tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-07-31 14:44:34 -04:00
Code Issues Packages Projects Releases Wiki Activity

ubatch : new splitting logic (#14217)

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov
2025-06-20 10:14:14 +03:00
committed by GitHub
parent 9eaa51e7f0
commit 4c9fdfbe15
19 changed files with 992 additions and 915 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/llama-hparams.cpp
View File

@@ -90,6 +90,10 @@ bool llama_hparams::is_recurrent(uint32_t il) const {
return recurrent_layer_arr[il];
}
uint32_t llama_hparams::n_pos_per_embd() const {
return rope_type == LLAMA_ROPE_TYPE_MROPE ? 4 : 1;
}
bool llama_hparams::is_swa(uint32_t il) const {
if (il < n_layer) {
return swa_layers[il];