mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-25 02:44:36 +00:00
server : fix crash when prompt exceeds context size (#3996)
This commit is contained in:
parent
34b0a08207
commit
d96ca7ded7
@ -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)
|
||||||
{
|
{
|
||||||
|
Loading…
Reference in New Issue
Block a user