mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2025-01-06 08:44:35 +00:00
server : update slot->prompt after restore
This commit is contained in:
parent
61a66f25ab
commit
6556c90171
@ -1778,6 +1778,9 @@ struct server_context {
|
|||||||
}
|
}
|
||||||
slot->cache_tokens.resize(token_count);
|
slot->cache_tokens.resize(token_count);
|
||||||
|
|
||||||
|
// TODO: maybe detokenize the slot->cache_tokens instead?
|
||||||
|
slot->prompt = string_format("[restored %d tokens from file]", (int) token_count);
|
||||||
|
|
||||||
const int64_t t_end = ggml_time_us();
|
const int64_t t_end = ggml_time_us();
|
||||||
const double t_restore_ms = (t_end - t_start) / 1000.0;
|
const double t_restore_ms = (t_end - t_start) / 1000.0;
|
||||||
|
|
||||||
|
Loading…
Reference in New Issue
Block a user