CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : handle models with missing EOS token (#8997)

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov
2024-08-12 10:21:50 +03:00
committed by GitHub
parent 4134999e01
commit 5ef07e25ac
1 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
examples/server/server.cpp
View File

@@ -631,6 +631,7 @@ struct server_context {
bool clean_kv_cache = true;
bool add_bos_token = true;
bool has_eos_token = false;
int32_t n_ctx; // total context for all clients / slots
@@ -693,7 +694,7 @@ struct server_context {
n_ctx = llama_n_ctx(ctx);
add_bos_token = llama_should_add_bos_token(model);
GGML_ASSERT(llama_add_eos_token(model) != 1);
has_eos_token = llama_add_eos_token(model) != 1;
return true;
}
@@ -1031,7 +1032,7 @@ struct server_context {
{
slot.sparams.logit_bias.clear();
if (json_value(data, "ignore_eos", false)) {
if (json_value(data, "ignore_eos", false) && has_eos_token) {
slot.sparams.logit_bias[llama_token_eos(model)] = -INFINITY;
}