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

add train params to specify memory size

Browse source
This commit is contained in:
xaedes 2023-05-30 16:06:20 +02:00
parent ad966da955
commit 1074a81e81
No known key found for this signature in database
GPG key ID: 30030EDD817EA2B1
1 changed files with 22 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

24
examples/baby-llama/baby-llama-text.cpp
View file

@ -2122,6 +2122,9 @@ struct train_params {
int adam_n_iter; int adam_n_iter;
float adam_alpha; float adam_alpha;
float adam_decay; float adam_decay;
int mem_model_gb;
int mem_compute_gb;
}; };
struct train_params get_default_train_params() { struct train_params get_default_train_params() {
@ -2164,6 +2167,9 @@ struct train_params get_default_train_params() {
params.adam_alpha = 1e-3; params.adam_alpha = 1e-3;
params.adam_decay = 1e-3; params.adam_decay = 1e-3;
params.mem_model_gb = 2;
params.mem_compute_gb = 32;
return params; return params;
} }
@ -2203,6 +2209,8 @@ void train_print_usage(int /*argc*/, char ** argv, const struct train_params * p
fprintf(stderr, " --adam-iter N Maximum number of Adam optimization iterations for each batch (default %d)\n", params->adam_n_iter); fprintf(stderr, " --adam-iter N Maximum number of Adam optimization iterations for each batch (default %d)\n", params->adam_n_iter);
fprintf(stderr, " --adam-alpha N Adam learning rate alpha (default %f)\n", params->adam_alpha); fprintf(stderr, " --adam-alpha N Adam learning rate alpha (default %f)\n", params->adam_alpha);
fprintf(stderr, " --adam-decay N AdamW weight decay. Values greater zero enable AdamW instead of regular Adam. (default %f)\n", params->adam_decay); fprintf(stderr, " --adam-decay N AdamW weight decay. Values greater zero enable AdamW instead of regular Adam. (default %f)\n", params->adam_decay);
fprintf(stderr, " --mem-model N Memory to allocate for model and cache in gigabytes. (default %d)\n", params->mem_model_gb);
fprintf(stderr, " --mem-compute N Memory to allocate for compute in gigabytes. (default %d)\n", params->mem_compute_gb);
fprintf(stderr, "\n"); fprintf(stderr, "\n");
} }
@ -2384,6 +2392,18 @@ bool train_params_parse(int argc, char ** argv, struct train_params * params) {
break; break;
} }
params->adam_decay = std::stof(argv[i]); params->adam_decay = std::stof(argv[i]);
} else if (arg == "--mem-model") {
if (++i >= argc) {
invalid_param = true;
break;
}
params->mem_model_gb = std::stoi(argv[i]);
} else if (arg == "--mem-compute") {
if (++i >= argc) {
invalid_param = true;
break;
}
params->mem_compute_gb = std::stoi(argv[i]);
} else if (arg == "-h" || arg == "--help") { } else if (arg == "-h" || arg == "--help") {
train_print_usage(argc, argv, &default_params); train_print_usage(argc, argv, &default_params);
exit(0); exit(0);
@ -2480,7 +2500,7 @@ int main(int argc, char ** argv) {
struct ggml_init_params lcparams; struct ggml_init_params lcparams;
lcparams.mem_size = 1024ll*1024ll*1024ll*2ll; lcparams.mem_size = 1024ll*1024ll*1024ll*((size_t) params.mem_model_gb);
lcparams.mem_buffer = NULL; lcparams.mem_buffer = NULL;
lcparams.no_alloc = false; lcparams.no_alloc = false;
@ -2536,7 +2556,7 @@ int main(int argc, char ** argv) {
printf("used_mem model+cache: %zu bytes\n", ggml_used_mem(model.ctx)); printf("used_mem model+cache: %zu bytes\n", ggml_used_mem(model.ctx));
// ggml_print_tensor_objects(model.ctx); // ggml_print_tensor_objects(model.ctx);
size_t compute_size = 1024ll*1024ll*1024ll*32ll; size_t compute_size = 1024ll*1024ll*1024ll*((size_t) params.mem_compute_gb);
uint8_t * compute_addr = new uint8_t[compute_size]; uint8_t * compute_addr = new uint8_t[compute_size];
GGML_ASSERT(train_tokens.size() > n_tokens);; GGML_ASSERT(train_tokens.size() > n_tokens);;