CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-27 08:21:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

Vulkan: add conv_transpose_2d operation (#16022)

Browse Source 
* Vulkan: add conv_transpose_2d operation

* Vulkan: fix typo in conv_transpose_2d shader(s0mp, s0L, s1mp, s1L)

* Vulkan: fix incorrect indentation in conv_transpose_2d shader

* Vulkan: add checking the push constants size limit and reuse conv2d_mm.comp for conv_transpose_2d operation

* Vulkan: revert the order of the index calculation and bound check in conv_2d shader

* Vulkan: explicity check push constants limit in supports_op() for conv_transpose_2d operation.

* Vulkan: remove unnecessary lower bound checks for H/W_idx in the conv_2d shader.
This commit is contained in:
Shin-myoung-serp
2025-09-22 17:04:01 +09:00
committed by GitHub
parent b2d980fce0
commit 96fdca043b
4 changed files with 229 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/test-backend-ops.cpp
View File

@@ -3969,6 +3969,10 @@ struct test_conv_transpose_2d : public test_case {
return VARS_TO_STR3(ne_input, ne_kernel, stride);
}
double max_nmse_err() override {
return 5e-4; // The default 1e-7 is too small for Vulkan.
}
test_conv_transpose_2d(std::array<int64_t, 4> ne_input = {10, 10, 3, 1}, // [input_width, input_height, input_channels, 1]
std::array<int64_t, 4> ne_kernel = {3, 3, 3, 1}, // [kernel_width, kernel_height, input_channels, 1]
int stride = 1)