Skip to content

Switch to emplace_back to avoid extra object #5291

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.

Sign up for GitHub

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

Merged
merged 1 commit into from
Feb 3, 2024
Merged

Switch to emplace_back to avoid extra object #5291

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions common/common.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -515,7 +515,7 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
invalid_param = true;
break;
}
params.lora_adapter.push_back(std::make_tuple(argv[i], 1.0f));
params.lora_adapter.emplace_back(argv[i], 1.0f);
params.use_mmap = false;
} else if (arg == "--lora-scaled") {
if (++i >= argc) {
Expand All @@ -527,7 +527,7 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
invalid_param = true;
break;
}
params.lora_adapter.push_back(std::make_tuple(lora_adapter, std::stof(argv[i])));
params.lora_adapter.emplace_back(lora_adapter, std::stof(argv[i]));
params.use_mmap = false;
} else if (arg == "--lora-base") {
if (++i >= argc) {
Expand Down Expand Up @@ -664,7 +664,7 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
invalid_param = true;
break;
}
params.antiprompt.push_back(argv[i]);
params.antiprompt.emplace_back(argv[i]);
} else if (arg == "-ld" || arg == "--logdir") {
if (++i >= argc) {
invalid_param = true;
Expand Down Expand Up @@ -880,7 +880,7 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
}

if (!params.kv_overrides.empty()) {
params.kv_overrides.emplace_back(llama_model_kv_override());
params.kv_overrides.emplace_back();
params.kv_overrides.back().key[0] = 0;
}

Expand Down
34 changes: 17 additions & 17 deletions examples/llama-bench/llama-bench.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -948,46 +948,46 @@ struct markdown_printer : public printer {

void print_header(const cmd_params & params) override {
// select fields to print
fields.push_back("model");
fields.push_back("size");
fields.push_back("params");
fields.push_back("backend");
fields.emplace_back("model");
fields.emplace_back("size");
fields.emplace_back("params");
fields.emplace_back("backend");
bool is_cpu_backend = test::get_backend() == "CPU" || test::get_backend() == "BLAS";
if (!is_cpu_backend) {
fields.push_back("n_gpu_layers");
fields.emplace_back("n_gpu_layers");
}
if (params.n_threads.size() > 1 || params.n_threads != cmd_params_defaults.n_threads || is_cpu_backend) {
fields.push_back("n_threads");
fields.emplace_back("n_threads");
}
if (params.n_batch.size() > 1 || params.n_batch != cmd_params_defaults.n_batch) {
fields.push_back("n_batch");
fields.emplace_back("n_batch");
}
if (params.type_k.size() > 1 || params.type_k != cmd_params_defaults.type_k) {
fields.push_back("type_k");
fields.emplace_back("type_k");
}
if (params.type_v.size() > 1 || params.type_v != cmd_params_defaults.type_v) {
fields.push_back("type_v");
fields.emplace_back("type_v");
}
if (params.main_gpu.size() > 1 || params.main_gpu != cmd_params_defaults.main_gpu) {
fields.push_back("main_gpu");
fields.emplace_back("main_gpu");
}
if (params.split_mode.size() > 1 || params.split_mode != cmd_params_defaults.split_mode) {
fields.push_back("split_mode");
fields.emplace_back("split_mode");
}
if (params.mul_mat_q.size() > 1 || params.mul_mat_q != cmd_params_defaults.mul_mat_q) {
fields.push_back("mul_mat_q");
fields.emplace_back("mul_mat_q");
}
if (params.no_kv_offload.size() > 1 || params.no_kv_offload != cmd_params_defaults.no_kv_offload) {
fields.push_back("no_kv_offload");
fields.emplace_back("no_kv_offload");
}
if (params.tensor_split.size() > 1 || params.tensor_split != cmd_params_defaults.tensor_split) {
fields.push_back("tensor_split");
fields.emplace_back("tensor_split");
}
if (params.use_mmap.size() > 1 || params.use_mmap != cmd_params_defaults.use_mmap) {
fields.push_back("use_mmap");
fields.emplace_back("use_mmap");
}
fields.push_back("test");
fields.push_back("t/s");
fields.emplace_back("test");
fields.emplace_back("t/s");

fprintf(fout, "|");
for (const auto & field : fields) {
Expand Down
4 changes: 2 additions & 2 deletions examples/main/main.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -352,12 +352,12 @@ int main(int argc, char ** argv) {
// in instruct mode, we inject a prefix and a suffix to each input by the user
if (params.instruct) {
params.interactive_first = true;
params.antiprompt.push_back("### Instruction:\n\n");
params.antiprompt.emplace_back("### Instruction:\n\n");
}
// similar for chatml mode
else if (params.chatml) {
params.interactive_first = true;
params.antiprompt.push_back("<|im_start|>user\n");
params.antiprompt.emplace_back("<|im_start|>user\n");
}

// enable interactive mode if interactive start is specified
Expand Down
8 changes: 4 additions & 4 deletions examples/perplexity/perplexity.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -881,7 +881,7 @@ static void hellaswag_score(llama_context * ctx, const gpt_params & params) {
size_t li = hs_cur.common_prefix;
for (int s = 0; s < 4; ++s) {
for (size_t j = hs_cur.common_prefix; j < hs_cur.seq_tokens[s].size() - 1; j++) {
eval_pairs.push_back(std::make_pair(hs_cur.i_batch + li++, hs_cur.seq_tokens[s][j + 1]));
eval_pairs.emplace_back(hs_cur.i_batch + li++, hs_cur.seq_tokens[s][j + 1]);
}
++li;
}
Expand Down Expand Up @@ -1159,13 +1159,13 @@ static void winogrande_score(llama_context * ctx, const gpt_params & params) {
const int last_1st = task.seq_tokens[0].size() - n_base1 > 1 ? 1 : 0;
size_t li = n_base1 - 1;
for (size_t j = n_base1-1; j < task.seq_tokens[0].size()-1-last_1st; ++j) {
eval_pairs.push_back(std::make_pair(task.i_batch + li++, task.seq_tokens[0][j+1]));
eval_pairs.emplace_back(task.i_batch + li++, task.seq_tokens[0][j+1]);
}
const auto& n_base2 = skip_choice ? task.n_base2 : task.common_prefix;
const int last_2nd = task.seq_tokens[1].size() - n_base2 > 1 ? 1 : 0;
li = task.seq_tokens[0].size() - task.common_prefix + n_base2 - 1;
for (size_t j = n_base2-1; j < task.seq_tokens[1].size()-1-last_2nd; ++j) {
eval_pairs.push_back(std::make_pair(task.i_batch + li++, task.seq_tokens[1][j+1]));
eval_pairs.emplace_back(task.i_batch + li++, task.seq_tokens[1][j+1]);
}
}
compute_logprobs(batch_logits.data(), n_vocab, workers, eval_pairs, eval_results);
Expand Down Expand Up @@ -1524,7 +1524,7 @@ static void multiple_choice_score(llama_context * ctx, const gpt_params & params
size_t li = cur_task.common_prefix;
for (int s = 0; s < int(cur_task.seq_tokens.size()); ++s) {
for (size_t j = cur_task.common_prefix; j < cur_task.seq_tokens[s].size() - 1; j++) {
eval_pairs.push_back(std::make_pair(cur_task.i_batch + li++, cur_task.seq_tokens[s][j + 1]));
eval_pairs.emplace_back(cur_task.i_batch + li++, cur_task.seq_tokens[s][j + 1]);
}
++li;
}
Expand Down
4 changes: 2 additions & 2 deletions examples/quantize-stats/quantize-stats.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -257,13 +257,13 @@ int main(int argc, char ** argv) {
invalid_param = true;
break;
}
params.include_layers.push_back(argv[i]);
params.include_layers.emplace_back(argv[i]);
} else if (arg == "-L" || arg == "--exclude-layer") {
if (++i >= argc) {
invalid_param = true;
break;
}
params.exclude_layers.push_back(argv[i]);
params.exclude_layers.emplace_back(argv[i]);
} else if (arg == "-t" || arg == "--type") {
if (++i >= argc) {
invalid_param = true;
Expand Down
4 changes: 2 additions & 2 deletions examples/quantize/quantize.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -208,13 +208,13 @@ int main(int argc, char ** argv) {
}
} else if (strcmp(argv[arg_idx], "--include-weights") == 0) {
if (arg_idx < argc-1) {
included_weights.push_back(argv[++arg_idx]);
included_weights.emplace_back(argv[++arg_idx]);
} else {
usage(argv[0]);
}
} else if (strcmp(argv[arg_idx], "--exclude-weights") == 0) {
if (arg_idx < argc-1) {
excluded_weights.push_back(argv[++arg_idx]);
excluded_weights.emplace_back(argv[++arg_idx]);
} else {
usage(argv[0]);
}
Expand Down
8 changes: 4 additions & 4 deletions examples/server/server.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1884,7 +1884,7 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
invalid_param = true;
break;
}
sparams.api_keys.push_back(argv[i]);
sparams.api_keys.emplace_back(argv[i]);
}
else if (arg == "--api-key-file")
{
Expand Down Expand Up @@ -2160,7 +2160,7 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
invalid_param = true;
break;
}
params.lora_adapter.push_back(std::make_tuple(argv[i], 1.0f));
params.lora_adapter.emplace_back(argv[i], 1.0f);
params.use_mmap = false;
}
else if (arg == "--lora-scaled")
Expand All @@ -2176,7 +2176,7 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
invalid_param = true;
break;
}
params.lora_adapter.push_back(std::make_tuple(lora_adapter, std::stof(argv[i])));
params.lora_adapter.emplace_back(lora_adapter, std::stof(argv[i]));
params.use_mmap = false;
}
else if (arg == "--lora-base")
Expand Down Expand Up @@ -2318,7 +2318,7 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
}
}
if (!params.kv_overrides.empty()) {
params.kv_overrides.emplace_back(llama_model_kv_override());
params.kv_overrides.emplace_back();
params.kv_overrides.back().key[0] = 0;
}

Expand Down
2 changes: 1 addition & 1 deletion tests/test-llama-grammar.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,7 +105,7 @@ int main()

for (auto rule : expected_rules)
{
parsed_grammar.rules.push_back({});
parsed_grammar.rules.emplace_back();
for (auto element : rule)
{
parsed_grammar.rules.back().push_back(element);
Expand Down