ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-26 16:14:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

WIP KQ binary mask: make it a parameter, turn on via command line

Browse Source 
It is a pain to implement binary mask to 32-bit value conversion on
NEON and AVX2, so I decided to make the binary mask optional

There is also a commented out (and not working) attempt for NEON
in this commit.
This commit is contained in:
Iwan Kawrakow
2024-08-28 15:01:02 +02:00
parent fe825ecbe4
commit 05f95229a7
5 changed files with 72 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
common/common.cpp
View File

@@ -808,6 +808,10 @@ bool gpt_params_find_arg(int argc, char ** argv, const std::string & arg, gpt_pa
params.flash_attn = true;
return true;
}
if (arg == "-bkq" || arg == "--binary-kq") {
params.binary_kq = true;
return true;
}
if (arg == "-co" || arg == "--color") {
params.use_color = true;
return true;
@@ -1442,6 +1446,7 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
options.push_back({ "*", " --keep N", "number of tokens to keep from the initial prompt (default: %d, -1 = all)", params.n_keep });
options.push_back({ "*", " --chunks N", "max number of chunks to process (default: %d, -1 = all)", params.n_chunks });
options.push_back({ "*", "-fa, --flash-attn", "enable Flash Attention (default: %s)", params.flash_attn ? "enabled" : "disabled" });
options.push_back({ "*", "-bkq, --binary-kq", "enable binary KQ mask (default: %s)", params.binary_kq ? "enabled" : "disabled" });
options.push_back({ "*", "-p, --prompt PROMPT", "prompt to start generation with\n"
"in conversation mode, this will be used as system prompt\n"
"(default: '%s')", params.prompt.c_str() });
@@ -2265,6 +2270,7 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
cparams.cb_eval_user_data = params.cb_eval_user_data;
cparams.offload_kqv = !params.no_kv_offload;
cparams.flash_attn = params.flash_attn;
cparams.binary_kq = params.binary_kq;
cparams.type_k = kv_cache_type_from_str(params.cache_type_k);
cparams.type_v = kv_cache_type_from_str(params.cache_type_v);
@@ -3261,6 +3267,7 @@ void yaml_dump_non_result_info(FILE * stream, const gpt_params & params, const l
fprintf(stream, "simple_io: %s # default: false\n", params.simple_io ? "true" : "false");
fprintf(stream, "cont_batching: %s # default: false\n", params.cont_batching ? "true" : "false");
fprintf(stream, "flash_attn: %s # default: false\n", params.flash_attn ? "true" : "false");
fprintf(stream, "binary_kq: %s # default: false\n", params.binary_kq ? "true" : "false");
fprintf(stream, "temp: %f # default: 0.8\n", sparams.temp);
const std::vector<float> tensor_split_vector(params.tensor_split, params.tensor_split + llama_max_devices());