public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-10-18 13:17:59 +02:00
Code Issues Packages Projects Releases Wiki Activity

talk-llama : sync llama.cpp

Browse Source
This commit is contained in:
Georgi Gerganov
2026-06-08 14:36:36 +03:00
parent 4df9a57df2
commit 84bd03a438
151 changed files with 3439 additions and 871 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/talk-llama/llama-model-loader.cpp
+8 -7
View File
@@ -146,7 +146,7 @@ namespace GGUFMeta {
const enum gguf_type arr_type = gguf_get_arr_type(ctx, k);
return ArrayInfo {
arr_type,
size_t(gguf_get_arr_n(ctx, k)),
gguf_get_arr_n(ctx, k),
arr_type == GGUF_TYPE_STRING ? nullptr : gguf_get_arr_data(ctx, k),
};
}
@@ -393,6 +393,7 @@ namespace GGUFMeta {
}
template bool llama_model_loader::get_arr<std::vector<std::string>>(enum llm_kv kid, std::vector<std::string> & result, bool required);
template bool llama_model_loader::get_arr<std::array<int32_t, 512>>(enum llm_kv kid, std::array<int32_t, 512> & result, bool required);
template<typename T>
bool llama_model_loader::get_key(const std::string & key, T & result, bool required) {
@@ -445,7 +446,7 @@ namespace GGUFMeta {
}
if (n > N_MAX) {
throw std::runtime_error(format("n > N_MAX: %u > %u for key %s", (uint32_t) n, (uint32_t) N_MAX, key.c_str()));
throw std::runtime_error(format("n > N_MAX: %u > %u for key %s", n, (uint32_t) N_MAX, key.c_str()));
}
if (gguf_get_kv_type(metadata, kid) == GGUF_TYPE_ARRAY) {
@@ -502,9 +503,9 @@ namespace GGUFMeta {
}
// TODO: this is not very clever - figure out something better
template bool llama_model_loader::get_key_or_arr<std::array<int, 4>>(enum llm_kv kid, std::array<int, 4> & result, uint32_t n, bool required);
template bool llama_model_loader::get_key_or_arr<std::array<int, 4>> (enum llm_kv kid, std::array<int, 4> & result, uint32_t n, bool required);
template bool llama_model_loader::get_key_or_arr<std::array<uint32_t, 512>>(enum llm_kv kid, std::array<uint32_t, 512> & result, uint32_t n, bool required);
template bool llama_model_loader::get_key_or_arr<std::array<float, 512>>(enum llm_kv kid, std::array<float, 512> & result, uint32_t n, bool required);
template bool llama_model_loader::get_key_or_arr<std::array<float, 512>>(enum llm_kv kid, std::array<float, 512> & result, uint32_t n, bool required);
llama_model_loader::llama_model_loader(
@@ -1050,10 +1051,10 @@ struct ggml_tensor * llama_model_loader::create_tensor(
if (it == ctx_map.end()) {
// one ggml context per buffer type
int max_n_tensors = n_tensors;
max_n_tensors += 1; // duplicated output tensor
max_n_tensors += hparams.n_layer*2; // duplicated rope freq tensors
max_n_tensors += 1; // duplicated output tensor
max_n_tensors += hparams.n_layer()*2; // duplicated rope freq tensors
if (files.empty()) {
max_n_tensors += hparams.n_layer*256; // this should be well above what any model actually uses
max_n_tensors += hparams.n_layer()*256; // this should be well above what any model actually uses
}
const size_t ctx_size = ggml_tensor_overhead()*max_n_tensors;