vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Server: fix seed for multiple slots (#6835)

Browse source 
* Server: add tests for consistent results

* sampling: separate rng per sampling context
This commit is contained in:
Johannes Gäßler 2024-04-24 11:08:36 +02:00 • committed by GitHub
parent c0d1b3e03e
commit 28103f4832
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
11 changed files with 145 additions and 30 deletions
Download patch file Download diff file Expand all files Collapse all files

3
examples/server/server.cpp
View file

@ -854,7 +854,7 @@ struct server_context {
slot.sparams.penalize_nl = json_value(data, "penalize_nl", default_sparams.penalize_nl);
slot.params.n_keep = json_value(data, "n_keep", slot.params.n_keep);
slot.params.n_discard = json_value(data, "n_discard", default_params.n_discard);
slot.params.seed = json_value(data, "seed", default_params.seed);
slot.sparams.seed = json_value(data, "seed", default_sparams.seed);
slot.sparams.n_probs = json_value(data, "n_probs", default_sparams.n_probs);
slot.sparams.min_keep = json_value(data, "min_keep", default_sparams.min_keep);
@ -1028,7 +1028,6 @@ struct server_context {
send_error(task, "Failed to parse grammar", ERROR_TYPE_INVALID_REQUEST);
return false;
}
llama_set_rng_seed(ctx, slot.params.seed);
}
slot.command = SLOT_COMMAND_LOAD_PROMPT;