Skip to content

tests : add test-model-random #14139

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 10 commits into
base: master
Choose a base branch
from
Draft

tests : add test-model-random #14139

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
9cd402c
tests : add test-model-random
compilade Jun 11, 2025
7657835
tests : fix overflow and memory leaks in test-model-random
compilade Jun 12, 2025
8fe213a
tests : avoid sprintf in test-model-random
compilade Jun 12, 2025
61f6429
Merge branch 'master' into compilade/test-model-random
compilade Jun 13, 2025
dfa3c18
tests : add LLAMA, LLAMA4, and GEMMA2 to test-model-random
compilade Jun 14, 2025
352703b
test-model-random : better default tensor initialization distribution
compilade Jun 17, 2025
04b8f51
Merge branch 'master' into compilade/test-model-random
compilade Jun 17, 2025
9d873d7
test-model-random : shuffle across sequences but not within
compilade Jun 18, 2025
ccb2bb9
test-model-random : show max error
compilade Jun 18, 2025
7c3f9c2
Merge branch 'master' into compilade/test-model-random
compilade Jun 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 3 additions & 4 deletions src/llama-model.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9622,9 +9622,7 @@ struct llm_build_starcoder2 : public llm_graph_context {
};

struct llm_build_mamba : public llm_graph_context {
const llama_model & model;

llm_build_mamba(const llama_model & model, const llm_graph_params & params, ggml_cgraph * gf) : llm_graph_context(params), model(model) {
llm_build_mamba(const llama_model & model, const llm_graph_params & params, ggml_cgraph * gf) : llm_graph_context(params) {
ggml_tensor * cur;
ggml_tensor * inpL;

Expand All @@ -9642,7 +9640,7 @@ struct llm_build_mamba : public llm_graph_context {
LLM_NORM_RMS, il);
cb(cur, "attn_norm", il);

cur = build_mamba_layer(rs_inp, gf, cur, ubatch, il);
cur = build_mamba_layer(rs_inp, gf, cur, model, ubatch, il);

if (il == n_layer - 1 && inp_out_ids) {
cur = ggml_get_rows(ctx0, cur, inp_out_ids);
Expand Down Expand Up @@ -9681,6 +9679,7 @@ struct llm_build_mamba : public llm_graph_context {
llm_graph_input_rs * inp,
ggml_cgraph * gf,
ggml_tensor * cur,
const llama_model & model,
const llama_ubatch & ubatch,
int il) const {
const auto * mctx_cur = static_cast<const llama_memory_recurrent_context *>(mctx);
Expand Down
1 change: 1 addition & 0 deletions tests/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -145,6 +145,7 @@ if (NOT WIN32 OR NOT BUILD_SHARED_LIBS)
llama_build_and_test(test-grammar-integration.cpp)
llama_build_and_test(test-llama-grammar.cpp)
llama_build_and_test(test-chat.cpp)
llama_build_and_test(test-model-random.cpp)
# TODO: disabled on loongarch64 because the ggml-ci node lacks Python 3.8
if (NOT ${CMAKE_SYSTEM_PROCESSOR} MATCHES "loongarch64")
llama_build_and_test(test-json-schema-to-grammar.cpp WORKING_DIRECTORY ${PROJECT_SOURCE_DIR})
Expand Down
Loading
Loading