EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

llama: store mrope data in KV cell (#16825)

Browse Source 
* llama: store mrope data in KV cell

* correct x,y ordering

* address review comments

* add consistency checks

* Update src/llama-kv-cache.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* add TODO

* fix asan error

* kv-cells : improve ext handling

* cont : fix headers

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Xuan-Son Nguyen
2025-10-29 18:09:18 +01:00
committed by GitHub
parent 10fcc41290
commit e3af5563bd
6 changed files with 144 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
src/llama-batch.h
View File

@@ -17,6 +17,16 @@ struct llama_ubatch {
return b_equal_seqs != 0;
}
// typical for M-RoPE cases:
// 0 - sequantial position of the tokens/embeddings in the sequence
// 1 - y position in the image
// 2 - x position in the image
// 3 - other
bool is_pos_2d() const {
// TODO @ngxson : we may need to check for model arch when more models use >1 positions
return n_pos >= 3;
}
uint32_t b_equal_seqs; // note: this is a boolean, but we use an int32_t for alignment
// otherwise address sanitizer complains
// TODO: whole_seqs for embeddings?
@@ -25,6 +35,7 @@ struct llama_ubatch {
uint32_t n_seq_tokens; // tokens per sequence set
uint32_t n_seqs; // sequence sets in the ubatch
uint32_t n_seqs_unq; // unique sequence ids in the ubatch
uint32_t n_pos; // number of position inputs for each token/embedding
// seq_id_unq: unique sequence ids in the ubatch
// seq_idx: indices of the unique sequence ids in the ubatch in [0, n_seqs_unq)
@@ -33,7 +44,7 @@ struct llama_ubatch {
// // size | idx | val
llama_token * token; // [n_tokens] | i | id, token
float * embd; // [n_embd, n_tokens] | i | embd
llama_pos * pos; // [n_tokens] | i | pos
llama_pos * pos; // [n_tokens*n_pos] | i | pos
int32_t * n_seq_id; // [n_tokens] | i | -
llama_seq_id ** seq_id; // [n_tokens] | s | s0, s1, seq_id
llama_seq_id * seq_id_unq; // [n_seqs_unq] | s | seq_id