919081919082b, 0x0819190819191919, 0x0819190819192b08, 0x08191908192b0819,
0x08191908192b1908, 0x081919082b080808, 0x081919082b08082b, 0x081919082b081919,
0x081919082b082b08, 0x081919082b190819, 0x081919082b191908, 0x081919082b2b0808,
0x0819191908080819, 0x0819191908081908, 0x081919190808192b, 0x0819191908082b19,
0x0819191908190808, 0x081919190819082b, 0x0819191908191919, 0x0819191908192b08,
0x08191919082b0819, 0x08191919082b1908, 0x0819191919080808, 0x081919191908082b,
0x0819191919081919, 0x0819191919082b08, 0x0819191919190819, 0x0819191919191908,
0x08191919192b0808, 0x081919192b080819, 0x081919192b081908, 0x081919192b190808,
0x0819192b08080808, 0x0819192b08081919, 0x0819192b08082b08, 0x0819192b08190819,
0x0819192b08191908, 0x0819192b082b0808, 0x0819192b19080819, 0x0819192b19081908,
0x0819192b19190808, 0x0819192b2b080808, 0x0819192b2b2b2b2b, 0x08192b0808080819,
0x08192b0808081908, 0x08192b080808192b, 0x08192b0808082b19, 0x08192b0808190808,
0x08192b0808191919, 0x08192b0808192b08, 0x08192b08082b0819, 0x08192b0819080808,
0x08192b081908082b, 0x08192b0819081919, 0x08192b0819082b08, 0x08192b0819190819,
0x08192b0819191908, 0x08192b08192b0808, 0x08192b082b080819, 0x08192b082b081908,
0x08192b1908080808, 0x08192b190808082b, 0x08192b1908081919, 0x08192b1908082b08,
0x08192b1908190819, 0x08192b1908191908, 0x08192b19082b0808, 0x08192b1919080819,
0x08192b1919081908, 0x08192b1919190808, 0x08192b19192b2b19, 0x08192b192b2b082b,
0x08192b2b08081908, 0x08192b2b08190808, 0x08192b2b19080808, 0x08192b2b1919192b,
0x082b080808080808, 0x082b08080808082b, 0x082b080808081919, 0x082b080808082b08,
0x082b080808190819, 0x082b080808191908, 0x082b08080819192b, 0x082b080808192b19,
0x082b0808082b0808, 0x082b0808082b1919, 0x082b0808082b2b2b, 0x082b080819080819,
0x082b080819081908, 0x082b080819190808, 0x082b08081919082b, 0x082b080819191919,
0x082b0808192b1908, 0x082b08082b080808, 0x082b08082b082b2b, 0x082b08082b191908,
0x082b08082b2b2b2b, 0x082b081908080819, 0x082b081908081908, 0x082b081908190808,
0x082b08190819082b, 0x082b081908191919, 0x082b0819082b0819, 0x082b081919080808,
0x082b08191908082b, 0x082b081919081919, 0x082b081919190819, 0x082b081919191908,
0x082b0819192b0808, 0x082b08192b080819, 0x082b08192b081908, 0x082b08192b190808,
0x082b082b08080808, 0x082b082b08082b2b, 0x082b082b082b082b, 0x082b082b082b2b08,
0x082b082b082b2b2b, 0x082b082b19081908, 0x082b082b19190808, 0x082b082b2b082b08,
0x082b082b2b082b2b, 0x082b082b2b2b2b08, 0x082b190808080819, 0x082b190808081908,
0x082b19080808192b, 0x082b190808082b19, 0x082b190808190808, 0x082b190808191919,
0x082b190808192b08, 0x082b1908082b0819, 0x082b1908082b1908, 0x082b190819080808,
0x082b19081908082b, 0x082b190819081919, 0x082b190819082b08, 0x082b190819190819,
0x082b190819191908, 0x082b1908192b0808, 0x082b19082b080819, 0x082b19082b081908,
0x082b19082b190808, 0x082b191908080808, 0x082b191908081919, 0x082b191908082b08,
0x082b191908190819, 0x082b191908191908, 0x082b1919082b0808, 0x082b191919080819,
0x082b191919081908, 0x082b191919190808, 0x082b1919192b192b, 0x082b19192b080808,
0x082b192b08080819, 0x082b192b08081908, 0x082b192b08190808, 0x082b192b19080808,
0x082b192b19192b19, 0x082b2b0808080808, 0x082b2b0808081919, 0x082b2b0808190819,
0x082b2b0808191908, 0x082b2b0819080819, 0x082b2b0819081908, 0x082b2b0819190808,
0x082b2b082b082b2b, 0x082b2b082b2b2b2b, 0x082b2b1908080819, 0x082b2b1908081908,
0x082b2b1908190808, 0x082b2b192b191919, 0x082b2b2b08082b2b, 0x082b2b2b082b082b,
0x082b2b2b192b1908, 0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819,
0x1908080808081908, 0x190808080808192b, 0x1908080808082b19, 0x1908080808190808,
0x190808080819082b, 0x1908080808191919, 0x1908080808192b08, 0x1908080808192b2b,
0x19080808082b0819, 0x19080808082b1908, 0x19080808082b192b, 0x1908080819080808,
0x190808081908082b, 0x1908080819081919, 0x1908080819082b08, 0x1908080819082b2b,
0x1908080819190819, 0x1908080819191908, 0x190808081919192b, 0x1908080819192b19,
0x19080808192b0808, 0x19080808192b082b, 0x19080808192b1919, 0x190808082b080819,
0x190808082b081908, 0x190808082b190808, 0x190808082b191919, 0x190808082b192b08,
0x190808082b2b0819, 0x190808082b2b1908, 0x1908081908080808, 0x190808190808082b,
0x1908081908081919, 0x1908081908082b08, 0x1908081908190819, 0x1908081908191908,
0x190808190819192b, 0x1908081908192b19, 0x19080819082b0808, 0x19080819082b082b,
0x19080819082b1919, 0x1908081919080819, 0x1908081919081908, 0x190808191908192b,
0x1908081919082b19, 0x1908081919190808, 0x190808191919082b, 0x1908081919191919,
0x1908081919192b08, 0x19080819192b0819, 0x19080819192b1908, 0x190808192b080808,
0x190808192b08082b, 0x190808192b081919, 0x190808192b082b08, 0x190808192b190819,
0x190808192b191908, 0x190808192b2b0808, 0x1908082b08080819, 0x1908082b08081908,
0x1908082b08190808, 0x1908082b0819082b, 0x1908082b08191919, 0x1908082b08192b08,
0x1908082b082b1908, 0x1908082b19080808, 0x1908082b19081919, 0x1908082b19082b08,
0x1908082b19190819, 0x1908082b19191908, 0x1908082b192b0808, 0x1908082b2b080819,
0x1908082b2b081908, 0x1908190808080808, 0x190819080808082b, 0x1908190808081919,
0x1908190808082b08, 0x1908190808082b2b, 0x1908190808190819, 0x1908190808191908,
0x190819080819192b, 0x1908190808192b19, 0x19081908082b0808, 0x19081908082b082b,
0x19081908082b1919, 0x19081908082b2b08, 0x1908190819080819, 0x1908190819081908,
0x190819081908192b, 0x1908190819082b19, 0x1908190819190808, 0x190819081919082b,
0x1908190819191919, 0x1908190819192b08, 0x19081908192b0819, 0x19081908192b1908,
0x190819082b080808, 0x190819082b08082b, 0x190819082b081919, 0x190819082b082b08,
0x190819082b190819, 0x190819082b191908, 0x190819082b2b0808, 0x1908191908080819,
0x1908191908081908, 0x190819190808192b, 0x1908191908082b19, 0x1908191908190808,
0x190819190819082b, 0x1908191908191919, 0x1908191908192b08, 0x19081919082b0819,
0x19081919082b1908, 0x1908191919080808, 0x190819191908082b, 0x1908191919081919,
0x1908191919082b08, 0x1908191919190819, 0x1908191919191908, 0x19081919192b0808,
0x19081919192b2b2b, 0x190819192b080819, 0x190819192b081908, 0x190819192b190808,
0x1908192b08080808, 0x1908192b0808082b, 0x1908192b08081919, 0x1908192b08082b08,
0x1908192b08190819, 0x1908192b08191908, 0x1908192b082b0808, 0x1908192b19080819,
0x1908192b19081908, 0x1908192b19190808, 0x1908192b2b080808, 0x1908192b2b2b1919,
0x19082b0808080819, 0x19082b0808081908, 0x19082b0808082b19, 0x19082b0808190808,
0x19082b080819082b, 0x19082b0808191919, 0x19082b0808192b08, 0x19082b08082b0819,
0x19082b08082b1908, 0x19082b0819080808, 0x19082b081908082b, 0x19082b0819081919,
0x19082b0819082b08, 0x19082b0819190819, 0x19082b0819191908, 0x19082b08192b0808,
0x19082b082b081908, 0x19082b082b190808, 0x19082b1908080808, 0x19082b190808082b,
0x19082b1908081919, 0x19082b1908082b08, 0x19082b1908190819, 0x19082b1908191908,
0x19082b19082b0808, 0x19082b1919080819, 0x19082b1919081908, 0x19082b1919190808,
0x19082b192b080808, 0x19082b192b19192b, 0x19082b2b08080819, 0x19082b2b08081908,
0x19082b2b08190808, 0x19082b2b19080808, 0x1919080808080808, 0x191908080808082b,
0x1919080808081919, 0x1919080808082b08, 0x1919080808190819, 0x1919080808191908,
0x191908080819192b, 0x1919080808192b19, 0x19190808082b0808, 0x19190808082b082b,
0x19190808082b1919, 0x19190808082b2b08, 0x1919080819080819, 0x1919080819081908,
0x191908081908192b, 0x1919080819082b19, 0x1919080819190808, 0x191908081919082b,
0x1919080819191919, 0x1919080819192b08, 0x19190808192b0819, 0x19190808192b1908,
0x191908082b080808, 0x191908082b08082b, 0x191908082b081919, 0x191908082b082b08,
0x191908082b190819, 0x191908082b191908, 0x1919081908080819, 0x1919081908081908,
0x191908190808192b, 0x1919081908082b19, 0x1919081908190808, 0x191908190819082b,
0x1919081908191919, 0x1919081908192b08, 0x19190819082b0819, 0x19190819082b1908,
0x1919081919080808, 0x191908191908082b, 0x1919081919081919, 0x1919081919082b08,
0x1919081919190819, 0x1919081919191908, 0x19190819192b0808, 0x191908192b080819,
0x191908192b081908, 0x191908192b190808, 0x1919082b08080808, 0x1919082b08081919,
0x1919082b08082b08, 0x1919082b08190819, 0x1919082b08191908, 0x1919082b082b0808,
0x1919082b19080819, 0x1919082b19081908, 0x1919082b19190808, 0x1919082b192b2b19,
0x1919082b2b080808, 0x1919190808080819, 0x1919190808081908, 0x191919080808192b,
0x1919190808082b19, 0x1919190808190808, 0x191919080819082b, 0x1919190808191919,
0x1919190808192b08, 0x19191908082b0819, 0x19191908082b1908, 0x1919190819080808,
0x191919081908082b, 0x1919190819081919, 0x1919190819082b08, 0x1919190819190819,
0x1919190819191908, 0x19191908192b0808, 0x191919082b080819, 0x191919082b081908,
0x191919082b190808, 0x1919191908080808, 0x191919190808082b, 0x1919191908081919,
0x1919191908082b08, 0x1919191908190819, 0x1919191908191908, 0x19191919082b0808,
0x1919191919080819, 0x1919191919081908, 0x1919191919190808, 0x191919192b080808,
0x1919192b08080819, 0x1919192b08081908, 0x1919192b08190808, 0x1919192b082b192b,
0x1919192b19080808, 0x19192b0808080808, 0x19192b080808082b, 0x19192b0808081919,
0x19192b0808082b08, 0x19192b0808190819, 0x19192b0808191908, 0x19192b08082b0808,
0x19192b0819080819, 0x19192b0819081908, 0x19192b0819190808, 0x19192b0819192b2b,
0x19192b082b080808, 0x19192b1908080819, 0x19192b1908081908, 0x19192b1908190808,
0x19192b1919080808, 0x19192b2b08080808, 0x19192b2b08192b19, 0x19192b2b2b081919,
0x19192b2b2b2b2b08, 0x192b080808080819, 0x192b080808081908, 0x192b08080808192b,
0x192b080808190808, 0x192b08080819082b, 0x192b080808191919, 0x192b080808192b08,
0x192b0808082b0819, 0x192b0808082b1908, 0x192b080819080808, 0x192b080819081919,
0x192b080819082b08, 0x192b080819190819, 0x192b080819191908, 0x192b0808192b0808,
0x192b08082b081908, 0x192b08082b190808, 0x192b081908080808, 0x192b08190808082b,
0x192b081908081919, 0x192b081908082b08, 0x192b081908190819, 0x192b081908191908,
0x192b0819082b0808, 0x192b081919080819, 0x192b081919081908, 0x192b081919190808,
0x192b08192b080808, 0x192b08192b192b19, 0x192b082b08081908, 0x192b082b08190808,
0x192b082b19080808, 0x192b082b1919192b, 0x192b082b2b2b0819, 0x192b190808080808,
0x192b190808081919, 0x192b190808082b08, 0x192b190808190819, 0x192b190808191908,
0x192b1908082b0808, 0x192b190819080819, 0x192b190819081908, 0x192b190819190808,
0x192b19082b080808, 0x192b191908080819, 0x192b191908081908, 0x192b191908190808,
0x192b191919080808, 0x192b191919082b2b, 0x192b1919192b2b08, 0x192b19192b19082b,
0x192b192b08080808, 0x192b192b2b191908, 0x192b2b0808080819, 0x192b2b0808081908,
0x192b2b0808190808, 0x192b2b08192b1919, 0x192b2b082b192b08, 0x192b2b1908080808,
0x192b2b19082b2b2b, 0x192b2b2b1908082b, 0x192b2b2b2b2b0819, 0x2b08080808080808,
0x2b0808080808082b, 0x2b08080808081919, 0x2b08080808082b08, 0x2b08080808190819,
0x2b08080808191908, 0x2b08080808192b19, 0x2b080808082b0808, 0x2b080808082b1919,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808, 0x2b0808081919082b,
0x2b08080819191919, 0x2b08080819192b08, 0x2b080808192b0819, 0x2b0808082b080808,
0x2b0808082b081919, 0x2b0808082b190819, 0x2b0808082b191908, 0x2b08081908080819,
0x2b08081908081908, 0x2b08081908082b19, 0x2b08081908190808, 0x2b0808190819082b,
0x2b08081908191919, 0x2b08081908192b08, 0x2b080819082b0819, 0x2b080819082b1908,
0x2b08081919080808, 0x2b0808191908082b, 0x2b08081919081919, 0x2b08081919082b08,
0x2b08081919190819, 0x2b08081919191908, 0x2b0808192b080819, 0x2b0808192b081908,
0x2b0808192b190808, 0x2b0808192b2b2b19, 0x2b08082b08080808, 0x2b08082b08081919,
0x2b08082b08082b2b, 0x2b08082b08190819, 0x2b08082b08191908, 0x2b08082b19080819,
0x2b08082b19081908, 0x2b08082b19190808, 0x2b08190808080819, 0x2b08190808081908,
0x2b0819080808192b, 0x2b08190808082b19, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190808192b08, 0x2b081908082b0819, 0x2b08190819080808,
0x2b0819081908082b, 0x2b08190819081919, 0x2b08190819082b08, 0x2b08190819190819,
0x2b08190819191908, 0x2b081908192b0808, 0x2b0819082b080819, 0x2b0819082b081908,
0x2b0819082b190808, 0x2b08191908080808, 0x2b0819190808082b, 0x2b08191908081919,
0x2b08191908082b08, 0x2b08191908190819, 0x2b08191908191908, 0x2b081919082b0808,
0x2b08191919080819, 0x2b08191919081908, 0x2b08191919190808, 0x2b0819192b080808,
0x2b0819192b082b2b, 0x2b08192b08080819, 0x2b08192b08081908, 0x2b08192b08190808,
0x2b08192b082b2b19, 0x2b08192b19080808, 0x2b082b0808080808, 0x2b082b0808081919,
0x2b082b0808190819, 0x2b082b0808191908, 0x2b082b0819080819, 0x2b082b0819081908,
0x2b082b0819190808, 0x2b082b082b2b082b, 0x2b082b1908080819, 0x2b082b1908081908,
0x2b082b1919080808, 0x2b082b19192b1919, 0x2b082b2b082b082b, 0x2b082b2b19192b08,
0x2b082b2b19192b2b, 0x2b082b2b2b08082b, 0x2b082b2b2b2b082b, 0x2b19080808080819,
0x2b19080808081908, 0x2b19080808082b19, 0x2b19080808190808, 0x2b1908080819082b,
0x2b19080808191919, 0x2b19080808192b08, 0x2b190808082b1908, 0x2b19080819080808,
0x2b1908081908082b, 0x2b19080819081919, 0x2b19080819082b08, 0x2b19080819190819,
0x2b19080819191908, 0x2b190808192b0808, 0x2b1908082b080819, 0x2b1908082b081908,
0x2b1908082b190808, 0x2b19081908080808, 0x2b19081908081919, 0x2b19081908190819,
0x2b19081908191908, 0x2b19081919080819, 0x2b19081919081908, 0x2b19081919190808,
0x2b19081919192b2b, 0x2b19082b08080819, 0x2b19082b08081908, 0x2b19082b08190808,
0x2b19082b19080808, 0x2b19082b2b2b192b, 0x2b19190808080808, 0x2b1919080808082b,
0x2b19190808081919, 0x2b19190808082b08, 0x2b19190808190819, 0x2b19190808191908,
0x2b191908082b0808, 0x2b19190819080819, 0x2b19190819081908, 0x2b19190819190808,
0x2b1919082b080808, 0x2b1919082b19192b, 0x2b19191908080819, 0x2b19191908081908,
0x2b19191908190808, 0x2b19191919080808, 0x2b1919192b192b08, 0x2b1919192b2b0819,
0x2b19192b08080808, 0x2b19192b1908192b, 0x2b19192b192b1908, 0x2b192b0808080819,
0x2b192b0808081908, 0x2b192b0808190808, 0x2b192b08082b192b, 0x2b192b0819080808,
0x2b192b082b2b2b19, 0x2b192b1908080808, 0x2b192b1919082b19, 0x2b192b191919082b,
0x2b192b2b2b190808, 0x2b2b080808080808, 0x2b2b080808081919, 0x2b2b080808082b2b,
0x2b2b080808191908, 0x2b2b0808082b082b, 0x2b2b0808082b2b2b, 0x2b2b080819080819,
0x2b2b080819081908, 0x2b2b080819190808, 0x2b2b08082b2b082b, 0x2b2b08082b2b2b2b,
0x2b2b081919080808, 0x2b2b0819192b1919, 0x2b2b082b0808082b, 0x2b2b082b08082b2b,
0x2b2b082b082b082b, 0x2b2b082b082b2b08, 0x2b2b082b082b2b2b, 0x2b2b082b2b08082b,
0x2b2b082b2b082b08, 0x2b2b082b2b082b2b, 0x2b2b082b2b2b2b08, 0x2b2b190808080819,
0x2b2b190808081908, 0x2b2b190808190808, 0x2b2b190819080808, 0x2b2b19082b082b19,
0x2b2b19082b2b1908, 0x2b2b191908080808, 0x2b2b191908192b19, 0x2b2b192b19190819,
0x2b2b2b0808082b2b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b082b, 0x2b2b2b1919191908,
0x2b2b2b192b08192b, 0x2b2b2b2b08082b08, 0x2b2b2b2b08082b2b, 0x2b2b2b2b082b0808,
0x2b2b2b2b082b082b, 0x2b2b2b2b082b2b08, 0x2b2b2b2b2b082b08, 0x2b2b2b2b2b2b2b2b,
};
constexpr constant static uint32_t iq3xxs_grid[256] = {
0x04040404, 0x04040414, 0x04040424, 0x04040c0c, 0x04040c1c, 0x04040c3e, 0x04041404, 0x04041414,
0x04041c0c, 0x04042414, 0x04043e1c, 0x04043e2c, 0x040c040c, 0x040c041c, 0x040c0c04, 0x040c0c14,
0x040c140c, 0x040c142c, 0x040c1c04, 0x040c1c14, 0x040c240c, 0x040c2c24, 0x040c3e04, 0x04140404,
0x04140414, 0x04140424, 0x04140c0c, 0x04141404, 0x04141414, 0x04141c0c, 0x04141c1c, 0x04141c3e,
0x04142c0c, 0x04142c3e, 0x04143e2c, 0x041c040c, 0x041c043e, 0x041c0c04, 0x041c0c14, 0x041c142c,
0x041c3e04, 0x04240c1c, 0x04241c3e, 0x04242424, 0x04242c3e, 0x04243e1c, 0x04243e2c, 0x042c040c,
0x042c043e, 0x042c1c14, 0x042c2c14, 0x04341c2c, 0x04343424, 0x043e0c04, 0x043e0c24, 0x043e0c34,
0x043e241c, 0x043e340c, 0x0c04040c, 0x0c04041c, 0x0c040c04, 0x0c040c14, 0x0c04140c, 0x0c04141c,
0x0c041c04, 0x0c041c14, 0x0c041c24, 0x0c04243e, 0x0c042c04, 0x0c0c0404, 0x0c0c0414, 0x0c0c0c0c,
0x0c0c1404, 0x0c0c1414, 0x0c14040c, 0x0c14041c, 0x0c140c04, 0x0c140c14, 0x0c14140c, 0x0c141c04,
0x0c143e14, 0x0c1c0404, 0x0c1c0414, 0x0c1c1404, 0x0c1c1c0c, 0x0c1c2434, 0x0c1c3434, 0x0c24040c,
0x0c24042c, 0x0c242c04, 0x0c2c1404, 0x0c2c1424, 0x0c2c2434, 0x0c2c3e0c, 0x0c34042c, 0x0c3e1414,
0x0c3e2404, 0x14040404, 0x14040414, 0x14040c0c, 0x14040c1c, 0x14041404, 0x14041414, 0x14041434,
0x14041c0c, 0x14042414, 0x140c040c, 0x140c041c, 0x140c042c, 0x140c0c04, 0x140c0c14, 0x140c140c,
0x140c1c04, 0x140c341c, 0x140c343e, 0x140c3e04, 0x14140404, 0x14140414, 0x14140c0c, 0x14140c3e,
0x14141404, 0x14141414, 0x14141c3e, 0x14142404, 0x14142c2c, 0x141c040c, 0x141c0c04, 0x141c0c24,
0x141c3e04, 0x141c3e24, 0x14241c2c, 0x14242c1c, 0x142c041c, 0x142c143e, 0x142c240c, 0x142c3e24,
0x143e040c, 0x143e041c, 0x143e0c34, 0x143e242c, 0x1c04040c, 0x1c040c04, 0x1c040c14, 0x1c04140c,
0x1c04141c, 0x1c042c04, 0x1c04342c, 0x1c043e14, 0x1c0c0404, 0x1c0c0414, 0x1c0c1404, 0x1c0c1c0c,
0x1c0c2424, 0x1c0c2434, 0x1c14040c, 0x1c14041c, 0x1c140c04, 0x1c14142c, 0x1c142c14, 0x1c143e14,
0x1c1c0c0c, 0x1c1c1c1c, 0x1c241c04, 0x1c24243e, 0x1c243e14, 0x1c2c0404, 0x1c2c0434, 0x1c2c1414,
0x1c2c2c2c, 0x1c340c24, 0x1c341c34, 0x1c34341c, 0x1c3e1c1c, 0x1c3e3404, 0x24040424, 0x24040c3e,
0x24041c2c, 0x24041c3e, 0x24042c1c, 0x24042c3e, 0x240c3e24, 0x24141404, 0x24141c3e, 0x24142404,
0x24143404, 0x24143434, 0x241c043e, 0x241c242c, 0x24240424, 0x24242c0c, 0x24243424, 0x242c142c,
0x242c241c, 0x242c3e04, 0x243e042c, 0x243e0c04, 0x243e0c14, 0x243e1c04, 0x2c040c14, 0x2c04240c,
0x2c043e04, 0x2c0c0404, 0x2c0c0434, 0x2c0c1434, 0x2c0c2c2c, 0x2c140c24, 0x2c141c14, 0x2c143e14,
0x2c1c0414, 0x2c1c2c1c, 0x2c240c04, 0x2c24141c, 0x2c24143e, 0x2c243e14, 0x2c2c0414, 0x2c2c1c0c,
0x2c342c04, 0x2c3e1424, 0x2c3e2414, 0x34041424, 0x34042424, 0x34042434, 0x34043424, 0x340c140c,
0x340c340c, 0x34140c3e, 0x34143424, 0x341c1c04, 0x341c1c34, 0x34242424, 0x342c042c, 0x342c2c14,
0x34341c1c, 0x343e041c, 0x343e140c, 0x3e04041c, 0x3e04042c, 0x3e04043e, 0x3e040c04, 0x3e041c14,
0x3e042c14, 0x3e0c1434, 0x3e0c2404, 0x3e140c14, 0x3e14242c, 0x3e142c14, 0x3e1c0404, 0x3e1c0c2c,
0x3e1c1c1c, 0x3e1c3404, 0x3e24140c, 0x3e24240c, 0x3e2c0404, 0x3e2c0414, 0x3e2c1424, 0x3e341c04,
};
constexpr constant static uint32_t iq3s_grid[512] = {
0x01010101, 0x01010103, 0x01010105, 0x0101010b, 0x0101010f, 0x01010301, 0x01010303, 0x01010305,
0x01010309, 0x0101030d, 0x01010501, 0x01010503, 0x0101050b, 0x01010707, 0x01010901, 0x01010905,
0x0101090b, 0x0101090f, 0x01010b03, 0x01010b07, 0x01010d01, 0x01010d05, 0x01010f03, 0x01010f09,
0x01010f0f, 0x01030101, 0x01030103, 0x01030105, 0x01030109, 0x01030301, 0x01030303, 0x0103030b,
0x01030501, 0x01030507, 0x0103050f, 0x01030703, 0x0103070b, 0x01030909, 0x01030d03, 0x01030d0b,
0x01030f05, 0x01050101, 0x01050103, 0x0105010b, 0x0105010f, 0x01050301, 0x01050307, 0x0105030d,
0x01050503, 0x0105050b, 0x01050701, 0x01050709, 0x01050905, 0x0105090b, 0x0105090f, 0x01050b03,
0x01050b07, 0x01050f01, 0x01050f07, 0x01070107, 0x01070303, 0x0107030b, 0x01070501, 0x01070505,
0x01070703, 0x01070707, 0x0107070d, 0x01070909, 0x01070b01, 0x01070b05, 0x01070d0f, 0x01070f03,
0x01070f0b, 0x01090101, 0x01090307, 0x0109030f, 0x01090503, 0x01090509, 0x01090705, 0x01090901,
0x01090907, 0x01090b03, 0x01090f01, 0x010b0105, 0x010b0109, 0x010b0501, 0x010b0505, 0x010b050d,
0x010b0707, 0x010b0903, 0x010b090b, 0x010b090f, 0x010b0d0d, 0x010b0f07, 0x010d010d, 0x010d0303,
0x010d0307, 0x010d0703, 0x010d0b05, 0x010d0f03, 0x010f0101, 0x010f0105, 0x010f0109, 0x010f0501,
0x010f0505, 0x010f050d, 0x010f0707, 0x010f0b01, 0x010f0b09, 0x03010101, 0x03010103, 0x03010105,
0x03010109, 0x03010301, 0x03010303, 0x03010307, 0x0301030b, 0x0301030f, 0x03010501, 0x03010505,
0x03010703, 0x03010709, 0x0301070d, 0x03010b09, 0x03010b0d, 0x03010d03, 0x03010f05, 0x03030101,
0x03030103, 0x03030107, 0x0303010d, 0x03030301, 0x03030309, 0x03030503, 0x03030701, 0x03030707,
0x03030903, 0x03030b01, 0x03030b05, 0x03030f01, 0x03030f0d, 0x03050101, 0x03050305, 0x0305030b,
0x0305030f, 0x03050501, 0x03050509, 0x03050705, 0x03050901, 0x03050907, 0x03050b0b, 0x03050d01,
0x03050f05, 0x03070103, 0x03070109, 0x0307010f, 0x03070301, 0x03070307, 0x03070503, 0x0307050f,
0x03070701, 0x03070709, 0x03070903, 0x03070d05, 0x03070f01, 0x03090107, 0x0309010b, 0x03090305,
0x03090309, 0x03090703, 0x03090707, 0x03090905, 0x0309090d, 0x03090b01, 0x03090b09, 0x030b0103,
0x030b0301, 0x030b0307, 0x030b0503, 0x030b0701, 0x030b0705, 0x030b0b03, 0x030d0501, 0x030d0509,
0x030d050f, 0x030d0909, 0x030d090d, 0x030f0103, 0x030f0107, 0x030f0301, 0x030f0305, 0x030f0503,
0x030f070b, 0x030f0903, 0x030f0d05, 0x030f0f01, 0x05010101, 0x05010103, 0x05010107, 0x0501010b,
0x0501010f, 0x05010301, 0x05010305, 0x05010309, 0x0501030d, 0x05010503, 0x05010507, 0x0501050f,
0x05010701, 0x05010705, 0x05010903, 0x05010907, 0x0501090b, 0x05010b01, 0x05010b05, 0x05010d0f,
0x05010f01, 0x05010f07, 0x05010f0b, 0x05030101, 0x05030105, 0x05030301, 0x05030307, 0x0503030f,
0x05030505, 0x0503050b, 0x05030703, 0x05030709, 0x05030905, 0x05030b03, 0x05050103, 0x05050109,
0x0505010f, 0x05050503, 0x05050507, 0x05050701, 0x0505070f, 0x05050903, 0x05050b07, 0x05050b0f,
0x05050f03, 0x05050f09, 0x05070101, 0x05070105, 0x0507010b, 0x05070303, 0x05070505, 0x05070509,
0x05070703, 0x05070707, 0x05070905, 0x05070b01, 0x05070d0d, 0x05090103, 0x0509010f, 0x05090501,
0x05090507, 0x05090705, 0x0509070b, 0x05090903, 0x05090f05, 0x05090f0b, 0x050b0109, 0x050b0303,
0x050b0505, 0x050b070f, 0x050b0901, 0x050b0b07, 0x050b0f01, 0x050d0101, 0x050d0105, 0x050d010f,
0x050d0503, 0x050d0b0b, 0x050d0d03, 0x050f010b, 0x050f0303, 0x050f050d, 0x050f0701, 0x050f0907,
0x050f0b01, 0x07010105, 0x07010303, 0x07010307, 0x0701030b, 0x0701030f, 0x07010505, 0x07010703,
0x07010707, 0x0701070b, 0x07010905, 0x07010909, 0x0701090f, 0x07010b03, 0x07010d07, 0x07010f03,
0x07030103, 0x07030107, 0x0703010b, 0x07030309, 0x07030503, 0x07030507, 0x07030901, 0x07030d01,
0x07030f05, 0x07030f0d, 0x07050101, 0x07050305, 0x07050501, 0x07050705, 0x07050709, 0x07050b01,
0x07070103, 0x07070301, 0x07070309, 0x07070503, 0x07070507, 0x0707050f, 0x07070701, 0x07070903,
0x07070907, 0x0707090f, 0x07070b0b, 0x07070f07, 0x07090107, 0x07090303, 0x0709030d, 0x07090505,
0x07090703, 0x07090b05, 0x07090d01, 0x07090d09, 0x070b0103, 0x070b0301, 0x070b0305, 0x070b050b,
0x070b0705, 0x070b0909, 0x070b0b0d, 0x070b0f07, 0x070d030d, 0x070d0903, 0x070f0103, 0x070f0107,
0x070f0501, 0x070f0505, 0x070f070b, 0x09010101, 0x09010109, 0x09010305, 0x09010501, 0x09010509,
0x0901050f, 0x09010705, 0x09010903, 0x09010b01, 0x09010f01, 0x09030105, 0x0903010f, 0x09030303,
0x09030307, 0x09030505, 0x09030701, 0x0903070b, 0x09030907, 0x09030b03, 0x09030b0b, 0x09050103,
0x09050107, 0x09050301, 0x0905030b, 0x09050503, 0x09050707, 0x09050901, 0x09050b0f, 0x09050d05,
0x09050f01, 0x09070109, 0x09070303, 0x09070307, 0x09070501, 0x09070505, 0x09070703, 0x0907070b,
0x09090101, 0x09090105, 0x09090509, 0x0909070f, 0x09090901, 0x09090f03, 0x090b010b, 0x090b010f,
0x090b0503, 0x090b0d05, 0x090d0307, 0x090d0709, 0x090d0d01, 0x090f0301, 0x090f030b, 0x090f0701,
0x090f0907, 0x090f0b03, 0x0b010105, 0x0b010301, 0x0b010309, 0x0b010505, 0x0b010901, 0x0b010909,
0x0b01090f, 0x0b010b05, 0x0b010d0d, 0x0b010f09, 0x0b030103, 0x0b030107, 0x0b03010b, 0x0b030305,
0x0b030503, 0x0b030705, 0x0b030f05, 0x0b050101, 0x0b050303, 0x0b050507, 0x0b050701, 0x0b05070d,
0x0b050b07, 0x0b070105, 0x0b07010f, 0x0b070301, 0x0b07050f, 0x0b070909, 0x0b070b03, 0x0b070d0b,
0x0b070f07, 0x0b090103, 0x0b090109, 0x0b090501, 0x0b090705, 0x0b09090d, 0x0b0b0305, 0x0b0b050d,
0x0b0b0b03, 0x0b0b0b07, 0x0b0d0905, 0x0b0f0105, 0x0b0f0109, 0x0b0f0505, 0x0d010303, 0x0d010307,
0x0d01030b, 0x0d010703, 0x0d010707, 0x0d010d01, 0x0d030101, 0x0d030501, 0x0d03050f, 0x0d030d09,
0x0d050305, 0x0d050709, 0x0d050905, 0x0d050b0b, 0x0d050d05, 0x0d050f01, 0x0d070101, 0x0d070309,
0x0d070503, 0x0d070901, 0x0d09050b, 0x0d090907, 0x0d090d05, 0x0d0b0101, 0x0d0b0107, 0x0d0b0709,
0x0d0b0d01, 0x0d0d010b, 0x0d0d0901, 0x0d0f0303, 0x0d0f0307, 0x0f010101, 0x0f010109, 0x0f01010f,
0x0f010501, 0x0f010505, 0x0f01070d, 0x0f010901, 0x0f010b09, 0x0f010d05, 0x0f030105, 0x0f030303,
0x0f030509, 0x0f030907, 0x0f03090b, 0x0f050103, 0x0f050109, 0x0f050301, 0x0f05030d, 0x0f050503,
0x0f050701, 0x0f050b03, 0x0f070105, 0x0f070705, 0x0f07070b, 0x0f070b07, 0x0f090103, 0x0f09010b,
0x0f090307, 0x0f090501, 0x0f090b01, 0x0f0b0505, 0x0f0b0905, 0x0f0d0105, 0x0f0d0703, 0x0f0f0101,
};
#define NGRID_IQ1S 512
constexpr constant static uint64_t iq1s_grid[NGRID_IQ1S] = {
0xffffffffffff0101, 0xffffffffff01ff00, 0xffffffffff010100, 0xffffffff00000000,
0xffffffff01ff00ff, 0xffffffff01ff0001, 0xffffffff0101ffff, 0xffffffff0101ff01,
0xffffff00ff000000, 0xffffff000000ff00, 0xffffff00000000ff, 0xffffff0000000100,
0xffffff0000010000, 0xffffff0001000000, 0xffffff01ffff00ff, 0xffffff01ff01ff00,
0xffffff01ff010100, 0xffffff0100000001, 0xffffff0101ffff00, 0xffffff0101ff0101,
0xffffff0101010100, 0xffff00ffff00ff01, 0xffff00ffff0000ff, 0xffff00ff00ff0100,
0xffff00ff0100ff00, 0xffff00ff010001ff, 0xffff0000ff0101ff, 0xffff000000ffff00,
0xffff000000000000, 0xffff00000001ff01, 0xffff000001000101, 0xffff0000010100ff,
0xffff0001ffff0100, 0xffff00010000ff00, 0xffff000100010101, 0xffff000101000000,
0xffff01ffffff0000, 0xffff01ffff01ffff, 0xffff01ffff010100, 0xffff01ff00000000,
0xffff01ff01ffffff, 0xffff01ff01ff0001, 0xffff01ff0101ffff, 0xffff01ff01010001,
0xffff0100ffffff01, 0xffff01000000ffff, 0xffff010000000100, 0xffff010001ff01ff,
0xffff010001000000, 0xffff0101ff000000, 0xffff0101000101ff, 0xffff010101ffff01,
0xffff01010101ff00, 0xff00ffffff000000, 0xff00ffff00ffff00, 0xff00ffff00000001,
0xff00ffff000001ff, 0xff00ffff01010000, 0xff00ff00ffff0000, 0xff00ff00ff00ff00,
0xff00ff00ff0000ff, 0xff00ff00ff000100, 0xff00ff00ff010001, 0xff00ff0000ff0001,
0xff00ff000000ffff, 0xff00ff0000000000, 0xff00ff000001ff00, 0xff00ff0000010100,
0xff00ff0001ff0000, 0xff00ff000100ff00, 0xff00ff0001000100, 0xff00ff01ff000000,
0xff00ff0100ff0000, 0xff00ff01000001ff, 0xff00ff0101010001, 0xff0000ff00000000,
0xff0000ff0001ff00, 0xff0000ff00010100, 0xff000000ffff0101, 0xff000000ff000000,
0xff000000ff01ff00, 0xff00000000ff0000, 0xff0000000000ff00, 0xff000000000000ff,
0xff00000000000000, 0xff00000000000001, 0xff00000000000100, 0xff0000000001ffff,
0xff00000000010000, 0xff00000001000000, 0xff00000001010100, 0xff000001ff00ff01,
0xff000001ff0100ff, 0xff00000100000000, 0xff0000010001ff00, 0xff00000101ff0100,
0xff0000010100ff00, 0xff0001ff00ff00ff, 0xff0001ff00000101, 0xff0001ff000100ff,
0xff0001ff01000000, 0xff000100ff0001ff, 0xff0001000000ff01, 0xff00010000000000,
0xff00010000010001, 0xff00010000010100, 0xff00010001ffff00, 0xff00010001ff0101,
0xff00010001010000, 0xff000101ffffffff, 0xff000101ff000101, 0xff00010101ff00ff,
0xff00010101000001, 0xff000101010100ff, 0xff01ffffff000101, 0xff01ffffff01ffff,
0xff01ffffff01ff01, 0xff01ffffff0101ff, 0xff01ffff00000000, 0xff01ffff01ff0001,
0xff01ffff0101ff01, 0xff01ff00ff000000, 0xff01ff0000ff0100, 0xff01ff000000ff01,
0xff01ff0000010000, 0xff01ff00010000ff, 0xff01ff01ff01ff00, 0xff01ff0100000101,
0xff0100ffffff0000, 0xff0100ffff010000, 0xff0100ff01ff00ff, 0xff0100ff01000100,
0xff0100ff010100ff, 0xff010000ffffff01, 0xff01000000000000, 0xff0100000101ff00,
0xff010001ffff00ff, 0xff010001ff000100, 0xff01000100ffff00, 0xff01000100010001,
0xff01000101ff0001, 0xff010001010001ff, 0xff0101ffffffffff, 0xff0101ffff01ffff,
0xff0101ffff010101, 0xff0101ff0000ff00, 0xff0101ff01010001, 0xff010100ff000000,
0xff010100ff01ff01, 0xff01010000ff0001, 0xff01010000000100, 0xff01010001000000,
0xff0101010100ffff, 0x00ffffff0000ff01, 0x00ffffff000000ff, 0x00ffffff00000100,
0x00ffffff00010000, 0x00ffff00ffff0001, 0x00ffff00ff0000ff, 0x00ffff00ff000100,
0x00ffff0000000000, 0x00ffff0001000100, 0x00ffff0001010001, 0x00ffff01ff00ff01,
0x00ffff0100ff0100, 0x00ffff010000ff00, 0x00ffff01000100ff, 0x00ffff0101ff00ff,
0x00ffff010101ff00, 0x00ff00ffffffffff, 0x00ff00ffffff01ff, 0x00ff00ffff000101,
0x00ff00ff00000000, 0x00ff00ff000101ff, 0x00ff00ff01010101, 0x00ff0000ff000000,
0x00ff0000ff01ffff, 0x00ff000000ff0000, 0x00ff00000000ff00, 0x00ff0000000000ff,
0x00ff000000000000, 0x00ff000000000001, 0x00ff000000000100, 0x00ff000000010000,
0x00ff000001ffff01, 0x00ff000001000000, 0x00ff0001ff000101, 0x00ff000100ffffff,
0x00ff000100000000, 0x00ff0001010001ff, 0x00ff01ffff000000, 0x00ff01ff0001ff00,
0x00ff01ff01ff0100, 0x00ff0100ff01ff01, 0x00ff010000ff00ff, 0x00ff010000ff0101,
0x00ff010000000000, 0x00ff010000010101, 0x00ff01000100ff00, 0x00ff010001010000,
0x00ff0101ffffff00, 0x00ff01010000ff01, 0x00ff010100000100, 0x00ff010101ff0000,
0x0000ffffffff0100, 0x0000ffffff00ff00, 0x0000ffffff0000ff, 0x0000ffffff010000,
0x0000ffff00000000, 0x0000ffff00010101, 0x0000ffff01ffff01, 0x0000ffff01000100,
0x0000ff00ff000000, 0x0000ff00ff01ff00, 0x0000ff00ff0101ff, 0x0000ff0000ff0000,
0x0000ff000000ff00, 0x0000ff00000000ff, 0x0000ff0000000000, 0x0000ff0000000001,
0x0000ff0000000100, 0x0000ff0000010000, 0x0000ff0001ffffff, 0x0000ff0001ff01ff,
0x0000ff0001000000, 0x0000ff000101ffff, 0x0000ff01ffff0101, 0x0000ff01ff010000,
0x0000ff0100000000, 0x0000ff0101000101, 0x000000ffffff0001, 0x000000ffff000000,
0x000000ff00ff0000, 0x000000ff0000ff00, 0x000000ff000000ff, 0x000000ff00000000,
0x000000ff00000001, 0x000000ff00000100, 0x000000ff00010000, 0x000000ff01000000,
0x000000ff0101ff00, 0x00000000ffff0000, 0x00000000ff00ff00, 0x00000000ff0000ff,
0x00000000ff000000, 0x00000000ff000001, 0x00000000ff000100, 0x00000000ff010000,
0x0000000000ffff00, 0x0000000000ff00ff, 0x0000000000ff0000, 0x0000000000ff0001,
0x0000000000ff0100, 0x000000000000ffff, 0x000000000000ff00, 0x000000000000ff01,
0x00000000000000ff, 0x0000000000000001, 0x00000000000001ff, 0x0000000000000100,
0x0000000000000101, 0x000000000001ff00, 0x00000000000100ff, 0x0000000000010000,
0x0000000000010001, 0x0000000000010100, 0x0000000001ff0000, 0x000000000100ff00,
0x00000000010000ff, 0x0000000001000000, 0x0000000001000001, 0x0000000001000100,
0x0000000001010000, 0x00000001ffff01ff, 0x00000001ff000000, 0x0000000100ff0000,
0x000000010000ff00, 0x00000001000000ff, 0x0000000100000000, 0x0000000100000001,
0x0000000100000100, 0x0000000100010000, 0x0000000101000000, 0x000001ffff00ff00,
0x000001ffff010001, 0x000001ffff0101ff, 0x000001ff00ffff01, 0x000001ff0000ffff,
0x000001ff00000000, 0x000001ff010000ff, 0x000001ff01010100, 0x00000100ffff0100,
0x00000100ff000000, 0x0000010000ff0000, 0x000001000000ff00, 0x00000100000000ff,
0x0000010000000000, 0x0000010000000001, 0x0000010000000100, 0x0000010000010000,
0x0000010001000000, 0x000001000101ff01, 0x00000101ffff0001, 0x00000101ff01ffff,
0x0000010100000000, 0x0000010101010100, 0x0001ffffff000000, 0x0001ffff00ffffff,
0x0001ffff00000100, 0x0001ffff0001ff00, 0x0001ffff01000000, 0x0001ff00ffffff00,
0x0001ff00ffff01ff, 0x0001ff00ff010000, 0x0001ff0000000000, 0x0001ff0000010001,
0x0001ff0001ff0000, 0x0001ff0001010100, 0x0001ff01ff0000ff, 0x0001ff01ff000001,
0x0001ff0100ffffff, 0x0001ff010001ffff, 0x0001ff01000101ff, 0x0001ff010100ff01,
0x000100ffff00ffff, 0x000100ffff00ff01, 0x000100ffff000100, 0x000100ff00000000,
0x000100ff000101ff, 0x000100ff01ff0101, 0x000100ff0100ffff, 0x000100ff01010101,
0x00010000ff000000, 0x00010000ff010100, 0x0001000000ff0000, 0x000100000000ff00,
0x00010000000000ff, 0x0001000000000000, 0x0001000000000001, 0x0001000000000100,
0x0001000000010000, 0x0001000001ffff01, 0x0001000001000000, 0x0001000100ff0101,
0x0001000100000000, 0x00010001010100ff, 0x000101ffffff01ff, 0x000101ffffff0101,
0x000101ff00010000, 0x000101ff01ff0000, 0x000101ff0100ff01, 0x00010100ffff0000,
0x0001010000000000, 0x000101000001ffff, 0x0001010000010101, 0x00010100010001ff,
0x00010101ff00ff00, 0x00010101ff010001, 0x0001010100ffffff, 0x0001010100ff01ff,
0x00010101000101ff, 0x0001010101ff0000, 0x000101010100ff01, 0x0001010101000101,
0x01ffffffffff0101, 0x01ffffffff01ffff, 0x01ffffffff01ff01, 0x01ffffffff0101ff,
0x01ffffffff010101, 0x01ffffff00000000, 0x01ffffff01ff01ff, 0x01ffffff01000101,
0x01ffffff0101ff01, 0x01ffffff010100ff, 0x01ffff000000ff00, 0x01ffff0000000001,
0x01ffff00000001ff, 0x01ffff0000010000, 0x01ffff0001ff0000, 0x01ffff01ffffffff,
0x01ffff01ffff01ff, 0x01ffff01ff000000, 0x01ffff01ff01ffff, 0x01ffff01ff0101ff,
0x01ffff010100ffff, 0x01ff00ffffff0000, 0x01ff00ffff010000, 0x01ff00ff00ffff01,
0x01ff0000ff0000ff, 0x01ff000000000000, 0x01ff00000001ff01, 0x01ff000001ffffff,
0x01ff000001010100, 0x01ff0001ffffff01, 0x01ff0001ff010001, 0x01ff000101ff0100,
0x01ff000101000001, 0x01ff0001010100ff, 0x01ff01ffff00ffff, 0x01ff01ff00010001,
0x01ff01ff01000000, 0x01ff01ff010101ff, 0x01ff0100ff000001, 0x01ff010000ffff00,
0x01ff010000000100, 0x01ff010001ff01ff, 0x01ff01000101ffff, 0x01ff0101ffff00ff,
0x01ff0101ffff0101, 0x01ff0101ff0101ff, 0x01ff010100010000, 0x0100ffff00ff00ff,
0x0100ffff00ff0001, 0x0100ffff00000100, 0x0100ffff0100ff00, 0x0100ff00ffff0000,
0x0100ff00ff00ffff, 0x0100ff00ff00ff01, 0x0100ff00ff000100, 0x0100ff00ff010000,
0x0100ff0000000000, 0x0100ff00000100ff, 0x0100ff0001ff0101, 0x0100ff0001010101,
0x0100ff0100ff00ff, 0x0100ff0100ff0001, 0x0100ff0100000100, 0x0100ff0100010001,
0x0100ff0101000000, 0x010000ffff00ff00, 0x010000ff0000ffff, 0x010000ff00000000,
0x010000ff010001ff, 0x010000ff01010001, 0x01000000ffffff00, 0x01000000ffff0101,
0x01000000ff000000, 0x01000000ff0100ff, 0x01000000ff010101, 0x0100000000ff0000,
0x010000000000ff00, 0x01000000000000ff, 0x0100000000000000, 0x0100000000000001,
0x0100000000000100, 0x0100000000010000, 0x0100000001000000, 0x0100000100000000,
0x01000001000101ff, 0x0100000101ffff01, 0x010001ffff000101, 0x010001ff00ff0100,
0x010001ff0000ff00, 0x010001ff000100ff, 0x010001ff01ffffff, 0x01000100ffff0000,
0x01000100ff0001ff, 0x0100010000000000, 0x010001000001ff00, 0x0100010001ff0000,
0x01000100010000ff, 0x0100010001000101, 0x01000101ff00ff01, 0x0100010100ff0100,
0x010001010000ffff, 0x0100010101010001, 0x0101ffffffff0101, 0x0101ffffff0001ff,
0x0101ffffff01ffff, 0x0101ffffff010101, 0x0101ffff00000000, 0x0101ffff0101ffff,
0x0101ffff010101ff, 0x0101ff00ff000000, 0x0101ff0000ff0100, 0x0101ff000000ff00,
0x0101ff0000010000, 0x0101ff00010000ff, 0x0101ff0001000001, 0x0101ff01ff010101,
0x0101ff0100000000, 0x0101ff010101ff00, 0x010100ffffff0000, 0x010100ffff010000,
0x010100ff00ff01ff, 0x010100ff000000ff, 0x010100ff00000101, 0x010100ff01ffff00,
0x01010000ffffff01, 0x01010000ff000100, 0x01010000ff01ff01, 0x0101000000000000,
0x01010000000100ff, 0x010100000101ff01, 0x01010001ffff0000, 0x01010001ff00ffff,
0x01010001ff010000, 0x0101000101ffffff, 0x0101000101ff01ff, 0x0101000101010101,
0x010101ffff01ffff, 0x010101ff00000000, 0x010101ff0001ff01, 0x010101ff0101ffff,
0x010101ff010101ff, 0x01010100ffffffff, 0x01010100ff000001, 0x010101000000ff00,
0x0101010001010000, 0x0101010100ff0001, 0x010101010001ff01, 0x010101010101ffff,
};
constexpr constant static uint8_t ksigns_iq2xs[128] = {
0, 129, 130, 3, 132, 5, 6, 135, 136, 9, 10, 139, 12, 141, 142, 15,
144, 17, 18, 147, 20, 149, 150, 23, 24, 153, 154, 27, 156, 29, 30, 159,
160, 33, 34, 163, 36, 165, 166, 39, 40, 169, 170, 43, 172, 45, 46, 175,
48, 177, 178, 51, 180, 53, 54, 183, 184, 57, 58, 187, 60, 189, 190, 63,
192, 65, 66, 195, 68, 197, 198, 71, 72, 201, 202, 75, 204, 77, 78, 207,
80, 209, 210, 83, 212, 85, 86, 215, 216, 89, 90, 219, 92, 221, 222, 95,
96, 225, 226, 99, 228, 101, 102, 231, 232, 105, 106, 235, 108, 237, 238, 111,
240, 113, 114, 243, 116, 245, 246, 119, 120, 249, 250, 123, 252, 125, 126, 255,
};
constexpr constant static uint8_t kmask_iq2xs[8] = {1, 2, 4, 8, 16, 32, 64, 128};
void kernel_mul_mv_iq2_xxs_f32_impl(
device const void * src0,
device const float * src1,
@ -6087,7 +5386,7 @@ template<typename block_q, short nl, void (*dequantize_func)(device const block_
void kernel_mul_mm_id_impl(
device const uchar * src0,
device const uchar * src1,
thread short * src1ids,
threadgroup short * src1ids,
device float * dst,
constant int64_t & ne00,
constant int64_t & ne02,
@ -6290,9 +5589,9 @@ kernel void kernel_mul_mm_id(
tgpig.z = tgpig.z%(ne12*ne13);
// row indices of src1 for expert id
int64_t _ne1 = 0;
short src1ids[512];
threadgroup short * src1ids = (threadgroup short *)(shared_memory + 8192);
int64_t _ne1 = 0;
for (int64_t i1 = 0; i1 < ne1; i1++) {
if (((device int32_t *) (ids + i1*nbi1))[idx] == id) {
src1ids[_ne1++] = i1;

887
ggml-quants.c
View File

File diff suppressed because it is too large Load Diff

46
ggml-quants.h
View File

@ -1,9 +1,9 @@
#pragma once
#include "ggml-impl.h"
// GGML internal header
#include "ggml-impl.h"
#include <stdint.h>
#include <stddef.h>
@ -261,6 +261,7 @@ void quantize_row_q4_K_reference(const float * GGML_RESTRICT x, block_q4_K * GGM
void quantize_row_q5_K_reference(const float * GGML_RESTRICT x, block_q5_K * GGML_RESTRICT y, int k);
void quantize_row_q6_K_reference(const float * GGML_RESTRICT x, block_q6_K * GGML_RESTRICT y, int k);
void quantize_row_q8_K_reference(const float * GGML_RESTRICT x, block_q8_K * GGML_RESTRICT y, int k);
void quantize_row_iq3_xxs_reference(const float * GGML_RESTRICT x, block_iq3_xxs * GGML_RESTRICT y, int k);
void quantize_row_iq4_nl_reference (const float * GGML_RESTRICT x, block_iq4_nl * GGML_RESTRICT y, int k);
void quantize_row_iq4_xs_reference (const float * GGML_RESTRICT x, block_iq4_xs * GGML_RESTRICT y, int k);
@ -280,6 +281,7 @@ void quantize_row_q4_K(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, in
void quantize_row_q5_K(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
void quantize_row_q6_K(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
void quantize_row_q8_K(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
void quantize_row_iq3_xxs(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
void quantize_row_iq4_nl (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
void quantize_row_iq4_xs (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
@ -300,6 +302,7 @@ void dequantize_row_q4_K(const block_q4_K * GGML_RESTRICT x, float * GGML_RESTRI
void dequantize_row_q5_K(const block_q5_K * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
void dequantize_row_q6_K(const block_q6_K * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
void dequantize_row_q8_K(const block_q8_K * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
void dequantize_row_iq2_xxs(const block_iq2_xxs * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
void dequantize_row_iq2_xs (const block_iq2_xs * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
void dequantize_row_iq2_s (const block_iq2_s * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
@ -321,6 +324,7 @@ void ggml_vec_dot_q3_K_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const voi
void ggml_vec_dot_q4_K_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_q5_K_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_q6_K_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_iq2_xxs_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_iq2_xs_q8_K (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_iq2_s_q8_K (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
@ -330,26 +334,26 @@ void ggml_vec_dot_iq4_nl_q8_0 (int n, float * GGML_RESTRICT s, size_t bs, const
void ggml_vec_dot_iq4_xs_q8_K (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
void ggml_vec_dot_iq3_s_q8_K (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
//
// Quantization utilizing an importance matrix (a.k.a. "Activation aWare Quantization")
//
size_t quantize_iq2_xxs(const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq2_xs (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq2_s (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq3_xxs(const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq1_s (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq4_nl (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq4_xs (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq3_s (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q2_K (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q3_K (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q4_K (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q5_K (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q6_K (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q4_0 (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q4_1 (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q5_0 (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_q5_1 (const float * src, void * dst, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
size_t quantize_iq2_xxs(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq2_xs (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq2_s (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq3_xxs(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq1_s (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq4_nl (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq4_xs (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_iq3_s (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q2_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q3_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q4_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q5_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q6_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q4_0(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q4_1(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q5_0(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q5_1(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
size_t quantize_q8_0(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int nrows, int n_per_row, const float * imatrix);
void iq2xs_init_impl(enum ggml_type type);
void iq2xs_free_impl(enum ggml_type type);

643
ggml-sycl.cpp
View File

@ -3144,6 +3144,8 @@ namespace dpct
} // COPY from DPCT head files
#define GGML_COMMON_IMPL_SYCL
#include "ggml-common.h"
static int g_ggml_sycl_debug=0;
#define GGML_SYCL_DEBUG(...) do{if(g_ggml_sycl_debug) printf(__VA_ARGS__);}while(0)
@ -3794,8 +3796,42 @@ void log_ggml_var_device(const char*name, float *src, size_t total_elements, boo
std::ofstream logfile;
logfile.open(filename);
for(size_t i=0; i<total_elements; i++){
logfile << local_buf[i] <<" ";
if((i+1)%20 ==0) logfile <<std::endl;
}
logfile <<std::endl;
logfile.close();
if(src_on_device) ggml_sycl_host_free(local_buf);
}
void log_ggml_var_device_fp16(const char*name, sycl::half *src, size_t total_elements, bool src_on_device){
if(!g_ggml_sycl_debug) return;
if(!src){
printf("GGML Tensor:%s skip to save for NULL pointer\n", name);
return;
}
char filename[1024];
sprintf(filename, "%s.txt", name);
printf("GGML Tensor:%s save to %s\n", name, filename);
size_t total_size = total_elements*sizeof(sycl::half);
sycl::half *local_buf = NULL;
if(src_on_device) {
local_buf = (sycl::half *) ggml_sycl_host_malloc(total_size);
ggml_sycl_set_device(g_main_device);
dpct::queue_ptr main_stream = g_syclStreams[g_main_device][0];
main_stream->memcpy(local_buf, src, total_size).wait();
}
else {
local_buf = (sycl::half *)src;
}
std::ofstream logfile;
logfile.open(filename);
for(size_t i=0; i<total_elements; i++){
logfile << local_buf[i] <<" ";
if((i+1)%20 ==0) logfile <<std::endl;
else logfile << local_buf[i] <<" ";
}
logfile <<std::endl;
logfile.close();
@ -4736,588 +4772,6 @@ static void dequantize_block_q6_K(const void * __restrict__ vx, dst_t * __restri
#endif
}
static dpct::global_memory<const uint64_t, 1>
iq2xxs_grid(sycl::range<1>(256),
{
0x0808080808080808, 0x080808080808082b, 0x0808080808081919,
0x0808080808082b08, 0x0808080808082b2b, 0x0808080808190819,
0x0808080808191908, 0x08080808082b0808, 0x08080808082b082b,
0x08080808082b2b08, 0x08080808082b2b2b, 0x0808080819080819,
0x0808080819081908, 0x0808080819190808, 0x0808080819192b08,
0x08080808192b0819, 0x08080808192b1908, 0x080808082b080808,
0x080808082b08082b, 0x080808082b082b2b, 0x080808082b2b082b,
0x0808081908080819, 0x0808081908081908, 0x0808081908190808,
0x0808081908191919, 0x0808081919080808, 0x080808192b081908,
0x080808192b192b08, 0x0808082b08080808, 0x0808082b0808082b,
0x0808082b082b082b, 0x0808082b2b08082b, 0x0808190808080819,
0x0808190808081908, 0x0808190808190808, 0x08081908082b0819,
0x08081908082b1908, 0x0808190819080808, 0x080819081908082b,
0x0808190819082b08, 0x08081908192b0808, 0x080819082b080819,
0x080819082b081908, 0x080819082b190808, 0x080819082b2b1908,
0x0808191908080808, 0x080819190808082b, 0x0808191908082b08,
0x08081919082b0808, 0x080819191908192b, 0x08081919192b2b19,
0x080819192b080808, 0x080819192b190819, 0x0808192b08082b19,
0x0808192b08190808, 0x0808192b19080808, 0x0808192b2b081908,
0x0808192b2b2b1908, 0x08082b0808080808, 0x08082b0808081919,
0x08082b0808082b08, 0x08082b0808191908, 0x08082b08082b2b08,
0x08082b0819080819, 0x08082b0819081908, 0x08082b0819190808,
0x08082b081919082b, 0x08082b082b082b08, 0x08082b1908081908,
0x08082b1919080808, 0x08082b2b0808082b, 0x08082b2b08191908,
0x0819080808080819, 0x0819080808081908, 0x0819080808190808,
0x08190808082b0819, 0x0819080819080808, 0x08190808192b0808,
0x081908082b081908, 0x081908082b190808, 0x081908082b191919,
0x0819081908080808, 0x0819081908082b08, 0x08190819082b0808,
0x0819081919190808, 0x0819081919192b2b, 0x081908192b080808,
0x0819082b082b1908, 0x0819082b19081919, 0x0819190808080808,
0x0819190808082b08, 0x08191908082b0808, 0x08191908082b1919,
0x0819190819082b19, 0x081919082b080808, 0x0819191908192b08,
0x08191919192b082b, 0x0819192b08080808, 0x0819192b0819192b,
0x08192b0808080819, 0x08192b0808081908, 0x08192b0808190808,
0x08192b0819080808, 0x08192b082b080819, 0x08192b1908080808,
0x08192b1908081919, 0x08192b192b2b0808, 0x08192b2b19190819,
0x082b080808080808, 0x082b08080808082b, 0x082b080808082b2b,
0x082b080819081908, 0x082b0808192b0819, 0x082b08082b080808,
0x082b08082b08082b, 0x082b0819082b2b19, 0x082b081919082b08,
0x082b082b08080808, 0x082b082b0808082b, 0x082b190808080819,
0x082b190808081908, 0x082b190808190808, 0x082b190819080808,
0x082b19081919192b, 0x082b191908080808, 0x082b191919080819,
0x082b1919192b1908, 0x082b192b2b190808, 0x082b2b0808082b08,
0x082b2b08082b0808, 0x082b2b082b191908, 0x082b2b2b19081908,
0x1908080808080819, 0x1908080808081908, 0x1908080808190808,
0x1908080808192b08, 0x19080808082b0819, 0x19080808082b1908,
0x1908080819080808, 0x1908080819082b08, 0x190808081919192b,
0x19080808192b0808, 0x190808082b080819, 0x190808082b081908,
0x190808082b190808, 0x1908081908080808, 0x19080819082b0808,
0x19080819192b0819, 0x190808192b080808, 0x190808192b081919,
0x1908082b08080819, 0x1908082b08190808, 0x1908082b19082b08,
0x1908082b1919192b, 0x1908082b192b2b08, 0x1908190808080808,
0x1908190808082b08, 0x19081908082b0808, 0x190819082b080808,
0x190819082b192b19, 0x190819190819082b, 0x19081919082b1908,
0x1908192b08080808, 0x19082b0808080819, 0x19082b0808081908,
0x19082b0808190808, 0x19082b0819080808, 0x19082b0819081919,
0x19082b1908080808, 0x19082b1919192b08, 0x19082b19192b0819,
0x19082b192b08082b, 0x19082b2b19081919, 0x19082b2b2b190808,
0x1919080808080808, 0x1919080808082b08, 0x1919080808190819,
0x1919080808192b19, 0x19190808082b0808, 0x191908082b080808,
0x191908082b082b08, 0x1919081908081908, 0x191908191908082b,
0x191908192b2b1908, 0x1919082b2b190819, 0x191919082b190808,
0x191919082b19082b, 0x1919191908082b2b, 0x1919192b08080819,
0x1919192b19191908, 0x19192b0808080808, 0x19192b0808190819,
0x19192b0808192b19, 0x19192b08192b1908, 0x19192b1919080808,
0x19192b2b08082b08, 0x192b080808081908, 0x192b080808190808,
0x192b080819080808, 0x192b0808192b2b08, 0x192b081908080808,
0x192b081919191919, 0x192b082b08192b08, 0x192b082b192b0808,
0x192b190808080808, 0x192b190808081919, 0x192b191908190808,
0x192b19190819082b, 0x192b19192b081908, 0x192b2b081908082b,
0x2b08080808080808, 0x2b0808080808082b, 0x2b08080808082b2b,
0x2b08080819080819, 0x2b0808082b08082b, 0x2b08081908081908,
0x2b08081908192b08, 0x2b08081919080808, 0x2b08082b08190819,
0x2b08190808080819, 0x2b08190808081908, 0x2b08190808190808,
0x2b08190808191919, 0x2b08190819080808, 0x2b081908192b0808,
0x2b08191908080808, 0x2b0819191908192b, 0x2b0819192b191908,
0x2b08192b08082b19, 0x2b08192b19080808, 0x2b08192b192b0808,
0x2b082b080808082b, 0x2b082b1908081908, 0x2b082b2b08190819,
0x2b19080808081908, 0x2b19080808190808, 0x2b190808082b1908,
0x2b19080819080808, 0x2b1908082b2b0819, 0x2b1908190819192b,
0x2b1908192b080808, 0x2b19082b19081919, 0x2b19190808080808,
0x2b191908082b082b, 0x2b19190819081908, 0x2b19191919190819,
0x2b192b082b080819, 0x2b192b19082b0808, 0x2b2b08080808082b,
0x2b2b080819190808, 0x2b2b08082b081919, 0x2b2b081908082b19,
0x2b2b082b08080808, 0x2b2b190808192b08, 0x2b2b2b0819190808,
0x2b2b2b1908081908,
});
static dpct::global_memory<const uint64_t, 1>
iq2xs_grid(sycl::range<1>(512),
{
0x0808080808080808, 0x080808080808082b, 0x0808080808081919,
0x0808080808082b08, 0x0808080808082b2b, 0x0808080808190819,
0x0808080808191908, 0x080808080819192b, 0x0808080808192b19,
0x08080808082b0808, 0x08080808082b082b, 0x08080808082b1919,
0x08080808082b2b08, 0x0808080819080819, 0x0808080819081908,
0x080808081908192b, 0x0808080819082b19, 0x0808080819190808,
0x080808081919082b, 0x0808080819191919, 0x0808080819192b08,
0x08080808192b0819, 0x08080808192b1908, 0x080808082b080808,
0x080808082b08082b, 0x080808082b081919, 0x080808082b082b08,
0x080808082b190819, 0x080808082b191908, 0x080808082b192b19,
0x080808082b2b0808, 0x0808081908080819, 0x0808081908081908,
0x080808190808192b, 0x0808081908082b19, 0x0808081908190808,
0x080808190819082b, 0x0808081908191919, 0x0808081908192b08,
0x0808081908192b2b, 0x08080819082b0819, 0x08080819082b1908,
0x0808081919080808, 0x080808191908082b, 0x0808081919081919,
0x0808081919082b08, 0x0808081919190819, 0x0808081919191908,
0x08080819192b0808, 0x08080819192b2b08, 0x080808192b080819,
0x080808192b081908, 0x080808192b190808, 0x0808082b08080808,
0x0808082b0808082b, 0x0808082b08081919, 0x0808082b08082b08,
0x0808082b08190819, 0x0808082b08191908, 0x0808082b082b0808,
0x0808082b19080819, 0x0808082b19081908, 0x0808082b19190808,
0x0808082b19191919, 0x0808082b2b080808, 0x0808082b2b082b2b,
0x0808190808080819, 0x0808190808081908, 0x080819080808192b,
0x0808190808082b19, 0x0808190808190808, 0x080819080819082b,
0x0808190808191919, 0x0808190808192b08, 0x08081908082b0819,
0x08081908082b1908, 0x0808190819080808, 0x080819081908082b,
0x0808190819081919, 0x0808190819082b08, 0x0808190819190819,
0x0808190819191908, 0x080819081919192b, 0x08081908192b0808,
0x080819082b080819, 0x080819082b081908, 0x080819082b190808,
0x0808191908080808, 0x080819190808082b, 0x0808191908081919,
0x0808191908082b08, 0x0808191908190819, 0x0808191908191908,
0x08081919082b0808, 0x0808191919080819, 0x0808191919081908,
0x0808191919190808, 0x08081919192b0819, 0x080819192b080808,
0x0808192b08080819, 0x0808192b08081908, 0x0808192b08190808,
0x0808192b082b192b, 0x0808192b19080808, 0x0808192b1908082b,
0x0808192b2b081908, 0x08082b0808080808, 0x08082b080808082b,
0x08082b0808081919, 0x08082b0808082b08, 0x08082b0808082b2b,
0x08082b0808190819, 0x08082b0808191908, 0x08082b08082b0808,
0x08082b08082b1919, 0x08082b0819080819, 0x08082b0819081908,
0x08082b0819190808, 0x08082b0819192b08, 0x08082b082b080808,
0x08082b082b2b0808, 0x08082b082b2b2b2b, 0x08082b1908080819,
0x08082b1908081908, 0x08082b1908190808, 0x08082b1919080808,
0x08082b192b080819, 0x08082b192b082b19, 0x08082b2b08080808,
0x08082b2b082b0808, 0x08082b2b082b2b08, 0x08082b2b2b19192b,
0x08082b2b2b2b0808, 0x0819080808080819, 0x0819080808081908,
0x081908080808192b, 0x0819080808082b19, 0x0819080808190808,
0x081908080819082b, 0x0819080808191919, 0x0819080808192b08,
0x08190808082b0819, 0x08190808082b1908, 0x0819080819080808,
0x081908081908082b, 0x0819080819081919, 0x0819080819082b08,
0x0819080819190819, 0x0819080819191908, 0x08190808192b0808,
0x08190808192b2b2b, 0x081908082b080819, 0x081908082b081908,
0x081908082b190808, 0x0819081908080808, 0x081908190808082b,
0x0819081908081919, 0x0819081908082b08, 0x0819081908190819,
0x0819081908191908, 0x08190819082b0808, 0x0819081919080819,
0x0819081919081908, 0x0819081919190808, 0x081908192b080808,
0x081908192b191908, 0x081908192b19192b, 0x0819082b08080819,
0x0819082b08081908, 0x0819082b0808192b, 0x0819082b08190808,
0x0819082b19080808, 0x0819082b192b0808, 0x0819190808080808,
0x081919080808082b, 0x0819190808081919, 0x0819190808082b08,
0x0819190808190819, 0x0819190808191908, 0x08191908082b0808,
0x0819190819080819, 0x0819190819081908, 0x0819190819082b19,
0x0819190819190808, 0x08191908192b1908, 0x081919082b080808,
0x0819191908080819, 0x0819191908081908, 0x0819191908190808,
0x0819191919080808, 0x0819192b08080808, 0x0819192b08191908,
0x0819192b19082b19, 0x08192b0808080819, 0x08192b0808081908,
0x08192b0808190808, 0x08192b080819082b, 0x08192b0819080808,
0x08192b0819191908, 0x08192b082b08192b, 0x08192b1908080808,
0x08192b1908081919, 0x08192b19192b192b, 0x08192b2b19190819,
0x08192b2b2b2b2b19, 0x082b080808080808, 0x082b08080808082b,
0x082b080808081919, 0x082b080808082b08, 0x082b080808082b2b,
0x082b080808190819, 0x082b080808191908, 0x082b0808082b0808,
0x082b080819080819, 0x082b080819081908, 0x082b080819190808,
0x082b08082b080808, 0x082b08082b2b0808, 0x082b081908080819,
0x082b081908081908, 0x082b081908190808, 0x082b081919080808,
0x082b081919082b08, 0x082b0819192b1919, 0x082b082b08080808,
0x082b082b082b082b, 0x082b082b2b080808, 0x082b082b2b2b2b08,
0x082b190808080819, 0x082b190808081908, 0x082b190808190808,
0x082b1908082b2b19, 0x082b190819080808, 0x082b191908080808,
0x082b191919080819, 0x082b19191919082b, 0x082b19192b192b19,
0x082b192b08080819, 0x082b192b08192b2b, 0x082b192b2b2b192b,
0x082b2b0808080808, 0x082b2b0808082b08, 0x082b2b0808082b2b,
0x082b2b08082b0808, 0x082b2b0819191919, 0x082b2b082b082b08,
0x082b2b082b2b082b, 0x082b2b19192b2b08, 0x082b2b192b190808,
0x082b2b2b08082b08, 0x082b2b2b082b0808, 0x082b2b2b2b08082b,
0x082b2b2b2b082b08, 0x082b2b2b2b082b2b, 0x1908080808080819,
0x1908080808081908, 0x190808080808192b, 0x1908080808082b19,
0x1908080808190808, 0x190808080819082b, 0x1908080808191919,
0x1908080808192b08, 0x19080808082b0819, 0x19080808082b1908,
0x1908080819080808, 0x190808081908082b, 0x1908080819081919,
0x1908080819082b08, 0x1908080819082b2b, 0x1908080819190819,
0x1908080819191908, 0x19080808192b0808, 0x19080808192b1919,
0x190808082b080819, 0x190808082b081908, 0x190808082b190808,
0x1908081908080808, 0x190808190808082b, 0x1908081908081919,
0x1908081908082b08, 0x1908081908190819, 0x1908081908191908,
0x19080819082b0808, 0x1908081919080819, 0x1908081919081908,
0x1908081919190808, 0x190808192b080808, 0x190808192b081919,
0x190808192b2b082b, 0x1908082b08080819, 0x1908082b08081908,
0x1908082b08190808, 0x1908082b0819082b, 0x1908082b082b2b19,
0x1908082b19080808, 0x1908190808080808, 0x190819080808082b,
0x1908190808081919, 0x1908190808082b08, 0x1908190808190819,
0x1908190808191908, 0x1908190808192b19, 0x19081908082b0808,
0x1908190819080819, 0x1908190819081908, 0x1908190819190808,
0x190819082b080808, 0x190819082b191908, 0x1908191908080819,
0x1908191908081908, 0x1908191908190808, 0x19081919082b1908,
0x1908191919080808, 0x190819192b192b2b, 0x1908192b08080808,
0x1908192b08082b2b, 0x1908192b19081908, 0x1908192b19190808,
0x19082b0808080819, 0x19082b0808081908, 0x19082b0808190808,
0x19082b0819080808, 0x19082b0819081919, 0x19082b0819191908,
0x19082b08192b082b, 0x19082b1908080808, 0x19082b1908190819,
0x19082b1919081908, 0x19082b1919190808, 0x19082b19192b2b19,
0x19082b2b08081908, 0x1919080808080808, 0x191908080808082b,
0x1919080808081919, 0x1919080808082b08, 0x1919080808190819,
0x1919080808191908, 0x19190808082b0808, 0x19190808082b2b08,
0x1919080819080819, 0x1919080819081908, 0x1919080819190808,
0x191908082b080808, 0x1919081908080819, 0x1919081908081908,
0x1919081908190808, 0x1919081908191919, 0x1919081919080808,
0x191908191908082b, 0x1919082b08080808, 0x1919082b19081908,
0x1919082b2b2b2b2b, 0x1919190808080819, 0x1919190808081908,
0x1919190808190808, 0x19191908082b0819, 0x1919190819080808,
0x19191908192b0808, 0x191919082b080819, 0x191919082b2b0819,
0x1919191908080808, 0x1919191908082b08, 0x191919192b080808,
0x191919192b082b08, 0x1919192b082b0819, 0x1919192b192b2b08,
0x1919192b2b2b0819, 0x19192b0808080808, 0x19192b0808191908,
0x19192b0819080819, 0x19192b0819190808, 0x19192b082b192b19,
0x19192b1908192b2b, 0x19192b1919080808, 0x19192b191908082b,
0x19192b2b2b081919, 0x192b080808080819, 0x192b080808081908,
0x192b080808190808, 0x192b080819080808, 0x192b080819191908,
0x192b0808192b082b, 0x192b08082b08192b, 0x192b08082b2b2b19,
0x192b081908080808, 0x192b082b082b1908, 0x192b082b19082b2b,
0x192b082b2b19082b, 0x192b190808080808, 0x192b19080819192b,
0x192b191908190808, 0x192b191919080808, 0x192b191919081919,
0x192b19192b2b1908, 0x192b2b0808080819, 0x192b2b08192b2b2b,
0x192b2b19082b1919, 0x192b2b2b0808192b, 0x192b2b2b19191908,
0x192b2b2b192b082b, 0x2b08080808080808, 0x2b0808080808082b,
0x2b08080808081919, 0x2b08080808082b08, 0x2b08080808190819,
0x2b08080808191908, 0x2b080808082b0808, 0x2b080808082b2b2b,
0x2b08080819080819, 0x2b08080819081908, 0x2b08080819190808,
0x2b0808082b080808, 0x2b0808082b08082b, 0x2b0808082b2b2b08,
0x2b0808082b2b2b2b, 0x2b08081908080819, 0x2b08081908081908,
0x2b0808190808192b, 0x2b08081908190808, 0x2b08081919080808,
0x2b08081919190819, 0x2b08081919192b19, 0x2b08082b08080808,
0x2b08082b082b0808, 0x2b08082b2b080808, 0x2b08082b2b08082b,
0x2b08082b2b2b0808, 0x2b08082b2b2b2b08, 0x2b08190808080819,
0x2b08190808081908, 0x2b08190808190808, 0x2b0819080819082b,
0x2b08190808191919, 0x2b08190819080808, 0x2b081908192b0808,
0x2b0819082b082b19, 0x2b08191908080808, 0x2b08191919081908,
0x2b0819192b2b1919, 0x2b08192b08192b08, 0x2b08192b192b2b2b,
0x2b082b0808080808, 0x2b082b0808082b08, 0x2b082b08082b1919,
0x2b082b0819192b2b, 0x2b082b082b080808, 0x2b082b082b08082b,
0x2b082b082b2b2b08, 0x2b082b190808192b, 0x2b082b2b082b082b,
0x2b082b2b2b080808, 0x2b082b2b2b082b08, 0x2b082b2b2b19192b,
0x2b082b2b2b2b2b08, 0x2b19080808080819, 0x2b19080808081908,
0x2b19080808190808, 0x2b19080819080808, 0x2b1908081919192b,
0x2b1908082b081908, 0x2b19081908080808, 0x2b190819082b082b,
0x2b190819192b1908, 0x2b19082b1919192b, 0x2b19082b2b082b19,
0x2b19190808080808, 0x2b19190808081919, 0x2b19190819081908,
0x2b19190819190808, 0x2b19190819192b08, 0x2b191919082b2b19,
0x2b1919192b190808, 0x2b1919192b19082b, 0x2b19192b19080819,
0x2b192b0819190819, 0x2b192b082b2b192b, 0x2b192b1919082b19,
0x2b192b2b08191919, 0x2b192b2b192b0808, 0x2b2b080808080808,
0x2b2b08080808082b, 0x2b2b080808082b08, 0x2b2b080808082b2b,
0x2b2b0808082b0808, 0x2b2b0808082b2b2b, 0x2b2b08082b2b0808,
0x2b2b081919190819, 0x2b2b081919192b19, 0x2b2b08192b2b192b,
0x2b2b082b08080808, 0x2b2b082b0808082b, 0x2b2b082b08082b08,
0x2b2b082b082b2b2b, 0x2b2b082b2b080808, 0x2b2b082b2b2b0808,
0x2b2b190819080808, 0x2b2b19082b191919, 0x2b2b192b192b1919,
0x2b2b192b2b192b08, 0x2b2b2b0808082b2b, 0x2b2b2b08082b0808,
0x2b2b2b08082b082b, 0x2b2b2b08082b2b08, 0x2b2b2b082b2b0808,
0x2b2b2b082b2b2b08, 0x2b2b2b1908081908, 0x2b2b2b192b081908,
0x2b2b2b192b08192b, 0x2b2b2b2b082b2b08, 0x2b2b2b2b082b2b2b,
0x2b2b2b2b2b190819, 0x2b2b2b2b2b2b2b2b,
});
static dpct::global_memory<const uint32_t, 1> iq3xxs_grid(
sycl::range<1>(256),
{
0x04040404, 0x04040414, 0x04040424, 0x04040c0c, 0x04040c1c, 0x04040c3e,
0x04041404, 0x04041414, 0x04041c0c, 0x04042414, 0x04043e1c, 0x04043e2c,
0x040c040c, 0x040c041c, 0x040c0c04, 0x040c0c14, 0x040c140c, 0x040c142c,
0x040c1c04, 0x040c1c14, 0x040c240c, 0x040c2c24, 0x040c3e04, 0x04140404,
0x04140414, 0x04140424, 0x04140c0c, 0x04141404, 0x04141414, 0x04141c0c,
0x04141c1c, 0x04141c3e, 0x04142c0c, 0x04142c3e, 0x04143e2c, 0x041c040c,
0x041c043e, 0x041c0c04, 0x041c0c14, 0x041c142c, 0x041c3e04, 0x04240c1c,
0x04241c3e, 0x04242424, 0x04242c3e, 0x04243e1c, 0x04243e2c, 0x042c040c,
0x042c043e, 0x042c1c14, 0x042c2c14, 0x04341c2c, 0x04343424, 0x043e0c04,
0x043e0c24, 0x043e0c34, 0x043e241c, 0x043e340c, 0x0c04040c, 0x0c04041c,
0x0c040c04, 0x0c040c14, 0x0c04140c, 0x0c04141c, 0x0c041c04, 0x0c041c14,
0x0c041c24, 0x0c04243e, 0x0c042c04, 0x0c0c0404, 0x0c0c0414, 0x0c0c0c0c,
0x0c0c1404, 0x0c0c1414, 0x0c14040c, 0x0c14041c, 0x0c140c04, 0x0c140c14,
0x0c14140c, 0x0c141c04, 0x0c143e14, 0x0c1c0404, 0x0c1c0414, 0x0c1c1404,
0x0c1c1c0c, 0x0c1c2434, 0x0c1c3434, 0x0c24040c, 0x0c24042c, 0x0c242c04,
0x0c2c1404, 0x0c2c1424, 0x0c2c2434, 0x0c2c3e0c, 0x0c34042c, 0x0c3e1414,
0x0c3e2404, 0x14040404, 0x14040414, 0x14040c0c, 0x14040c1c, 0x14041404,
0x14041414, 0x14041434, 0x14041c0c, 0x14042414, 0x140c040c, 0x140c041c,
0x140c042c, 0x140c0c04, 0x140c0c14, 0x140c140c, 0x140c1c04, 0x140c341c,
0x140c343e, 0x140c3e04, 0x14140404, 0x14140414, 0x14140c0c, 0x14140c3e,
0x14141404, 0x14141414, 0x14141c3e, 0x14142404, 0x14142c2c, 0x141c040c,
0x141c0c04, 0x141c0c24, 0x141c3e04, 0x141c3e24, 0x14241c2c, 0x14242c1c,
0x142c041c, 0x142c143e, 0x142c240c, 0x142c3e24, 0x143e040c, 0x143e041c,
0x143e0c34, 0x143e242c, 0x1c04040c, 0x1c040c04, 0x1c040c14, 0x1c04140c,
0x1c04141c, 0x1c042c04, 0x1c04342c, 0x1c043e14, 0x1c0c0404, 0x1c0c0414,
0x1c0c1404, 0x1c0c1c0c, 0x1c0c2424, 0x1c0c2434, 0x1c14040c, 0x1c14041c,
0x1c140c04, 0x1c14142c, 0x1c142c14, 0x1c143e14, 0x1c1c0c0c, 0x1c1c1c1c,
0x1c241c04, 0x1c24243e, 0x1c243e14, 0x1c2c0404, 0x1c2c0434, 0x1c2c1414,
0x1c2c2c2c, 0x1c340c24, 0x1c341c34, 0x1c34341c, 0x1c3e1c1c, 0x1c3e3404,
0x24040424, 0x24040c3e, 0x24041c2c, 0x24041c3e, 0x24042c1c, 0x24042c3e,
0x240c3e24, 0x24141404, 0x24141c3e, 0x24142404, 0x24143404, 0x24143434,
0x241c043e, 0x241c242c, 0x24240424, 0x24242c0c, 0x24243424, 0x242c142c,
0x242c241c, 0x242c3e04, 0x243e042c, 0x243e0c04, 0x243e0c14, 0x243e1c04,
0x2c040c14, 0x2c04240c, 0x2c043e04, 0x2c0c0404, 0x2c0c0434, 0x2c0c1434,
0x2c0c2c2c, 0x2c140c24, 0x2c141c14, 0x2c143e14, 0x2c1c0414, 0x2c1c2c1c,
0x2c240c04, 0x2c24141c, 0x2c24143e, 0x2c243e14, 0x2c2c0414, 0x2c2c1c0c,
0x2c342c04, 0x2c3e1424, 0x2c3e2414, 0x34041424, 0x34042424, 0x34042434,
0x34043424, 0x340c140c, 0x340c340c, 0x34140c3e, 0x34143424, 0x341c1c04,
0x341c1c34, 0x34242424, 0x342c042c, 0x342c2c14, 0x34341c1c, 0x343e041c,
0x343e140c, 0x3e04041c, 0x3e04042c, 0x3e04043e, 0x3e040c04, 0x3e041c14,
0x3e042c14, 0x3e0c1434, 0x3e0c2404, 0x3e140c14, 0x3e14242c, 0x3e142c14,
0x3e1c0404, 0x3e1c0c2c, 0x3e1c1c1c, 0x3e1c3404, 0x3e24140c, 0x3e24240c,
0x3e2c0404, 0x3e2c0414, 0x3e2c1424, 0x3e341c04,
});
static dpct::global_memory<const uint32_t, 1> iq3s_grid(
sycl::range<1>(512),
{
0x01010101, 0x01010103, 0x01010105, 0x0101010b, 0x0101010f, 0x01010301, 0x01010303, 0x01010305,
0x01010309, 0x0101030d, 0x01010501, 0x01010503, 0x0101050b, 0x01010707, 0x01010901, 0x01010905,
0x0101090b, 0x0101090f, 0x01010b03, 0x01010b07, 0x01010d01, 0x01010d05, 0x01010f03, 0x01010f09,
0x01010f0f, 0x01030101, 0x01030103, 0x01030105, 0x01030109, 0x01030301, 0x01030303, 0x0103030b,
0x01030501, 0x01030507, 0x0103050f, 0x01030703, 0x0103070b, 0x01030909, 0x01030d03, 0x01030d0b,
0x01030f05, 0x01050101, 0x01050103, 0x0105010b, 0x0105010f, 0x01050301, 0x01050307, 0x0105030d,
0x01050503, 0x0105050b, 0x01050701, 0x01050709, 0x01050905, 0x0105090b, 0x0105090f, 0x01050b03,
0x01050b07, 0x01050f01, 0x01050f07, 0x01070107, 0x01070303, 0x0107030b, 0x01070501, 0x01070505,
0x01070703, 0x01070707, 0x0107070d, 0x01070909, 0x01070b01, 0x01070b05, 0x01070d0f, 0x01070f03,
0x01070f0b, 0x01090101, 0x01090307, 0x0109030f, 0x01090503, 0x01090509, 0x01090705, 0x01090901,
0x01090907, 0x01090b03, 0x01090f01, 0x010b0105, 0x010b0109, 0x010b0501, 0x010b0505, 0x010b050d,
0x010b0707, 0x010b0903, 0x010b090b, 0x010b090f, 0x010b0d0d, 0x010b0f07, 0x010d010d, 0x010d0303,
0x010d0307, 0x010d0703, 0x010d0b05, 0x010d0f03, 0x010f0101, 0x010f0105, 0x010f0109, 0x010f0501,
0x010f0505, 0x010f050d, 0x010f0707, 0x010f0b01, 0x010f0b09, 0x03010101, 0x03010103, 0x03010105,
0x03010109, 0x03010301, 0x03010303, 0x03010307, 0x0301030b, 0x0301030f, 0x03010501, 0x03010505,
0x03010703, 0x03010709, 0x0301070d, 0x03010b09, 0x03010b0d, 0x03010d03, 0x03010f05, 0x03030101,
0x03030103, 0x03030107, 0x0303010d, 0x03030301, 0x03030309, 0x03030503, 0x03030701, 0x03030707,
0x03030903, 0x03030b01, 0x03030b05, 0x03030f01, 0x03030f0d, 0x03050101, 0x03050305, 0x0305030b,
0x0305030f, 0x03050501, 0x03050509, 0x03050705, 0x03050901, 0x03050907, 0x03050b0b, 0x03050d01,
0x03050f05, 0x03070103, 0x03070109, 0x0307010f, 0x03070301, 0x03070307, 0x03070503, 0x0307050f,
0x03070701, 0x03070709, 0x03070903, 0x03070d05, 0x03070f01, 0x03090107, 0x0309010b, 0x03090305,
0x03090309, 0x03090703, 0x03090707, 0x03090905, 0x0309090d, 0x03090b01, 0x03090b09, 0x030b0103,
0x030b0301, 0x030b0307, 0x030b0503, 0x030b0701, 0x030b0705, 0x030b0b03, 0x030d0501, 0x030d0509,
0x030d050f, 0x030d0909, 0x030d090d, 0x030f0103, 0x030f0107, 0x030f0301, 0x030f0305, 0x030f0503,
0x030f070b, 0x030f0903, 0x030f0d05, 0x030f0f01, 0x05010101, 0x05010103, 0x05010107, 0x0501010b,
0x0501010f, 0x05010301, 0x05010305, 0x05010309, 0x0501030d, 0x05010503, 0x05010507, 0x0501050f,
0x05010701, 0x05010705, 0x05010903, 0x05010907, 0x0501090b, 0x05010b01, 0x05010b05, 0x05010d0f,
0x05010f01, 0x05010f07, 0x05010f0b, 0x05030101, 0x05030105, 0x05030301, 0x05030307, 0x0503030f,
0x05030505, 0x0503050b, 0x05030703, 0x05030709, 0x05030905, 0x05030b03, 0x05050103, 0x05050109,
0x0505010f, 0x05050503, 0x05050507, 0x05050701, 0x0505070f, 0x05050903, 0x05050b07, 0x05050b0f,
0x05050f03, 0x05050f09, 0x05070101, 0x05070105, 0x0507010b, 0x05070303, 0x05070505, 0x05070509,
0x05070703, 0x05070707, 0x05070905, 0x05070b01, 0x05070d0d, 0x05090103, 0x0509010f, 0x05090501,
0x05090507, 0x05090705, 0x0509070b, 0x05090903, 0x05090f05, 0x05090f0b, 0x050b0109, 0x050b0303,
0x050b0505, 0x050b070f, 0x050b0901, 0x050b0b07, 0x050b0f01, 0x050d0101, 0x050d0105, 0x050d010f,
0x050d0503, 0x050d0b0b, 0x050d0d03, 0x050f010b, 0x050f0303, 0x050f050d, 0x050f0701, 0x050f0907,
0x050f0b01, 0x07010105, 0x07010303, 0x07010307, 0x0701030b, 0x0701030f, 0x07010505, 0x07010703,
0x07010707, 0x0701070b, 0x07010905, 0x07010909, 0x0701090f, 0x07010b03, 0x07010d07, 0x07010f03,
0x07030103, 0x07030107, 0x0703010b, 0x07030309, 0x07030503, 0x07030507, 0x07030901, 0x07030d01,
0x07030f05, 0x07030f0d, 0x07050101, 0x07050305, 0x07050501, 0x07050705, 0x07050709, 0x07050b01,
0x07070103, 0x07070301, 0x07070309, 0x07070503, 0x07070507, 0x0707050f, 0x07070701, 0x07070903,
0x07070907, 0x0707090f, 0x07070b0b, 0x07070f07, 0x07090107, 0x07090303, 0x0709030d, 0x07090505,
0x07090703, 0x07090b05, 0x07090d01, 0x07090d09, 0x070b0103, 0x070b0301, 0x070b0305, 0x070b050b,
0x070b0705, 0x070b0909, 0x070b0b0d, 0x070b0f07, 0x070d030d, 0x070d0903, 0x070f0103, 0x070f0107,
0x070f0501, 0x070f0505, 0x070f070b, 0x09010101, 0x09010109, 0x09010305, 0x09010501, 0x09010509,
0x0901050f, 0x09010705, 0x09010903, 0x09010b01, 0x09010f01, 0x09030105, 0x0903010f, 0x09030303,
0x09030307, 0x09030505, 0x09030701, 0x0903070b, 0x09030907, 0x09030b03, 0x09030b0b, 0x09050103,
0x09050107, 0x09050301, 0x0905030b, 0x09050503, 0x09050707, 0x09050901, 0x09050b0f, 0x09050d05,
0x09050f01, 0x09070109, 0x09070303, 0x09070307, 0x09070501, 0x09070505, 0x09070703, 0x0907070b,
0x09090101, 0x09090105, 0x09090509, 0x0909070f, 0x09090901, 0x09090f03, 0x090b010b, 0x090b010f,
0x090b0503, 0x090b0d05, 0x090d0307, 0x090d0709, 0x090d0d01, 0x090f0301, 0x090f030b, 0x090f0701,
0x090f0907, 0x090f0b03, 0x0b010105, 0x0b010301, 0x0b010309, 0x0b010505, 0x0b010901, 0x0b010909,
0x0b01090f, 0x0b010b05, 0x0b010d0d, 0x0b010f09, 0x0b030103, 0x0b030107, 0x0b03010b, 0x0b030305,
0x0b030503, 0x0b030705, 0x0b030f05, 0x0b050101, 0x0b050303, 0x0b050507, 0x0b050701, 0x0b05070d,
0x0b050b07, 0x0b070105, 0x0b07010f, 0x0b070301, 0x0b07050f, 0x0b070909, 0x0b070b03, 0x0b070d0b,
0x0b070f07, 0x0b090103, 0x0b090109, 0x0b090501, 0x0b090705, 0x0b09090d, 0x0b0b0305, 0x0b0b050d,
0x0b0b0b03, 0x0b0b0b07, 0x0b0d0905, 0x0b0f0105, 0x0b0f0109, 0x0b0f0505, 0x0d010303, 0x0d010307,
0x0d01030b, 0x0d010703, 0x0d010707, 0x0d010d01, 0x0d030101, 0x0d030501, 0x0d03050f, 0x0d030d09,
0x0d050305, 0x0d050709, 0x0d050905, 0x0d050b0b, 0x0d050d05, 0x0d050f01, 0x0d070101, 0x0d070309,
0x0d070503, 0x0d070901, 0x0d09050b, 0x0d090907, 0x0d090d05, 0x0d0b0101, 0x0d0b0107, 0x0d0b0709,
0x0d0b0d01, 0x0d0d010b, 0x0d0d0901, 0x0d0f0303, 0x0d0f0307, 0x0f010101, 0x0f010109, 0x0f01010f,
0x0f010501, 0x0f010505, 0x0f01070d, 0x0f010901, 0x0f010b09, 0x0f010d05, 0x0f030105, 0x0f030303,
0x0f030509, 0x0f030907, 0x0f03090b, 0x0f050103, 0x0f050109, 0x0f050301, 0x0f05030d, 0x0f050503,
0x0f050701, 0x0f050b03, 0x0f070105, 0x0f070705, 0x0f07070b, 0x0f070b07, 0x0f090103, 0x0f09010b,
0x0f090307, 0x0f090501, 0x0f090b01, 0x0f0b0505, 0x0f0b0905, 0x0f0d0105, 0x0f0d0703, 0x0f0f0101,
});
static dpct::global_memory<const uint64_t, 1> iq1s_grid(
sycl::range<1>(512),
{
0xffffffffffff0101, 0xffffffffff01ff00, 0xffffffffff010100, 0xffffffff00000000,
0xffffffff01ff00ff, 0xffffffff01ff0001, 0xffffffff0101ffff, 0xffffffff0101ff01,
0xffffff00ff000000, 0xffffff000000ff00, 0xffffff00000000ff, 0xffffff0000000100,
0xffffff0000010000, 0xffffff0001000000, 0xffffff01ffff00ff, 0xffffff01ff01ff00,
0xffffff01ff010100, 0xffffff0100000001, 0xffffff0101ffff00, 0xffffff0101ff0101,
0xffffff0101010100, 0xffff00ffff00ff01, 0xffff00ffff0000ff, 0xffff00ff00ff0100,
0xffff00ff0100ff00, 0xffff00ff010001ff, 0xffff0000ff0101ff, 0xffff000000ffff00,
0xffff000000000000, 0xffff00000001ff01, 0xffff000001000101, 0xffff0000010100ff,
0xffff0001ffff0100, 0xffff00010000ff00, 0xffff000100010101, 0xffff000101000000,
0xffff01ffffff0000, 0xffff01ffff01ffff, 0xffff01ffff010100, 0xffff01ff00000000,
0xffff01ff01ffffff, 0xffff01ff01ff0001, 0xffff01ff0101ffff, 0xffff01ff01010001,
0xffff0100ffffff01, 0xffff01000000ffff, 0xffff010000000100, 0xffff010001ff01ff,
0xffff010001000000, 0xffff0101ff000000, 0xffff0101000101ff, 0xffff010101ffff01,
0xffff01010101ff00, 0xff00ffffff000000, 0xff00ffff00ffff00, 0xff00ffff00000001,
0xff00ffff000001ff, 0xff00ffff01010000, 0xff00ff00ffff0000, 0xff00ff00ff00ff00,
0xff00ff00ff0000ff, 0xff00ff00ff000100, 0xff00ff00ff010001, 0xff00ff0000ff0001,
0xff00ff000000ffff, 0xff00ff0000000000, 0xff00ff000001ff00, 0xff00ff0000010100,
0xff00ff0001ff0000, 0xff00ff000100ff00, 0xff00ff0001000100, 0xff00ff01ff000000,
0xff00ff0100ff0000, 0xff00ff01000001ff, 0xff00ff0101010001, 0xff0000ff00000000,
0xff0000ff0001ff00, 0xff0000ff00010100, 0xff000000ffff0101, 0xff000000ff000000,
0xff000000ff01ff00, 0xff00000000ff0000, 0xff0000000000ff00, 0xff000000000000ff,
0xff00000000000000, 0xff00000000000001, 0xff00000000000100, 0xff0000000001ffff,
0xff00000000010000, 0xff00000001000000, 0xff00000001010100, 0xff000001ff00ff01,
0xff000001ff0100ff, 0xff00000100000000, 0xff0000010001ff00, 0xff00000101ff0100,
0xff0000010100ff00, 0xff0001ff00ff00ff, 0xff0001ff00000101, 0xff0001ff000100ff,
0xff0001ff01000000, 0xff000100ff0001ff, 0xff0001000000ff01, 0xff00010000000000,
0xff00010000010001, 0xff00010000010100, 0xff00010001ffff00, 0xff00010001ff0101,
0xff00010001010000, 0xff000101ffffffff, 0xff000101ff000101, 0xff00010101ff00ff,
0xff00010101000001, 0xff000101010100ff, 0xff01ffffff000101, 0xff01ffffff01ffff,
0xff01ffffff01ff01, 0xff01ffffff0101ff, 0xff01ffff00000000, 0xff01ffff01ff0001,
0xff01ffff0101ff01, 0xff01ff00ff000000, 0xff01ff0000ff0100, 0xff01ff000000ff01,
0xff01ff0000010000, 0xff01ff00010000ff, 0xff01ff01ff01ff00, 0xff01ff0100000101,
0xff0100ffffff0000, 0xff0100ffff010000, 0xff0100ff01ff00ff, 0xff0100ff01000100,
0xff0100ff010100ff, 0xff010000ffffff01, 0xff01000000000000, 0xff0100000101ff00,
0xff010001ffff00ff, 0xff010001ff000100, 0xff01000100ffff00, 0xff01000100010001,
0xff01000101ff0001, 0xff010001010001ff, 0xff0101ffffffffff, 0xff0101ffff01ffff,
0xff0101ffff010101, 0xff0101ff0000ff00, 0xff0101ff01010001, 0xff010100ff000000,
0xff010100ff01ff01, 0xff01010000ff0001, 0xff01010000000100, 0xff01010001000000,
0xff0101010100ffff, 0x00ffffff0000ff01, 0x00ffffff000000ff, 0x00ffffff00000100,
0x00ffffff00010000, 0x00ffff00ffff0001, 0x00ffff00ff0000ff, 0x00ffff00ff000100,
0x00ffff0000000000, 0x00ffff0001000100, 0x00ffff0001010001, 0x00ffff01ff00ff01,
0x00ffff0100ff0100, 0x00ffff010000ff00, 0x00ffff01000100ff, 0x00ffff0101ff00ff,
0x00ffff010101ff00, 0x00ff00ffffffffff, 0x00ff00ffffff01ff, 0x00ff00ffff000101,
0x00ff00ff00000000, 0x00ff00ff000101ff, 0x00ff00ff01010101, 0x00ff0000ff000000,
0x00ff0000ff01ffff, 0x00ff000000ff0000, 0x00ff00000000ff00, 0x00ff0000000000ff,
0x00ff000000000000, 0x00ff000000000001, 0x00ff000000000100, 0x00ff000000010000,
0x00ff000001ffff01, 0x00ff000001000000, 0x00ff0001ff000101, 0x00ff000100ffffff,
0x00ff000100000000, 0x00ff0001010001ff, 0x00ff01ffff000000, 0x00ff01ff0001ff00,
0x00ff01ff01ff0100, 0x00ff0100ff01ff01, 0x00ff010000ff00ff, 0x00ff010000ff0101,
0x00ff010000000000, 0x00ff010000010101, 0x00ff01000100ff00, 0x00ff010001010000,
0x00ff0101ffffff00, 0x00ff01010000ff01, 0x00ff010100000100, 0x00ff010101ff0000,
0x0000ffffffff0100, 0x0000ffffff00ff00, 0x0000ffffff0000ff, 0x0000ffffff010000,
0x0000ffff00000000, 0x0000ffff00010101, 0x0000ffff01ffff01, 0x0000ffff01000100,
0x0000ff00ff000000, 0x0000ff00ff01ff00, 0x0000ff00ff0101ff, 0x0000ff0000ff0000,
0x0000ff000000ff00, 0x0000ff00000000ff, 0x0000ff0000000000, 0x0000ff0000000001,
0x0000ff0000000100, 0x0000ff0000010000, 0x0000ff0001ffffff, 0x0000ff0001ff01ff,
0x0000ff0001000000, 0x0000ff000101ffff, 0x0000ff01ffff0101, 0x0000ff01ff010000,
0x0000ff0100000000, 0x0000ff0101000101, 0x000000ffffff0001, 0x000000ffff000000,
0x000000ff00ff0000, 0x000000ff0000ff00, 0x000000ff000000ff, 0x000000ff00000000,
0x000000ff00000001, 0x000000ff00000100, 0x000000ff00010000, 0x000000ff01000000,
0x000000ff0101ff00, 0x00000000ffff0000, 0x00000000ff00ff00, 0x00000000ff0000ff,
0x00000000ff000000, 0x00000000ff000001, 0x00000000ff000100, 0x00000000ff010000,
0x0000000000ffff00, 0x0000000000ff00ff, 0x0000000000ff0000, 0x0000000000ff0001,
0x0000000000ff0100, 0x000000000000ffff, 0x000000000000ff00, 0x000000000000ff01,
0x00000000000000ff, 0x0000000000000001, 0x00000000000001ff, 0x0000000000000100,
0x0000000000000101, 0x000000000001ff00, 0x00000000000100ff, 0x0000000000010000,
0x0000000000010001, 0x0000000000010100, 0x0000000001ff0000, 0x000000000100ff00,
0x00000000010000ff, 0x0000000001000000, 0x0000000001000001, 0x0000000001000100,
0x0000000001010000, 0x00000001ffff01ff, 0x00000001ff000000, 0x0000000100ff0000,
0x000000010000ff00, 0x00000001000000ff, 0x0000000100000000, 0x0000000100000001,
0x0000000100000100, 0x0000000100010000, 0x0000000101000000, 0x000001ffff00ff00,
0x000001ffff010001, 0x000001ffff0101ff, 0x000001ff00ffff01, 0x000001ff0000ffff,
0x000001ff00000000, 0x000001ff010000ff, 0x000001ff01010100, 0x00000100ffff0100,
0x00000100ff000000, 0x0000010000ff0000, 0x000001000000ff00, 0x00000100000000ff,
0x0000010000000000, 0x0000010000000001, 0x0000010000000100, 0x0000010000010000,
0x0000010001000000, 0x000001000101ff01, 0x00000101ffff0001, 0x00000101ff01ffff,
0x0000010100000000, 0x0000010101010100, 0x0001ffffff000000, 0x0001ffff00ffffff,
0x0001ffff00000100, 0x0001ffff0001ff00, 0x0001ffff01000000, 0x0001ff00ffffff00,
0x0001ff00ffff01ff, 0x0001ff00ff010000, 0x0001ff0000000000, 0x0001ff0000010001,
0x0001ff0001ff0000, 0x0001ff0001010100, 0x0001ff01ff0000ff, 0x0001ff01ff000001,
0x0001ff0100ffffff, 0x0001ff010001ffff, 0x0001ff01000101ff, 0x0001ff010100ff01,
0x000100ffff00ffff, 0x000100ffff00ff01, 0x000100ffff000100, 0x000100ff00000000,
0x000100ff000101ff, 0x000100ff01ff0101, 0x000100ff0100ffff, 0x000100ff01010101,
0x00010000ff000000, 0x00010000ff010100, 0x0001000000ff0000, 0x000100000000ff00,
0x00010000000000ff, 0x0001000000000000, 0x0001000000000001, 0x0001000000000100,
0x0001000000010000, 0x0001000001ffff01, 0x0001000001000000, 0x0001000100ff0101,
0x0001000100000000, 0x00010001010100ff, 0x000101ffffff01ff, 0x000101ffffff0101,
0x000101ff00010000, 0x000101ff01ff0000, 0x000101ff0100ff01, 0x00010100ffff0000,
0x0001010000000000, 0x000101000001ffff, 0x0001010000010101, 0x00010100010001ff,
0x00010101ff00ff00, 0x00010101ff010001, 0x0001010100ffffff, 0x0001010100ff01ff,
0x00010101000101ff, 0x0001010101ff0000, 0x000101010100ff01, 0x0001010101000101,
0x01ffffffffff0101, 0x01ffffffff01ffff, 0x01ffffffff01ff01, 0x01ffffffff0101ff,
0x01ffffffff010101, 0x01ffffff00000000, 0x01ffffff01ff01ff, 0x01ffffff01000101,
0x01ffffff0101ff01, 0x01ffffff010100ff, 0x01ffff000000ff00, 0x01ffff0000000001,
0x01ffff00000001ff, 0x01ffff0000010000, 0x01ffff0001ff0000, 0x01ffff01ffffffff,
0x01ffff01ffff01ff, 0x01ffff01ff000000, 0x01ffff01ff01ffff, 0x01ffff01ff0101ff,
0x01ffff010100ffff, 0x01ff00ffffff0000, 0x01ff00ffff010000, 0x01ff00ff00ffff01,
0x01ff0000ff0000ff, 0x01ff000000000000, 0x01ff00000001ff01, 0x01ff000001ffffff,
0x01ff000001010100, 0x01ff0001ffffff01, 0x01ff0001ff010001, 0x01ff000101ff0100,
0x01ff000101000001, 0x01ff0001010100ff, 0x01ff01ffff00ffff, 0x01ff01ff00010001,
0x01ff01ff01000000, 0x01ff01ff010101ff, 0x01ff0100ff000001, 0x01ff010000ffff00,
0x01ff010000000100, 0x01ff010001ff01ff, 0x01ff01000101ffff, 0x01ff0101ffff00ff,
0x01ff0101ffff0101, 0x01ff0101ff0101ff, 0x01ff010100010000, 0x0100ffff00ff00ff,
0x0100ffff00ff0001, 0x0100ffff00000100, 0x0100ffff0100ff00, 0x0100ff00ffff0000,
0x0100ff00ff00ffff, 0x0100ff00ff00ff01, 0x0100ff00ff000100, 0x0100ff00ff010000,
0x0100ff0000000000, 0x0100ff00000100ff, 0x0100ff0001ff0101, 0x0100ff0001010101,
0x0100ff0100ff00ff, 0x0100ff0100ff0001, 0x0100ff0100000100, 0x0100ff0100010001,
0x0100ff0101000000, 0x010000ffff00ff00, 0x010000ff0000ffff, 0x010000ff00000000,
0x010000ff010001ff, 0x010000ff01010001, 0x01000000ffffff00, 0x01000000ffff0101,
0x01000000ff000000, 0x01000000ff0100ff, 0x01000000ff010101, 0x0100000000ff0000,
0x010000000000ff00, 0x01000000000000ff, 0x0100000000000000, 0x0100000000000001,
0x0100000000000100, 0x0100000000010000, 0x0100000001000000, 0x0100000100000000,
0x01000001000101ff, 0x0100000101ffff01, 0x010001ffff000101, 0x010001ff00ff0100,
0x010001ff0000ff00, 0x010001ff000100ff, 0x010001ff01ffffff, 0x01000100ffff0000,
0x01000100ff0001ff, 0x0100010000000000, 0x010001000001ff00, 0x0100010001ff0000,
0x01000100010000ff, 0x0100010001000101, 0x01000101ff00ff01, 0x0100010100ff0100,
0x010001010000ffff, 0x0100010101010001, 0x0101ffffffff0101, 0x0101ffffff0001ff,
0x0101ffffff01ffff, 0x0101ffffff010101, 0x0101ffff00000000, 0x0101ffff0101ffff,
0x0101ffff010101ff, 0x0101ff00ff000000, 0x0101ff0000ff0100, 0x0101ff000000ff00,
0x0101ff0000010000, 0x0101ff00010000ff, 0x0101ff0001000001, 0x0101ff01ff010101,
0x0101ff0100000000, 0x0101ff010101ff00, 0x010100ffffff0000, 0x010100ffff010000,
0x010100ff00ff01ff, 0x010100ff000000ff, 0x010100ff00000101, 0x010100ff01ffff00,
0x01010000ffffff01, 0x01010000ff000100, 0x01010000ff01ff01, 0x0101000000000000,
0x01010000000100ff, 0x010100000101ff01, 0x01010001ffff0000, 0x01010001ff00ffff,
0x01010001ff010000, 0x0101000101ffffff, 0x0101000101ff01ff, 0x0101000101010101,
0x010101ffff01ffff, 0x010101ff00000000, 0x010101ff0001ff01, 0x010101ff0101ffff,
0x010101ff010101ff, 0x01010100ffffffff, 0x01010100ff000001, 0x010101000000ff00,
0x0101010001010000, 0x0101010100ff0001, 0x010101010001ff01, 0x010101010101ffff,
});
static dpct::global_memory<const uint8_t, 1> ksigns_iq2xs(
sycl::range<1>(128),
{
0, 129, 130, 3, 132, 5, 6, 135, 136, 9, 10, 139, 12,
141, 142, 15, 144, 17, 18, 147, 20, 149, 150, 23, 24, 153,
154, 27, 156, 29, 30, 159, 160, 33, 34, 163, 36, 165, 166,
39, 40, 169, 170, 43, 172, 45, 46, 175, 48, 177, 178, 51,
180, 53, 54, 183, 184, 57, 58, 187, 60, 189, 190, 63, 192,
65, 66, 195, 68, 197, 198, 71, 72, 201, 202, 75, 204, 77,
78, 207, 80, 209, 210, 83, 212, 85, 86, 215, 216, 89, 90,
219, 92, 221, 222, 95, 96, 225, 226, 99, 228, 101, 102, 231,
232, 105, 106, 235, 108, 237, 238, 111, 240, 113, 114, 243, 116,
245, 246, 119, 120, 249, 250, 123, 252, 125, 126, 255,
});
static dpct::global_memory<const uint64_t, 1>
ksigns64(sycl::range<1>(128),
{
0x0000000000000000, 0xff000000000000ff, 0xff0000000000ff00,
0x000000000000ffff, 0xff00000000ff0000, 0x0000000000ff00ff,
0x0000000000ffff00, 0xff00000000ffffff, 0xff000000ff000000,
0x00000000ff0000ff, 0x00000000ff00ff00, 0xff000000ff00ffff,
0x00000000ffff0000, 0xff000000ffff00ff, 0xff000000ffffff00,
0x00000000ffffffff, 0xff0000ff00000000, 0x000000ff000000ff,
0x000000ff0000ff00, 0xff0000ff0000ffff, 0x000000ff00ff0000,
0xff0000ff00ff00ff, 0xff0000ff00ffff00, 0x000000ff00ffffff,
0x000000ffff000000, 0xff0000ffff0000ff, 0xff0000ffff00ff00,
0x000000ffff00ffff, 0xff0000ffffff0000, 0x000000ffffff00ff,
0x000000ffffffff00, 0xff0000ffffffffff, 0xff00ff0000000000,
0x0000ff00000000ff, 0x0000ff000000ff00, 0xff00ff000000ffff,
0x0000ff0000ff0000, 0xff00ff0000ff00ff, 0xff00ff0000ffff00,
0x0000ff0000ffffff, 0x0000ff00ff000000, 0xff00ff00ff0000ff,
0xff00ff00ff00ff00, 0x0000ff00ff00ffff, 0xff00ff00ffff0000,
0x0000ff00ffff00ff, 0x0000ff00ffffff00, 0xff00ff00ffffffff,
0x0000ffff00000000, 0xff00ffff000000ff, 0xff00ffff0000ff00,
0x0000ffff0000ffff, 0xff00ffff00ff0000, 0x0000ffff00ff00ff,
0x0000ffff00ffff00, 0xff00ffff00ffffff, 0xff00ffffff000000,
0x0000ffffff0000ff, 0x0000ffffff00ff00, 0xff00ffffff00ffff,
0x0000ffffffff0000, 0xff00ffffffff00ff, 0xff00ffffffffff00,
0x0000ffffffffffff, 0xffff000000000000, 0x00ff0000000000ff,
0x00ff00000000ff00, 0xffff00000000ffff, 0x00ff000000ff0000,
0xffff000000ff00ff, 0xffff000000ffff00, 0x00ff000000ffffff,
0x00ff0000ff000000, 0xffff0000ff0000ff, 0xffff0000ff00ff00,
0x00ff0000ff00ffff, 0xffff0000ffff0000, 0x00ff0000ffff00ff,
0x00ff0000ffffff00, 0xffff0000ffffffff, 0x00ff00ff00000000,
0xffff00ff000000ff, 0xffff00ff0000ff00, 0x00ff00ff0000ffff,
0xffff00ff00ff0000, 0x00ff00ff00ff00ff, 0x00ff00ff00ffff00,
0xffff00ff00ffffff, 0xffff00ffff000000, 0x00ff00ffff0000ff,
0x00ff00ffff00ff00, 0xffff00ffff00ffff, 0x00ff00ffffff0000,
0xffff00ffffff00ff, 0xffff00ffffffff00, 0x00ff00ffffffffff,
0x00ffff0000000000, 0xffffff00000000ff, 0xffffff000000ff00,
0x00ffff000000ffff, 0xffffff0000ff0000, 0x00ffff0000ff00ff,
0x00ffff0000ffff00, 0xffffff0000ffffff, 0xffffff00ff000000,
0x00ffff00ff0000ff, 0x00ffff00ff00ff00, 0xffffff00ff00ffff,
0x00ffff00ffff0000, 0xffffff00ffff00ff, 0xffffff00ffffff00,
0x00ffff00ffffffff, 0xffffffff00000000, 0x00ffffff000000ff,
0x00ffffff0000ff00, 0xffffffff0000ffff, 0x00ffffff00ff0000,
0xffffffff00ff00ff, 0xffffffff00ffff00, 0x00ffffff00ffffff,
0x00ffffffff000000, 0xffffffffff0000ff, 0xffffffffff00ff00,
0x00ffffffff00ffff, 0xffffffffffff0000, 0x00ffffffffff00ff,
0x00ffffffffffff00, 0xffffffffffffffff,
});
//#endif
static dpct::global_memory<const uint8_t, 1>
kmask_iq2xs(sycl::range<1>(8), {1, 2, 4, 8, 16, 32, 64, 128});
template<typename dst_t>
static void dequantize_block_iq2_xxs(const void * __restrict__ vx, dst_t * __restrict__ yy,
const sycl::nd_item<3> &item_ct1,
@ -14699,7 +14153,7 @@ inline void ggml_sycl_op_mul_mat_sycl(
src1_ptr, dpct::library_data_t::real_half, ne10, &beta_f16,
dst_f16.get(), dpct::library_data_t::real_half, ldc,
dpct::library_data_t::real_half)));
g_sycl_handles[id]->wait();
const to_fp32_sycl_t to_fp32_sycl = ggml_get_to_fp32_sycl(GGML_TYPE_F16);
to_fp32_sycl(dst_f16.get(), dst_dd_i, row_diff*src1_ncols, stream);
}
@ -14732,6 +14186,7 @@ inline void ggml_sycl_op_mul_mat_sycl(
dpct::get_value(&alpha, *g_sycl_handles[id]), src0_ddf_i, ne00,
src1_ddf1_i, ne10, dpct::get_value(&beta, *g_sycl_handles[id]),
dst_dd_i, ldc)));
g_sycl_handles[id]->wait();
}
(void) dst;
(void) src1_ddq_i;
@ -15868,8 +15323,8 @@ static void ggml_sycl_mul_mat_batched_sycl(const ggml_tensor *src0,
sycl_pool_alloc<sycl::half> dst_f16;
char * dst_t;
dpct::library_data_t cu_compute_type = dpct::library_data_t::real_half;
dpct::library_data_t cu_data_type = dpct::library_data_t::real_half;
dpct::library_data_t cu_compute_type = dpct::library_data_t::real_float;
dpct::library_data_t cu_data_type = dpct::library_data_t::real_float;
// dst strides
size_t nbd2 = dst->nb[2];
@ -15881,15 +15336,13 @@ static void ggml_sycl_mul_mat_batched_sycl(const ggml_tensor *src0,
const float alpha_f32 = 1.0f;
const float beta_f32 = 0.0f;
const void * alpha = &alpha_f16;
const void * beta = &beta_f16;
const void * alpha = &alpha_f32;
const void * beta = &beta_f32;
// TODO: Renable (dst->op_params[0] =! GGML_PREC_DEFAULT) pathway
// once oneMKL open source supports half, half, float, float: datatypes
dst_t = (char *) dst_f16.alloc(ne_dst);
// oneMKL open source supports half, half, float, float: datatypes
nbd2 /= sizeof(float) / sizeof(sycl::half);
nbd3 /= sizeof(float) / sizeof(sycl::half);
dst_t = (char *) dst_ddf;
GGML_ASSERT(ne12 % ne02 == 0);
GGML_ASSERT(ne13 % ne03 == 0);
@ -15929,6 +15382,7 @@ static void ggml_sycl_mul_mat_batched_sycl(const ggml_tensor *src0,
nb11 / nb10, nb12 / nb10, beta,
(char *)dst_t, cu_data_type, ne01, nb2 / nb0,
ne12 * ne13, cu_compute_type)));
g_sycl_handles[g_main_device]->wait();
} else {
const int ne23 = ne12*ne13;
@ -15959,7 +15413,7 @@ static void ggml_sycl_mul_mat_batched_sycl(const ggml_tensor *src0,
nb02, nb03, nb12_scaled, nb13_scaled,
nbd2, nbd3, r2, r3, item_ct1);
});
});
}).wait();
}
SYCL_CHECK(CHECK_TRY_ERROR(dpct::gemm_batch(
*g_sycl_handles[g_main_device], oneapi::mkl::transpose::trans,
@ -15970,11 +15424,10 @@ static void ggml_sycl_mul_mat_batched_sycl(const ggml_tensor *src0,
dpct::library_data_t::real_half, nb11 / nb10, beta,
(void **)(ptrs_dst.get() + 0 * ne23), cu_data_type, ne01, ne23,
cu_compute_type)));
g_sycl_handles[g_main_device]->wait();
}
#endif
const to_fp32_sycl_t to_fp32_sycl = ggml_get_to_fp32_sycl(GGML_TYPE_F16);
to_fp32_sycl(dst_f16.get(), dst_ddf, ne_dst, main_stream);
}
catch (sycl::exception const &exc) {
std::cerr << exc.what() << "Exception caught at file:" << __FILE__

86931
ggml-vulkan-shaders.hpp
View File

File diff suppressed because it is too large Load Diff

1940
ggml-vulkan.cpp
View File

File diff suppressed because it is too large Load Diff

1
ggml-vulkan.h
View File

@ -10,6 +10,7 @@ extern "C" {
#define GGML_VK_NAME "Vulkan"
#define GGML_VK_MAX_DEVICES 16
GGML_API void ggml_vk_instance_init(void);
GGML_API void ggml_vk_init_cpu_assist(void);
GGML_API void ggml_vk_preallocate_buffers_graph_cpu_assist(struct ggml_tensor * node);

723
ggml.c
View File

@ -1841,6 +1841,8 @@ static const char * GGML_OP_NAME[GGML_OP_COUNT] = {
"FLASH_ATTN",
"FLASH_FF",
"FLASH_ATTN_BACK",
"SSM_CONV",
"SSM_SCAN",
"WIN_PART",
"WIN_UNPART",
"GET_REL_POS",
@ -1863,7 +1865,7 @@ static const char * GGML_OP_NAME[GGML_OP_COUNT] = {
"CROSS_ENTROPY_LOSS_BACK",
};
static_assert(GGML_OP_COUNT == 74, "GGML_OP_COUNT != 74");
static_assert(GGML_OP_COUNT == 76, "GGML_OP_COUNT != 76");
static const char * GGML_OP_SYMBOL[GGML_OP_COUNT] = {
"none",
@ -1929,6 +1931,8 @@ static const char * GGML_OP_SYMBOL[GGML_OP_COUNT] = {
"flash_attn(x)",
"flash_ff(x)",
"flash_attn_back(x)",
"ssm_conv(x)",
"ssm_scan(x)",
"win_part(x)",
"win_unpart(x)",
"get_rel_pos(x)",
@ -1951,7 +1955,7 @@ static const char * GGML_OP_SYMBOL[GGML_OP_COUNT] = {
"cross_entropy_loss_back(x,y)",
};
static_assert(GGML_OP_COUNT == 74, "GGML_OP_COUNT != 74");
static_assert(GGML_OP_COUNT == 76, "GGML_OP_COUNT != 76");
static_assert(GGML_OP_POOL_COUNT == 2, "GGML_OP_POOL_COUNT != 2");
@ -2154,7 +2158,10 @@ void ggml_numa_init(enum ggml_numa_strategy numa_flag) {
getcpu_ret = getcpu(&current_cpu, &g_state.numa.current_node);
#else
// old glibc doesn't have a wrapper for this call. Fall back on direct syscall
getcpu_ret = syscall(SYS_getcpu,&current_cpu,&g_state.numa.current_node);
# if !defined(SYS_getcpu) && defined(SYS_get_cpu)
# define SYS_getcpu SYS_get_cpu // some older glibc versions use this name
# endif
getcpu_ret = syscall(SYS_getcpu, &current_cpu, &g_state.numa.current_node);
#endif
if (g_state.numa.n_nodes < 1 || g_state.numa.total_cpus < 1 || getcpu_ret != 0) {
@ -6151,6 +6158,108 @@ struct ggml_tensor * ggml_flash_attn_back(
return result;
}
// ggml_ssm_conv
struct ggml_tensor * ggml_ssm_conv(
struct ggml_context * ctx,
struct ggml_tensor * s,
struct ggml_tensor * x,
struct ggml_tensor * c,
struct ggml_tensor * sq) {
GGML_ASSERT(ggml_is_3d(s));
GGML_ASSERT(ggml_is_matrix(x));
GGML_ASSERT(ggml_is_matrix(c));
GGML_ASSERT(ggml_is_matrix(sq));
GGML_ASSERT(sq->type == GGML_TYPE_I32);
const int64_t d_conv = c->ne[0];
const int64_t d_inner = c->ne[1];
const int64_t n_tokens = x->ne[1];
const int64_t n_kv = s->ne[2];
GGML_ASSERT( s->ne[0] == d_conv - 1);
GGML_ASSERT( s->ne[1] == d_inner);
GGML_ASSERT( x->ne[0] == d_inner);
GGML_ASSERT(sq->ne[0] == n_kv);
GGML_ASSERT(sq->ne[1] == n_tokens);
bool is_node = false;
if (s->grad || x->grad || c->grad || sq->grad) {
GGML_ASSERT(false); // TODO: implement
is_node = true;
}
// 2-in-1 concatenated x and conv_states, {d_inner, n_tokens} with {d_conv, d_inner, n_kv}
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, (d_inner*n_tokens) + (d_conv*d_inner*n_kv));
result->op = GGML_OP_SSM_CONV;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
result->src[0] = s;
result->src[1] = x;
result->src[2] = c;
result->src[3] = sq;
return result;
}
// ggml_ssm_scan
struct ggml_tensor * ggml_ssm_scan(
struct ggml_context * ctx,
struct ggml_tensor * s,
struct ggml_tensor * x,
struct ggml_tensor * dt,
struct ggml_tensor * A,
struct ggml_tensor * B,
struct ggml_tensor * C,
struct ggml_tensor * sq) {
GGML_ASSERT(ggml_is_contiguous(s));
GGML_ASSERT(ggml_is_contiguous(x));
GGML_ASSERT(ggml_is_contiguous(dt));
GGML_ASSERT(ggml_is_contiguous(A));
GGML_ASSERT(sq->type == GGML_TYPE_I32);
GGML_ASSERT(B->nb[0] == ggml_type_size(B->type));
GGML_ASSERT(C->nb[0] == ggml_type_size(C->type));
GGML_ASSERT(ggml_are_same_shape(x, dt));
{
const int64_t d_state = s->ne[0];
const int64_t d_inner = s->ne[1];
const int64_t n_tokens = x->ne[1];
GGML_ASSERT(x->ne[0] == d_inner);
GGML_ASSERT(A->ne[0] == d_state);
GGML_ASSERT(A->ne[1] == d_inner);
GGML_ASSERT(B->ne[0] == d_state);
GGML_ASSERT(B->ne[1] == n_tokens);
GGML_ASSERT(C->ne[0] == d_state);
GGML_ASSERT(C->ne[1] == n_tokens);
}
bool is_node = false;
if (s->grad || x->grad || dt->grad || A->grad || B->grad || C->grad || sq->grad) {
GGML_ASSERT(false); // TODO: implement
is_node = true;
}
// 2-in-1 concatenated y and ssm_states, {d_inner, n_tokens} with {d_state, d_inner, n_kv}
struct ggml_tensor * result = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, ggml_nelements(x) + ggml_nelements(s));
result->op = GGML_OP_SSM_SCAN;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
result->src[0] = s;
result->src[1] = x;
result->src[2] = dt;
result->src[3] = A;
result->src[4] = B;
result->src[5] = C;
result->src[6] = sq;
return result;
}
// ggml_win_part
struct ggml_tensor * ggml_win_part(
@ -14768,6 +14877,257 @@ static void ggml_compute_forward_flash_attn_back(
}
}
// ggml_compute_forward_ssm_conv
static void ggml_compute_forward_ssm_conv_f32(
const struct ggml_compute_params * params,
struct ggml_tensor * dst) {
if (params->type == GGML_TASK_TYPE_INIT || params->type == GGML_TASK_TYPE_FINALIZE) {
return;
}
const struct ggml_tensor * src0 = dst->src[0]; // conv_state
const struct ggml_tensor * src1 = dst->src[1]; // x
const struct ggml_tensor * src2 = dst->src[2]; // conv1d.weight
const struct ggml_tensor * src3 = dst->src[3]; // state_seq
const int ith = params->ith;
const int nth = params->nth;
const int nc = src2->ne[0]; // d_conv
const int nr = src0->ne[1]; // d_inner
const int n_t = src1->ne[1]; // n_tokens
const int n_kv = src0->ne[2]; // max number of sequences in the batch
GGML_ASSERT((nr*n_t) + (nc*nr*n_kv) == ggml_nelements(dst));
GGML_ASSERT(src0->nb[0] == sizeof(float));
GGML_ASSERT(src1->nb[0] == sizeof(float));
GGML_ASSERT(src2->nb[0] == sizeof(float));
GGML_ASSERT(src3->nb[0] == sizeof(int32_t));
GGML_ASSERT(src0->nb[1] == src0->ne[0]*sizeof(float));
// for use with the destination state offset between sequences
GGML_ASSERT(src2->nb[2] == src2->ne[1]*src2->ne[0]*sizeof(float));
// rows per thread
const int dr = (nr + nth - 1)/nth;
// row range for this thread
const int ir0 = dr*ith;
const int ir1 = MIN(ir0 + dr, nr);
const int ir = ir1 - ir0;
if (n_kv > 1) {
// multiple sequences means it's hard to know when it's the first time a state is read,
// so copy them all over to the destination, just to be sure.
for (int i3 = 0; i3 < n_kv; ++i3) {
float * s0 = (float *) ((char *) src0->data + ir0*(src0->nb[1]) + i3*(src0->nb[2]));
float * s = (float *) ((char *) dst->data + ir0*(src2->nb[1]) + i3*(src2->nb[2]) + nr*n_t*sizeof(float));
// can't use memcpy because of d_conv vs d_conv - 1
for (int i1 = 0; i1 < ir; ++i1) {
for (int i0 = 0; i0 < nc - 1; ++i0) {
// copy s0 to last (d_conv - 1) columns of s
s[1 + i0 + i1*nc] = s0[i0 + i1*(nc - 1)];
}
}
}
}
for (int i2 = 0; i2 < n_t; ++i2) {
int32_t * sq = (int32_t *) ((char *) src3->data + i2*(src3->nb[1])); // {n_kv, n_tokens}
float * x = (float *) ((char *) dst->data + ir0*sizeof(float) + i2*(nr*sizeof(float))); // {d_inner, n_tokens}
float * s = (float *) ((char *) dst->data + ir0*(src2->nb[1]) + sq[0]*(src2->nb[2]) + nr*n_t*sizeof(float)); // {d_conv, d_inner, n_kv}
float * s0; // {d_conv - 1, d_inner, n_kv}
float * x0 = (float *) ((char *) src1->data + ir0*(src1->nb[0]) + i2*(src1->nb[1])); // {d_inner, n_tokens}
float * c = (float *) ((char *) src2->data + ir0*(src2->nb[1])); // {d_conv, d_inner}
int ne0s0;
GGML_ASSERT(0 <= sq[0] && sq[0] < n_kv);
// avoid needing to copy the state for the first token
if (i2 == 0) {
s0 = (float *) ((char *) src0->data + ir0*(src0->nb[1]) + sq[0]*(src0->nb[2])); // {d_conv - 1, d_inner, n_kv}
ne0s0 = src0->ne[0];
} else {
// the source is the last (d_conv - 1) columns of the destination
s0 = s + 1;
ne0s0 = nc;
}
// d_inner
for (int i1 = 0; i1 < ir; ++i1) {
// shift state left
for (int i0 = 0; i0 < nc - 1; ++i0) {
s[i0 + i1*nc] = s0[i0 + i1*ne0s0];
}
// insert x on the last column
s[(nc - 1) + i1*nc] = x0[i1];
}
// handle copies when there are multiple output states
for (int i3 = 1; i3 < n_kv; ++i3) {
int32_t seq = sq[i3];
if (0 <= seq && seq < n_kv) {
float * s1 = s + (seq - sq[0])*nc*nr;
memcpy(s1, s, nc*ir*sizeof(float));
} else {
// stop at negative or too big seq_ids
break;
}
}
// it seems a little faster when this is separate from the state shift
for (int i1 = 0; i1 < ir; ++i1) {
// rowwise dot product
float sumf = 0.0f;
for (int i0 = 0; i0 < nc; ++i0) {
int i = i0 + i1*nc;
sumf += s[i] * c[i];
}
x[i1] = sumf;
}
}
}
static void ggml_compute_forward_ssm_conv(
const struct ggml_compute_params * params,
struct ggml_tensor * dst) {
switch (dst->src[0]->type) {
case GGML_TYPE_F32:
{
ggml_compute_forward_ssm_conv_f32(params, dst);
} break;
default:
{
GGML_ASSERT(false);
} break;
}
}
// ggml_compute_forward_ssm_scan
static void ggml_compute_forward_ssm_scan_f32(
const struct ggml_compute_params * params,
struct ggml_tensor * dst) {
if (params->type == GGML_TASK_TYPE_INIT || params->type == GGML_TASK_TYPE_FINALIZE) {
return;
}
const struct ggml_tensor * src0 = dst->src[0]; // s
const struct ggml_tensor * src1 = dst->src[1]; // x
const struct ggml_tensor * src2 = dst->src[2]; // dt
const struct ggml_tensor * src3 = dst->src[3]; // A
const struct ggml_tensor * src4 = dst->src[4]; // B
const struct ggml_tensor * src5 = dst->src[5]; // C
const struct ggml_tensor * src6 = dst->src[6]; // sq
const int ith = params->ith;
const int nth = params->nth;
const int64_t nc = src0->ne[0]; // d_state
const int64_t nr = src0->ne[1]; // d_inner
const int64_t n_t = src1->ne[1]; // number of tokens in the batch
const int64_t n_kv = src0->ne[2]; // max number of sequences in the batch
GGML_ASSERT(ggml_nelements(src1) + ggml_nelements(src0) == ggml_nelements(dst));
GGML_ASSERT(src0->nb[0] == sizeof(float));
GGML_ASSERT(src1->nb[0] == sizeof(float));
GGML_ASSERT(src2->nb[0] == sizeof(float));
GGML_ASSERT(src3->nb[0] == sizeof(float));
GGML_ASSERT(src4->nb[0] == sizeof(float));
GGML_ASSERT(src5->nb[0] == sizeof(float));
// required for the dot product between s and C, and when copying the states
GGML_ASSERT(src0->nb[1] == src0->ne[0]*sizeof(float));
// required for per-sequence offsets for states
GGML_ASSERT(src0->nb[2] == src0->ne[0]*src0->ne[1]*sizeof(float));
// required to get correct offset for state destination (i.e. src1->nb[2])
GGML_ASSERT(src1->nb[2] == src1->ne[0]*src1->ne[1]*sizeof(float));
// rows per thread
const int dr = (nr + nth - 1)/nth;
// row range for this thread
const int ir0 = dr*ith;
const int ir1 = MIN(ir0 + dr, nr);
const int ir = ir1 - ir0;
if (n_kv > 1) {
// it's hard to know if the source states have already been copied
// when there are multiple, so copy them already.
for (int i3 = 0; i3 < n_kv; ++i3) {
float * s0 = (float *) ((char *) src0->data + ir0*(src0->nb[1]) + i3*(src0->nb[2]));
float * s = (float *) ((char *) dst->data + ir0*(src0->nb[1]) + i3*(src0->nb[2]) + src1->nb[2]);
memcpy(s, s0, nc*ir*sizeof(float));
}
}
for (int i2 = 0; i2 < n_t; ++i2) {
int32_t * sq = (int32_t *) ((char *) src6->data + i2*(src6->nb[1])); // {n_kv, n_tokens}
float * y = (float *) ((char *) dst->data + ir0*(src1->nb[0]) + i2*(src1->nb[1])); // {d_inner, n_tokens}
float * s = (float *) ((char *) dst->data + ir0*(src0->nb[1]) + sq[0]*(src0->nb[2]) + src1->nb[2]); // {d_state, d_inner, n_kv}
float * s0;
float * x = (float *) ((char *) src1->data + ir0*(src1->nb[0]) + i2*(src1->nb[1])); // {d_inner, n_tokens}
float * dt = (float *) ((char *) src2->data + ir0*(src2->nb[0]) + i2*(src2->nb[1])); // {d_inner, n_tokens}
float * A = (float *) ((char *) src3->data + ir0*(src3->nb[1])); // {d_state, d_inner}
float * B = (float *) ((char *) src4->data + i2*(src4->nb[1])); // {d_state, n_tokens}
float * C = (float *) ((char *) src5->data + i2*(src5->nb[1])); // {d_state, n_tokens}
GGML_ASSERT(0 <= sq[0] && sq[0] < n_kv);
// avoid needing to copy the state for the first token
if (i2 == 0) {
s0 = (float *) ((char *) src0->data + ir0*(src0->nb[1]) + sq[0]*(src0->nb[2])); // {d_state, d_inner, n_kv}
} else {
// otherwise the source is the same as the destination
s0 = s;
}
// d_inner
for (int i1 = 0; i1 < ir; ++i1) {
// ref: https://github.com/state-spaces/mamba/blob/34076d664838588a3c97727b263478ab9f621a07/mamba_ssm/ops/triton/selective_state_update.py#L78
float dt_soft_plus = dt[i1] <= 20.0f ? log1pf(expf(dt[i1])) : dt[i1];
float x_dt = x[i1] * dt_soft_plus;
float sumf = 0.0f;
// d_state
for (int i0 = 0; i0 < nc; ++i0) {
int i = i0 + i1*nc;
// state = prev_state * dA + dB * x
float state = (s0[i] * expf(dt_soft_plus * A[i])) + (B[i0] * x_dt);
// y = rowwise_dotprod(state, C)
sumf += state * C[i0];
s[i] = state;
}
y[i1] = sumf;
}
// handle copies when there are multiple output states
for (int i3 = 1; i3 < n_kv; ++i3) {
int32_t seq = sq[i3];
if (0 <= seq && seq < n_kv) {
float * s1 = s + (seq - sq[0])*nc*nr;
memcpy(s1, s, nc*ir*sizeof(float));
} else {
// stop at negative or too big seq_ids
break;
}
}
}
}
static void ggml_compute_forward_ssm_scan(
const struct ggml_compute_params * params,
struct ggml_tensor * dst) {
switch (dst->src[0]->type) {
case GGML_TYPE_F32:
{
ggml_compute_forward_ssm_scan_f32(params, dst);
} break;
default:
{
GGML_ASSERT(false);
} break;
}
}
// ggml_compute_forward_win_part
static void ggml_compute_forward_win_part_f32(
@ -15827,6 +16187,14 @@ static void ggml_compute_forward(struct ggml_compute_params * params, struct ggm
bool masked = t != 0;
ggml_compute_forward_flash_attn_back(params, masked, tensor);
} break;
case GGML_OP_SSM_CONV:
{
ggml_compute_forward_ssm_conv(params, tensor);
} break;
case GGML_OP_SSM_SCAN:
{
ggml_compute_forward_ssm_scan(params, tensor);
} break;
case GGML_OP_WIN_PART:
{
ggml_compute_forward_win_part(params, tensor);
@ -16881,6 +17249,11 @@ static void ggml_compute_backward(struct ggml_context * ctx, struct ggml_tensor
{
GGML_ASSERT(false); // not supported
} break;
case GGML_OP_SSM_CONV:
case GGML_OP_SSM_SCAN:
{
GGML_ASSERT(false); // TODO: not implemented
} break;
case GGML_OP_WIN_PART:
case GGML_OP_WIN_UNPART:
case GGML_OP_UNARY:
@ -17587,6 +17960,11 @@ static int ggml_get_n_tasks(struct ggml_tensor * node, int n_threads) {
{
n_tasks = n_threads;
} break;
case GGML_OP_SSM_CONV:
case GGML_OP_SSM_SCAN:
{
n_tasks = n_threads;
} break;
case GGML_OP_WIN_PART:
case GGML_OP_WIN_UNPART:
case GGML_OP_GET_REL_POS:
@ -19781,133 +20159,6 @@ void ggml_quantize_free(void) {
ggml_critical_section_end();
}
size_t ggml_quantize_q4_0(const float * src, void * dst, int n, int k, int64_t * hist) {
assert(k % QK4_0 == 0);
const int nb = k / QK4_0;
for (int b = 0; b < n; b += k) {
block_q4_0 * restrict y = (block_q4_0 *) dst + b/QK4_0;
quantize_row_q4_0_reference(src + b, y, k);
for (int i = 0; i < nb; i++) {
for (int j = 0; j < QK4_0; j += 2) {
const uint8_t vi0 = y[i].qs[j/2] & 0x0F;
const uint8_t vi1 = y[i].qs[j/2] >> 4;
hist[vi0]++;
hist[vi1]++;
}
}
}
return (n/QK4_0*sizeof(block_q4_0));
}
size_t ggml_quantize_q4_1(const float * src, void * dst, int n, int k, int64_t * hist) {
assert(k % QK4_1 == 0);
const int nb = k / QK4_1;
for (int b = 0; b < n; b += k) {
block_q4_1 * restrict y = (block_q4_1 *) dst + b/QK4_1;
quantize_row_q4_1_reference(src + b, y, k);
for (int i = 0; i < nb; i++) {
for (int j = 0; j < QK4_1; j += 2) {
const uint8_t vi0 = y[i].qs[j/2] & 0x0F;
const uint8_t vi1 = y[i].qs[j/2] >> 4;
hist[vi0]++;
hist[vi1]++;
}
}
}
return (n/QK4_1*sizeof(block_q4_1));
}
size_t ggml_quantize_q5_0(const float * src, void * dst, int n, int k, int64_t * hist) {
assert(k % QK5_0 == 0);
const int nb = k / QK5_0;
for (int b = 0; b < n; b += k) {
block_q5_0 * restrict y = (block_q5_0 *)dst + b/QK5_0;
quantize_row_q5_0_reference(src + b, y, k);
for (int i = 0; i < nb; i++) {
uint32_t qh;
memcpy(&qh, &y[i].qh, sizeof(qh));
for (int j = 0; j < QK5_0; j += 2) {
const uint8_t vh0 = ((qh & (1u << (j/2 + 0 ))) >> (j/2 + 0 )) << 4;
const uint8_t vh1 = ((qh & (1u << (j/2 + 16))) >> (j/2 + 12));
// cast to 16 bins
const uint8_t vi0 = ((y[i].qs[j/2] & 0x0F) | vh0) / 2;
const uint8_t vi1 = ((y[i].qs[j/2] >> 4) | vh1) / 2;
hist[vi0]++;
hist[vi1]++;
}
}
}
return (n/QK5_0*sizeof(block_q5_0));
}
size_t ggml_quantize_q5_1(const float * src, void * dst, int n, int k, int64_t * hist) {
assert(k % QK5_1 == 0);
const int nb = k / QK5_1;
for (int b = 0; b < n; b += k) {
block_q5_1 * restrict y = (block_q5_1 *)dst + b/QK5_1;
quantize_row_q5_1_reference(src + b, y, k);
for (int i = 0; i < nb; i++) {
uint32_t qh;
memcpy(&qh, &y[i].qh, sizeof(qh));
for (int j = 0; j < QK5_1; j += 2) {
const uint8_t vh0 = ((qh & (1u << (j/2 + 0 ))) >> (j/2 + 0 )) << 4;
const uint8_t vh1 = ((qh & (1u << (j/2 + 16))) >> (j/2 + 12));
// cast to 16 bins
const uint8_t vi0 = ((y[i].qs[j/2] & 0x0F) | vh0) / 2;
const uint8_t vi1 = ((y[i].qs[j/2] >> 4) | vh1) / 2;
hist[vi0]++;
hist[vi1]++;
}
}
}
return (n/QK5_1*sizeof(block_q5_1));
}
size_t ggml_quantize_q8_0(const float * src, void * dst, int n, int k, int64_t * hist) {
assert(k % QK8_0 == 0);
const int nb = k / QK8_0;
for (int b = 0; b < n; b += k) {
block_q8_0 * restrict y = (block_q8_0 *)dst + b/QK8_0;
quantize_row_q8_0_reference(src + b, y, k);
for (int i = 0; i < nb; i++) {
for (int j = 0; j < QK8_0; ++j) {
const int8_t vi = y[i].qs[j];
hist[vi/16 + 8]++;
}
}
}
return (n/QK8_0*sizeof(block_q8_0));
}
bool ggml_quantize_requires_imatrix(enum ggml_type type) {
return
type == GGML_TYPE_IQ2_XXS ||
@ -19915,177 +20166,52 @@ bool ggml_quantize_requires_imatrix(enum ggml_type type) {
type == GGML_TYPE_IQ1_S;
}
size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst, int start,
int nrows, int n_per_row, int64_t * hist, const float * imatrix) {
size_t ggml_quantize_chunk(
enum ggml_type type,
const float * src,
void * dst,
int start,
int nrows,
int n_per_row,
const float * imatrix) {
const int n = nrows * n_per_row;
if (ggml_quantize_requires_imatrix(type)) {
GGML_ASSERT(imatrix != NULL);
}
GGML_ASSERT(start % type_traits[type].blck_size == 0);
GGML_ASSERT(start % n_per_row == 0);
ggml_quantize_init(type); // this is noop if already initialized
const size_t start_row = start / n_per_row;
const size_t row_size = ggml_row_size(type, n_per_row);
size_t result = 0;
int n = nrows * n_per_row;
switch (type) {
case GGML_TYPE_Q4_0:
{
GGML_ASSERT(start % QK4_0 == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q4_0(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q4_1:
{
GGML_ASSERT(start % QK4_1 == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q4_1(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q5_0:
{
GGML_ASSERT(start % QK5_0 == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q5_0(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q5_1:
{
GGML_ASSERT(start % QK5_1 == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q5_1(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q8_0:
{
GGML_ASSERT(start % QK8_0 == 0);
block_q8_0 * block = (block_q8_0*)dst + start / QK8_0;
result = ggml_quantize_q8_0(src + start, block, n, n, hist);
} break;
case GGML_TYPE_Q2_K:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q2_K(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q3_K:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q3_K(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q4_K:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q4_K(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q5_K:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q5_K(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_Q6_K:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_q6_K(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ2_XXS:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
GGML_ASSERT(imatrix);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq2_xxs(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ2_XS:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
GGML_ASSERT(imatrix);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq2_xs(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ3_XXS:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq3_xxs(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ3_S:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq3_s(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ2_S:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq2_s(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ1_S:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq1_s(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ4_NL:
case GGML_TYPE_Q4_0: result = quantize_q4_0(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q4_1: result = quantize_q4_1(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q5_0: result = quantize_q5_0(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q5_1: result = quantize_q5_1(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q8_0: result = quantize_q8_0(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q2_K: result = quantize_q2_K(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q3_K: result = quantize_q3_K(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q4_K: result = quantize_q4_K(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q5_K: result = quantize_q5_K(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_Q6_K: result = quantize_q6_K(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ2_XXS: result = quantize_iq2_xxs(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ2_XS: result = quantize_iq2_xs (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ3_XXS: result = quantize_iq3_xxs(src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ3_S: result = quantize_iq3_s (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ2_S: result = quantize_iq2_s (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ1_S: result = quantize_iq1_s (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
case GGML_TYPE_IQ4_NL: result = quantize_iq4_nl (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
#if QK_K == 64
case GGML_TYPE_IQ4_XS:
#endif
{
GGML_ASSERT(start % QK4_NL == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq4_nl(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
#if QK_K != 64
case GGML_TYPE_IQ4_XS:
{
GGML_ASSERT(start % QK_K == 0);
GGML_ASSERT(start % n_per_row == 0);
size_t start_row = start / n_per_row;
size_t row_size = ggml_row_size(type, n_per_row);
result = quantize_iq4_xs(src + start, (char *)dst + start_row * row_size, nrows, n_per_row, hist, imatrix);
GGML_ASSERT(result == row_size * nrows);
} break;
case GGML_TYPE_IQ4_XS: result = quantize_iq4_nl (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
#else
case GGML_TYPE_IQ4_XS: result = quantize_iq4_xs (src + start, (char *) dst + start_row * row_size, nrows, n_per_row, imatrix); break;
#endif
case GGML_TYPE_F16:
{
@ -20102,6 +20228,9 @@ size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst, i
default:
assert(false);
}
GGML_ASSERT(result == nrows * row_size);
return result;
}

42
ggml.h
View File

@ -472,6 +472,8 @@ extern "C" {
GGML_OP_FLASH_ATTN,
GGML_OP_FLASH_FF,
GGML_OP_FLASH_ATTN_BACK,
GGML_OP_SSM_CONV,
GGML_OP_SSM_SCAN,
GGML_OP_WIN_PART,
GGML_OP_WIN_UNPART,
GGML_OP_GET_REL_POS,
@ -1728,6 +1730,23 @@ extern "C" {
struct ggml_tensor * c0,
struct ggml_tensor * c1);
GGML_API struct ggml_tensor * ggml_ssm_conv(
struct ggml_context * ctx,
struct ggml_tensor * s,
struct ggml_tensor * x,
struct ggml_tensor * c,
struct ggml_tensor * sq);
GGML_API struct ggml_tensor * ggml_ssm_scan(
struct ggml_context * ctx,
struct ggml_tensor * s,
struct ggml_tensor * x,
struct ggml_tensor * dt,
struct ggml_tensor * A,
struct ggml_tensor * B,
struct ggml_tensor * C,
struct ggml_tensor * sq);
// partition into non-overlapping windows with padding if needed
// example:
// a: 768 64 64 1
@ -2175,25 +2194,18 @@ extern "C" {
GGML_API void ggml_quantize_init(enum ggml_type type);
GGML_API void ggml_quantize_free(void);
// 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);
// some quantization type cannot be used without an importance matrix
GGML_API bool ggml_quantize_requires_imatrix(enum ggml_type type);
// calls ggml_quantize_init internally (i.e. can allocate memory)
GGML_API size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst,
int start, int nrows, int n_per_row, int64_t * hist, const float * imatrix);
GGML_API size_t ggml_quantize_chunk(
enum ggml_type type,
const float * src,
void * dst,
int start,
int nrows,
int n_per_row,
const float * imatrix);
//
// gguf

1207
ggml_vk_generate_shaders.py
View File

File diff suppressed because it is too large Load Diff

41
gguf-py/gguf/constants.py
View File

@ -61,6 +61,12 @@ class Keys:
SCALING_ORIG_CTX_LEN = "{arch}.rope.scaling.original_context_length"
SCALING_FINETUNED = "{arch}.rope.scaling.finetuned"
class SSM:
CONV_KERNEL = "{arch}.ssm.conv_kernel"
INNER_SIZE = "{arch}.ssm.inner_size"
STATE_SIZE = "{arch}.ssm.state_size"
TIME_STEP_RANK = "{arch}.ssm.time_step_rank"
class Tokenizer:
MODEL = "tokenizer.ggml.model"
LIST = "tokenizer.ggml.tokens"
@ -113,6 +119,7 @@ class MODEL_ARCH(IntEnum):
MINICPM = auto()
GEMMA = auto()
STARCODER2 = auto()
MAMBA = auto()
class MODEL_TENSOR(IntEnum):
@ -144,6 +151,13 @@ class MODEL_TENSOR(IntEnum):
ATTN_Q_NORM = auto()
ATTN_K_NORM = auto()
LAYER_OUT_NORM = auto()
SSM_IN = auto()
SSM_CONV1D = auto()
SSM_X = auto()
SSM_DT = auto()
SSM_A = auto()
SSM_D = auto()
SSM_OUT = auto()
MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
@ -171,6 +185,7 @@ MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
MODEL_ARCH.MINICPM: "minicpm",
MODEL_ARCH.GEMMA: "gemma",
MODEL_ARCH.STARCODER2: "starcoder2",
MODEL_ARCH.MAMBA: "mamba",
}
TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
@ -202,6 +217,13 @@ TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
MODEL_TENSOR.FFN_DOWN_EXP: "blk.{bid}.ffn_down.{xid}",
MODEL_TENSOR.FFN_UP_EXP: "blk.{bid}.ffn_up.{xid}",
MODEL_TENSOR.LAYER_OUT_NORM: "blk.{bid}.layer_output_norm",
MODEL_TENSOR.SSM_IN: "blk.{bid}.ssm_in",
MODEL_TENSOR.SSM_CONV1D: "blk.{bid}.ssm_conv1d",
MODEL_TENSOR.SSM_X: "blk.{bid}.ssm_x",
MODEL_TENSOR.SSM_DT: "blk.{bid}.ssm_dt",
MODEL_TENSOR.SSM_A: "blk.{bid}.ssm_a",
MODEL_TENSOR.SSM_D: "blk.{bid}.ssm_d",
MODEL_TENSOR.SSM_OUT: "blk.{bid}.ssm_out",
}
MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
@ -543,6 +565,19 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
MODEL_TENSOR.FFN_DOWN,
MODEL_TENSOR.FFN_UP,
],
MODEL_ARCH.MAMBA: [
MODEL_TENSOR.TOKEN_EMBD,
MODEL_TENSOR.OUTPUT_NORM,
MODEL_TENSOR.OUTPUT,
MODEL_TENSOR.ATTN_NORM,
MODEL_TENSOR.SSM_IN,
MODEL_TENSOR.SSM_CONV1D,
MODEL_TENSOR.SSM_X,
MODEL_TENSOR.SSM_DT,
MODEL_TENSOR.SSM_A,
MODEL_TENSOR.SSM_D,
MODEL_TENSOR.SSM_OUT,
],
# TODO
}
@ -734,6 +769,12 @@ KEY_ROPE_SCALING_FACTOR = Keys.Rope.SCALING_FACTOR
KEY_ROPE_SCALING_ORIG_CTX_LEN = Keys.Rope.SCALING_ORIG_CTX_LEN
KEY_ROPE_SCALING_FINETUNED = Keys.Rope.SCALING_FINETUNED
# SSM
KEY_SSM_CONV_KERNEL = Keys.SSM.CONV_KERNEL
KEY_SSM_INNER_SIZE = Keys.SSM.INNER_SIZE
KEY_SSM_STATE_SIZE = Keys.SSM.STATE_SIZE
KEY_SSM_TIME_STEP_RANK = Keys.SSM.TIME_STEP_RANK
# tokenization
KEY_TOKENIZER_MODEL = Keys.Tokenizer.MODEL
KEY_TOKENIZER_LIST = Keys.Tokenizer.LIST

12
gguf-py/gguf/gguf_writer.py
View File

@ -382,6 +382,18 @@ class GGUFWriter:
def add_rope_scaling_finetuned(self, value: bool) -> None:
self.add_bool(Keys.Rope.SCALING_FINETUNED.format(arch=self.arch), value)
def add_ssm_conv_kernel(self, value: int) -> None:
self.add_uint32(Keys.SSM.CONV_KERNEL.format(arch=self.arch), value)
def add_ssm_inner_size(self, value: int) -> None:
self.add_uint32(Keys.SSM.INNER_SIZE.format(arch=self.arch), value)
def add_ssm_state_size(self, value: int) -> None:
self.add_uint32(Keys.SSM.STATE_SIZE.format(arch=self.arch), value)
def add_ssm_time_step_rank(self, value: int) -> None:
self.add_uint32(Keys.SSM.TIME_STEP_RANK.format(arch=self.arch), value)
def add_tokenizer_model(self, model: str) -> None:
self.add_string(Keys.Tokenizer.MODEL, model)

46
gguf-py/gguf/tensor_mapping.py
View File

@ -20,6 +20,9 @@ class TensorNameMap:
"wte", # gpt2
"transformer.embd.wte", # phi2
"model.tok_embeddings", # internlm2
"model.embedding", # mamba-qbert
"backbone.embedding", # mamba
"backbone.embeddings", # mamba-hf
),
# Token type embeddings
@ -44,7 +47,7 @@ class TensorNameMap:
# Output
MODEL_TENSOR.OUTPUT: (
"embed_out", # gptneox
"lm_head", # gpt2 mpt falcon llama-hf baichuan qwen
"lm_head", # gpt2 mpt falcon llama-hf baichuan qwen mamba
"output", # llama-pth bloom internlm2
"word_embeddings_for_head", # persimmon
"lm_head.linear", # phi2
@ -61,6 +64,8 @@ class TensorNameMap:
"language_model.encoder.final_layernorm", # persimmon
"model.final_layernorm", # persimmon
"lm_head.ln", # phi2
"model.norm_f", # mamba-qbert
"backbone.norm_f", # mamba
),
# Rope frequencies
@ -86,6 +91,8 @@ class TensorNameMap:
"transformer.h.{bid}.ln", # phi2
"model.layers.layers.{bid}.norm", # plamo
"model.layers.{bid}.attention_norm", # internlm2
"model.layers.{bid}.norm", # mamba-qbert
"backbone.layers.{bid}.norm", # mamba
),
# Attention norm 2
@ -282,7 +289,42 @@ class TensorNameMap:
MODEL_TENSOR.LAYER_OUT_NORM: (
"encoder.layer.{bid}.output.LayerNorm", # bert
"encoder.layers.{bid}.norm2", # nomic-bert
)
),
MODEL_TENSOR.SSM_IN: (
"model.layers.{bid}.in_proj",
"backbone.layers.{bid}.mixer.in_proj",
),
MODEL_TENSOR.SSM_CONV1D: (
"model.layers.{bid}.conv1d",
"backbone.layers.{bid}.mixer.conv1d",
),
MODEL_TENSOR.SSM_X: (
"model.layers.{bid}.x_proj",
"backbone.layers.{bid}.mixer.x_proj",
),
MODEL_TENSOR.SSM_DT: (
"model.layers.{bid}.dt_proj",
"backbone.layers.{bid}.mixer.dt_proj",
),
MODEL_TENSOR.SSM_A: (
"model.layers.{bid}.A_log",
"backbone.layers.{bid}.mixer.A_log",
),
MODEL_TENSOR.SSM_D: (
"model.layers.{bid}.D",
"backbone.layers.{bid}.mixer.D",
),
MODEL_TENSOR.SSM_OUT: (
"model.layers.{bid}.out_proj",
"backbone.layers.{bid}.mixer.out_proj",
),
}
mapping: dict[str, tuple[MODEL_TENSOR, str]]

2
grammars/json.gbnf
View File

@ -15,7 +15,7 @@ array ::=
string ::=
"\"" (
[^"\\] |
[^"\\\x7F\x00-\x1F] |
"\\" (["\\/bfnrt] | "u" [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F]) # escapes
)* "\"" ws

2
grammars/json_arr.gbnf
View File

@ -24,7 +24,7 @@ array ::=
string ::=
"\"" (
[^"\\] |
[^"\\\x7F\x00-\x1F] |
"\\" (["\\/bfnrt] | "u" [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F]) # escapes
)* "\"" ws

810
llama.cpp
View File

File diff suppressed because it is too large Load Diff

8
llama.h
View File

@ -235,6 +235,7 @@ extern "C" {
uint32_t seed; // RNG seed, -1 for random
uint32_t n_ctx; // text context, 0 = from model
uint32_t n_batch; // prompt processing maximum batch size
uint32_t n_parallel; // number of parallel sequences (i.e. distinct states for recurrent models)
uint32_t n_threads; // number of threads to use for generation
uint32_t n_threads_batch; // number of threads to use for batch processing
@ -376,6 +377,7 @@ extern "C" {
LLAMA_API uint32_t llama_n_ctx (const struct llama_context * ctx);
LLAMA_API uint32_t llama_n_batch (const struct llama_context * ctx);
LLAMA_API uint32_t llama_n_max_seq (const struct llama_context * ctx);
LLAMA_API enum llama_vocab_type llama_vocab_type(const struct llama_model * model);
LLAMA_API enum llama_rope_type llama_rope_type (const struct llama_model * model);
@ -502,7 +504,7 @@ extern "C" {
// seq_id < 0 : match any sequence
// p0 < 0 : [0, p1]
// p1 < 0 : [p0, inf)
LLAMA_API void llama_kv_cache_seq_rm(
LLAMA_API bool llama_kv_cache_seq_rm(
struct llama_context * ctx,
llama_seq_id seq_id,
llama_pos p0,
@ -641,6 +643,10 @@ extern "C" {
// n_threads_batch is the number of threads used for prompt and batch processing (multiple tokens)
LLAMA_API void llama_set_n_threads(struct llama_context * ctx, uint32_t n_threads, uint32_t n_threads_batch);
// Set whether to use causal attention or not
// If set to true, the model will only attend to the past tokens
LLAMA_API void llama_set_causal_attn(struct llama_context * ctx, bool causal_attn);
// Set abort callback
LLAMA_API void llama_set_abort_callback(struct llama_context * ctx, ggml_abort_callback abort_callback, void * abort_callback_data);

2
scripts/compare-llama-bench.py
View File

@ -18,7 +18,7 @@ except ImportError as e:
KEY_PROPERTIES = [
"cpu_info", "gpu_info", "n_gpu_layers", "main_gpu", "cuda", "opencl", "metal", "gpu_blas",
"blas", "model_filename", "model_type", "model_size", "model_n_params", "n_batch", "n_threads",
"type_k", "type_v", "no_kv_offload", "mul_mat_q", "tensor_split", "n_prompt", "n_gen"
"type_k", "type_v", "no_kv_offload", "tensor_split", "n_prompt", "n_gen"
]
# Properties that are boolean and are converted to Yes/No for the table:

2
scripts/sync-ggml-am.sh
View File

@ -94,6 +94,7 @@ if [ -f $SRC_LLAMA/ggml-src.patch ]; then
# src/ggml-alloc.c -> ggml-alloc.c
# src/ggml-backend-impl.h -> ggml-backend-impl.h
# src/ggml-backend.c -> ggml-backend.c
# src/ggml-common.h -> ggml-common.h
# src/ggml-cuda.cu -> ggml-cuda.cu
# src/ggml-cuda.h -> ggml-cuda.h
# src/ggml-impl.h -> ggml-impl.h
@ -126,6 +127,7 @@ if [ -f $SRC_LLAMA/ggml-src.patch ]; then
-e 's/src\/ggml-alloc\.c/ggml-alloc.c/g' \
-e 's/src\/ggml-backend-impl\.h/ggml-backend-impl.h/g' \
-e 's/src\/ggml-backend\.c/ggml-backend.c/g' \
-e 's/src\/ggml-common\.h/ggml-common.h/g' \
-e 's/src\/ggml-cuda\.cu/ggml-cuda.cu/g' \
-e 's/src\/ggml-cuda\.h/ggml-cuda.h/g' \
-e 's/src\/ggml-impl\.h/ggml-impl.h/g' \

2
scripts/sync-ggml.last
View File

@ -1 +1 @@
8695910a39102609073d0e099aa7c97d6bcb3bf9
43a6d4af1971ee2912ff7bc2404011ff327b6a60

1
scripts/sync-ggml.sh
View File

@ -4,6 +4,7 @@ cp -rpv ../ggml/src/ggml.c ./ggml.c
cp -rpv ../ggml/src/ggml-alloc.c ./ggml-alloc.c
cp -rpv ../ggml/src/ggml-backend-impl.h ./ggml-backend-impl.h
cp -rpv ../ggml/src/ggml-backend.c ./ggml-backend.c
cp -rpv ../ggml/src/ggml-common.h ./ggml-common.h
cp -rpv ../ggml/src/ggml-cuda.cu ./ggml-cuda.cu
cp -rpv ../ggml/src/ggml-cuda.h ./ggml-cuda.h
cp -rpv ../ggml/src/ggml-impl.h ./ggml-impl.h

1
tests/.gitignore vendored
View File

@ -1,3 +1,4 @@
*
!*.*
*.o
ggml-common.h

3
tests/test-backend-ops.cpp
View File

@ -53,7 +53,6 @@ static void init_tensor_uniform(ggml_tensor * tensor, float min = -1.0f, float m
} else if (ggml_is_quantized(tensor->type) || tensor->type == GGML_TYPE_F16) {
GGML_ASSERT(size % ggml_blck_size(tensor->type) == 0);
std::vector<uint8_t> dataq(ggml_row_size(tensor->type, size));
int64_t hist[16];
std::vector<float> imatrix(tensor->ne[0], 1.0f); // dummy importance matrix
const float * im = imatrix.data();
if (!ggml_quantize_requires_imatrix(tensor->type)) {
@ -63,7 +62,7 @@ static void init_tensor_uniform(ggml_tensor * tensor, float min = -1.0f, float m
im = nullptr;
}
}
ggml_quantize_chunk(tensor->type, data.data(), dataq.data(), 0, size/tensor->ne[0], tensor->ne[0], hist, im);
ggml_quantize_chunk(tensor->type, data.data(), dataq.data(), 0, size/tensor->ne[0], tensor->ne[0], im);
ggml_backend_tensor_set(tensor, dataq.data(), 0, dataq.size());
} else if (tensor->type == GGML_TYPE_I8 || tensor->type == GGML_TYPE_I16 || tensor->type == GGML_TYPE_I32) {
// This is going to create some weird integers though.