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

remove trailing whitespace

Browse source
This commit is contained in:
xaedes 2023-05-11 20:03:18 +02:00
parent 581e5eb954
commit b9ef08ccab
No known key found for this signature in database
GPG key ID: 30030EDD817EA2B1
1 changed files with 38 additions and 38 deletions
Download patch file Download diff file Expand all files Collapse all files

76
examples/baby-llama/baby-llama.cpp
View file

@ -825,15 +825,15 @@ struct ggml_tensor * forward_batch(
// compute the transposed [N, n_embd] V matrix // compute the transposed [N, n_embd] V matrix
// wv shape [n_embd, n_embd, 1, 1] // wv shape [n_embd, n_embd, 1, 1]
// Vcur shape [N, n_embd, n_batch, 1] // Vcur shape [N, n_embd, n_batch, 1]
struct ggml_tensor * Vcur = ggml_cont(ctx0, struct ggml_tensor * Vcur = ggml_cont(ctx0,
ggml_permute(ctx0, ggml_permute(ctx0,
ggml_reshape_3d(ctx0, ggml_reshape_3d(ctx0,
ggml_mul_mat(ctx0, ggml_mul_mat(ctx0,
model->layers[il].wv, model->layers[il].wv,
cur), cur),
n_embd, N, n_batch), n_embd, N, n_batch),
1, 0, 2, 3)); 1, 0, 2, 3));
assert_shape_3d(Vcur, N, n_embd, n_batch); assert_shape_3d(Vcur, N, n_embd, n_batch);
// kv_self.k shape [n_embd * n_ctx * n_batch * n_layer] // kv_self.k shape [n_embd * n_ctx * n_batch * n_layer]
@ -852,12 +852,12 @@ struct ggml_tensor * forward_batch(
ggml_build_forward_expand(gf, ggml_cpy(ctx0, Vcur, v)); ggml_build_forward_expand(gf, ggml_cpy(ctx0, Vcur, v));
} //*/ } //*/
kc = ggml_set_2d(ctx0, kc, kc = ggml_set_2d(ctx0, kc,
ggml_reshape_2d(ctx0, Kcur, n_embd*N, n_batch), ggml_reshape_2d(ctx0, Kcur, n_embd*N, n_batch),
ggml_element_size(kc)*n_embd*n_ctx, ggml_element_size(kc)*n_embd*n_ctx,
(ggml_element_size(kc)*n_embd)*(il*n_batch*n_ctx + n_past)); (ggml_element_size(kc)*n_embd)*(il*n_batch*n_ctx + n_past));
vc = ggml_set_2d(ctx0, vc, vc = ggml_set_2d(ctx0, vc,
ggml_reshape_2d(ctx0, Vcur, N*n_embd, n_batch), ggml_reshape_2d(ctx0, Vcur, N*n_embd, n_batch),
ggml_element_size(vc)*n_ctx*n_embd, ggml_element_size(vc)*n_ctx*n_embd,
ggml_element_size(vc)*(n_past + il*n_embd*n_batch*n_ctx)); ggml_element_size(vc)*(n_past + il*n_embd*n_batch*n_ctx));
@ -878,10 +878,10 @@ struct ggml_tensor * forward_batch(
struct ggml_tensor * K = struct ggml_tensor * K =
ggml_permute(ctx0, ggml_permute(ctx0,
ggml_reshape_4d(ctx0, ggml_reshape_4d(ctx0,
ggml_view_3d(ctx0, ggml_view_3d(ctx0,
kc, kc,
n_embd, n_embd,
(n_past + N), (n_past + N),
n_batch, n_batch,
n_embd*ggml_element_size(kc), n_embd*ggml_element_size(kc),
n_ctx*n_embd*ggml_element_size(kc), n_ctx*n_embd*ggml_element_size(kc),
@ -1036,7 +1036,7 @@ struct ggml_tensor * forward_batch(
{ {
// inpL shape [n_vocab,N,n_batch,1] // inpL shape [n_vocab,N,n_batch,1]
inpL = ggml_reshape_3d(ctx0, inpL = ggml_reshape_3d(ctx0,
inpL, inpL,
n_vocab, N, n_batch); n_vocab, N, n_batch);
assert_shape_3d(inpL, n_vocab, N, n_batch); assert_shape_3d(inpL, n_vocab, N, n_batch);
@ -1346,23 +1346,23 @@ void sample_softmax_batch(struct ggml_context * ctx, struct ggml_tensor * logits
GGML_ASSERT(n_vocab == probs->ne[0]); GGML_ASSERT(n_vocab == probs->ne[0]);
GGML_ASSERT(n_tokens == probs->ne[1]); GGML_ASSERT(n_tokens == probs->ne[1]);
GGML_ASSERT(n_batch == probs->ne[2]); GGML_ASSERT(n_batch == probs->ne[2]);
for (int k=0; k<n_batch; ++k) { for (int k=0; k<n_batch; ++k) {
struct ggml_tensor * best_samples_k = ggml_view_1d(ctx, struct ggml_tensor * best_samples_k = ggml_view_1d(ctx,
best_samples, best_samples,
best_samples->ne[0], best_samples->ne[0],
k*best_samples->nb[1]); k*best_samples->nb[1]);
struct ggml_tensor * logits_k = ggml_view_2d(ctx, struct ggml_tensor * logits_k = ggml_view_2d(ctx,
logits, logits,
logits->ne[0], logits->ne[0],
logits->ne[1], logits->ne[1],
logits->nb[1], logits->nb[1],
k*logits->nb[2]); k*logits->nb[2]);
struct ggml_tensor * probs_k = ggml_view_2d(ctx, struct ggml_tensor * probs_k = ggml_view_2d(ctx,
probs, probs,
probs->ne[0], probs->ne[0],
probs->ne[1], probs->ne[1],
probs->nb[1], probs->nb[1],
k*probs->nb[2]); k*probs->nb[2]);
sample_softmax(logits_k, probs_k, best_samples_k); sample_softmax(logits_k, probs_k, best_samples_k);
} }
@ -1436,15 +1436,15 @@ void get_example_targets_batch(struct ggml_context * ctx, int example_id, struct
GGML_ASSERT(n_batch == targets->ne[2]); GGML_ASSERT(n_batch == targets->ne[2]);
for (int k=0; k<n_batch; ++k) { for (int k=0; k<n_batch; ++k) {
struct ggml_tensor * tokens_input_k = ggml_view_1d(ctx, struct ggml_tensor * tokens_input_k = ggml_view_1d(ctx,
tokens_input, tokens_input,
tokens_input->ne[0], tokens_input->ne[0],
k*tokens_input->nb[1]); k*tokens_input->nb[1]);
struct ggml_tensor * targets_k = ggml_view_2d(ctx, struct ggml_tensor * targets_k = ggml_view_2d(ctx,
targets, targets,
targets->ne[0], targets->ne[0],
targets->ne[1], targets->ne[1],
targets->nb[1], targets->nb[1],
k*targets->nb[2]); k*targets->nb[2]);
get_example_targets(example_id*n_batch + k, tokens_input_k, targets_k); get_example_targets(example_id*n_batch + k, tokens_input_k, targets_k);
} }