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

ggml : add IQ2 to test-backend-ops + refactoring

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-01-14 13:15:30 +02:00
parent bee938da74
commit bc0bb3009c
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
7 changed files with 99 additions and 64 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml-backend.c
View file

@ -692,6 +692,8 @@ GGML_CALL static bool ggml_backend_cpu_graph_compute(ggml_backend_t backend, str
GGML_CALL static bool ggml_backend_cpu_supports_op(ggml_backend_t backend, const struct ggml_tensor * op) {
switch (op->op) {
case GGML_OP_CPY:
return op->type != GGML_TYPE_IQ2_XXS && op->type != GGML_TYPE_IQ2_XS; // missing type_traits.from_float
case GGML_OP_MUL_MAT:
return op->src[1]->type == GGML_TYPE_F32 || op->src[1]->type == ggml_internal_get_type_traits(op->src[0]->type).vec_dot_type;
default: