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

fix softmax in baby-llama example

Browse source
This commit is contained in:
xaedes 2023-05-01 20:02:48 +02:00
parent 8fde656d24
commit 29a0f8b940
No known key found for this signature in database
GPG key ID: 30030EDD817EA2B1
1 changed files with 21 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

26
examples/baby-llama/baby-llama.cpp
View file

@ -520,10 +520,16 @@ void sample_softmax(struct ggml_tensor * logits, struct ggml_tensor * probs, str
ggml_set_i32_1d(best_samples, i, k); ggml_set_i32_1d(best_samples, i, k);
} }
} }
float psum = 0;
for (int k = 0; k < logits->ne[0]; ++k) { for (int k = 0; k < logits->ne[0]; ++k) {
float logit = ggml_get_f32_1d(logits, i * logits->ne[0] + k); float logit = ggml_get_f32_1d(logits, i * logits->ne[0] + k);
float p = expf(logit - max_logit); float p = (logit == -INFINITY) ? 0 : expf(logit - max_logit);
ggml_set_i32_1d(probs, i * probs->ne[0] + k, p); psum += p;
ggml_set_f32_1d(probs, i * probs->ne[0] + k, p);
}
for (int k = 0; k < logits->ne[0]; ++k) {
float p = ggml_get_f32_1d(probs, i*probs->ne[0] + k);
ggml_set_f32_1d(probs, i * probs->ne[0] + k, p / psum);
} }
} }
} }
@ -532,7 +538,7 @@ void print_probs(struct ggml_tensor * probs) {
assert(probs->n_dims == 2); assert(probs->n_dims == 2);
for (int i=0; i<probs->ne[1]; ++i) { for (int i=0; i<probs->ne[1]; ++i) {
for (int k = 0; k < probs->ne[0]; ++k) { for (int k = 0; k < probs->ne[0]; ++k) {
float p = ggml_get_f32_1d(probs, i*probs->ne[1] + k); float p = ggml_get_f32_1d(probs, i*probs->ne[0] + k);
printf(" %.1f", p); printf(" %.1f", p);
} }
printf("\n"); printf("\n");
@ -600,13 +606,13 @@ int main(int argc, char ** argv) {
int token = (int)(z*(float)(model.hparams.n_vocab-1)); int token = (int)(z*(float)(model.hparams.n_vocab-1));
for (int k = 0; k < token; ++k) { for (int k = 0; k < token; ++k) {
printf(" "); printf(" ");
ggml_set_f32_1d(targets, i*model.hparams.n_vocab + k, -1.0f); ggml_set_f32_1d(targets, i*model.hparams.n_vocab + k, 0.0f);
} }
printf("X"); printf("X");
ggml_set_f32_1d(targets, i*model.hparams.n_vocab + token, +1.0f); ggml_set_f32_1d(targets, i*model.hparams.n_vocab + token, +1.0f);
for (int k = token+1; k < model.hparams.n_vocab; ++k) { for (int k = token+1; k < model.hparams.n_vocab; ++k) {
printf(" "); printf(" ");
ggml_set_f32_1d(targets, i*model.hparams.n_vocab + k, -1.0f); ggml_set_f32_1d(targets, i*model.hparams.n_vocab + k, 0.0f);
} }
printf("\n"); printf("\n");
ggml_set_i32_1d(tokens_input, i, token); ggml_set_i32_1d(tokens_input, i, token);
@ -626,6 +632,11 @@ int main(int argc, char ** argv) {
float error_before_opt = ggml_get_f32_1d(e, 0); float error_before_opt = ggml_get_f32_1d(e, 0);
sample_softmax(logits, before_opt_probs, before_opt_best_samples); sample_softmax(logits, before_opt_probs, before_opt_best_samples);
printf("probabilities before optimization:\n");
print_probs(before_opt_probs);
printf("best samples before optimization:\n");
print_tokens(before_opt_best_samples, model.hparams.n_vocab);
struct ggml_opt_params opt_params = ggml_opt_default_params(GGML_OPT_ADAM); struct ggml_opt_params opt_params = ggml_opt_default_params(GGML_OPT_ADAM);
ggml_opt(ctx0, opt_params, e); ggml_opt(ctx0, opt_params, e);
// //
@ -638,11 +649,6 @@ int main(int argc, char ** argv) {
printf("error_before_opt: %.2f\n", error_before_opt); printf("error_before_opt: %.2f\n", error_before_opt);
printf("error_after_opt: %.2f\n", error_after_opt); printf("error_after_opt: %.2f\n", error_after_opt);
printf("probabilities before optimization:\n");
print_probs(before_opt_probs);
printf("best samples before optimization:\n");
print_tokens(before_opt_best_samples, model.hparams.n_vocab);
printf("probabilities after optimization:\n"); printf("probabilities after optimization:\n");
print_probs(after_opt_probs); print_probs(after_opt_probs);
printf("best samples after optimization:\n"); printf("best samples after optimization:\n");