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

llama : allow for user specified embedding pooling type (#5849)

Browse source 
* allow for user specified pooling type

* llama : use enum types over int

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Douglas Hanley 2024-03-03 04:40:27 -06:00 • committed by GitHub
parent 87c2e8b279
commit 475df1d6cf
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 60 additions and 29 deletions
Download patch file Download diff file Expand all files Collapse all files

13
common/common.cpp
View file

@ -335,6 +335,16 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
break;
}
params.yarn_beta_slow = std::stof(argv[i]);
} else if (arg == "--pooling") {
if (++i >= argc) {
invalid_param = true;
break;
}
std::string value(argv[i]);
/**/ if (value == "none") { params.pooling_type = LLAMA_POOLING_TYPE_NONE; }
else if (value == "mean") { params.pooling_type = LLAMA_POOLING_TYPE_MEAN; }
else if (value == "cls") { params.pooling_type = LLAMA_POOLING_TYPE_CLS; }
else { invalid_param = true; break; }
} else if (arg == "--defrag-thold" || arg == "-dt") {
if (++i >= argc) {
invalid_param = true;
@ -1014,6 +1024,8 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
printf(" --yarn-attn-factor N YaRN: scale sqrt(t) or attention magnitude (default: 1.0)\n");
printf(" --yarn-beta-slow N YaRN: high correction dim or alpha (default: %.1f)\n", params.yarn_beta_slow);
printf(" --yarn-beta-fast N YaRN: low correction dim or beta (default: %.1f)\n", params.yarn_beta_fast);
printf(" --pooling {none,mean,cls}\n");
printf(" pooling type for embeddings, use model default if unspecified\n");
printf(" -dt N, --defrag-thold N\n");
printf(" KV cache defragmentation threshold (default: %.1f, < 0 - disabled)\n", params.defrag_thold);
printf(" --ignore-eos ignore end of stream token and continue generating (implies --logit-bias 2-inf)\n");
@ -1296,6 +1308,7 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
cparams.yarn_beta_fast = params.yarn_beta_fast;
cparams.yarn_beta_slow = params.yarn_beta_slow;
cparams.yarn_orig_ctx = params.yarn_orig_ctx;
cparams.pooling_type = params.pooling_type;
cparams.defrag_thold = params.defrag_thold;
cparams.offload_kqv = !params.no_kv_offload;