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

llama : llama_perf + option to disable timings during decode (#9355)

Browse source 
* llama : llama_perf + option to disable timings during decode

ggml-ci

* common : add llama_arg

* Update src/llama.cpp

Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>

* perf : separate functions in the API

ggml-ci

* perf : safer pointer handling + naming update

ggml-ci

* minor : better local var name

* perf : abort on invalid sampler pointer

ggml-ci

---------

Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>
This commit is contained in:
Georgi Gerganov 2024-09-13 09:53:38 +03:00 • committed by GitHub
parent bd35cb0ae3
commit 0abc6a2c25
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
23 changed files with 135 additions and 91 deletions
Download patch file Download diff file Expand all files Collapse all files

8
common/arg.cpp
View file

@ -720,6 +720,14 @@ gpt_params_context gpt_params_parser_init(gpt_params & params, llama_example ex,
params.prompt = value;
}
));
add_opt(llama_arg(
{"--no-perf"},
format("disable internal libllama performance timings (default: %s)", params.no_perf ? "true" : "false"),
[](gpt_params & params) {
params.no_perf = true;
params.sparams.no_perf = true;
}
).set_env("LLAMA_ARG_NO_PERF"));
add_opt(llama_arg(
{"-f", "--file"}, "FNAME",
"a file containing the prompt (default: none)",