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

add m-rope testcase to test-backend-ops

Browse source
This commit is contained in:
HimariO 2024-12-09 22:13:11 +08:00
parent 3ba7664de9
commit b24ab8634f
1 changed files with 34 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

38
tests/test-backend-ops.cpp
View file

@ -2201,7 +2201,15 @@ struct test_rope : public test_case {
ggml_set_name(a, "a");
}
ggml_tensor * pos = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, ne_a[2]);
const bool is_mrope = mode & GGML_ROPE_TYPE_MROPE;
const bool is_vision = mode == GGML_ROPE_TYPE_VISION;
ggml_tensor * pos;
if (is_mrope || is_vision) {
pos = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, ne_a[2] * 4);
} else {
pos = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, ne_a[2]);
}
ggml_set_name(pos, "pos");
ggml_tensor * freq = nullptr;
@ -2210,7 +2218,20 @@ struct test_rope : public test_case {
ggml_set_name(freq, "freq");
}
ggml_tensor * out = ggml_rope_ext(ctx, a, pos, freq, n_dims, mode, 0, 10000.0f, fs, ef, af, 1.0f, 1.0f);
ggml_tensor * out;
if (is_mrope) {
if (is_vision) {
GGML_ASSERT(n_dims/4 > 0);
int rope_sections[4] = {n_dims/4, n_dims/4, 0, 0}; // Vision-RoPE only use first two dimension for image (x, y) coordinate
out = ggml_rope_multi(ctx, a, pos, freq, n_dims/2, rope_sections, mode, 0, 10000.0f, fs, ef, af, 1.0f, 1.0f);
} else {
GGML_ASSERT(n_dims/3 > 0);
int rope_sections[4] = {n_dims/3, n_dims/3, n_dims/3, 0};
out = ggml_rope_multi(ctx, a, pos, freq, n_dims, rope_sections, mode, 0, 10000.0f, fs, ef, af, 1.0f, 1.0f);
}
} else {
out = ggml_rope_ext(ctx, a, pos, freq, n_dims, mode, 0, 10000.0f, fs, ef, af, 1.0f, 1.0f);
}
ggml_set_name(out, "out");
return out;
@ -2220,11 +2241,12 @@ struct test_rope : public test_case {
for (ggml_tensor * t = ggml_get_first_tensor(ctx); t != NULL; t = ggml_get_next_tensor(ctx, t)) {
if (t->type == GGML_TYPE_I32) {
// pos
std::vector<int> data(ne_a[2]);
for (int i = 0; i < ne_a[2]; i++) {
const int num_pos_ids = (mode & GGML_ROPE_TYPE_MROPE) ? ne_a[2] * 4 : ne_a[2];
std::vector<int> data(num_pos_ids);
for (int i = 0; i < num_pos_ids; i++) {
data[i] = rand() % n_ctx;
}
ggml_backend_tensor_set(t, data.data(), 0, ne_a[2] * sizeof(int));
ggml_backend_tensor_set(t, data.data(), 0, num_pos_ids * sizeof(int));
} else {
if (t->ne[0] == n_dims/2) {
// frequency factors in the range [0.9f, 1.1f]
@ -3766,6 +3788,12 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_eval() {
test_cases.emplace_back(new test_rope(type, { 80, 32, 2, 1}, 32, 2, 512, fs, ef, af, ff, v)); // neox (phi-2)
}
if (all) {
test_cases.emplace_back(new test_rope(type, {128, 12, 2, 1}, 128, GGML_ROPE_TYPE_MROPE, 512, fs, ef, af, ff, v)); // rope_multi,m-rope (qwen2vl 2B)
test_cases.emplace_back(new test_rope(type, {128, 28, 2, 1}, 128, GGML_ROPE_TYPE_MROPE, 512, fs, ef, af, ff, v)); // rope_multi,m-rope (qwen2vl 7B)
test_cases.emplace_back(new test_rope(type, { 80, 16, 2, 1}, 80, GGML_ROPE_TYPE_VISION, 512, fs, ef, af, ff, v)); // rope_multi,m-rope (qwen2vl ViT)
}
test_cases.emplace_back(new test_rope(type, { 64, 128, 2, 1}, 64, 2, 512, fs, ef, af, ff, v)); // neox (falcon 40B)
}
}