nicolas
/
ik_llama_opt
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Synch DFlash Tokens ID (#2303) 

* llama: synchronize DFlash argmax readback

* llama: synchronize DFlash token access
This commit is contained in:
Samuel Oliveira Alves 2026-08-13 13:02:08 -03:00 committed by GitHub
parent 37d82c2313
commit cf711918e2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/llama.cpp
View File

@ -11729,6 +11729,8 @@ float * llama_get_logits_ith(struct llama_context * ctx, int32_t i) {
} }
llama_token llama_get_dflash_draft_token_ith(struct llama_context * ctx, int32_t i) { llama_token llama_get_dflash_draft_token_ith(struct llama_context * ctx, int32_t i) {
llama_synchronize(ctx);
if ((size_t) i >= ctx->dflash.draft_tokens.size()) { if ((size_t) i >= ctx->dflash.draft_tokens.size()) {
return LLAMA_TOKEN_NULL; return LLAMA_TOKEN_NULL;
} }