-
Notifications
You must be signed in to change notification settings - Fork 23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
add llama perf test case #555
Conversation
SharzyL
commented
May 3, 2024
- [nix] use single-float abi compilerrt
- [tests] fix t1.ld DDR map
- [emurt] support uart/print, export header
- [cases] add perf.llama
checkpoint_bin = fetchurl { | ||
url = "https://huggingface.co/karpathy/tinyllamas/resolve/main/stories15M.bin"; | ||
sha256 = "sha256-zVkGRNljhnorbloRB/UfrWY8QdecFJ++y7sflfqB9Jo="; | ||
}; | ||
|
||
tokenizer_bin = fetchurl { | ||
url = "https://github.com/karpathy/llama2.c/raw/b3c4b6c3c4bbff42e5211293280307019368ccb5/tokenizer.bin"; | ||
sha256 = "sha256-UKUu+CLunoPeXOnQvgoCWnc9AZQ39Ytf+dyvsGPs42E="; | ||
}; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
oh my
tests/perf/llama/run.c
Outdated
#if defined _WIN32 | ||
#include "win.h" | ||
#else |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
remove it?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Just following the upstream but totally ok to remove
https://github.com/karpathy/llama2.c/blob/master/run.c#L10-L12
ssize_t file_size; // size of the checkpoint file in bytes | ||
} Transformer; | ||
|
||
void malloc_run_state(RunState* s, Config* p) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we may need to allocate them to specific memory(SRAM) range in the future.
nice! |
please add documentation to run the test case. |
RTL still has some bug running llama. Will be fixed in the following PRs |
The llama still hang at spike side, in order to not block others commit, Let get this into master branch, and debug it later. |