CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-04 09:32:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : fix crash when prompt exceeds context size (#3996)

Browse Source
This commit is contained in:
Alexey Parfenov
2023-11-11 05:48:21 +00:00
committed by GitHub
parent 34b0a08207
commit d96ca7ded7
1 changed files with 29 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
examples/server/server.cpp
View File

@@ -1557,15 +1557,6 @@ struct llama_server_context
slot.num_prompt_tokens = prompt_tokens.size(); slot.num_prompt_tokens = prompt_tokens.size();
if (!slot.params.cache_prompt)
{
llama_sampling_reset(slot.ctx_sampling);
slot.n_past = 0;
slot.num_prompt_tokens_processed = slot.num_prompt_tokens;
}
else
{
if (slot.params.n_keep < 0) if (slot.params.n_keep < 0)
{ {
slot.params.n_keep = slot.num_prompt_tokens; slot.params.n_keep = slot.num_prompt_tokens;
@@ -1595,6 +1586,15 @@ struct llama_server_context
GGML_ASSERT(slot.num_prompt_tokens < slot.n_ctx); GGML_ASSERT(slot.num_prompt_tokens < slot.n_ctx);
} }
if (!slot.params.cache_prompt)
{
llama_sampling_reset(slot.ctx_sampling);
slot.n_past = 0;
slot.num_prompt_tokens_processed = slot.num_prompt_tokens;
}
else
{
// push the prompt into the sampling context (do not apply grammar) // push the prompt into the sampling context (do not apply grammar)
for (auto &token : prompt_tokens) for (auto &token : prompt_tokens)
{ {