Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-13 02:10:39 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'upstream' into concedo_experimental

Browse Source 
# Conflicts:
#	docs/backend/SYCL.md
#	docs/backend/snapdragon/developer.md
#	examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp
#	examples/sycl/run-llama2.sh
#	examples/sycl/start-svr.sh
#	examples/sycl/test.sh
#	examples/sycl/win-run-llama2.bat
#	examples/sycl/win-start-svr.bat
#	examples/sycl/win-test.bat
#	ggml/CMakeLists.txt
#	ggml/src/ggml-et/et-kernels/src/ssm_scan_f32.c
#	ggml/src/ggml-et/ggml-et-ops.cpp
#	ggml/src/ggml-et/ggml-et-ops.h
#	ggml/src/ggml-sycl/ssm_scan.cpp
#	ggml/src/ggml-webgpu/ggml-webgpu.cpp
#	ggml/src/ggml-webgpu/wgsl-shaders/ssm_scan.wgsl
#	scripts/bench-models.sh
#	scripts/snapdragon/adb/run-bench.sh
#	scripts/snapdragon/adb/run-cli.sh
#	scripts/snapdragon/adb/run-completion.sh
#	scripts/snapdragon/adb/run-mtmd.sh
#	scripts/snapdragon/windows/run-bench.ps1
#	scripts/snapdragon/windows/run-cli.ps1
#	scripts/snapdragon/windows/run-completion.ps1
#	scripts/snapdragon/windows/run-mtmd.ps1
#	scripts/sync-ggml.last
#	scripts/sync_vendor.py
#	tests/CMakeLists.txt
#	tests/test-backend-ops.cpp
#	tests/test-chat.cpp
#	tests/test-jinja.cpp
#	tests/test-llama-archs.cpp
#	tools/cli/README.md
#	tools/completion/README.md
#	tools/llama-bench/README.md
#	tools/server/README.md
This commit is contained in:
Concedo
2026-08-15 22:50:04 +08:00
parent 531cac9cd3 adb55e5148
commit 82e15c08bf
60 changed files with 1781 additions and 447 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-model-loader.cpp
+18 -10
View File
@@ -316,15 +316,19 @@ namespace GGUFMeta {
struct GGUFMeta::ArrayInfo arr_info =
GGUFMeta::GKV<GGUFMeta::ArrayInfo>::get_kv(ctx, kid);
bool type_ok = false;
switch (arr_info.gt) {
case GGUF_TYPE_UINT32:
case GGUF_TYPE_INT32: GGML_ASSERT((std::is_same<T, int32_t>::value) ||
(std::is_same<T, uint32_t>::value)); break;
case GGUF_TYPE_FLOAT32: GGML_ASSERT((std::is_same<T, float>::value)); break;
case GGUF_TYPE_STRING: GGML_ASSERT((std::is_same<T, std::string>::value)); break;
case GGUF_TYPE_INT32: type_ok = (std::is_same<T, int32_t>::value) ||
(std::is_same<T, uint32_t>::value); break;
case GGUF_TYPE_FLOAT32: type_ok = (std::is_same<T, float>::value); break;
case GGUF_TYPE_STRING: type_ok = (std::is_same<T, std::string>::value); break;
default:
throw std::runtime_error(format("%s is not a string/float32/uint32/int32 array", key.c_str()));
}
if (!type_ok) {
throw std::runtime_error(format("%s has wrong array element type %s", key.c_str(), gguf_type_name(arr_info.gt)));
}
if constexpr (std::is_same<T, std::string>::value) {
const size_t n_items = gguf_get_arr_n(ctx, kid);
@@ -357,16 +361,20 @@ namespace GGUFMeta {
struct GGUFMeta::ArrayInfo arr_info =
GGUFMeta::GKV<GGUFMeta::ArrayInfo>::get_kv(ctx, kid);
bool type_ok = false;
switch (arr_info.gt) {
case GGUF_TYPE_BOOL:
case GGUF_TYPE_UINT32:
case GGUF_TYPE_INT32: GGML_ASSERT((std::is_same<T, int32_t>::value) ||
(std::is_same<T, uint32_t>::value)); break;
case GGUF_TYPE_FLOAT32: GGML_ASSERT((std::is_same<T, float>::value)); break;
case GGUF_TYPE_STRING: GGML_ASSERT((std::is_same<T, std::string>::value)); break;
case GGUF_TYPE_INT32: type_ok = (std::is_same<T, int32_t>::value) ||
(std::is_same<T, uint32_t>::value); break;
case GGUF_TYPE_FLOAT32: type_ok = (std::is_same<T, float>::value); break;
case GGUF_TYPE_STRING: type_ok = (std::is_same<T, std::string>::value); break;
default:
throw std::runtime_error(format("%s is not a string/float32/uint32/int32 array", key.c_str()));
}
if (!type_ok) {
throw std::runtime_error(format("%s has wrong array element type %s", key.c_str(), gguf_type_name(arr_info.gt)));
}
if (arr_info.length > N_MAX) {
throw std::runtime_error(format("array length %u for key %s exceeds max %u", (uint32_t) arr_info.length, key.c_str(), (uint32_t) N_MAX));
@@ -1003,7 +1011,7 @@ static bool weight_buft_supported(const llama_hparams & hparams, ggml_tensor * w
ggml_tensor * B = ggml_new_tensor_4d(ctx, GGML_TYPE_F32, d_state, n_group, n_seq_tokens, n_seqs);
ggml_tensor * C = ggml_new_tensor_4d(ctx, GGML_TYPE_F32, d_state, n_group, n_seq_tokens, n_seqs);
ggml_tensor * ids = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, n_seqs);
op_tensor = ggml_ssm_scan(ctx, s, x, dt, w, B, C, ids);
op_tensor = ggml_ssm_scan(ctx, s, x, dt, w, B, C, ids, /*K=*/1);
} break;
case GGML_OP_RWKV_WKV6:
{
@@ -1179,7 +1187,7 @@ struct ggml_tensor * llama_model_loader::create_tensor(
if (use_mmap) {
static std::once_flag once;
std::call_once(once, [] {
LLAMA_LOG_WARN("llama_model_loader: tensor overrides to CPU are used with mmap enabled - consider using --no-mmap for better performance\n");
LLAMA_LOG_WARN("llama_model_loader: tensor overrides to CPU are used with mmap enabled - consider using --load-mode none for better performance\n");
});
}
} else {